Poly(urethane-isocyanurate) Networks
Transcript of Poly(urethane-isocyanurate) Networks
Poly(urethane-isocyanurate) NetworksSynthesis, Properties and Applications
Piet Driest
Poly(urethane-isocyanurate) Netw
orks Piet D
riest
POLY(URETHANE-ISOCYANURATE)
NETWORKS
Synthesis, properties and applications
Pieter Job Driest
POLY(URETHANE-ISOCYANURATE)
NETWORKS
Synthesis, properties and applications
DISSERTATION
to obtain
the degree of doctor at the Universiteit Twente,
on the authority of the rector magnificus,
Prof. dr. T.T.M. Palstra,
on account of the decision of the graduation committee
to be publicly defended
on Friday 31 January 2020 at 12.45 hr.
by
Pieter Job Driest
born on the 6th of September, 1987
in Amsterdam, the Netherlands
This dissertation has been approved by:
Supervisors
Prof. dr. D.W. Grijpma
Prof. dr. D. Stamatialis
Cover design: Douwe Oppewal Printed by: Ipskamp Printing Lay-out: Piet Driest
ISBN: 978-90-365-4934-9
DOI: 10.3990/1.9789036549349
© 2020 Pieter Job Driest, The Netherlands. All rights reserved. No parts of
this thesis may be reproduced, stored in a retrieval system or transmitted in
any form or by any means without permission of the author. Alle rechten
voorbehouden. Niets uit deze uitgave mag worden vermenigvuldigd, in
enige vorm of op enige wijze, zonder voorafgaande schriftelijke
toestemming van de auteur.
Graduation Committee:
Chairman / Secretary Prof. dr. J.L. Herek
Supervisors: Prof. dr. D.W. Grijpma
Prof. dr. D. Stamatialis
Committee Members: Dr. D.J. Dijkstra
Prof. dr. G.J. Vancso
Prof. dr. P.J. Dijkstra
Prof. dr. K.U. Loos
Prof. dr. L.H. Koole
Imagination is not only the uniquely human capacity to envision that which
is not, and therefore the fount of all invention and innovation. In its
arguably most transformative and revelatory capacity, it is the power that
enables us to empathise with humans whose experiences we have never
shared.
J.K. Rowling
Harvard Commencement Speech │ June 5, 2008
Table of Contents
Chapter 1 General Introduction: Scope, Aims and Outline
of this Thesis
10
Chapter 2 Poly(urethane-isocyanurate)s: Chemistry,
Network Formation and Applications
18
Chapter 3 Aliphatic Isocyanurates and Polyisocyanurate
Networks
50
Chapter 4 The Influence of Molecular Interactions on the
Viscosity of Aliphatic Polyisocyanate Precursors
for Polyurethane Coatings
70
Chapter 5 The Trimerization of Isocyanate-functionalized
Prepolymers: an Effective Method for
Synthesizing Well-defined Polymer Networks
92
Chapter 6 Combinatorial Networks by Trimerization of
Mixtures of NCO-functionalized Prepolymers
110
Chapter 7 Structure-property Relations in Semi-crystalline
Combinatorial Poly(urethane-isocyanurate)
Hydrogels
138
Chapter 8 Poly(ethylene glycol) Poly(urethane-
isocyanurate) Type Hydrogels as a New Class of
Materials for Contact Lens Applications
154
Summary and Outlook 184
Acknowledgements | Dankwoord | Danksagung 194
List of Publications 204
Chapter 1 - General Introduction: Scope,
Aims and Outline of this Thesis
12
1.1 General Introduction
2019 has been pronounced the International Year of the Periodic Table, a
milestone year in chemistry in which we celebrate the 150th anniversary of
Dimitri Mendeleev’s discovery of the periodic table of the elements.1 Not
only did his ordering of the elements known at the time based on their
atomic weights and general properties lead him to (correctly!) predict which
ones were missing, it also paved the way for the creation of an international,
unified chemical language. This concept was formalized 50 years later when
the International Union of Pure and Applied Chemistry (IUPAC, celebrating
its 100th birthday in 2019) came to life. “A neutral and objective scientific
organization, IUPAC was established in 1919 by academic and industrial
chemists who shared a common goal – to unite a fragmented, global
chemistry community for the advancement of the chemical sciences via
collaboration and the free exchange of scientific information. Throughout its
long history, IUPAC has fulfilled that goal through the creation of a common
language and the standardization of processes and procedures.”2
Today, 150 years after Mendeleev first published his ordering of the
elements and 100 years after chemists worldwide united behind a shared
cause, this mission is as relevant as ever. The challenges in the field of
chemistry have obviously changed over time and presently include subjects
like climate change, the depletion of the earth’s natural reserves and the
upcoming of new scientific fields for which the ethical guidelines have not
been written yet.3 Relevant chemical topics are explicitly taken up in the
United Nations’ Sustainable Development Goals for 2030 and include the
responsible use of resources and the sustainable development and
production of materials (nrs. 9 and 12).4 Clearly, the present-day global
chemistry community formally created in 1919 will play an important role in
dealing with these issues. But where the challenges have changed over time,
the solution has not. As before, international collaboration, a free exchange
of scientific information and a common language are the key to realizing our
ambitions.
13
The Marie Curie ITN project “TheLink” is an EU-funded international,
interdisciplinary research project which aims to contribute to this ideal. As
part of this project, the work presented in this thesis was performed in the
labs of Covestro Deutschland AG (formerly Bayer MaterialScience) and was
scientifically coordinated by the University of Twente. With this approach,
we mean to illustrate the possibilities for synergetic collaboration between
industrial and academic research partners. Facilitating the free exchange of
scientific information, all research output presented in this thesis has been
(or will be) published Open Access. Finally, by bringing together researchers
with different nationalities, native tongues and cultural backgrounds, we
mean to contribute to the (further) development of an international,
common language in chemistry and material science.
1.2 Scope, Aims and Outline of this Thesis
Scope and Aims
Throughout this thesis, a recurring aim is to generate a link between
conventionally separated research communities. This may concern a
separation in the research objective (industrial / academic), the research
approach (computational / experimental) or the research discipline
(chemistry / physics / biology). We aim to establish this by performing
fundamental, scientific research based on industrially relevant chemical
processes and materials. Also, different scientific approaches and/or
disciplines are combined in each of the studies presented in this thesis.
The main subject of this work is the investigation and development
of poly(urethane-isocyanurate) (PUI) polymer networks for biomedical
applications. In the first part of this thesis, the viscosity of liquid aliphatic
isocyanurates is investigated. These compounds are important industrial
precursors in the conventional 2-component synthesis of PUI networks.5 By
isolating series of isocyanate-functional and non-functional model
compounds in high purity, the molecular interactions that dictate the
viscosity of these compounds are investigated. The experimental results are
complemented by a detailed computational study on the same series of
14
model compounds. Using this combined experimental and theoretical
approach, we aim to provide a first framework of the molecular interactions
that dictate the viscosity of liquid aliphatic isocyanurates.
In the second part of this thesis, focus is gradually shifted from raw
materials towards products. Network formation and the molecular network
structure of PUI type polymer networks are investigated. We develop a new
synthesis method for PUI type polymer networks, aiming to address some
of the drawbacks that are typically associated with the conventional 2-
component synthesis of these materials. Also here, we aim to provide
fundamental scientific understanding of industrially relevant materials and
make a correlation between the molecular structure of the PUI networks
and the physical and biological properties observed macroscopically.
In the third part of this thesis, we explore the possibilities and limits
of the new synthesis method in terms of resulting material properties and
potential applications. Specifically, new PUI type networks are synthesized,
whereby we target useful (combinations of) material properties for
biomedical applications. Relevant properties on which we report include
water uptake, crystallinity, elastic modulus, tensile strength and toughness.
These properties are investigated both in the dry state and in the water-
swollen state. Continuously, we aim to relate the observed macroscopic
material properties to the molecular polymer network structure. Finally, the
potential application of several new PUI networks as contact lens materials
is explored.
Outline
A general introduction to isocyanate- and polyurethane chemistry is
provided in Chapter 2. Emphasis is put on aliphatic poly(urethane-
isocyanurate) (PUI) networks, including their synthesis, general material
properties and typical applications. Conventional and newly developed
synthesis strategies are reviewed. The trimerization reaction of isocyanates
is shown to play a key role throughout these processes.
In Chapters 3 and 4, the viscosity of liquid isocyanate-functional
aliphatic isocyanurates is investigated. These compounds are important
precursors in the conventional 2-component synthesis of PUI networks.
15
Using a combined computational and experimental approach, model
compounds are used to investigate which molecular interactions play a role
in the observed macroscopic trends.
In Chapter 5, a new synthesis method is presented for synthesizing
well-defined PUI networks based on the trimerization of isocyanate-
functionalized prepolymers. The molecular network structure is investigated
and compared to PUI polymer networks synthesized by conventional
synthesis methods.
In Chapter 6, combinatorial PUI networks, synthesized by the
trimerization of mixtures of isocyanate-functionalized prepolymers in
solution, are investigated aiming to achieve unique material properties.6 Of
specific interest is the polymerization of mixtures of hydrophilic and
hydrophobic network components into the same combinatorial network.7
The mechanical and thermal properties of the resulting networks are
investigated both in the dry state and in the water-swollen state. Specific
relevant material properties for biomedical applications, such as toughness
in the hydrated state, are assessed.
In Chapter 7, a structure-property-relations study is presented for
the combinatorial PUI hydrogel materials. By varying the crystallizable
hydrophobic content of the combinatorial PUI networks, the degree of
crystallinity of the hydrogel in the hydrated state is correlated to the
toughness of the hydrogel in the hydrated state.
As poly(ethylene glycol) (PEG) is a hydrophilic and low-biofouling
polymer, PUI networks based on PEG are expected to be suitable materials
for contact lens applications.8 In Chapter 8, a series of these PEG PUI type
hydrogels is synthesized and characterized. Material properties relevant for
contact lens applications are investigated, such as transparency, elastic
modulus and toughness in the water-swollen state, refractive index,
biocompatibility and protein adsorption.
16
References
1. Mendelejeff, D. Über die Beziehungen der Eigenshaften zu den Atomgewichten der Elemente. Zeitschrift für Chemie 12, 405–406 (1869).
2. IUPAC. The International Union of Pure & Applied Chemistry. Available at: https://iupac.org/who-we-are/. (Accessed: 3rd August 2019)
3. United Nations Environment. Frontiers 2018/19: Emerging Issues of Environmental Concern. (2019).
4. Transforming our world : the 2030 Agenda for Sustainable Development. (United Nations General Assembly, 2015).
5. Avtomonov, E., Danielmeier, K., Driest, P. J. & Al., E. Chapters 3 and 4: Chemical Principles and Coating Technology Principles. in Polyurethanes − Coatings, Adhesives, Sealants (eds. Meier-Westhues, U., Danielmeier, K., Kruppa, P. & Squiller, E. P.) (Vincentz-Network Verlag, 2019).
6. Zant, E., Bosman, M. J. & Grijpma, D. W. Combinatorial synthesis of photo-crosslinked biodegradable networks. J. Appl. Biomater. Funct. Mater. 10, 197–202 (2012).
7. Zant, E. & Grijpma, D. W. Tough biodegradable mixed-macromer networks and hydrogels by photo-crosslinking in solution. Acta Biomater. 31, 80–88 (2016).
8. Musgrave, C. S. A. & Fang, F. Contact Lens Materials: A Materials Science Perspective. Materials (Basel). 12, 261 (2019).
17
Chapter 2 - Poly(urethane-
isocyanurate)s: Chemistry, Network
Formation and Applications
Adapted from:
E. Avtomonov1, K. Danielmeier1, P. Driest1,2 et al.; Chapters 3 and 4 in:
Polyurethanes – Coatings, Adhesives and Sealants (Editors: U. Meier-
Westhues et al.); Vincentz-network Verlag; Hannover, Germany, 2019
1Covestro Deutschland AG, CAS-Global R&D, 51373 Leverkusen, Germany
2Technical Medical Centre, Faculty of Science and Technology, Department
of Biomaterials Science and Technology, University of Twente, P.O. Box 217,
7500 AE Enschede, The Netherlands
20
2.1 Introduction
“A world without polyurethanes is hard to imagine these days. Polyurethane
chemistry has become an established technology worldwide in various
industrial applications. For innovative coatings, adhesives and sealants, it
plays a significant role and has been in many cases the key to new technology
implementation. Polyurethanes enable innovation and technological
advancement and help to develop products that meet market needs,
devising efficient and environmentally friendly manufacturing processes and
asserting a position in the global competitive environment.”
From the foreword of Polyurethanes – Coatings, Adhesives and Sealants (U.
Meier-Westhues et al.)1
2.2 Isocyanate- and Polyurethane Chemistry
Synthesis, structure and reactivity of the isocyanate group
As one of the oldest known functional groups in organic chemistry, the
isocyanate functional group has been studied extensively.2–4 Many different
synthesis routes have been described over the years for obtaining
isocyanates, as well as reactions involving them. Conventionally, the
industrial production of isocyanates is performed via the phosgenation of
the corresponding amines. The phosgenation reaction is sketched in Scheme
2.1. Although alternative production pathways of isocyanates are
increasingly investigated, these are generally limited to use in the academic
field. The most commonly adopted alternative synthesis pathways (i.e. not
involving phosgene) are by the Curtius, Hofmann and Lossen
rearrangements, which all involve carbonyl nitrene intermediates.5
21
Scheme 2.1. Synthesis of isocyanates by phosgenation.
Although the isocyanate functional group has been known for over
170 years, still new details are being discovered on a regular basis regarding
its chemical structure and reactivity. With two adjacent double bonds
connecting a nitrogen-, carbon- and oxygen atom, the isocyanate group falls
in the class of heterocumulenes. Resulting from the high electronegativity
of the adjacent heteroatoms, illustrated by the resonance structures shown
in Scheme 2.2, the central carbon atom possesses a relatively large positive
charge. Consequently, this carbon atom is a strong electrophile. In fact, the
high electrophilicity of the central carbon atom is the main rationalization
that isocyanates are employed so widely.
Scheme 2.2. Resonance structures of the isocyanate functional group.
Adapted from Delebecq et al.6
Surprisingly, despite extensive studies, still an active debate is
ongoing regarding the exact molecular structure of the isocyanate functional
group.6,7 For example, Sacher demonstrated that the isocyanate group of
phenyl isocyanate is positioned perpendicular to the benzene ring plane and
demonstrates distinctive cis/trans conformational isomerism (shown in
Scheme 2.3).8 Borisov et al. on the other hand described a completely planar
structure for the phenyl isocyanate molecule.9
22
Scheme 2.3. Perpendicular conformation and cis/trans conformational
isomerism of phenyl isocyanate reported by Sacher et al. Adapted from
Delebecq et al.6
These structural differences are significant, since they influence both the
spatial availability of the carbon centre as well as the electronic structure of
the functional group. The latter is a result of the (lack of) possible
interactions between the nitrogen lone-pair and the connecting moiety (e.g.
the aromatic ring system). Thereby, both the relative reactivities of the two
double bonds (the N=C and C=O double bond, respectively) as well as the
absolute reactivity of the functional group as a whole depend on the specific
electronic and spatial structure of the isocyanate group.
Aside from the specific structure of the isocyanate group itself, the
reactivity of isocyanates is strongly dependent on the moiety to which the
group is attached. Generally speaking, electron-withdrawing groups reduce
the electron density on the carbon atom and thereby enhance the reactivity
of the isocyanate group. This results in the following trend in reactivity:
ClSO2NCO > RSO2NCO (R = alkyl, aryl) > O=P(NCO)3 > aryl-NCO (p-NO2C6H4−
> p-ClC6H4− > p-CH3C6H4− > p-CH3OC6H4−) > alkyl-NCO.6,10 Most relevant in
this regard is the differentiation between aromatic and aliphatic
isocyanates, of which aromatic isocyanates are generally the more reactive
species. The higher reactivity of aromatic isocyanates can be considered
both an advantage (in terms of reaction rate and conversion) and a
disadvantage (in terms of selectivity and reduced control over (un)desired
side reactions). As a result, aromatic and aliphatic isocyanates are typically
used in different application fields and are rarely interchangeable. An
23
overview of commonly used industrial isocyanates is shown in Scheme 2.4:
methylenediphenyl diisocyanate (MDI) and toluene diisocyanate (TDI) are
the most common aromatic diisocyanates, while hexamethylene
diisocyanate (HDI) and isophorone diisocyanate (IPDI) are the most common
aliphatic diisocyanates.
Scheme 2.4. Examples of commonly used industrial aromatic and aliphatic
isocyanates: methylenediphenyl diisocyanate (MDI); toluene diisocyanate
(TDI); hexamethylene diisocyanate (HDI) and isophorone diisocyanate (IPDI).
Reactions of isocyanates with active hydrogen containing compounds:
formation of urethanes, ureas and equivalents
As mentioned previously, the electrophilicity of the central carbon atom is
one of the reasons that isocyanates are so useful. However, ultimately, the
combination of the electrophilic carbon atom with the adjacent nucleophilic
nitrogen atom is what makes isocyanates unique. It is owing to this specific
structure that isocyanates readily react with active hydrogen containing
compounds.6 The exploitation of this reactivity has led to the birth of what
would become a major class of industrial materials today:
polyurethanes.11,12
The most widely-adopted organic active hydrogen containing
reaction partners of isocyanates are alcohols and amines. Thiols and
carboxylic acids are used as well, but to a lesser extent. Besides organic
reaction partners, the reaction of isocyanates with water is also widely
adopted, especially in the production of foam materials. The chemical
equations of the most common isocyanate reactions with active hydrogen
containing compounds are shown in Table 2.1.
24
Table 2.1. Common active hydrogen containing reaction partners of
isocyanates and the resulting reaction products.
Reaction Partner / Product
Chemical Equation
Alcohol / Urethane
Amine / Urea
Thiol / Thio-urethane
Carboxylic Acid / Amide
Water / Urea
Urethane / Allophanat
e
Urea / Biuret
25
The resulting carbonyl compounds are generally thermodynamically
stable compounds, owing to their resonance-stabilized structure analogous
to the carboxamide structure.13 Common examples of isocyanate reaction
products are carbamate esters (generally referred to as urethanes) in the
case of reaction with an alcohol and ureas in the case of reaction with an
amine. Reactions of isocyanates with thiols or carboxylic acids result in the
formation of thiourethanes or amides, respectively. The reaction of an
isocyanate with water results in the formation of the unstable carbamic acid
moiety. After release of CO2 a free amine is formed, which typically instantly
reacts with a second isocyanate group to form the stable urea moiety.
Some of the reaction products of isocyanates still contain active
hydrogens. Also these can react with a second isocyanate moiety. These
reactions tend to occur less readily and generally require elevated
temperatures and/or specific catalysts. The most common such reactions
are the addition of an isocynate to a urethane group resulting in an
allophanate moiety and the addition of an isocyanate to a urea group
resulting in a biuret moiety.
Owing to the huge versatility of isocyanate reaction products,
catalysis plays a very important role in isocyanate chemistry. In this regard,
a high selectivity for a specific reaction product is often a more important
property of a catalyst than a reduction in reaction time. For the reactions of
isocyanates with active hydrogen containing compounds, two types of
catalysts are commonly employed. These are (organic) tertiary amines14,15
and organometallic salts (e.g. organo-lead16, -tin17, -zirconium18,19, -
magnesium20, or -iron21,22 compounds).6 Of these, dibutyltin dilaureate
(typically abbreviated DBTL or DBTDL) is historically the most widely
adopted catalyst for the urethanization reaction and is still often used as a
benchmark catalyst. In the case of DBTL, the combination of a high
selectivity for the urethanization reaction with a high catalyst activity is
considered attractive.23 Due to its high toxicity, the industrial use of DBTL is
more and more being reduced though. Thus far, no substitute catalyst with
the same level of selectivity and activity has been reported.
26
Reactions exclusively involving isocyanates: formation of uretdiones
(dimers), isocyanurates (trimers) and equivalents
Another important class of isocyanate reactions is comprised by reactions
exclusively involving isocyanates (in other words, isocyanates reacting with
themselves). The formation of cyclic dimers (uretdiones) and trimers
(isocyanurate or iminooxadiazinedione) are industrially the most important
examples. When reactions involving the addition or release of CO2 are also
taken into account, even more species become available. These include the
formation of carbodiimid, uretonimin and oxadiazinetrione. Finally,
isocyanates can polymerize to form linear polymeric or cyclooligomeric
structures, typically referred to as 1-nylons or α-nylons.24 The most common
reactions and reactions products exclusively involving isocyanates are
shown in Table 2.2.
Of these, the most widely-used reaction is the trimerization
reaction, which results in the formation of a trifunctional isocyanurate ring.
Owing to the symmetrical ring structure and the hyperconjugated π-system
parallel to the ring plane, isocyanurates are thermodynamically highly stable
compounds.25,26 On top of that, or arguably because of that, isocyanurates
form relatively easily, especially considering that three independent reacting
moieties are required in the formation. The trifunctional ring-structure and
relative ease of formation make isocyanurates ideal structures for the
synthesis of polyurethane networks. Because of the high thermodynamic
stability of these compounds, isocyanurates are especially attractive for
applications where chemical and/or thermal resistance are required.
Examples include flame-retardant construction panels (mostly aromatic
isocyanurates), industrial coatings and biomedical applications (mostly
aliphatic isocyanurates), which will be discussed in section 2.4.
27
Table 2.2. Common reactions and reaction products exclusively involving
isocyanate groups.
Formation of Reaction Equation
Uretdione
(dimer)27
Isocyanurate
(symmetrical
trimer)28
Iminooxadiazine
-dione
(asymmetrical
trimer)2930
Carbodiimid31
Uretonimin31
Oxadiazine-
trione32
Linear polymer
(α-nylon)24
28
Because of the wide variety of possible reaction products starting
from a single chemical moiety, catalysis and the choice of reaction
conditions plays an even more important role in reactions exclusively
involving isocyanates than in the reactions with active hydrogen containing
compounds discussed previously. Generally, nucleophilic, typically anionic
catalysts are used for reactions exclusively involving isocyanates. For
example, catalysts commonly used for the trimerization of isocyanates
include (alkali) metal alkoxides3, metal carboxylates33, (hetero)carbenes34,
trialkyl phosphines35 and rare-earth metal complexes33. Several examples of
commonly used isocyanate trimerization catalysts are shown in Scheme 2.5.
Scheme 2.5. Examples of commonly used isocyanate trimerization catalysts:
a) potassium methoxide3; b) potassium acetate33; c) tin octoate33; d) (CO2-
protected) N-carbene34; e) trialkyl phosphines35
2.3 Network Formation and Polyurethane Materials
The combination of the high reactivity of the starting material and the
thermodynamic stability of the reaction products makes isocyanates ideal
reactants for the production of polymeric materials. Especially the reaction
of (poly)isocyanates with poly(alcohols), resulting in polyurethane (PU) type
polymers has been of major industrial significance during the past 80
years.11,12 Many synthesis strategies have been developed over the years for
the production of PU materials. These generally involve the formation of a
3-dimensional polymer network of some sort. A main distinction can be
made between physical PU networks (also referred to as phase segregating
thermoplastic polyurethanes (TPUs)) on the one hand and chemically
crosslinked (covalent) PU networks on the other hand.
29
In TPU materials, a 3-dimensional physical polymer network is
formed based on the phase-segregation of urethane-group-containing hard
segments.36,37 Owing to the linear nature of the polymer chains, these
materials can be (re)processed at elevated temperatures. The durability of
these materials tends to be less than that of covalent PU networks though.
Since this thesis focusses on covalent PU networks, the reader is redirected
to the literature for a more detailed discussion on TPU materials.6,36–39
Covalent PU networks (or simply, PU networks) are widely adopted
materials owing to their generally excellent mechanical, optical, thermal and
chemical properties. Multiple approaches exist to effectively synthesize PU
networks, each having its advantages and disadvantages. The most common
PU network synthesis strategies, their benefits and limitations, as well as the
material properties of the resulting PU materials will be discussed below.
Two-component mixing of (poly)isocyanates and (poly)alcohols
Historically, the most widely adopted way of synthesizing PU networks is by
a 2-component mixing approach (also referred to as a 2K approach,
abbreviated from the German “2-Komponent”). In this approach, an
isocyanate-containing component is mixed with an alcohol containing
component upon application of the PU network. Once mixed, the free
isocyanate and alcohol groups react to form urethane groups. When both
components have an average functionality ≥2 and at least one of the
components has an average functionality ≥3, an infinte 3-dimensional
covalent PU network is formed.40,41 The synthesis of a PU network by a 2-
component mixing approach is schematically sketched in Scheme 2.6.
In 2K PU formulations, the (typically aliphatic) (poly)isocyanate is
often the component with a functionality ≥3. The reason for this is that high-
functionality aliphatic (poly)isocyanate oligomers are readily synthesized
based on the isocyanate addition reactions discussed previously in section
2.2.27 Specifically the trimerization of aliphatic diisocyanates such as HDI or
IPDI has been of great industrial significance during the past decades.29 The
resulting isocyanate-functional isocyanurates are broadly applied as PU
crosslinking components.
30
Scheme 2.6. Schematic representation of the synthesis of a PU network by a
2-component mixing approach.
The advantages of a 2-component approach are mostly the
relatively straightforward synthesis of the precursors with good control over
the precursor properties. As a result, 2K formulations are often the most
economic formulations. Also, because the reactive components are kept
physically separated until application, 2K formulations generally display a
beneficial combination of long shelf-life and short curing times.
The main disadvantage of a 2K approach is the requirement of
mixing of the two components upon application. Although this may seem
trivial, it has several important implications. Firstly, the final material
properties strongly depend on the mixing of components in the correct
stoichiometry. This requires a certain level of understanding of the
technology and/or clear instructions for the person applying the
formulation. As application is typically not performed by the manufacturer
but rather by an (industrial or private) consumer, 2K PU formulations are
generally not considered the most end-user-friendly approach.
Secondly, on a more fundamental level, a 2K approach requires the
two components to be readily miscible. This limits the scope of possible
combinations of precursors and thereby the range of accessible material
properties. More importantly though, adequate mixing requires the two
components to be of comparable and preferably low viscosity. However, the
synthesis of (poly)isocyanate oligomeric precursors is generally associated
with a strong increase in viscosity. For example, at room temperature the
viscosity of an industrial hexamethylene diisocyanate (HDI) trimer mixture is
on the order of 1000-3000 mPa·s while monomeric HDI has a viscosity of
only 2 mPa·s.29,42 As a result, a search for low-viscosity high-functionality (≥3)
aliphatic (poly)isocyanates has been going on for decades.42–45 A
31
fundamental understanding of the mechanisms dictating the viscosity of
aliphatic (poly)isocyanate precursors is lacking though. A framework for the
intermolecular interactions dictating the viscosity of aliphatic isocyanurates
is presented in Chapter 3 and 4 of this thesis.46,47
Thirdly, the PU networks resulting from a 2-component crosslinking
mechanism are typically poorly defined on a molecularl level. In an industrial
setting, this is generally not considered a problem. On an acdemic level
though, accurate structure-property relations studies require molecularly
well-defined polymer networks. The development of a method for
synthsizing well-defined poly(urethane-isocyanurate) networks is presented
in Chapter 5 of this thesis.48
Alternative crosslinking mechanisms: 1- or 1.5-component formulations,
blocked isocyanates and polyaddition reactions
During the past decades, several alternative PU crosslinking methods have
been developed to improve on or replace the 2-component crosslinking
approach. The first strategy is by the use of so-called blocked isocyanates.49
In this case, the reactive isocyanate groups of a (poly)isocyanate mixture are
reacted with a monofunctional active hydrogen containing compound called
a blocking agent. The resulting compounds are thermodynamically stable at
low temperature, and can be mixed with conventional OH-functional
components and/or additives without crosslinking. At elevated
temperatures, the blocking reaction is reversed, whereby the small-
molecular blocking agent is released and the reactive isocyanate group is
regenerated in-situ. The isocyanate groups subsequently react with the OH-
functional reaction partner in the conventional manner, resulting in a 3-
dimensional covalent PU network. The blocking agent typically evaporates
from the formulation. The general synthetic approach and several common
blocking agents are shown in Scheme 2.7.
32
Scheme 2.7. Blocked isocyanates are adopted to realize 1-component PU
fomulations; a) blocking and deblocking reactions using a blocking agent (H-
Bl). The reactive isocyanate group is generated in-situ by heating and
subsequent release of the blocking agent; b) examples of common blocking
agents.
The main advantage of this technique is the possibility for 1-component PU
formulations. This means that PU raw material manufacturers can control a
larger part of the production and application process, such as the correct
mixing of components. The resulting 1-component PU systems are more
enduser-friendly as adequate application relies less on the chemical
expertise of the applier. Also, because the final formulation does not contain
any free isocyanate groups, blocked isocyanate precursors are typically
more friendly to natural environments.
There are several disadvantages associated with blocked isocyanate
technologies though. Firstly, blocked isocyanate systems are more difficult
to cure than 2K systems. They require elevated temperatures to achieve a
full curing of the PU network, which limits the scope of application. Also, the
curing times of blocked isocyanate systems are typically longer compared to
their 2K equivalents. Secondly, the release of the blocking agent upon curing
is considered a disadvantage. This can potentially have a negative impact on
the final material properties (e.g. in terms of blistering or color change).
33
Also, the amount of volatile organic compounds (VOCs) that are released
upon application is increased, which has a negative environmental impact
and potentially introduces health and safety risks. Thirdly, because of the
extra blocking step and specialty chemicals involved, blocked isocyanate
systems are typically more expensive than their 2K equivalents.
Another alternative crosslinking mechanism increasingly used in industry is
crosslinking by trimerization.50 In this case, the trimerization reaction of
isocyanates is employed for the final crosslinking step instead of for the
synthesis of (poly)isocyanate-functional oligomer precursors.51,52 The
incorporation of conventional OH-functional components is either
performed beforehand (Scheme 2.8a) or is omitted (Scheme 2.8b).
Scheme 2.8. Increasingly, isocyanate addition reactions (e.g. trimerization)
are being used as the crosslinking reaction instead of for the synthesis of
isocyanate-functional oligomeric precursors; a) synthesis of an NCO-
functional prepolymer and subsequent crosslinking by trimerization; b)
synthesis and crosslinking of an isocyanate-functional precursor by
trimerization, resulting in a dense polyisocyanurate network known as
TrimTrim. No OH-functional component is incorporated in this case, so the
final network does not contain any urethane-bonds.
34
There are many advantages associated with this type of crosslinking
mechanism. Firstly, the technology allows the formulation of 1-component
PU systems. More often though, formulations are prepared to which only a
suitable catalyst has to be added.48 These are also referred to as 1.5-
component systems. What makes crosslinking by trimerization unique, is
that no (further) addition or release of chemicals is required to achieve
network formation. This is a clear advantage over the other technologies
described previously, which all rely on either the addition or release of
chemical compounds in stoichiometric amounts. Secondly, owing to the 1-
component nature of the precursor, reaction mixtures are typically
homogeneous in every stage of the reaction. The miscibility issues that are
sometimes encountered for 2K formulations are thus less frequent when
crosslinking by trimerization. Thirdly, because the molecular weight of the
precursors (e.g. an NCO-functional prepolymer) is typically higher than those
of oligomeric (poly)isocyanate precursors, the concentration of free
isocyanate groups is relatively low.
The disadvantages to crosslinking by trimerization are two-fold.
Firstly, when incorporating an OH-functional component, urethane groups
are present in the 1(.5)-component precursor. Since viscosity strongly
increases with the presence and increasing relative content of urethane
groups, the viscosity of the 1(.5)-component precursors is typically high.53
Secondly, the relatively low concentration of free isocyanate groups reduces
activity. As a result, curing times are typically significantly longer than for 2K
PU equivalents. Also, the trimerization reaction typically requires elevated
temperatures, which limits the scope of application.
Some of the disadvantages of the various crosslinking mechanisms discussed
above can be adressed by the use of a so-called dual cure system, as shown
in Scheme 2.9. In this case, PU crosslinking chemistry (e.g. crosslinking by
the trimerization of isozyanates) is combined with a second, complementary
type of crosslinking chemistry (e.g. the free radical crosslinking of
(meth)acrylates under UV-exposure).54 Precursors for such dual cure
systems contain both types of functional groups (e.g. isocyanates and
35
(meth)acrylates). The two complementary crosslinking mechanisms can be
activated either simultaneously or sequentially.
Scheme 2.9. A dual cure mechanisms can be adopted to achieve fast initial
crosslinking whilst still obtaining the good material properties associated
with PU chemistry in a post-curing step (top). Alternatively, a dual cure
mechanism can be used to the increase the network density of a PU network
after crosslinking (bottom).
The main advantage of using a dual cure system is the possibility to optimize
between curing time and final material properties. This is possible because
the crosslinking reaction of acrylates typically occurs faster than the
crosslinking reactions used in PU chemistry (e.g. trimerization).
Consequently, a covalent polymer network can be formed in the early stages
of application (i.e. on a timescale of seconds or minutes, which is difficult to
realize when relying only on the trimerization of isocyanates). The slower
trimerization reaction can then be used as a post-curing step to enhance the
material properties. Alternatively, crosslinking by trimerization can be
performed first, after which the crosslinking of (meth)acrylates can be used
as a post-curing step to increase network density.
36
Overall, many PU crosslinking mechanisms have been developed over the
years, each having their specific advantages and disadvantages. Novel PU
crosslinking technologies are still actively investigated though. One such
novel PU crosslinking technique, which is based on the trimerization of NCO-
functionalized prepolymers with a narrow molecular weight distribution, is
presented in Chapter 5 of this thesis.48
Material properties of polyurethane networks
Owing to the broad range of possible starting materials and reaction
products, PU materials are largely characterized by versatility. Generally
speaking, PUs are considered high-performance materials, well-known for
their excellent mechanical, optical, chemical and biomedical properties.55
These include high ultimate tensile strengths and high toughness values56,
high transparency and low coloration, high chemical and thermal resistance
and good biocompatibility57. Compared to other types of polymer networks
(e.g. UV-cured (meth)acrylate networks), PU networks are generally
considered tougher and more durable.55,58
PU materials containing isocyanurate rings as the crosslinking unit
(poly(urethane-isocyanurate)s or PUIs) especially demonstrate excellent
material properties, most notably in terms of thermal stability, optical
quality, durability and toughness.59–61 The high performance of such PUI type
materials is generally attributed to the thermodynamic stability of the
isocyanurate ring structure, as discussed previously in section 2.2.25,26 As a
result, PUI networks typically demonstrate improved chemcial and thermal
resistance compared to PU networks containing other types of crosslinking
units.33 The syntheses and material properties of novel PUI networks are
presented in Chapters 5, 6, 7 and 8 of this thesis.48,62–64
To illustrate the high resilience of materials crosslinked by
isocyanurate rings, the glass transition temperatures of various dense
polyisocyanurate networks are shown in Scheme 2.10 and compared to an
analogous 2K PU network that does not contain isocyanurate rings.
37
Scheme 2.10. The thermal properties such as the glasss transition
temperature (Tg) of dense isocyanurate networks (TrimTrim) can easily be
tailored over a broad range of values by combining different types of
isocyanate-functional precursors. In this example, the relative content of
IPDI-trimer in an HDI-trimer/IPDI-trimer mixture is varied. These are
compared to a conventional 2K PU network based on mixing an industrial
HDI-trimer with an industrial polyacrylate polyol.
2.4 Applications of Polyurethane Materials
Owing to the wide versatility and widely varying material properties, PU
materials are found in a broad range of applications. PU application fields
are generally characterized by a high demand on mechanical, optical and/or
biological properties, which is where PU materials are distinctive from other
material types. Since this thesis focusses on the use of aliphatic PU networks,
the most important application fields of aliphatic PU networks will be
discussed below. For an overview of the use of other types of PU materials,
such as TPUs or networks based on aromatic PUs, the reader is directed to
the literature.39,65
Coatings, adhesives and sealants
Historically, the largest application field of aliphatic PU networks is the
coatings, adhesives and sealants industry. Of these, the coatings field is by
far the largest, with an annual global production of PU resins of
38
approximately 1.600 kilotons.66 This compares to an annual global
production of PU resins for the combined adhesives and sealants industries
of approximately 0.8 kilotons.66
The use of PU resins for coatings is widespread and includes wood
coating, metal coating (e.g. coil coating, can coating and protective marine
coating), automotive (re)finish and transportation coating, plastic coating,
glass coating, textile and leather coating, paper coating and coating of
construction materials (e.g. walls and floors). The distribution (per volume)
of coating materials over the various application fields in 2017 is shown in
Scheme 2.11.
Scheme 2.11. Formulated coating volumes per application in 2017 (industrial
and architectural). Adapted from U. Meier-Westhues et al.66
Compared to other coating technologies, PU coatings are generally
appreciated for their high versatility and performance. Especially the high
toughness, high chemical, thermal and mechanical resistance, high
durability and good optical appearance associated with PU networks are
attractive properties for coatings applications. In Scheme 2.12, the results of
a benchmark study on the performance of various industrial clear coat
39
systems for automotive applications are shown.67 Clearly, the PU-based
technologies are superior compared to the other coating technologies in the
market.
Scheme 2.12. Clear coat benchmark study showing the performance levels of
various industrial clear coat systems with regard to scratch and acid
resistance: two-component high-solid polyurethane (2K PU), one-component
PU modified thermosetting acrylics (1K PU), one-component carbamate-
modified thermosetting acrylics (1K carbamate TSA), one-component acid
epoxy (1K Acid-Epoxy), one-component silane-modified thermosetting
acrylics (1K Si-TSA) and one-component thermosetting acrylics (1K TSA)
systems were tested. 2K PU systems are the overall best-performing
materials on the marktet. Adapted from U. Meier-Westhues et al.67
High-performance coatings based on PU technologies are found
across many application fields with completely different demands on
material properties. In Scheme 2.13, several examples of PU-based coating
applicaitions are shown, illustrating the broad versatility of PUs. These
include industrial soft-feel plastic coatings, marine protective metal coatings
and decorative glass coatings. The common denominator is the high demand
on durability and optical appearance.
40
Scheme 2.13. Examples of PU coating applications include, but are not
limited to: a) industrial soft-feel plastic coatings; b) marine protective
coatings; c) decorative glass coatings. Adapted from U. Meier-Westhues et
al 67
Polyurethanes for biomedical applications
An important application field for PU materials is the field of biomedical
materials and devices.55,68 Also in this field, PU materials are recognized for
their mechanical properties and durability, this time combined with their
generally good biocompatibility.69–71 As a result, PUs are increasingly used as
biomaterials, with applications including: artificial heart valves, vascular
protheses, pericardial patches, artificial blood pumps, vascular intra-aortic
balloons, catheter and cannula coating, sutures, artificial ligaments, wound
dressings, shock absorbing elements and pacemakers.72 The high toughness
of PU materials has even made it possible to replace some conventional
load-bearing materials, such as metal (alloy)s and ceramics.73 A recently
developed aliphatic PU-based wound dressing is shown in Scheme 2.14.
41
Scheme 2.14. Schematic view and application of an aliphatic PU-based
wound dressing. Adapted from U. Meier-Westhues et al.74
In Chapter 8 of this thesis, a novel potential biomedical application
for PUI networks is explored.64 Making use of the beneficial combination of
high toughness, excellent optical properties and good biocompatibility, the
potential application of PUI network hydrogels as contact lens materials is
investigated.
2.5 Conclusions The isocyanate group has a unique chemical structure, containing both an
electrophilic and nucleophilic reaction centre. The presence of a carbonyl
structure adjacent to a nitrogen heteroatom results in the useful
combination of high reactivity towards active hydrogen containing
compounds and thermodynamic (resonance-)stability of the resulting
carbonyl reaction products. The most important examples are the reactions
with alcohols and amines, resulting in urethane- and urea moieties,
respectively. Additionally, many reactions are possible that exclusively
involve isocyanates. The most important one of these is the symmetrical
trimerization of isocyanates, yielding the thermodynamically highly stable
trifunctional isocyanurate ring structure.
Owing to the high reactivity of the starting material and the
thermodynamic stability of the resulting structures, isocyanates are ideal
reactants for the production of polymer materials. Crosslinked aliphatic
polyurethane (PU) networks comprise an industrially significant class of
materials and many strategies exist to synthesize them. A 2-component (2K)
42
mixing approach, in which high-functionality (poly)isocyanate- and
(poly)alcohol components are mixed upon application, is the most widely-
adopted. This approach offers the easiest synthesis of starting materials and
is often the most economic. The largest disadvantage is the need to mix the
two components upon application. This makes the technique less end-user-
friendly and requires the two components to be of comparable and
preferably low viscosity. Several alternative crosslinking techniques
presently exist, the most promising of which is crosslinking by the
trimerization of isocyanates.
As PU materials are characterized by their versatility, so are their
application fields. For crosslinked aliphatic PU materials, the industrially
most significant application field is the coatings, adhesives and sealants
industry. The high toughness and thermodynamic stability typically
associated with PU materials result in a superior durability and thermal,
mechanical and chemical resistance compared to other coating techniques.
The superior properties of PU materials are also recognized in emerging
application fields. An example of a more recently developed area of
application is the use of PUs as biomedical materials. Also in this relatively
new field of application, PU materials are considered high-performing and
durable compared to existing techniques.
43
2.6 References 1. Meier-Westhues, U., Danielmeier, K., Kruppa, P. & Squiller, E. P.
Foreword. in Polyurethanes − Coatings, Adhesives, Sealants (eds. Meier-Westhues, U., Danielmeier, K., Kruppa, P. & Squiller, E. P.) (Vincentz-Network Verlag, 2019).
2. Wurtz, A. Reserches sur les éthers cyanique et leurs dérivés. Compt. Rend. 27, 241−243 (1848).
3. Arnold, R. G., Nelson, J. A. & Verbanc, J. J. Recent Advances In Isocyanate Chemistry. Chem. Rev. 57, 47–76 (1957).
4. Ozaki, S. Recent Advances in Isocyanate Chemistry. Chem. Rev. 72, 457–496 (1972).
5. Rauk, A. & Alewood, P. F. A theoretical study of the Curtius rearrangement. The electronic structures and interconversions of the CHNO species. Can. J. Chem. 55, 1498–1510 (1977).
6. Delebecq, E., Pascault, J. P., Boutevin, B. & Ganachaud, F. On the versatility of urethane/urea bonds: Reversibility, blocked isocyanate, and non-isocyanate polyurethane. Chem. Rev. 113, 80–118 (2013).
7. Caraculacu, A. A. & Coseri, S. Isocyanates in polyaddition processes. Structure and reaction mechanisms. Prog. Polym. Sci. 26, 799–851 (2001).
8. Sacher, E. A re-examination of the polyurethane reaction. J. Macromol. Sci. Part B 16, 525–538 (1979).
9. Borisov, E. V., Vasyanina, L. K., Bondarenko, S. P. & Zabrodin, V. B. Electronic structure of the conformation and properties of phenyl isocyanate. Zh. Fiz. Khim. 51, 1930 (1977).
10. Ozaki, S. Recent Advances in Isocyanate Chemistry. J. Synth. Org. Chem. Japan 28, 17–40 (1970).
11. Bayer, O. Verfahren zur Herstellung von Polyurethanen bzw. Polyharnstoffen. (1937).
12. Bayer, O. Das Di-Isocyanat-Polyadditionsverfahren (Polyurethane). Angew. Chemie 59, 257–272 (1947).
13. Volkov, A., Tinnis, F., Slagbrand, T., Trillo, P. & Adolfsson, H. Chemoselective reduction of carboxamides. Chem. Soc. Rev. 45, 6685–6697 (2016).
14. Chang, M.-C. & Chen, S.-A. Kinetics and mechanism of urethane reactions: Phenyl isocyanate–alcohol systems. J. Polym. Sci. Part A Polym. Chem. 25, 2543–2559 (1987).
15. Baker, J. W. & Holdsworth, J. B. 135. The mechanism of aromatic side-chain reactions with special reference to the polar effects of substituents. Part XIII. Kinetic examination of the reaction of aryl
44
isocyanates with methyl alcohol. J. Chem. Soc. 713 (1947). doi:10.1039/jr9470000713
16. Rand, L., Thir, B., Reegen, S. L. & Frisch, K. C. Kinetics of alcohol–isocyanate reactions with metal catalysts. J. Appl. Polym. Sci. 9, 1787–1795 (1965).
17. Han, Q. & Urban, M. W. Kinetics and mechanisms of catalyzed and noncatalyzed reactions of OH and NCO in acrylic polyol-1,6-hexamethylene diisocyanate (HDI) polyurethanes. VI. J. Appl. Polym. Sci. 86, 2322–2329 (2002).
18. Blank, W. J., He, Z. A. & Hessell, E. T. Catalysis of the isocyanate-hydroxyl reaction by non-tin catalysts. Prog. Org. Coatings 35, 19–29 (1999).
19. Petrak, S., Shadurka, V. & Binder, W. H. Cleavage of blocked isocyanates within amino-type resins: Influence of metal catalysis on reaction pathways in model systems. Prog. Org. Coatings 66, 296–305 (2009).
20. Gertzmann, R. & Gürtler, C. A catalyst system for the formation of amides by reaction of carboxylic acids with blocked isocyanates. Tetrahedron Lett. 46, 6659–6662 (2005).
21. Ahmad, I., Zaidi, J. H., Hussain, R. & Munir, A. Synthesis, characterization and thermal dissociation of 2-butoxyethanol-blocked diisocyanates and their use in the synthesis of isocyanate-terminated prepolymers. Polym. Int. 56, 1521–1529 (2007).
22. Korah Bina, C., Kannan, K. G. & Ninan, K. N. DSC study on the effect of isocyanates and catalysts on the HTPB cure reaction. J. Therm. Anal. Calorim. 78, 753–760 (2004).
23. Silva, A. L. & Bordado, J. C. Recent developments in polyurethane catalysis: Catalytic mechanisms review. Catal. Rev. - Sci. Eng. 46, 31–51 (2004).
24. King, C. Cyclopolymerization of Aliphatic 1,2-Diisocyanates. J. Am. Chem. Soc. 86, 437–440 (1964).
25. Heift, D., Benko, Z., Grützmacher, H., Jupp, A. R. & Goicoechea, J. M. Cyclo-oligomerization of isocyanates with Na(PH2) or Na(OCP) as ‘p-’ anion sources. Chem. Sci. 6, 4017–4024 (2015).
26. Okumoto, S. & Yamabe, S. A computational study of base-catalyzed reactions between isocyanates and epoxides affording 2-oxazolidones and isocyanurates. J. Comput. Chem. 22, 316–326 (2001).
27. Laas, H. J., Halpaap, R. & Pedain, J. Zur Synthese aliphatischer Polyisocyanate − Lackpolyisocyanate mit Biuret-, Isocyanurat- oder Uretdionstruktur. J. Prakt. Chem./Chem.-Ztg. 336, 185−200 (1994).
45
28. Bock, M., Pedain, J. & Uerdingen, W. Process for preparing polyisocyanates with isocyanate groups and their use as isocyanate components in polyurethane varnishes. (1994).
29. Richter, F., Pedain, J., Mertes, H. & Dieris, C.-G. Isocyanate trimers and mixtures of isocyanate trimers, production and use thereof. 13–14 (1997).
30. Decostanzi, M., Auvergne, R., Darroman, E., Boutevin, B. & Caillol, S. Reactivity and kinetics of HDI-iminooxadiazinedione: Application to polyurethane synthesis. Eur. Polym. J. 96, 443–451 (2017).
31. Wagner, K., Findeisen, K., Schäfer, W. & Dietrich, W. α,ω-Diisocyanatocarbodiimides, -Polycarbodiimides, and Their Derivatives. Angew. Chemie Int. Ed. English 20, 819–830 (1981).
32. Scholl, H.-J. Process for the preparation of modified isocyanurate-containing polyisocyanates and use of them. (1990).
33. Flipsen, T. A. C., Steendam, R., Pennings, A. J. & Hadziioannou, G. A novel thermoset polymer optical fiber. Adv. Mater. 8, 45–48 (1996).
34. Bantu, B. et al. CO2 and SnII Adducts of N-Heterocyclic Carbenes as Delayed-Action Catalysts for Polyurethane Synthesis. Chem. - A Eur. J. 15, 3103–3109 (2009).
35. Richter, F. U. 1,4,2-Diazaphospholidine-3,5-diones and Related Compounds: A Lecture on Unpredictability in Catalysis. Chem. - A Eur. J. 15, 5200–5202 (2009).
36. Yilgor, I. & Yilgor, E. Structure-morphology-property behavior of segmented thermoplastic polyurethanes and polyureas prepared without chain extenders. Polym. Rev. 47, 487–510 (2007).
37. Yilgör, I., Yilgör, E. & Wilkes, G. L. Critical parameters in designing segmented polyurethanes and their effect on morphology and properties: A comprehensive review. Polymer (Guildf). 58, A1–A36 (2015).
38. Sami, S. et al. Understanding the influence of hydrogen bonding and diisocyanate symmetry on the morphology and properties of segmented polyurethanes and polyureas: Computational and experimental study. Polymer (Guildf). 55, 4563–4576 (2014).
39. Engels, H.-W. et al. Polyurethanes: Versatile Materials and Sustainable Problem Solvers for Today’s Challenges. Angew. Chemie Int. Ed. 52, 9422–9441 (2013).
40. Dusek, K. Networks from Telechelic Polymers: Theory and Application to Polyurethanes. in Telechelic Polymers: Synthesis and Applications 289–360 (CRC Press, 1989).
41. Erdödi, G. & Iván, B. Novel Amphiphilic Conetworks Composed of Telechelic Poly(ethylene oxide) and Three-Arm Star
46
Polyisobutylene. 16, 959–962 (2004). 42. Renz, H. & Bruchmann, B. Pathways targeting solvent-free PUR
coatings. Prog. Org. Coatings 43, 32–40 (2001). 43. Chattopadhyay, D. K. & Raju, K. V. S. N. Structural engineering of
polyurethane coatings for high performance applications. Prog. Polym. Sci. 32, 352–418 (2007).
44. Ludewig, M., Weikard, J. & Stockel, N. Allophanate-containing modified polyurethanes (US 2006/0205911 A1). (2006).
45. Woicik, R. T., Goldstein, S. L., Barnowski, H. G. & Chandalis, K. B. Novel Super Low Viscosity Aliphatic Isocyanate Crosslinker for Polyurethane Coatings. J. Cell. Plast. 27, 64–64 (1991).
46. Driest, P. J. et al. Aliphatic isocyanurates and polyisocyanurate networks. Polym. Adv. Technol. 28, 1299–1304 (2017).
47. Driest, P. J. et al. The Influence of Molecular Interactions on the Viscosity of Aliphatic Polyisocyanate Precursors for Polyurethane Coatings. ECS Conf. Proc. (2019).
48. Driest, P. J., Dijkstra, D. J., Stamatialis, D. & Grijpma, D. W. The Trimerization of Isocyanate-Functionalized Prepolymers : An Effective Method for Synthesizing Well-Defined Polymer Networks. Macromol. Rapid Commun. 40, 1800867 (2019).
49. Tork, L., Rottmaier, L. & Hohne, W. Preparations and processes for finishing leather and for coating textiles. (1988).
50. Samborska-Skowron, R. & Balas, A. An overview of developments in poly(urethane-isocyanurates) elastomers. Polym. Adv. Technol. 13, 653–662 (2002).
51. Sasaki, N., Yokoyama, T. & Tanaka, T. Properties of isocyanurate-type crosslinked polyurethanes. J. Polym. Sci. Polym. Chem. Ed. 11, 1765–1779 (1973).
52. Špírková, M., Budinski-Simendic, J., Ilavský, M., Špaček, P. & Dušek, K. Formation of poly(urethane-isocyanurate) networks from poly(oxypropylene)diols and diisocyanate. Polym. Bull. 31, 83–88 (1993).
53. Rimdusit, S., Bangsen, W. & Kasemsiri, P. Chemorheology and thermomechanical characteristics of benzoxazine-urethane copolymers. J. Appl. Polym. Sci. 121, 3669–3678 (2011).
54. Konuray, O. et al. State of the Art in Dual-Curing Acrylate Systems. Polymers (Basel). 10, 178 (2018).
55. Akindoyo, J. O. et al. Polyurethane types, synthesis and applications – a review. RSC Adv. 6, 114453–114482 (2016).
56. Petrović, Z. S. & Ferguson, J. Polyurethane elastomers. Prog. Polym. Sci. 16, 695–836 (1991).
47
57. Burel, F., Poussard, L., Tabrizian, M., Merhi, Y. & Bunel, C. The influence of isocyanurate content on the bioperformance of hydrocarbon-based polyurethanes. J. Biomater. Sci. Polym. Ed. 19, 525–540 (2008).
58. Zant, E., Bosman, M. J. & Grijpma, D. W. Combinatorial synthesis of photo-crosslinked biodegradable networks. J. Appl. Biomater. Funct. Mater. 10, 197–202 (2012).
59. Fabbri, P. et al. Highly stable plastic optical fibre amplifiers containing [Eu(btfa)3(MeOH)(bpeta)]: A luminophore able to drive the synthesis of polyisocyanates. Polymer (Guildf). 55, 488–494 (2014).
60. Moritsugu, M., Sudo, A. & Endo, T. Cyclotrimerization of diisocyanates toward high-performance networked polymers with rigid isocyanurate structure: Combination of aromatic and aliphatic diisocyanates for tunable flexibility. J. Polym. Sci. Part A Polym. Chem. 51, 2631–2637 (2013).
61. Preis, E., Schindler, N., Adrian, S. & Scherf, U. Microporous Polymer Networks Made by Cyclotrimerization of Commercial, Aromatic Diisocyanates. ACS Macro Lett. 4, 1268–1272 (2015).
62. Driest, P. J., Dijkstra, D. J., Stamatialis, D. & Grijpma, D. W. Tough Combinatorial Poly(urethane-isocyanurate) Polymer Networks and Hydrogels Synthesized by the Trimerization of Mixtures of NCO-Prepolymers. to be Submitt. (Chapter 6 this Thesis) (2019).
63. Driest, P. J., Dijkstra, D. J., Stamatialis, D. & Grijpma, D. W. Structure-Property-Relations in Semi-crystalline Combinatorial Poly(urethane-isocyanurate) Hydrogels. to be Submitt. (Chapter 7 this Thesis) (2019).
64. Driest, P. J., Allijn, I. E., Dijkstra, D. J., Stamatialis, D. & Grijpma, D. W. Poly(ethylene glycol) Poly(urethane-isocyanurate) Type Hydrogels as a New Class of Materials for Contact Lens Applications. to be Submitt. (Chapter 8 this Thesis) (2019).
65. Brereton, G. et al. Polyurethanes. in Ullmann’s Encyclopedia of Industrial Chemistry 1–76 (Wiley-VCH Verlag GmbH & Co. KGaA, 2019). doi:10.1002/14356007.a21_665.pub3
66. Chapter 2: Economic aspects and market analysis. in Polyurethanes − Coatings, Adhesives, Sealants (eds. Meier-westhues, U., Danielmeier, K., Kruppa, P. & Squiller, E. P.) (Vincentz-Network Verlag, 2019).
67. Chapter 5: Coatings. in Polyurethanes − Coatings, Adhesives, Sealants (eds. Meier-Westhues, U., Danielmeier, K., Kruppa, P. & Squiller, E. P.) (Vincentz-Network Verlag, 2019).
48
68. Cooper, S. L. & Guan, J. Advances in polyurethane biomaterials. (Woodhead Publishing, 2016).
69. Zhou, B., Hu, Y., Li, J. & Li, B. Chitosan/phosvitin antibacterial films fabricated via layer-by-layer deposition. Int. J. Biol. Macromol. 64, 402–8 (2014).
70. Zhou, X., Zhang, T., Guo, D. & Gu, N. A facile preparation of poly(ethylene oxide)-modified medical polyurethane to improve hemocompatibility. Colloids Surfaces A Physicochem. Eng. Asp. 441, 34–42 (2014).
71. Wang, Y., Hong, Q., Chen, Y., Lian, X. & Xiong, Y. Surface properties of polyurethanes modified by bioactive polysaccharide-based polyelectrolyte multilayers. Colloids Surfaces B Biointerfaces 100, 77–83 (2012).
72. Khatoon, H. & Ahmad, S. Polyurethane: A Versatile Scaffold for Biomedical Applications. Significances Bioeng. Biosci. 2, (2018).
73. Middleton, J. C. & Tipton, A. J. Synthetic biodegradable polymers as orthopedic devices. Biomaterials 21, 2335–2346 (2000).
74. Chapter 8: New areas of application for polyurethanes. in Polyurethanes − Coatings, Adhesives, Sealants (eds. Meier-Westhues, U., Danielmeier, K., Kruppa, P. & Squiller, E. P.) (Vincentz-Network Verlag, 2019).
49
Chapter 3 - Aliphatic isocyanurates and
polyisocyanurate networks
P.J. Driest1,2, V. Lenzi3, L.S.A. Marques3, M.M.D. Ramos3, D.J. Dijkstra1, F.U.
Richter1, D. Stamatialis2 and D.W. Grijpma2,4
1Covestro, CAS-Global R&D, 51365 Leverkusen, Germany
2MIRA Institute for Biomedical Technology and Technical Medicine,
Department of Biomaterials Science and Technology, University of Twente,
P.O. Box 217, 7500 AE Enschede, The Netherlands
3Department/Centre of Physics of University of Minho, campus de Gualtar,
4710-057 Braga, Portugal
4University Medical Center Groningen, W.J. Kolff Institute, Department of
Biomedical Engineering, P.O. Box 196, 7600 AD Groningen, The Netherlands
Published in:
Polymers for Advanced Technologies 28 (2017), 1299–1304
Abstract
The production, processing and
application of aliphatic isocyanate
(NCO) based thermosets such as
polyurethane coatings and
adhesives are generally limited by
the surprisingly high viscosity of tri-
and higher-functionality
isocyanurates. These compounds
are essential crosslinking additives
for network formation. However, the mechanism by which these high
viscosities are caused is not yet understood.
In this work, model aliphatic isocyanurates were synthesized and
isolated in high purity (>99%) and their viscosities were accurately
determined. It was shown that the presence of the NCO group has a strong
influence on the viscosity of the system. From density functional theory
(DFT) calculations a novel and significant bimolecular binding potential of -
8.7 kJ/mol was identified between NCO groups and isocyanurate rings,
confirming the important role of the NCO group. This NCO-to-ring
interaction was proposed to be the root cause for the high viscosities
observed for NCO-functional isocyanurate systems. MD simulations carried
out to further confirm this influence also suggest that the NCO-to-ring
interaction causes a significant additional contribution to viscosity. Finally,
model functional isocyanurates were further reacted into densely
crosslinked polyisocyanurate networks which showed interesting material
properties.
52
3.1 Introduction The current work is aimed at better understanding the viscosity
characteristics of isocyanate (NCO) based crosslinkers, an essential
parameter in their production and processing. Isocyanates, as primary
ingredients of polyurethanes, have been of wide interest in both industry
and academia ever since polyurethanes were invented.1-4 Aliphatic
diisocyanates such as 1,6-hexamethylenediisocyanate (HDI) and its
derivatives are produced on a large scale worldwide. They are well-known
for their high performance regarding UV resistance, as well as their excellent
mechanical, chemical, optical and thermal properties in polyurethane
coatings and adhesives. Of particular significance in the performance of
these materials are the higher-functionality trimerized isocyanurate forms
of the diisocyanates, key additives as cross-linking units for network
formation. The chemical and physical stability of these cross-linking units,
resulting from the deep thermodynamic minimum of the isocyanurate ring
structure5,6, is key in the high performance of modern polyurethane coatings
and adhesives. Furthermore, the possibility of trimerizing such trifunctional
isocyanurate additives in a pure form has been investigated. This results in
the formation of highly stable densely crosslinked polyisocyanurate
networks, which show interesting material properties. In particular the good
optical properties and high chemical and thermal stability of these materials
are attractive qualities.7-10
Given that the same basic starting materials are used in the
preparation of many isocyanate-based thermosets, including well-
established polyurethane coatings, adhesives and densely crosslinked
polyisocyanurate networks, a good understanding of the physical behavior
of these precursors would be beneficial. However, a major problem
encountered in the use of aliphatic functional isocyanurates is their
relatively high viscosity, which complicates their production, processing and
application. Consequently, the search for low-viscosity high-functionality
(≥3) aliphatic isocyanates has been a topic of continuous interest for the past
decades.11-13 Surprisingly, however, very little is known about the
fundamental causes of the increase of viscosity in such systems, and the
underlying chemistry is hardly understood. Although surprisingly high
53
viscosity values have been observed and reported for aliphatic
isocyanurates (e.g., 1000-3000 mPa∙s at room temperature for industrial
HDI-based isocyanurates, compared to 2 mPa∙s for monomeric HDI13,14), a
chemical explanation for the cause of such high values has not yet been
presented. Combining the need for a better understanding of the physical
behavior of these precursors with a further investigation of the
polyisocyanurate networks made from them, we set out with three main
objectives:
1. to quantify accurately the viscosity increase upon trimerization of
aliphatic diisocyanates into NCO-functional aliphatic isocyanurates;
essential precursors for many isocyanate-based thermosets;
2. to investigate the underlying chemistry to explain the unexpectedly
high viscosities observed for NCO-functional aliphatic isocyanurates;
and
3. to further react model isocyanurates into densely crosslinked
polyisocyanurate networks and characterize the physical and
thermal properties of the resulting materials.
To investigate the viscosity characteristics of functional (network-forming)
isocyanurates, a series of model diisocyanates with varying carbon chain
lengths based on and including HDI was prepared. Also, a similar series of
monofunctional isocyanates, yielding non-functional (non-network-
forming) isocyanurates, was included as a reference. The relevant structures
and general synthetic approach are presented in Scheme 3.1.
Scheme 3.1. Structure and general synthesis of model isocyanurates.
54
3.2 Experimental Details
Materials and methods
α-ω-Linear diisocyanates were taken from internal industrial production
(Covestro Deutschland AG) and vacuum distilled before use. n-
Butylisocyanate and n-hexylisocyanate were purchased at Aldrich and
distilled before use. n-Pentylisocyanate and n-heptylisocyanate were
synthesized by reaction of corresponding amines (Sigma-Aldrich) with 1,6-
hexamethylenediisocyanate. Dodecylbenzene sulfonic acid was supplied by
Sigma-Aldrich and used as received, as were all other chemicals described.
Viscosity measurements were performed at RT on an Anton-Paar MCR501
rheometer using either a 50 mm cone-plate setup (CP50-2, d = 0.214 mm)
or a CC27 beaker-cup setup, in rotational mode at increasing shear rate from
0.1-100 s-1 (0.1-1000 s-1 for beaker-cup). DSC measurements were
performed under nitrogen on 10 mg samples on a Perkin-Elmer Kalorimeter
DSC-7 in 2 heating runs from -20°C-200°C by 20°C/min. TGA measurements
were performed under nitrogen on 4 mg samples on a Perkin-Elmer
Mikrothermowaage TGA-7 from RT to 600°C by 20°C/min. NMR-spectra
were collected on a Bruker Avance III - 700. FTIR-spectra were recorded on
a Bruker FT-IR Spektrometer Tensor II with Platinum-ATR-unit with
diamondcrystal.
Synthesis of n-pentylisocyanate and n-heptylisocyanate
In a typical experiment: in a 1000 mL 2-neck round-bottom flask with
distillation bridge, 15,45 g of n-alkylamine was dripped into 412,7 g of pure
HDI at 80 °C under formation of white deposit. The mixture was heated to
170°C and pressure was gradually reduced to 50 mbar. After 4 h distillation
was complete. The distillate fraction was redistilled at 175°C under
atmospheric pressure, collecting the pure products as transparent liquid,
analyzed by 1H-NMR and FTIR.
Synthesis of functional isocyanurates
In a typical experiment: in a cylindrical flask equipped with a continuous
Near InfraRed (NIR) probe-head (Bruker NIR Vector 22/N), 150 g of α-ω-
55
linear diisocyanate was heated to 60°C and 0.15 wt% of catalyst solution (1
wt% of N,N-methyl-butyl-pyrrolidiniumhydroxide in isopropanol) was
added, until trimerization started, followed by continuous NIR and
temperature (for 1,4-butyldiisocyanate, N,N-dimethyltrimethylsilylamine
was used as catalyst, at 120°C). After 5-20% reduction of NCO absorption-
intensity in NIR (30-90 min), an equimolar amount, compared to active
catalyst, of dodecylbenzene sulfonic acid was added to quench the reaction
(1,4-butanediol for the reaction of 1,4-butyldiisocyanate). The mixture was
transferred to a dripping funnel and evaporated using a thin-film short-path
evaporator at 150 °C and 1∙10-1 mbar, collecting the resin as highly viscous
brown oil. Reusing the distillate as starting material and supplementing with
freshly distilled diisocyanate, this procedure was repeated 6 times,
combining all resins. The combined resin-fraction was then distilled in a thin-
film short-path evaporator at 220°C and <10-7 mbar, collecting the distillate
as a very slightly greenish transparent viscous oil. The distillate was
evaporated using a thin-film short-path evaporator at 160 °C and <10-7 mbar,
collecting the pure product as resin as a very slightly greenish transparent
viscous oil, characterized by (numbers for HDI-trimer): 1H-NMR (700 MHz,
C6D6): δ = 3.78 (2H, t), 2.54 (2H, t), 1.53 (2H, quin), 1.03-0.92 (6H, m) ppm; 13C-NMR (700 MHz, C6D6): δ = 149.01, 122.92, 42.78, 42.60, 31.07, 27.99,
26.17, 26.16 ppm and gel permeation chromatography (GPC).
Synthesis of non-functional isocyanurates
In a typical experiment: in a 100 mL round-bottom flask, 34,63 g of n-alkyl
isocyanate was heated to 60°C. Dropwise, 2,46 g of catalyst solution (1 wt%
of N,N-methyl-butyl-pyrrolidiniumhydroxide in isopropanol) was added,
until trimerisation started, followed by FTIR and temperature. In 2-4 h the
reaction was run to completion, followed by disappearance of the 2275 cm-
1 NCO-band in FTIR. The mixture was trap-to-trap distilled under reduced
pressure (<10-7 mbar) at 200°C and products were isolated as intermediate
distillate fraction as transparent oils, analyzed by (numbers for n-
hexylisocyanate trimer): 1H-NMR (700 MHz, C6D6): δ = 3.82 (2H, t), 1.63 (2H,
quin), 1.22-1.14 (6H, m), 0.82 (3H, t) ppm; 13C-NMR (700 MHz, C6D6): δ =
56
149.06, 42.98, 31.73, 28.26, 26.72, 22.91, 14.19 ppm and gel permeation
chromatography (GPC).
Synthesis of polyisocyanurate networks
In a typical experiment: for the trifunctional isocyanurate, to 100 g of
isocyanurate precursor, 3.140 g of catalyst solution (0.148 g KOAc, 0.396 g
18-crown-6-ether, 2.596 g diethyleneglycol) was added and mixed in a
Hauschild Engineering DAC 150.1 FVZ Speemixer at 2000 rpm for 60 sec. The
mixture was poured out into 6.3 cm aluminium lids in 4, 8 or 16 g samples
and kept at 180°C for 15min. Materials were collected as rigid transparent
thermosets, showing slight yellowing and in some cases formation of a few
small bubbles. For difunctional isocyanates, mixtures were allowed to react
in a glass vial under nitrogen for 1-2 h before similar pouring and annealing.
Caution: Because of the strongly exothermic character of the trimerization
reaction, care should be taken with dissipation of heat, especially during the
pretrimerisation of difunctional diisocyanates.
Quantum calculations details
All DFT calculations were performed with Gaussian ‘09 software15 using a
B3-LYP hybrid functional and the 6-31+g(d,p) basis set, and have been
carried out using the ultrafine integration grid. The accuracy of the
calculations was improved considering empirical dispersion corrections, as
prescribed by the Grimme’s D3 scheme16 with the Becke-Johnson
damping.17 For all molecules, stable geometries were obtained and
characterized in terms of the relevant geometrical parameters (bonds,
angles, dihedrals and impropers).
Free energy calculations
The free energy of binding was calculated as follows:
∆𝐺𝑏𝑖𝑛𝑑𝑖𝑛𝑔 = 𝐺𝑑𝑖𝑚𝑒𝑟 − 𝐺𝑖𝑠𝑜𝑙𝑎𝑡𝑒𝑑
57
where 𝐺𝑑𝑖𝑚𝑒𝑟 is the Gibbs free energy of a bimolecular complex and
𝐺𝑖𝑠𝑜𝑙𝑎𝑡𝑒𝑑 is twice the Gibbs free energy of an isolated molecule in its lowest
energy configuration. The exact formula for 𝐺 is:
𝐺 = 𝐸0 + 𝐸𝑍𝑃𝐸 + 𝐸𝑣𝑖𝑏 + 𝐸𝑟𝑜𝑡 + 𝑘𝑏𝑇 − 𝑇𝑆
where the different contributions are from electronic energy 𝐸0, the zero-
point energy correction 𝐸𝑍𝑃𝐸 , the vibrational and rotational energies 𝐸𝑣𝑖𝑏
and 𝐸𝑟𝑜𝑡, the thermal contribution 𝑘𝑏𝑇 and the entropic term 𝑇𝑆. All of
these terms were calculated explicitly using frequency calculations. For the
bimolecular complexes, the Basis Set Superposition Error contribution has
been calculated and removed from the final free energy.
Molecular Dynamics Simulations
1) Force field parametrization: A fully flexible all-atom representation of the
molecules was adopted and the GAFF18 force field was used and
parametrized using the data from the quantum calculations. In particular,
bond angles and lengths at equilibrium were slightly tuned to match the
results from DFT. Partial charges on each atom were obtained using the RESP
method.19 To account for charge screening effects, all the coulomb
interactions were scaled by a factor of 2. To check the validity of the force
field, the interaction energies calculated by molecular mechanics were
compared to those calculated by DFT for various configurations of the
bimolecular complexes along a specified reaction coordinate, reported in
Figure S1 (Supporting Information). There is a slight underestimation of the
well depth by molecular mechanics, but the long-range part of the
interaction energy is very well described. As further validation, the
equilibrium density obtained for the HDI trimer liquid at 300 K and 1 Atm,
using molecular dynamics simulations, has been compared to the
experimental density at 300 K and 1 atm. An agreement of up to 1% was
found.
2) MD calculations Setup: MD simulations were performed using the
LAMMPS20 software package. The Velocity Verlet integration method has
been used. A timestep of 1 femtosecond has been used, along with the
58
RESPA algorithm21, using an inner timestep of 0.125 fs to keep the energy
drift under control. A general cut-off of 13 Angstroms for pair forces
calculations was considered. Above that cut-off, long range electrostatic
interactions have been evaluated using the Ewald Particle Mesh (PPPM)
method.22 For each molecular liquid six independent runs were performed.
Preparation of the system consisted of a high temperature (480 K) run of 1
ns to randomize the initial conditions, followed by an equilibration step
towards the target temperature. This was done under constant temperature
and pressure conditions, implemented by the Nosé-Hoover23 thermostat
and Parrinello-Rahman24 barostat. Input files for data production runs were
chosen to match both the equilibrium density and the target temperature.
Production runs consisted of 5 ns runs (5 million steps) under NVE (constant
energy, no thermo- and barostatting) conditions.
Calculation of Viscosity
From the production runs stress tensor components were stored every 10
steps and further post-processed to give the viscosity. Using the Green-
Kubo25 formalism, viscosities were obtained through integration of the
stress autocorrelation function:
𝜂 =𝑉
10𝑘𝑇∫ 𝑑𝑡⟨𝑃𝑥𝑦(𝑡)𝑃𝑥𝑦(0)⟩
𝜏
0
where angle brackets denote an ensemble average. To calculate the integral,
a time window of length 𝜏 was defined and different time origins were used
to minimize the integration error.26 To improve accuracy, averages from
different pressure tensor components were considered and averages over 6
independent runs were performed.
59
3.3 Results and Discussion
Synthesis of Isocyanurates
All syntheses were performed based on previously described methods27,28,
using an N,N-methyl-butyl-pyrrolidiniumhydroxide catalyst, as summarized
in Scheme 3.1. Because even small amounts of impurities can cause
deviations in viscosity measurements, extreme care was taken throughout
with the purification and isolation of isocyanurate products. On top of the
already high sensitivity of isocyanates to undergo various side reactions,
difunctional isocyanates (R = NCO) will inevitably form higher molecular
weight oligomers, because of the similar reactivities of the starting material
and product. The optimal conditions for isolating the functionalized trimers
in pure form were hence obtained by early quenching of the reaction.
Specifically, at about 10% conversion of the NCO groups (determined by
continuous near-infraRed (NIR) spectroscopy) the catalyst was quenched by
addition of an equimolar amount of a suitable catalyst poison. Because
isocyanates tend to exhibit reactivity toward most column-chromatography
media, the generally preferred purification technique is distillation.
Therefore, despite the high molecular weights of the products, workup of
the reaction products was done by means of a multiple-step thin-film
distillation process under very low pressures (<10-7 mbar) using a short-path
evaporator, as described below. In a typical experiment, first all unreacted
monomer was removed by thin-film evaporation at 150 °C and 10-1 mbar.
Then, the targeted trimers were separated from higher molecular weight
oligomers by thin-film distillation at 220 °C and <10-7 mbar, thereby giving
the products as the distillate. Finally, from this distillate, any reformed
monomer and other low-boiling impurities formed during the purification
process were removed by similar thin-film evaporation at 160 °C and <10-7
mbar, yielding the product in pure form as the residue.
During trimerization of the monofunctional isocyanates (R = H) of
the reference series, oligomerization does not occur, so the reactions could
be carried out to completion (as determined followed by the disappearance
of the 2275 cm-1 NCO band in the FTIR spectrum). Subsequently, the reaction
products were purified in a fashion to that described previously, in a
60
multistep trap-to-trap distillation process under very low pressures (<10-7
mbar). The products were collected as the intermediate distillate fractions.
All products were isolated in >99% purity with the exception of the sample
with n = 4 and R = NCO (>98%), as confirmed by 700-MHz 1H NMR and 13C
NMR spectroscopies and gel permeation chromatography (GPC), in most
cases limited only by the detection level of the equipment (data not shown).
Because of the strong emphasis on purity, no yields were determined.
Quantification of the Viscosities of Isocyanurates
For the isolated trimers, viscosities were measured on an Anton Paar
MCR501 rheometer using a 50 mm cone-plate (CP50-2) setup. For
validation, several samples were also measured using a more consuming but
more precise CC27 beaker-cup setup, yielding similar results. The measured
viscosity values are depicted in Figure 3.1 as a function of the carbon chain
length of the isocyanate reactants.
Figure 3.1. Viscosities of model isocyanurates.
61
As can be clearly seen from the figure, the viscosities of all functionalized
samples (R = NCO) were significantly higher than those of the non-
functionalized samples (R = H), while also increasing for higher NCO and
isocyanurate contents (i.e. lower value of n). Within the present series, the
side chains are assumed to be too short to show any macromolecular
behavior and no significant contribution to viscosity is expected from
entanglements. Because the molecular shapes of the two series are
practically identical, differing only in the fact that the linear NCO groups are
slightly more rigid than alkyl chain ends, no significant effects are expected
based on molecular shape either. Hence, any difference between the two
series can be attributed entirely to the presence/absence of the NCO group
or, more specifically, to electronic/chemical interactions involving the NCO
group. However, it should be noted that in monomeric (di)isocyanates such
as HDI, NCO groups are abundantly present as well, although here this does
not lead to increased viscosities. The specific chemical mechanism by which
the presence of the NCO group influences viscosity in these mixtures seems
to be somewhat complicated. To investigate this influence in more detail,
density functional theory (DFT) calculations were performed on the same
series of model compounds.
DFT Calculations of Bimolecular Interaction Potentials
To investigate the underlying mechanism causing the unexpectedly high
viscosities of functional aliphatic isocyanurates, quantum mechanical
calculations and viscosity simulations were carried out. To study the
molecular binding between NCO groups and the carbonyls of adjacent
isocyanurate rings, as well as between adjacent NCO groups, several model
bimolecular assembly systems were simulated using density functional
theory (DFT). For the NCO-to-ring interaction (Figure 3.2a), surprisingly, a
stable configuration was found. A significant free energy of binding of -8.7
kJ/mol was calculated for the NCO-to-ring interaction under standard
conditions (298 K, 1 atm). This is clearly above the thermal energy threshold.
However, the free energy of binding of the NCO-NCO bimolecular complex
(Figure 3.2b) was found to be positive (+23.1 kJ/mol), clearly suggesting the
absence of any relevant interaction between NCO groups themselves.
62
Figure 3.2. Bimolecular complexes used in the free energy calculations. (a)
The NCO-to-ring interaction (b) the NCO-NCO interaction. Possible
intermolecular binding sites involving Oxygen (large spheres) and Hydrogen
(small spheres) atoms are highlighted. Corresponding distances in (a) are: a
= 2.51 Å, b = 2.55 Å, c = 2.54 Å, d = 2.38 Å.
To verify the influence of these bimolecular interactions on the viscosity of
the system, MD simulations were carried out for both model series using the
Green-Kubo25 formalism. All calculated viscosity values were in agreement
with experimental results, showing the same trends. From the MD
simulations, it was found that, indeed, the additional electronic interaction
potential between NCO groups and isocyanurate rings gives rise to an
increase in the final viscosity.
Based on the combined experimental and computational results, it
is proposed that the newly identified bimolecular NCO-to-ring interaction
potential is the main cause of the high viscosities observed in functional
aliphatic isocyanurate systems. As a result, increased viscosities should be
expected only when NCO groups and isocyanurate rings are simultaneously
present. This is perfectly supported by the experimental data. Furthermore,
viscosity is expected to increase with increasing relative contents of NCO
63
groups and isocyanurate rings. This was also effectively demonstrated
experimentally. Due to the lack of interaction between adjacent NCO
groups, as indicated by the binding free energy calculations, no additional
contribution to viscosity is expected in systems containing only NCO groups.
This was also indeed observed, noting the very low viscosities of monomeric
(di)isocyanates such as HDI. Given that these interactions involve only the
functional groups, it is expected that the proposed mechanism will be
applicable to a much wider range of materials than just these model systems
based on linear aliphatic diisocyanates. In fact, in any system where both
NCO groups and isocyanurate rings are simultaneously present, this
interaction should be considered.
Synthesis and Characterization of Densely Crosslinked Isocyanurate
Networks
For the most common functional model precursor (HDI, n = 6) initial
attempts to prepare densely crosslinked networks by complete
trimerization were carried out. The resulting thermosets were investigated
in terms of mechanical and thermal properties. A comparison was made
between the use of two different catalysts (potassium acetate (KOAc) and
tin(II) 2-ethylhexanoate (SnOct)) and the difference between using the
difunctional diisocyanate or the trifunctional isocyanurate as the starting
material was investigated. The physical properties of the obtained materials
are reported in Table 3.1. All samples were collected as rigid transparent
thermosets, exhibiting slight yellowing, and in some cases, the formation of
a few small bubbles, depending on the conditions and catalyst (see Picture
3.1).
64
Picture 3.1. Photograph of HDI-based densely crosslinked polyisocyanurate
samples.
For these preliminary results it is noted that the glass transition
temperatures Tg are somewhat lower than those reported in the literature
(140 °C7,8), and that they exhibit some slight variation between samples.
Based on both the literature and our own findings, the differences in
physical properties can most likely be attributed to the degree and quality
of network formation. Because the network density increases rapidly during
the reaction, the probability of trimerizing three NCO groups quickly
decreases and complete network formation is difficult to achieve.
Furthermore, diffusion of the catalyst quickly becomes restricted, limiting
the rate of reaction even further. Based on these preliminary results,
thermosets prepared from NCO-functional isocyanurates as the starting
material seem to show better physical properties than those prepared from
the linear diisocyanates. Future work will be needed to find optimized
processing conditions.
65
Table 3.1. Physical properties of densely cross-linked isocyanurate networks
Onset of
mass loss
[°C]
Decom-
position
[°C]
Mass loss
[%]
Tga
[°C]
E-mod
[GPa]
σb
[MPa]
εb
[%]
150
°C
250
°C
HDI
trimerb
275 440 0.1 0.5 118 2.1c 60c 5.5c
HDIb 95 285 0.6 1.7 108 – – –
HDId 95 280 1.4 2.6 122 – – –
aGlass transition temperature. bCatalyst: potassium acetate (KOAc) in diethylene glycol (DEG). cCovestro internal report. dCatalyst: tin(II) 2-ethylhexanoate (SnOct).
3.4 Conclusions Model aliphatic isocyanurates of varying carbon chain length were
synthesized and isolated in high purities (>99%). The viscosities of these
model systems were accurately quantified, and viscosities were found to be
significantly higher for NCO-functional isocyanurates than for non-
functional isocyanurates. Viscosities also increased with increasing relative
contents of NCO groups and isocyanurate rings. Through DFT calculations, a
newly identified strong intermolecular binding potential was found to be
present between the NCO groups and the isocyanurate-ring carbonyl
groups. Also a less significant interaction between the NCO groups
themselves was identified. At standard conditions (298 K, 1 atm), a binding
free energy of -8.7 kJ/mol was calculated for the NCO-to-ring interaction.
This value is clearly above the thermal energy threshold. Molecular
dynamics simulations further confirmed the influence of this NCO-to-ring
interaction as a contributor to the viscosity of the system.
Based on the combined experimental and computational results, it
is proposed that this NCO-to-ring interaction is the main cause of the high
viscosities generally observed in functional aliphatic isocyanurate systems.
66
Consequently, the presence and relative content of the combination of NCO
groups and isocyanurate rings has a significant effect on increasing the
viscosity of such systems.
Finally, densely crosslinked HDI-based isocyanurate networks were
synthesized. These were found to exhibit interesting but varying mechanical
and thermal properties depending on catalyst use and the degree of
crosslinking achieved.
3.5 Acknowledgements This project has received funding from the European Union’s Horizon 2020
research and innovation program under the Marie Skłodowska-Curie Grant
Agreement no. 642890 (http://thelink-project.eu/) and it was partially
supported by the Portuguese Foundation for Science and Technology (FCT)
in the framework of the Strategic Funding UID/FIS/04650/2013, and under
the project Search-ON2: NORTE-07-0162-FEDER-000086.
67
3.6 References 1. A. Wurtz, Compt. Rend. 1848; 27, 241−243
2. O. Bayer, Angew. Chem. 1947; 59, 257−272
3. E. Delebecq, J.-P. Pascault, B. Boutevin, F. Ganachaud, Chem. Rev.
2013; 113, 80−118
4. 4 I. Yilgor, E. Yilgor, Polym. Rev. 2007; 47, 487–510
5. 5 D. Heift Z. Benkó, H. Grützmacher, A.R. Jupp, J.M. Goicoechea, Chem.
Sci. 2015; 6, 4017–4024
6. 6 S. Okumotoj, S. Yamabe, Comput. Chem. 2001; 22, 316-326
7. 7 T.A.C. Flipsen, R. Steendam, A.J. Pennings, G. Hadziioannou, Adv.
Mater. 1996; 8, 45-48
8. 8 P. Fabbri, S. Mohammad Poor, L. Ferrari, L. Rovati, S. Borsacchi, M.
Geppi, P.P. Lima, L.D. Carlos, Polymer 2014; 55, 488−494
9. 9 M. Moritsugu, A. Sudo, T. Endo, J. Polym. Sci. Part A: Polym Chem.
2013; 51, 2631−2637
10. 10 E. Preis N. Schindler, S. Adrian, U. Scherf, ACS Macro Lett. 2015; 4,
1268−1272
11. 11 D.K. Chattopadhyay, K.V.S.N. Raju, Prog. Polym. Sci. 2007; 32, 352-
418
12. 12 M. Ludewig, J. Weikard, N. Stockel, US 20060205911 A1, 2006
13. 13 H. Renz et al., Prog. Org. Coat. 2001; 43, 32–40.
14. 14 F.U. Richter, J. Pedain, H. Mertes, C.-G. Dieris, EP 0 798 299 B1, 1997,
13-14
15. 15 Gaussian ’09, Rev E01, Gaussian, Inc., Wallingford CT 2009
(www.gaussian.com)
16. 16 S. Grimme, J Antony, S. Ehrlich, H Krieg, J. Chem. Phys. 2010;
132,154104:1-19
17. 17 S. Grimme, S. Ehrlich. L. Goerigk, , J. Comput. Chem. 2011; 32, 1456-
1465
18. 18 J. Wang, R. M. Wolf, J. M. Caldwell, P. A. Kollman, D. A. Case, J. Comp.
Chem. 2004; 25, 1157-1174
19. 19 C. S. Bayly, P. Cieplak, W. Cornell, P. A. Kollman, J. Phys. Chem. 1993;
97, 10269-10280
20. 20 S. Plimpton, J. Comp. Phys. 1995; 117, 1-19
68
21. 21 M. Tuckerman, B. J. Berne, G. J. Martyna, J Chem Phys, 1992; 97,
1990-2001
22. 22 P. Ewald, Ann. Phys. 1921; 369 (3), 253–287
23. 23 W. G. Hoover, Phys. Rev. A 1985; 31 (3),1695–1697
24. 24 M. Parrinello, A. Rahman, J. Appl. Phys. 1981; 52, 7182-7190
25. 25 J. P. Hansen and I. R. McDonald, in Theory of Simple Liquids, Elsevier
Inc., 2006, Chapter 7
26. 26 D. C. Rapaport, in The Art of Molecular Dynamics Simulaton,
Cambridge University Press, 2004, Chapter 5
27. 27 Z. Pusztai, G.Vlád, A. Bodor, I.T. Horváth, H.-J. Laas, R. Halpaap, F.U.
Richter, Angew. Chem. 2006; 118, 113 –116
28. 28 H.-J. Laas, R. Halpaap, J. Pedain, J. Prakt. Chem. 1994; 336, 196–198
69
Chapter 4 - The Influence of Molecular
Interactions on the Viscosity of Aliphatic
Polyisocyanate Precursors for
Polyurethane Coatings
P.J. Driest1,2, V. Lenzi3, L.S.A. Marques3, M.M.D. Ramos3, F.U. Richter1, D.J.
Dijkstra1, D.W. Grijpma2
1Covestro Deutschland AG, CAS-Global R&D, 51373 Leverkusen, Germany
2Technical Medical Centre, Faculty of Science and Technology, Department
of Biomaterials Science and Technology, University of Twente P.O. Box 217,
7500AE Enschede, The Netherlands
3Department/Centre of Physics, University of Minho, Campus de Gualtar,
Braga 4710-057, Portugal
Published in:
ECS Conference Proceedings 2019
Abstract For the production, processing and application of aliphatic isocyanate (NCO)
based materials such as polyurethane coatings and adhesives, the viscosity
of NCO precursors is a key property. However, despite this importance,
surprisingly little is known about the underlying intermolecular interactions
influencing the viscosity these materials. In this work, a combined
experimental and theoretical study is presented investigating the various
intermolecular interactions present in aliphatic polyisocyanates and their
influence on viscosity.
By the use of ultrahigh-vacuum (<10-7mbar) short-path thin-film
distillation, series of functional and non-functional isocyanurate model
compounds were isolated in very high purity. The viscosities of these model
compounds were accurately determined and reported. It was shown that
intermolecular electronic effects involving the NCO group have a significant
influence on the viscosity of the liquid. In consequence, three different
intermolecular interactions involving NCO groups and isocyanurate rings
were proposed: NCO-NCO interactions, Ring-Ring interactions, and NCO-
Ring interactions.
To investigate these intermolecular interactions in more detail, ab
initio calculations were performed on representative small molecular model
compounds. Based on the results of the ab initio calculations, larger scale
molecular dynamics simulations could then be performed. From these
simulations, the contribution of each of the intermolecular interactions to
the viscosity of the liquid could be identified. It was shown that multiple
cooperative NCO-Ring interactions lead to the highest intermolecular
interaction energies, and consequently also to the highest viscosities.
Finally, using the Green-Kubo formalism and the recently developed GAFF-
IC forcefield, the viscosities of the model compounds used in the
experimental study were computationally predicted and turned out to be in
excellent agreement with the experimental data. Thereby, a complete
model is provided, explaining the previously poorly understood trends
regarding the viscosity of isocyanurate-based polyisocyanates.
72
4.1 Introduction Here we present a summary of the recent advancements in the study of the
viscosity of aliphatic polyisocyanates, as well as the various underlying
intermolecular interactions affecting it. A combined theoretical and
experimental approach is taken, investigating in detail the structure-
property relations of aliphatic polyisocyanates. Aliphatic polyisocyanates
form an important class of raw materials for the coatings, adhesives and
sealants industry, with a total annual production of ca. 242.000 tons.1,2 For
the production, the processing and especially the application of aliphatic
polyisocyanates, the viscosity of the liquid precursors is a key parameter.3–5
Surprisingly, however, very little is known about the underlying chemical
principles that dictate the viscosity of aliphatic polyisocyanate liquids.
Although some (mostly empirical) know-how exists within the industry, little
is reported in the scientific literature on the chemical principles and specific
(inter)molecular interactions playing a role.6 One of the things that is
specifically poorly understood, is why isocyanurate-based polyisocyanates
(also referred to as “trimers”) systematically seem to have a significantly
higher viscosity than similar oligomeric polyisocyanates based on other
NCO-derivatives, such as uretdione, allophanate and/or
iminooxadiazinedione (the latter also being referred to as “asymmetrical
trimer”).7 However, from a materials point of view, particularly the trimer-
based oligomers are of high interest and generally provide the best material
properties when it comes to e.g. heat-resistance, chemical resistance and/or
mechanical resistance of the cured coating or adhesive.8 Therefore, a more
fundamental understanding of the intermolecular interactions found in
aliphatic polyisocyanates in general, and in trimer-based polyisocyanates
specifically, and how these affect the viscosity of the liquid, would be very
useful.
Various intermolecular interactions in aliphatic polyisocyanates have been
identified and investigated in detail, together with their influence on
viscosity.9,10 For this purpose, a series of isocyanurate model compounds
with varying carbon chain lengths was designed, based on and including the
NCO-functional hexamethylenediisocyanate (HDI) trimer. To allow for
73
quantitative comparison within and between series, the model compounds
only concern the pure trimer fractions, i.e. the structures for which m = 1 as
drawn in Figure 4.1.
Figure 4.1, top. The trimerization of diisocyanates under the formation of one
or more isocyanurate ring(s).
Figure 4.1, bottom. A sketch representing the composition of a hypothetical
crude polyisocyanate reaction mixture. The trimerization reaction is
terminated prematurely by neutralizing the trimerization catalyst, after
which the unreacted monomer fraction is completely removed by thin-film
short-path evaporation. The specific composition of the resulting
“Polyisocyanate Mixture” depends on the conversion and is typically
reported as the average NCO% of the mixture. A.u. stands for arbitrary unit.
As a comparison, the equivalent series of non-functional tris(n-
alkyl)isocyanurates (i.e. the trimers of the respective monoisocyanates), as
well as the monomeric mono- and diisocyanates, were also included in the
study. In this way it was possible to isolate and investigate separately all the
74
relevant intermolecular interactions. A distinction was made between
isocyanate-isocyanate (NCO-NCO) interactions (present in the monomeric
mono- and diisocyanates), isocyanurate-isocyanurate (Ring-Ring)
interactions (present in the non-functional tris(n-alkyl)isocyanurates), and
isocyanate-isocyanurate (NCO-Ring) interactions (together with the first
two, present in the NCO-functional isocyanurates).
The first challenge, experimentally, is to isolate the model compounds in
high purity. This is particularly difficult for the functional isocyanurate model
compounds, for reasons sketched out in the bottom part of Figure 4.1.
Because during the trimerization reaction both the starting material and the
reaction product(s) contain equally reactive NCO groups, inevitably a
distribution is formed of trimers and increasing amounts of higher order
oligomers. Therefore, the reaction needs to be terminated prematurely by
neutralizing the trimerization catalyst, after which the unreacted monomer
fraction can be distilled off by thin-film short-path evaporation. The resulting
“Polyisocyanate Mixture” consists of a distribution of oligomers, the specific
composition depending on the degree of conversion during trimerization. In
order to isolate the pure trimer fractions from these mixtures of oligomers,
ultrahigh-vacuum thin-film short-path distillation was used to distil off each
trimer fraction as the distillate, followed by further purification.6 Using the
pure trimer fractions, the viscosities of the model compounds were
accurately determined. By comparing the viscosities of the functional
isocyanurates to those of the respective non-functional isocyanurates and
monomeric isocyanates, it was possible to obtain an estimation of the
contribution of each specific intermolecular interaction to viscosity.
To quantify more accurately the contribution of each of the various
intermolecular interactions, Density Functional Theory (DFT) and Symmetry-
Adapted Perturbation Theory (SAPT) calculations were performed. To keep
computational costs reasonable and isolate the bare interactions, smaller
representative model compounds were chosen for this part of the study.
Using the results from the quantum mechanical calculations as a starting
point, larger scale Molecular Dynamics (MD) simulations could then be
performed on the original series of model compounds, using the recently
75
developed GAFF-IC forcefield.10 Thereby, a complete overview is provided of
the various intermolecular interactions at play. Finally, the viscosities of the
model compounds were predicted using the Green-Kubo formalism and
compared to viscosities that were experimentally obtained.
4.2 Experimental Details
All experimental and computational details regarding this work are
described in the literature. For the details regarding the synthesis,
purification and characterization of the model compounds, the reader is
directed here.6 For the computational details regarding the DFT and ab initio
interaction energy calculations, the liquid state potential of mean force
(PMF) calculations and the pair correlation functions, the reader is directed
here.9 For the computational details regarding the modified GAFF forcefield,
GAFF-IC, and the resulting predictions of macroscopic physical properties,
including viscosity, the reader is directed here.10
76
4.3 Results and Discussion
Synthesis, Purification and Characterization of Isocyanate / Isocyanurate
Model Compounds
As previously described, we recently succeeded in isolating a series of pure
NCO-functional isocyanurate model compounds in high purity.6 First,
polyisocyanate mixtures of varying carbon chain lengths were synthesized
by the partial trimerization of the respective diisocyanate, followed by
removal of the unreacted monomer fraction by thin-film short-path
evaporation. Next, by using an ultrahigh-vacuum (<10-7mbar) thin-film
short-path evaporator operated at 220°C as the key step, it was possible to
distill off the pure trimer fractions from the respective polyisocyanate
mixtures as the distillate. Any reformed monomer or other low molecular
species that may have formed during the harsh conditions of the distillation
process were subsequently removed under <10-7mbar and at 160°C, yielding
the pure NCO-functional trimer fractions. As a comparison, the
corresponding non-functional tris(n-alkyl)isocyanurates (i.e. the trimers of
the corresponding monoisocyanates) were also synthesized and isolated.
Although no higher order oligomers are formed during the trimerization of
monoisocyanates, the isolation process was performed under the same
conditions as for the functional isocyanurates. All model compounds were
isolated in >99% purity (with the exception of the functional isocyanurate
with n=4 (>98%)), as confirmed by 1H-NMR, 13C-NMR and Size Exclusion
Chromatography (SEC), in most cases limited only by the detection level of
the equipment.
77
Figure 4.2. The viscosities at 23°C of the high-purity functional (R=NCO) and
non-functional (R=H) isocyanurate model compounds (i.e. the pure trimers),
with carbon chain lengths varying from 4 to 7. The R=NCO; n=6 point
corresponds to the HDI trimer.6
Next, the viscosities of all model compounds were accurately determined
using an Anton Paar MCR501 rheometer and a 50mm cone-plate setup. The
measured viscosities of all model compounds are reported in Figure 4.2. As
can be clearly seen from the figure, the viscosities of the functional
isocyanurates are systematically higher than those of the non-functional
isocyanurates. Considering that all molecules in this study are of comparable
molecular weight, it is not expected that differences in molecular weight
(alone) are the cause for such large differences in viscosity. Furthermore,
counter-intuitively, the viscosity increases with decreasing molecular weight
within the functional isocyanurate series. Likely, intermolecular non-
covalent effects are playing a role as well, apparently dictated by the
presence or absence of NCO groups. However, in the case of the monomeric
isocyanates, very low viscosities are found despite the abundant presence
of NCO groups. In order to gain more insight into the specific non-covalent
interactions involving the NCO groups and isocyanurate rings, quantum
78
mechanical calculations on NCO group- and isocyanurate-ring-containing
model compounds were performed.
Interaction Energy Calculations of Various Dimer Complexes of Isocyanate
/ Isocyanurate Model Compounds
In order to identify and quantify the various intermolecular interactions
existing between NCO groups and isocyanurate rings, a series of simple,
small molecular model compounds containing at least one NCO group,
isocyanurate ring, or both, was evaluated. The molecules used in this
computational study were methyl isocyanate (MIC), ethyl isocyanate (EIC),
ethylene diisocyanate (EDI), propyl isocyanate (PIC), propylene diisocyanate
(GDI; the acronym chosen with regard to the glutaryl backbone), hexyl
isocyanate (HIC), hexamethylene diisocyanate (HDI), trimethyl isocyanurate
(3MIC), trimethyl iminooxadiazinedione (A3MIC; the “asymmetrical trimer”
of methyl isocyanate), triethyl isocyanurate (3EIC) and tris(3-
isocyanatopropyl) isocyanurate (3GDI; the symmetrical trimer of GDI). To
analyze separately the different types of intermolecular interactions, various
bimolecular (dimer) complexes of these model compounds were formed.
The composition of the various dimer complexes, together with their
calculated intermolecular interaction energies by two computational
methods are reported in Table 4.1.9 The dimer complexes are organized into
three categories, based on the type of intermolecular interaction that is
present: NCO-NCO, NCO-Ring or Ring-Ring.
79
Table 4.1: Interaction energies of model dimer complexes, calculated at the
B3LYP-D3BJ level of theory and SAPT2+ methods.9 In some cases, more than
one stable configuration is possible. Three representative dimer complexes
(underlined in the first column) are shown on the right for each category.
Dimer ID
#
Interaction
type
EDFT
[kcal/mol]
ESAPT2+
[kcal/mol]
Representative
Dimer Complex
MIC-
MIC 1 NCO-NCO −5.81 −5.89
EIC-EIC 2 −6.06 −6.47
EDI-EDI 3 −6.68 −7.17
PIC-PIC 4 −6.26 −6.78
PIC-GDI 5 −7.18 −7.61
GDI-GDI 6 −7.85 −8.21
HIC-HIC 7 −6.37 −6.98
HIC-HDI 8 −6.68 −5.84
HDI-HDI 9 −6.92 −7.45
MIC-
3MIC 10 NCO-Ring −7.08 −7.55
3MIC-
EIC 11 −7.38 −8.11
3MIC-
PIC 12 −7.71 −8.46
EIC-
3EIC 13 −7.02 −7.91
3GDI-
3GDI (a) 14 −13.49 n/a
3GDI-
3GDI (b) 15 −17.58 n/a
80
Dimer ID
#
Interaction
type
EDFT
[kcal/mol]
ESAPT2+
[kcal/mol]
Representative
Dimer Complex
3MIC-
3MIC 16 Ring-Ring −13.75 −14.18
A3MIC-
A3MIC
(a)
17 −11.64 −13.13
A3MIC-
A3MIC
(b)
18 −12.36 −13.60
A3MIC-
3MIC 19 −13.16 −14.40
3EIC-
3EIC (a) 20 −12.13 −13.72
3EIC-
3EIC (b) 21 −13.08 −15.31
3EIC-
3EIC (c) 22 −10.02 −10.84
It was found that the intermolecular interaction energies are the weakest
for NCO-NCO interactions, ranging from -5.81 kcal/mol to -7.85 kcal/mol.
The various Ring-Ring interactions are clearly stronger, showing significant
interaction energies of up to -13.75 kcal/mol. As a comparison, this is ca.
three times higher than the calculated intermolecular interaction energy of
a hydrogen bond between two water molecules in vacuum.11 The NCO-Ring
interactions are more elusive in nature. Interaction energies in the range of
-7 to -8 kcal/mol were found when a single interaction is present, i.e. on the
same order of magnitude as the NCO-NCO interactions. However,
interaction energies increasing up to -17.58 kcal/mol were found when
multiple, cooperative NCO-Ring interactions are possible, such as between
a pair of (functional) GDI-trimers: 3GDI-3GDI. Although it was not possible
to obtain SAPT results for this specific dimer complex, based on DFT
calculations it was possible to quantify the cooperative effect of the multiple
81
NCO-Ring interactions more accurately. As shown in Figure 4.3, a 3GDI-3GDI
dimer complex was modified by sequentially replacing NCO groups with
hydrogen atoms, whose positions were then relaxed while keeping all other
atoms fixed. As reported in Figure 4.3, the removal of an NCO group taking
part in an NCO-Ring binding (i.e. NCO groups labeled 1 and 3) leads to a
drastic reduction in the total interaction energy of 6.46 kcal/mol and 5.72
kcal/mol, respectively. The removal of a more distant NCO group on the
other hand, such as the one labeled 2 in Figure 4.3, reduces the total
interaction energy by only 1.52 kcal/mol. Finally, to confirm that the NCO-
Ring interactions are indeed the main cause for the high total interaction
energies, the case is shown in which the three outermost NCO groups,
labeled 3* in Figure 4.3, are replaced by hydrogen atoms. In this case, the
total interaction energy is only reduced by a mere 1.07 kcal/mol. This clearly
shows that the NCO groups taking part in NCO-Ring interactions are
responsible for the high interaction energies.
Figure 4.3. Interaction energies of a 3GDI-3GDI dimer complex as a function
of the replacement of specific NCO groups by hydrogen atoms. Full (red)
circles highlight potential NCO-Ring interactions, while dashed (black) circles
indicate the peripheral NCO groups. On the left, a representation of the
dimer complex is shown, along with the NCO groups being substituted. The
3* point represents the dimer complex in which only the three outermost
NCO groups, labelled 3*, are substituted.9
82
Towards the Liquid State: Molecular Dynamics (MD) Simulations of
Isocyanate / Isocyanurate Molecular Liquids
To make the step from model compounds in the vacuum to more realistic
conditions in which isocyanates are commonly processed, i.e. the liquid
state, larger scale molecular dynamics simulations were performed. For this,
the recently developed modified GAFF forcefield, GAFF-IC, was used,
optimized for accurately predicting macroscopic physical properties of
aliphatic (poly)isocyanates.10 To obtain estimates of the binding free
energies in the liquid state, the Potential of Mean Force (PMF) curves of
three selected dimer complexes, representative of the three types of
interactions, were calculated and are shown in Figure 4.4.8
Methylisocyanate (MIC) was chosen as the solvent in this case. In agreement
with the results obtained in vacuum, the NCO-NCO and single NCO-Ring free
energies are the smallest, with local minima less than -0.2 kcal/mol. The
3MIC-3MIC dimer complex on the other hand, capable of forming the
stronger Ring-Ring interactions, shows a strong free energy profile. The most
important features are a deep minimum around 3.6 Å of -0.81 kcal/mol, and
two shallower local minima at 5.4 Å and 8.2 Å of −0.22 kcal/mol and −0.18
kcal/mol, respectively. In other words, systems containing isocyanurate
rings can form strongly bound pairs at short intermolecular distances, but
only moderately bound pairs at longer intermolecular distances.
Unfortunately, due to the high conformational complexity, it was not
possible to obtain PMF curves for the 3GDI-3GDI dimer complex to
investigate the effect of multiple, cooperative NCO-Ring interactions.
83
Figure 4.4: PMF curves obtained at 298 K and 1 atm, in a pure MIC liquid
solution, for three different model compound dimer complexes, representing
the three possible contact types: MIC-MIC representing NCO-NCO
interactions, MIC-3MIC representing single NCO-Ring interactions and 3MIC-
3MIC representing Ring-Ring interactions.9 Due to the high conformational
complexity, it was not possible to obtain PMF curves for the 3GDI-3GDI dimer
complex.
Extending the model one step further towards more realistic liquid
(poly)isocyanate systems, the next step was to perform molecular dynamics
simulations on the larger functional and non-functional isocyanurate model
compounds representing real (poly)isocyanate liquids. Now, the same series
of model isocyanurates as presented in the experimental study were used.
First, the intermolecular ring-ring pair correlation functions g(r) were
calculated both for the functional and non-functional isocyanurates, which
are reported in Figure 4.5. It is seen that for the non-functional isocyanurate
liquids, only capable of forming Ring-Ring interactions, the pair correlation
functions approximately reflect the PMF curve of the 3MIC-3MIC dimer
84
complex reported in Figure 4.4. The most important features are the two
maxima around 4Å and 6Å, corresponding to molecular pairs involved in
direct Ring-Ring interactions. However, for the functional isocyanurates, the
pair correlation functions look very different, with a significant reduction of
the g(r) at short distances. This means that when NCO groups are present,
direct Ring-Ring contacts are less frequent. Based on the data shown
previously in Table 4.1 and Figure 4.3, it is expected that this reduction in
Ring-Ring interactions can be attributed to NCO groups undergoing NCO-
Ring interactions, thereby hindering the rings for further contacts.
Next, the ring-ring pair correlation functions were decomposed for
two representative isocyanurates (the functional and non-functional
isocyanurates for which n=5) by calculating, for each molecular pair, the
relative tilt angle of the two rings. The angular decomposition is shown in
the bottom part of Figure 4.5. For the non-functional isocyanurate, it was
found that the rings have a tendency to orient in a parallel fashion, especially
at short intermolecular distances, optimizing the orientation for direct Ring-
Ring interactions. The functional isocyanurate on the other hand, has a much
less pronounced orientation profile, with fewer rings found in a parallel
orientation. The chain length of the alkylene side-chains does not seem to
have a very significant effect on the pair correlation functions.
85
Figure 4.5. Intermolecular ring-ring pair correlation functions g(r) and an
angular decomposition g(r,θ) for the case of n=5 as a function of the ring-
ring pair distance r and tilt angle θ for non-functional (left) and functional
(right) isocyanurates.9
86
Testing the Model: Predicting the Viscosities of Isocyanate / Isocyanurate
Model Compounds
Using the recently developed modified GAFF forcefield, GAFF-IC, optimized
for accurately predicting macroscopic physical properties of aliphatic
(poly)isocyanates,10 the effect of multiple, cooperative NCO-Ring
interactions on viscosity was investigated. Due to the observed high
viscosities at 23 ºC, which prevent any meaningful calculation with the
Green-Kubo formula, the MD simulations and respective experimental
measurements were performed at 85°C in this case. The resulting predicted
viscosities are reported in Figure 4.6, together with the respective
experimentally obtained values. As seen in the figure, the newly developed
model was perfectly capable of reproducing the experimentally found data
of the model compounds. The key trends, such as the large difference
between functional and non-functional isocyanurates and increasing
viscosity for decreasing molecular weight functional isocyanurates were
confirmed.
Furthermore, the effects of the various intermolecular interactions
described in the previous sections are clearly visible, providing an
explanation for the observed trends. Firstly, in systems only capable of
forming NCO-NCO interactions (i.e. isocyanate monomers), the lowest
intermolecular interaction energies were found, and consequently the
lowest viscosities are expected (data not shown in the figure). Second, in
systems only capable of forming Ring-Ring interactions (i.e. the non-
functional isocyanurates), slightly higher interaction energies were found,
and thus slightly higher viscosities are expected. However, since the Ring-
Ring interactions were shown to not be capable of forming strong
intermolecular interactions over large distances, the contribution of this
type of binding to viscosity is expected to be only moderate. Finally, in
systems capable of forming multiple, cooperative NCO-Ring interactions (i.e.
the functional isocyanurates), the highest interaction energies were found,
and consequently the highest viscosities are expected. Moreover, because
of its cooperative nature, the presence of the NCO-Ring binding is expected
to have the highest contribution to viscosity.
87
Figure 4.6. The viscosities of isocyanurate model compounds. The solid
markers represent the experimentally measured viscosities6, while the open
markers represent the calculated viscosities using the Green-Kubo formalism
and the modified GAFF-IC force field.10
When comparing these expected trends with the actual viscosity values
(both calculated and experimentally measured), it is seen that these are
exactly supported. Even the initially counter-intuitive trend of increasing
viscosity with decreasing molecular weight within the functional
isocyanurate series can be explained by the proposed model of multiple,
cooperative NCO-Ring interactions, being more concentrated in the shorter-
chain isocyanurate liquids. Thereby, a first overview is provided of the
relevant intermolecular interactions in aliphatic polyisocyanates, and how
these affect the viscosity of the liquid.
88
4.4 Conclusions For the investigation of the viscosity of aliphatic isocyanurate-based
polyisocyanates, a combined experimental and theoretical study was
performed. For this purpose, series of functional and non-functional
isocyanurate model compounds were designed. These model compounds
were synthesized and by the use of ultrahigh-vacuum (<10-7mbar) short-
path thin-film distillation were isolated in very high purity.6 The viscosities
of these model compounds were accurately determined and reported. It
was shown that intermolecular non-covalent effects involving the NCO
group have a significant influence on the viscosity of the liquid.
Subsequently, three different intermolecular interactions involving NCO
groups and isocyanurate rings were proposed: NCO-NCO interactions, Ring-
Ring interactions, and NCO-Ring interactions.
To investigate these intermolecular interactions in more detail, ab
initio and DFT calculations were performed on representative small
molecular model compounds. It was shown that NCO-NCO interactions are
the weakest, ranging from -5.81 to -7.85 kcal/mol. Ring-Ring interactions are
stronger, showing significant interaction energies of up to -13.75 kcal/mol.
Finally, NCO-Ring interactions were shown to be less predictable in nature,
with interaction energies between -7 to -8 kcal/mol for single interactions,
but with interaction energies increasing up to -17.58 kcal/mol when
multiple, cooperative interactions are possible.
Based on the results of the ab initio and DFT calculations, larger
scale molecular dynamics simulations were performed. From these
simulations, the contribution of each of the intermolecular interactions to
the viscosity of the liquid was identified. Using the Green-Kubo formalism
and the recently developed GAFF-IC force field, the viscosities of the model
compounds used in the experimental study were computationally predicted
and turned out to be in excellent agreement with the experimental data.
Thereby, a first overview is provided of the relevant intermolecular
interactions found in aliphatic polyisocyanates, by which the previously
poorly understood trends regarding the viscosity of isocyanurate-based
polyisocyanates could be explained.
89
4.5 Acknowledgements This project has received funding from the European Union’s Horizon 2020
research and innovation program under the Marie Skłodowska-Curie Grant
Agreement no. 642890 (http://thelink-project.eu/) and it was partially
supported by the Portuguese Foundation for Science and Technology (FCT)
in the framework of the Strategic Funding UID/FIS/04650/2013, by the FCT
grant SFRH/BD/128666/2017 and under the project Search-ON2: NORTE-07-
0162-FEDER-000086.
90
4.6 References 1. Meier-Westhues, U., Danielmeier, K., Kruppa, P. & Squiller, E. P.
Polyurethanes − Coatings, Adhesives, Sealants. (Vincentz Network,
Hannover, 2019).
2. Golling, F. E. et al. Polyurethanes for coatings and
adhesives - chemistry and applications. Polym. Int. (2018).
doi:10.1002/pi.5665
3. Chattopadhyay, D. K. & Raju, K. V. S. N. Structural engineering of
polyurethane coatings for high performance applications. Prog.
Polym. Sci. 32, 352–418 (2007).
4. Ludewig, M., Weikard, J. & Stockel, N. Allophanate-containing
modified polyurethanes (US 2006/0205911 A1). (2006).
5. Renz, H. & Bruchmann, B. Pathways targeting solvent-free PUR
coatings. Prog. Org. Coatings 43, 32–40 (2001).
6. Driest, P. J. et al. Aliphatic isocyanurates and polyisocyanurate
networks. Polym. Adv. Technol. 28, 1299–1304 (2017).
7. Widemann, M. et al. Structure-Property Relations in Oligomers of
Linear Aliphatic Diisocyanates. ACS Sustain. Chem. Eng. 6, 9753–
9759 (2018).
8. Behnken, G., Hecking, A. & Sánchez, B. V. High performance enabled
by nature − New polyurethane crosslinkers with significant bio-
based content. Eur. Coatings J. 46–50 (2016).
9. Lenzi, V., Driest, P. J., Dijkstra, D. J., Ramos, M. M. D. & Marques, L.
S. A. Investigation on the intermolecular interactions in aliphatic
isocyanurate liquids: revealing the importance of dispersion. J. Mol.
Liq. (2019). doi:10.1016/j.molliq.2019.01.165 (in press)
10. Lenzi, V., Driest, P. J., Dijkstra, D. J., Ramos, M. M. D. & Marques, L.
S. A. GAFF-IC: realistic viscosities for isocyanate molecules with a
GAFF-based force field. Mol. Simul. 45, 207–214 (2019).
11. Steiner, T. The Hydrogen Bond in the Solid State. Angew. Chemie Int.
Ed. 41, 48–76 (2002).
91
Chapter 5 - The Trimerization of
Isocyanate-functionalized Prepolymers:
an Effective Method for Synthesizing
Well-defined Polymer Networks
P.J. Driest1,2, D.J. Dijkstra1, D. Stamatialis2, D.W. Grijpma2
1Covestro Deutschland AG, CAS-Global R&D, 51373 Leverkusen, Germany
2Technical Medical Centre, Faculty of Science and Technology, Department
of Biomaterials Science and Technology, University of Twente P.O. Box 217,
7500AE Enschede, The Netherlands
Published in:
Macromol. Rapid Commun. 40 (2019), 1800867
Abstract
For the study of polymer networks,
having access to polymer networks
with a controlled and well-defined
microscopic network structure is of
great importance. However, typically,
such networks are difficult to
synthesize. In this work, a simple,
effective and widely-applicable
method is presented for synthesizing
polymer networks with a well-defined network structure. This is done by the
functionalization of polymeric diols using a diisocyanate, and their
subsequent trimerization.
Using hexamethylene diisocyanate and hydroxyl-group-terminated
poly(ε-caprolactone) and poly(ethylene glycol), it was shown that both
hydrophobic and hydrophilic poly(urethane-isocyanurate) networks with a
well-defined network structure can readily be synthesized. By using in-situ
ATR-FTIR, it was shown that the trimerization of isocyanate endgroups is
clearly the predominant reaction pathway of network formation, supporting
the proposed mechanism and network structure. The resulting networks
possessed excellent mechanical properties, in both the dry and in the wet
state.
94
5.1 Introduction In modern materials chemistry, chemically crosslinked polymer networks
play a significant role. To understand and control the physical properties of
these materials, the relation between the microscopic, molecular structure
of the polymer network and the macroscopic material properties is key.
Therefore the study of these structure-property relations has received much
attention over the years.[1-3] For an accurate study, it is essential to have
access to well-defined polymer networks with a variable network structure
that can be prepared in a controlled manner. However, most of the common
synthesis methods for polymer networks lead to random network structures
which, at a molecular level, are poorly defined.[4,5] Consequently, methods
for synthesizing microscopically well-defined polymer networks are of
special interest,[6-8] but typically involve complex synthesis and/or are
limited to specific systems only.[9,10] Therefore, a simple and widely
applicable method for synthesizing well-defined polymer networks would be
of great use.
Scheme 5.1. The isocyanate reactions discussed: a) reaction with an alcohol
under formation of a urethane bond (in this case uncatalyzed), b) the
trimerization of isocyanates resulting in the formation of an isocyanurate-
ring, c) the mechanism of the trimerization reaction catalyzed by a
nucleophile (Nu-).[12]
In polyurethane chemistry, a well-known means of network formation is by
the trimerization reaction of isocyanates.[11] In this reaction, shown in
95
Schemes 5.1b and 5.1c, three isocyanate-groups undergo a step-wise
addition reaction forming an isocyanurate-ring, typically catalyzed by a
potent nucleophile (Nu-).[12] Since the trimerization of isocyanates occurs
readily and controllably, forming well-defined trifunctional isocyanurate-
rings, this reaction is an ideal candidate for the synthesis of well-defined
polymer networks. Surprisingly, the full potential of this reaction towards
this end has not been exploited yet.
Scheme 5.2. Simplified graphical representation of the discussed synthesis
methods for poly(urethane-isocyanurate) networks: a) trimerization
during/after step-growth polyurethanization (also referred to as the
“prepolymer two-stage technique”, see reference 13) leads to a broad
distribution in molecular weight between crosslinks (Mc); b) 2-Component
polyurethane network formation leads to a poorly defined functionality of
the crosslinking units, as well as to stoichiometry issues; whereas c) the
trimerization of NCO-functionalized prepolymers leads to a narrow
distribution of molecular weight between crosslinks as well as a well-defined
functionality of the crosslinking units of exactly 3.
96
Currently, there are two methods that are commonly adopted to
prepare polyurethane networks on the basis of isocyanate trimerization. In
the first method, shown in Scheme 5.2a, the trimerization reaction is
introduced as an intended side-reaction during polyurethanization.[13] In this
case, a (typically oligomeric) difunctional alcohol is reacted with a slight
molar excess of monomeric diisocyanate (typical molar ratios for NCO:OH lie
between 1.1:1 and 2:1). The remaining isocyanates are then trimerized
(either simultaneously or afterwards), resulting in a poly(urethane-
isocyanurate) network.[14 In such a network, the crosslinking units are well-
defined and trifunctional. However, as a result of the step-growth-character
of the polyurethanization reaction, the molecular weight between crosslinks
(Mc) has a broad molecular-weight-distribution with a dispersity index (Đ =
𝑀𝑤/𝑀𝑛) approaching a value of 2 and is therefore poorly-defined.[15]
In the second method, shown in Scheme 5.2b, a monomeric
diisocyanate is trimerized in advance into a functional polyisocyanate
mixture (sometimes referred to as “trimer” or “hardener”). During this
reaction, both the reactants and the reaction-products contain (equally)
reactive isocyanate-groups. Therefore, this reaction can only be carried out
to low conversion. After removal of the unreacted monomer fraction, the
resulting polyisocyanate mixture consists of a distribution of trimers and
higher-order oligomers, with a corresponding distribution in functionality
that is on average larger than 3.[16] To form the poly(urethane-isocyanurate)
network, this polyisocyanate mixture is reacted with a polymeric diol as a 2-
component system. In the resulting network, the molecular weight between
crosslinks is well-defined by the characteristics of the polymeric diol used.
However, in this case the functionality of the crosslinking units is poorly-
defined, due to the distribution in functionality of the polyisocyanate
mixture. Also, the 2-component character of the network formation step
often leads to stoichiometry issues. In this light, it is worth noting that other
2-component approaches for synthesizing well-defined polyurethane
networks have been described as well.[17,18] These are typically based on
reacting a macromolecular di-/polyol with a multifunctional isocyanate, or a
macromolecular polyisocyanate with a multifunctional alcohol or amine.
Since these networks do not contain isocyanurate-rings, and hence are not
97
poly(uretahane-isocyanurate) networks, further details will not be discussed
in this paper.
Here we propose a new method for poly(urethane-isocyanurate)
network formation that combines the benefits of the two previously
described techniques, see Scheme 5.2c. First, a polymeric diol (with a narrow
molecular-weight-distribution) is functionalized using a large excess of
monomeric diisocyanate. After removal of the unreacted excess of
monomeric diisocyanate, an NCO-functionalized prepolymer with a narrow
molecular-weight-distribution is obtained. This functionalized prepolymer is
then trimerized to form the poly(urethane-isocyanurate) network. In this
case, the molecular weight between crosslinks is expected to be well-
defined (i.e. displaying a narrow molecular-weight-distribution), while at the
same time all the crosslinking units are well-defined and trifunctional.
Furthermore, since the network formation step is based on a single-
component reactive system, no stoichiometry issues arise. A simple and
effective method with a wide applicability (i.e. to any polymeric diol) for
synthesizing well-defined polymer networks is thus provided. It is worth
noting that we chose the term “well-defined” networks deliberately here, as
the terms “model” or “ideal/perfect” networks which are also found in the
literature include addressing the potential formation of trapped
entanglements and/or macrocycles.
In this paper, two well-known polymeric diols (poly(ε-caprolactone)
(PCL) and poly(ethylene glycol) (PEG)) of different molecular weights were
functionalized using hexamethylene diisocyanate (HDI), and used to
respectively prepare well-defined hydrophobic and hydrophilic polymer
networks.
98
5.2 Experimental Details
Materials and methods
Poly(ethylene glycol) (𝑀𝑛 4 and 10 kg/mol) were purchased from Sigma-
Aldrich. Poly(ε-caprolactone) (𝑀𝑛 4 and 8 kg/mol) were kindly supplied by
Perstorp Chemicals GmbH. Hexamethylene diisocyanate was supplied by
Covestro Deutschland AG. All other chemicals were purchased from Sigma-
Aldrich.
All chemical reactions were conducted under a nitrogen
atmosphere using standard Schlenck-line techniques. Differential scanning
calorimetry (DSC) measurements were performed under nitrogen on ca.
10mg samples on a Perkin-Elmer Calorimeter DSC-7 in 2 heating runs from -
100°C to +150°C with a heating rate of 20 K/min. and a cooling rate of 20
K/min. NMR spectra were collected on a Bruker Advance III-700 in C6D6.
FTIR-spectra were recorded on a Bruker FTIR Spectrometer Tensor II with
Platinum-ATR-unit with diamond crystal. Size exclusion chromatography
(SEC) was performed according to DIN 55672-1:2016-03, on 4 PSS SDV
Analytical columns (2x 100Å, 5µm; 2x 1000Å, 5µm) using an Agilent 1100
Series pump and an Agilent 1200 Series UV Detector (230 nm) with
tetrahydrofuran as the elution solvent at 40°C and 1.00 mL/min. Residual
monomeric diisocyanate contents were determined by gas chromatography
(GC) according to DIN EN ISO 10283 on an Agilent Technologies 6890N
system using a 15.0m DB17 column and tetradecane as an internal standard.
Tensile tests were performed in triplicate at room temperature using a Zwick
Z0.5 tensile tester equipped with a 500N load cell at 200mm/min. with a
grip-to-grip separation of 35mm on strips of ca 60 x 4 x 0.5mm cut from a
cured film.
For swelling and solvent uptake measurements, three samples
measuring approximately 10 x 20 x 0.5mm were cut from a cured film and
weighed (denoted 𝑚𝑖). Each sample was then extracted with chloroform for
24h while periodically refreshing the solvent. Next, the sample was dried,
initially in air and subsequently under vacuum until constant weight
(denoted 𝑚𝑒). The gel content was then quantified using:
99
𝑔𝑒𝑙 𝑐𝑜𝑛𝑡𝑒𝑛𝑡 =𝑚𝑒
𝑚𝑖× 100%
Each sample was then swollen in excess chloroform or water for at least 24h,
blotted dry and weighed. Chloroform and water uptake was defined as:
𝑠𝑜𝑙𝑣𝑒𝑛𝑡 𝑢𝑝𝑡𝑎𝑘𝑒 =𝑚𝑠 − 𝑚𝑒
𝑚𝑒× 100%
where 𝑚𝑠 is the weight of an extracted specimen after swelling in the
specified solvent.
Synthesis of NCO-functionalized prepolymers
All polymeric diols were dried prior to use by azeotropic distillation in
toluene, which was subsequently removed under reduced pressure. In a
typical experiment, 0.25g of dibutyl phosphate was added to 100 g of dried
polymeric diol (kept in the melt using a heat gun). This mixture was added
dropwise to HDI at 100°C, using a molar ratio of NCO:OH of 15:1, and
subsequently kept at 100°C for 3h. The resulting mixture was then
transferred to a short-path thin-film evaporator, which was operated at a
reduced pressure of 1∙10-2mbar at 140°C. The functionalized prepolymers
were collected as clear, colorless, viscous resins that crystallized upon
cooling. Products were analyzed by SEC, GC and 1H-NMR (600 MHz, C6D6,
representative data for PEG-4k): δ = 4.60 (2H, brs), 4.26 (4H, t), 3.50 (410H,
m), 2.97 (4H, q), 2.57 (4H, t), 1.12 (4H, t), 1.02 (4H, t), 0.88 (8H, m) ppm.
Synthesis of poly(urethane-isocyanurate) networks by the trimerization of
NCO-functionalized prepolymers
In a typical experiment, 12 droplets of Sn(II)Oct2 (corresponding to ca.
0.75wt% relative to the polymer) were added to 8g of NCO-functionalized
prepolymer at 90°C. After mixing for 30 seconds in a Hauschild Speedmixer
DAC150FVZ, the mixture was brought into a mold consisting of a 0.5mm-
thick polycarbonate frame clamped between two silanized glass plates and
kept at 90°C for 24h. Cured samples were collected as flexible transparent
films, of which PCL-8k-, PEG-4k- and PEG-10k-based films slowly turned
100
opaque upon cooling. The films were analyzed by extraction, swelling,
tensile testing and DSC experiments. From the DSC measurements, melting
enthalpy values were obtained from the area under the melting peak. The
degree of crystallinity was then defined as:
𝑑𝑒𝑔𝑟𝑒𝑒 𝑜𝑓 𝑐𝑟𝑦𝑠𝑡𝑎𝑙𝑙𝑖𝑛𝑖𝑡𝑦 =∆𝐻𝑓
∆𝐻𝑓,𝑐𝑟𝑦𝑠× 100%
where ∆𝐻𝑓 is the melting enthalpy of the sample and ∆𝐻𝑓,𝑐𝑟𝑦𝑠 is the melting
enthalpy of a 100%-crystalline sample, for which values were taken from the
literature as 139.5 J/g for PCL[19] and 197 J/g for PEG[20], respectively. For
ATR-FTIR measurements, the catalyst was added to a 50wt%-solution of
functionalized prepolymer in dichloromethane, after which the solvent was
allowed to evaporate at RT overnight. A small sample was then placed
directly onto the ATR-crystal and kept at 80°C for 3h.
101
5.3 Results and Discussion
Synthesis of NCO-functionalized prepolymers
Since commercial polymeric diols may contain traces of base as received, a
small amount of dibutyl phosphate was always added prior to reaction to
neutralize the reaction mixture. Based on 1H-NMR, it was shown that the
functionalization of the polymeric PCL and PEG diols with HDI succeeded
selectively and quantitatively. Specifically, no allophanate formation was
observed. From the 1H-NMR-spectra, the average molecular weight of the
functional prepolymers could also be calculated. In order to synthesize
functionalized prepolymers with a minimally affected polydispersity index Đ
with respect to the polymeric diol, a large molar excess of diisocyanate of
15:1 was used. Using SEC, it was shown that during the functionalization
with HDI, the polydispersity index Đ of the polymers typically only increased
by about 0.05 compared to the respective polymeric diol. This shows that
the synthesis of functionalized prepolymers with a narrow molecular-
weight-distribution close to 1 is possible; where the final polydispersity is
mostly dependent on the quality of the starting material. The resulting
polydispersity indices of the prepolymers after functionalization are
reported in Table 5.1. Finally, it was shown by GC that the unreacted excess
of monomeric diisocyanate, as well as any traces of unreacted dibutyl
phosphate, were effectively removed by short-path thin-film distillation,
resulting in functionalized prepolymers with a residual HDI content lower
than 0.2wt% in all cases. The characteristics of all functionalized
prepolymers are reported in Table 5.1.
Formation of poly(urethane-isocyanurate) networks by the trimerization
of NCO-functionalized prepolymers
To investigate the proposed mechanism of network formation, the
trimerization reaction was monitored in-situ over time using ATR-FTIR. In
Figure 5.1, the FTIR-spectra of the formation of a PEG-4k network are shown
as an example. It is seen that the absorption peak at 2270cm-1, which can be
attributed to the NCO-stretching vibration, gradually disappears over time
until completely absent. More importantly, an absorption peak
102
simultaneously appears at 1690cm-1. This peak can be attributed to the
absorption of the carbonyl group of the isocyanurate ring. Noting that in
isocyanate chemistry, typically, no free carbonyl absorptions other than that
of the isocyanurate ring appear at wave numbers lower than 1700cm-1,
trimerization of isocyanates is clearly taking place.[21] Furthermore, it should
be noted that the rest of the spectrum remains essentially unchanged
throughout the course of the reaction with regard to the shape and intensity
of the peaks. This also includes the urethane-carbonyl peak at 1721 cm-1.
Although potential side-reactions such as allophanate and/or uretdione
(dimer) formation cannot be completely ruled out, the trimerization of the
isocyanate endgroups is clearly the predominant reaction pathway of
network formation. Based on the proposed mechanism, this results in fully
crosslinked networks with well-defined trifunctional crosslinking units, as
sketched in Scheme 5.2c. It should be noted that the formation of trapped
entanglements (and to a very small extent macrocycle formation) cannot be
ruled out though.
Figure 5.1. ATR-FTIR spectra of the formation of a PEG-4k network by
trimerization. "Start. Mat." shows the spectrum of the NCO-functionalized
prepolymer without catalyst, whereas "x min." shows the spectrum of the
reaction mixture after addition of the catalyst, evaporation of the solvent
and curing at 80°C for x minutes.
Table 5.1. Characterization of the NCO-functionalized prepolymers and their corresponding networks (numbers in parentheses are standard deviations).
Polymer NCO-Functionalized
Prepolymer Network
Mn [kg/ mol]
Đ [-]
HDI Content
[wt%]
Gel Content [wt%]
CHCl3 Uptake [wt%]
Water Uptake [wt%]
Tg
[°C]
Tm,p
[°C]
Emod,dry [MPa]
σb,dry
[MPa]
εb,dry
[%]
Emod,wet [MPa]
σb,wet
[MPa]
εb,wet
[%]
PCL-4k 3.8 1.29 0.05 99.2 (0.2)
408.1 (17.7)
2.0 (0.7)
-53.0 +27.6 6.6
(0.8) 4.1
(0.4) 109 (13)
6.8 (0.5)
2.7 (0.2)
62 (7)
PCL-8k 7.7 1.43 0.17 98.9 (0.3)
507.4 (2.2)
1.0 (0.2)
-55.7 +40.6 136.6 (9.5)
20.9 (4.5)
224 (30)
142.9 (-)a
22.9 (-)a
225 (-)a
PEG-4k 4.0 1.15 0.03 100 (0.0)
598.4 (7.6)
219.7 (0.5)
-47.6 +40.6 187.7 (14.4)
13.6 (2.4)
184 (63)
3.8 (0.1)b
0.9 (0.1)b
32 (3)b
PEG-10k 10.1 1.08 0.05 96.6 (1.5)
1265.9 (155.5)
508.0 (74.3)
-49.4 +55.7 165.2 (0.4)b
30.9 (3.1)b
700 (10)b
1.0 (0.3)
0.6 (0.1)
106 (10)
an=1 bn=2
104
Material properties of well-defined poly(urethane-isocyanurate) networks
The material properties of the synthesized networks are also reported in
Table 5.1. For all networks, high gel contents close to 100% were
determined. Together with the FTIR results discussed before, this shows that
the trimerization reaction was completed to a high degree of conversion,
with fully crosslinked networks as a result. As shown by the presence of a
peak melting temperature (Tm,p) in the DSC experiments, all networks were
semi-crystalline upon cooling. The cured PCL-4k and PCL-8k networks
possessed degrees of crystallinity of 24.9% and 29.4%, respectively, while
the PEG-4k and PEG-10k networks possessed degrees of crystallinity of
39.3% and 54.4%, respectively. It was seen that the peak melting
temperatures and degrees of crystallinity decreased upon functionalization
with HDI, as well as upon crosslinking.[18] This can be attributed to the
introduction of HDI moieties upon functionalization and to a decrease in
freedom of movement of the polymer chains upon covalent crosslinking.
Despite that, the degrees of crystallinity of the networks were still relatively
high, suggesting both a homogeneous network structure and a certain
remaining degree of freedom of movement of the polymer chains and/or of
the crosslinking units within that structure. These suggestions are both
supported by the proposed network structure. As a result of the crystallinity
and high degree of crosslinking, the networks showed excellent mechanical
properties. The PCL-4k and PCL-8k networks possessed E-moduli of 6.6 and
136.6 MPa, stress-at-break values (σb,dry) of 4.1 and 20.9 MPa and strain-at-
break values (εb,dry) of 109% and 224%, respectively (see Table 5.1). The PEG-
4k and PEG-10k networks possessed E-moduli of 187.7 and 165.2 MPa,
stress-at-break values of 13.6 and 30.9 MPa and strain-at-break values of
184% and 700%, respectively.
The water uptake of the networks differed significantly. The
networks based on the hydrophobic PCL polymers barely swelled in water,
whereas the hydrophilic PEG networks strongly swelled in water and can be
classified as hydrogels. The mechanical properties of the PCL networks in the
wet state did not differ much from those in the dry state, due to the low
water uptake of these networks. In the wet state, the PCL-4k and PCL-8k
networks possessed E-moduli of 6.8 and 142.9 MPa, stress-at-break values
105
(σb,wet) of 2.7 and 22.9 MPa and strain-at-break values (εb,wet) of 62% and
225%, respectively. The mechanical properties of the PEG networks on the
other hand were clearly affected by equilibration in water, leading to high
water uptakes of 219.7% for PEG-4k and 508.0% for PEG-10k. Nevertheless,
the tensile properties of the PEG networks could be determined without
difficulty even in the water-swollen state. In the wet state, the PEG-4k and
PEG-10k networks possessed E-moduli of 3.8 and 1.0 MPa, stress-at-break
values of 0.9 and 0.6 MPa and strain-at-break values of 32% and 106%,
respectively. Both in the dry state and in the wet state, the toughness of all
the networks was significantly higher than those of comparable non-
poly(urethane-isocyanurate) based networks, e.g. obtained by a
functionalization with acrylates and subsequent crosslinking under UV
radiation.[22] Likely the improved mechanical properties can be attributed to
the introduction of the urethane groups and/or isocyanurate rings. Also, the
more homogeneous network structure could contribute to improved
mechanical properties. Based on the combination of the high water uptake
and high toughness in the wet state, the hydrophilic materials could find
application in a variety of fields, including tissue engineering, biomedical
devices and/or drug delivery.
5.4 Conclusions A simple, effective and widely-applicable method was developed for
synthesizing polymer networks with a well-defined network structure. Upon
functionalization of hydroxyl-group-terminated PCL and PEG polymers with
hexamethylene diisocyanate, it was shown that well-defined hydrophobic
and hydrophilic networks can readily be synthesized by trimerization of the
prepolymers. The proposed mechanism for network formation was
supported by in-situ ATR-FTIR, extraction and thermal measurements. It was
shown that the resulting networks possessed excellent mechanical
properties in terms of elongation-at-break, stress-at-break and toughness,
in both the dry and in the wet state.
106
5.5 Acknowledgements This project has received funding from the European Union’s Horizon 2020
research and innovation program, under the Marie Skłodowska-Curie Grant
Agreement no. 642890 (http://thelink-project.eu/).
The support of the labs of Dr. Frank Richter and the analytical
departments of Covestro Deutschland AG Leverkusen is greatly
acknowledged.
107
5.6 References
[1] M. Urban, C. Craver, Structure-Property Relations in Polymers,
American Chemical Society, 1993.
[2] T. Le, V. C. Epa, F. Burdent, D. Winkler, Chem. Rev. 2012, 112, 2889-
2919.
[3] J. Mark, B. Erman, Rubberlike Elasticity - A Molecular Primer, Wiley,
New-York, USA 1988.
[4] J. Bastide, L. Leibler, Macromolecules, 1988, 21, 2647-2649.
[5] T. Norisuye, N. Masui, Y. Kida, D. Ikuta, E. Kokufuta, S. Ito, S. Panyukov,
M. Shibayama, Polymer, 2002, 43, 5289–5297.
[6] Y. Akagi, T. Katashima, Y. Katsumoto, K. Fujii, T. Matsunaga, U.-I.
Chung, M. Shibayama, T. Sakai, Macromolecules, 2011, 44, 5817–5821.
[7] J. Mark, R. Rahalkar, J. Sullivan, J. Chem. Phys., 1979, 70, 1794.
[8] M. Malkoch, R. Vestberg, N. Gupta, L. Mespouille, P. Dubois, A. Mason,
J. Hedrick, Q. Liao, C. Frank, K. Kingsbury, C. Hawker, Chem. Comm.,
2006, 26, 2774–2776.
[9] J. Mark, M. Llorente, J. Am. Chem. Soc., 1980, 102 (2), 632-636.
[10] T. Sakai, T. Matsunaga, Y. Yamamoto, C. Ito, R. Yoshida, S. Suzuki, N.
Sasaki, M. Shibayama, U.-I. Chung, Macromolecules, 2008, 41, 5379-
5384.
[11] N. Sasaki, T. Yokoyama, T. Tanaka, J. Polym. Sci. Polym. Chem., 1973,
11, 1765-1779.
[12] E. Delebecq, J.-P. Pascault, B. Boutevin, F. Ganachaud, Chem. Rev.,
2013, 113, 80-118.
[13] M. Spírkova, J. Budinski-Simendic, M. Ilavský, P. Spacek, K. Dusek,
Polym. Bull., 1993, 31, 83-88.
[14] R. Samborska-Skowron, A. Balas, Polym. Adv. Technol., 2002, 13, 653-
662.
[15] P. Flory, J. Am. Chem. Soc., 1936, 58, 1877-1885.
108
[16] P.J. Driest, V. Lenzi, L.S.A. Marques, M.M.D. Ramos, D.J. Dijkstra, F.U.
Richter, D. Stamatialis, D.W. Grijpma, Polym. Adv. Technol., 2017, 28,
1299-1304.
[17] K. Dusek, Chapter 12 in Telechelic Polymers: Synthesis and
Applications, 1989, 289-360.
[18] G. Erdödi, B. Iván, Chem. Mater., 2004, 16, 959-962.
[19] G. Crescenzi, G. Manzini, G. Calzolari, C. Borri, Eur. Polym. Mater.,
1972, 8, 449.
[20] B. Wunderlich, Thermal Analysis, Academic Press, 1990.
[21] A. Mishra, D. Chattopadhyay, B. Sreedhar, K. Raju, Prog. Org. Coat.,
2006, 55, 231-243.
[22] E. Zant, M.J. Bosman, D.W. Grijpma, J. Appl. Biomater. Function.
Mater., 2012, 10, 197-202.
109
Chapter 6 - Tough Combinatorial
Poly(urethane-isocyanurate) Polymer
Networks and Hydrogels Synthesized by
the Trimerization of Mixtures of NCO-
Prepolymers
P.J. Driest1,2, D.J. Dijkstra1, D. Stamatialis2, D.W. Grijpma2
1Covestro Deutschland AG, CAS-Global R&D, 51373 Leverkusen, Germany
2Technical Medical Centre, Faculty of Science and Technology, Department
of Biomaterials Science and Technology, University of Twente P.O. Box 217,
7500AE Enschede, The Netherlands
Resubmitted
Abstract The development of tough hydrogels
is an essential but challenging topic in
biomaterials research that has
received much attention over the
past years. By the combinatorial
synthesis of polymer networks and
hydrogels based on prepolymers with
different properties, new materials
with widely varying characteristics
and unexpected properties may be identified. In this paper, we report on
the properties of combinatorial poly(urethane-isocyanurate) (PUI) type
polymer networks that were synthesized by the trimerization of mixtures of
NCO-functionalized poly(ethylene glycol) (PEG), poly(propylene gylcol)
(PPG), poly(ε-caprolactone) (PCL) and poly(trimethylene carbonate) (PTMC)
prepolymers in solution. The resulting polymer networks showed widely
varying material properties.
Combinatorial PUI networks containing at least one hydrophilic PEG
component showed high water uptakes of >100wt%. The resulting hydrogels
demonstrated elastic moduli of up to 10.1 MPa, ultimate tensile strengths
of up to 9.8 MPa, elongation at break values of up to 624.0% and toughness
values of up to 53.4 MJ·m-3. These values are exceptionally high and show
that combinatorial PUI hydrogels are among the toughest hydrogels
reported in the literature. Also, the simple two-step synthesis and wide
range of suitable starting materials make this synthesis method more
versatile and widely applicable than the existing methods for synthesizing
tough hydrogels.
Finally, it is shown that the presence of a hydrophobic network
component significantly enhances the toughness and tensile strength of a
combinatorial PUI hydrogel in the hydrated state. This enhancement is the
largest when the hydrophobic network component is crystallizable in
nature. It is shown that the PUI hydrogels containing a crystallizable
hydrophobic network component are semi-crystalline in the water-swollen
state.
112
6.1 Introduction During the past decades, hydrogels have played an increasingly important
role in the development of biomaterials.1 The interesting material properties
of hydrogels, including a high water uptake (>100wt%), typical
biocompatibility and physical properties resembling the natural extracellular
matrix, have been exploited in a wide range of biomedical applications, such
as drug-delivery, tissue engineering, injectable fillers and many more.2
However, the poor toughness typically associated with hydrogel materials is
still a drawback and has limited the exploitation of the full potential of
hydrogel materials so far. As a result, promising applications as load-bearing
biomaterials largely remain out of reach. Consequently, the development of
tough hydrogels has received much attention over the past years.3–5
Presently, several approaches exist for synthesizing tough
hydrogels, which include (interpenetrating) double network (DN)
hydrogels3,6,7, combinatorial hydrogels8–10, thermoplastic polyurethane
(TPU) hydrogels11,12, (nano)composite hydrogels13–15, rotaxane-based
hydrogels16 and physical interaction enhanced hydrogels17,18. Of these, the
interpenetrating double network (DN) approach, schematically sketched in
Scheme 6.1a, is presently the most widely-studied technique leading to the
best-performing materials.7,19–25 Although the double network approach has
its advantages in terms of mechanical performance, there are still some
drawbacks associated with the technique. These are mostly found in terms
of practicality and applicability. For example, typically a multi-step
sequential crosslinking-swelling-crosslinking procedure is needed to obtain
the interpenetrating double network structure that is required to achieve
the desired material properties.26 In general, in terms of type and molecular
weight of possible starting materials and the resulting control over the final
material properties, the range of possibilities is considered limited in a DN
approach.
113
Scheme 6.1. Schematic representation of the discussed synthesis methods
used for obtaining tough hydrogels: a) interpenetrating Double Network
(DN) formation by repeated photocrosslinking and swelling steps; b) the
combinatorial photocrosslinking of mixtures of (meth)acrylate-
functionalized prepolymers; c) the combinatorial trimerization of mixtures of
NCO-functionalized prepolymers.
A promising alternative approach for synthesizing tough hydrogels is based
on the combinatorial (photo)crosslinking of mixtures of hydrophobic and
hydrophilic (meth)acrylate-functionalized prepolymers, as sketched in
Scheme 6.1b.8–10 Although these combinatorial hydrogels did not yet
demonstrate the same level of toughness or tensile strength as the best-
performing DN hydrogels, they did match those of various biological
materials, most notably cartilage.10 Meanwhile, the relative simplicity of the
synthesis method, the wide range of possible starting materials and the
range and control over the final material properties are a clear advantage of
a combinatorial synthesis approach.27
114
We recently developed a method to synthesize well-defined
poly(urethane-isocyanurate) (PUI) type polymer networks, that is based on
the trimerization of NCO-functionalized prepolymers.28 Similarly to
photocrosslinked combinatorial networks, it was shown that these PUI type
polymer networks are readily synthesized as well. Also, the synthetic
approach allows for the use of a wide range of starting materials (i.e. any di-
or polyol) and allows for a wide range and control over final material
properties. Interestingly, it was found that the hydrophilic single-network
PUI materials based on poly(ethylene glycol) (PEG) prepolymers already
demonstrated unexpectedly high mechanical strength and toughness in the
hydrated state.28 Therefore, we set out to investigate the properties of
combinatorial PUI networks and hydrogels, synthesized by the trimerization
of mixtures of NCO-prepolymers as illustrated in Scheme 6.1c.
In this paper, the synthesis and characterization of a variety of
combinatorial PUI polymer networks and hydrogels is investigated, using
combinations of several hydrophilic and hydrophobic polymeric diols that
are often used in biomaterials research of different molecular weight as the
starting materials: i.e. poly(ethylene glycol) (PEG), poly(propylene gylcol)
(PPG), poly(ε-caprolactone) (PCL), and poly(trimethylene carbonate)
(PTMC). In the molecular weight range used, PEG is a crystallizable
hydrophilic polymer with a glass transition temperature (Tg) between -24°C
and -37°C and a peak melting temperature (Tm) between 58°C and 66°C.29,30
PCL is a crystallizable hydrophobic polymer with a Tg and Tm of approximately
-60°C and 59°C, respectively, while PPG and PTMC are amorphous
hydrophobic polymers with Tgs of approximately -60°C and -15°C,
respectively.31–33
115
6.2 Experimental Details
Materials and methods
Poly(ethylene glycol)s (𝑀𝑛 4 and 10 kg/mol), stannous octoate (Sn(II)Oct2),
1,6-hexanediol and propylene carbonate were purchased from Sigma-
Aldrich. Poly(ε-caprolactone)s (𝑀𝑛 4 and 8 kg/mol) were kindly supplied by
Perstorp Chemicals GmbH. Trimethylene carbonate (TMC) was kindly
supplied by ForYou Medical Devices Co Ltd. Hexamethylene diisocyanate
(HDI) and poly(propylene glycols) (𝑀𝑛 4 and 8 kg/mol) were supplied by
Covestro Deutschland AG. All other chemicals were purchased from Sigma-
Aldrich.
All reactions were conducted under a nitrogen atmosphere using
standard Schlenck-line techniques. Differential scanning calorimetry (DSC)
measurements were performed under nitrogen on ca. 10 mg samples on a
Perkin-Elmer Calorimeter DSC-7. Samples in the dry state were measured in
2 heating runs from -100°C to +150°C with a heating rate of 20 °C/min and a
cooling rate of 20 °C/min. Samples in the hydrated state were measured in
a high-pressure DSC-pan in 1 heating run from +5°C to +80°C with a heating
rate of 20 °C/min. The glass transition temperature(s) (Tg), peak melting
temperature(s) (Tm,p) and melting enthalpy value(s) (∆𝐻𝑚, obtained from
the area under the melting peak) were determined based on the last heating
run. NMR-spectra were collected on a Bruker Advance III-700 in the specified
solvent. FTIR-spectra were recorded on a Bruker FTIR Spectrometer Tensor
II with Platinum-ATR-unit with diamond crystal. Size exclusion
chromatography (SEC) was performed according to DIN 55672-1:2016-03,
on 4 PSS SDV Analytical columns (2x 100Å, 5µm; 2x 1000Å, 5µm) using an
Agilent 1100 Series pump and an Agilent 1200 Series UV Detector (230 nm)
with tetrahydrofuran as the elution solvent at 40°C and 1.00 mL/min.
Residual monomeric diisocyanate contents were determined by gas
chromatography (GC) according to DIN EN ISO 10283 on an Agilent
Technologies 6890N system using a 15.0m DB17 column and tetradecane as
an internal standard. Tensile tests were performed in triplicate at room
temperature using a Zwick Z0.5 tensile tester equipped with a 500N load cell
at 200mm/min. with a grip-to-grip separation of 35mm on strips of
116
approximately 60 x 4 x 0.5mm cut from a cured film. Elastic moduli and strain
values were derived from the initial grip-to-grip separation and are
therefore indicative only. The toughness values of the materials were
calculated from the work at break value, normalized for the dimensions of
the specimen.
For swelling measurements, three samples measuring
approximately 10 x 20 x 0.5mm were cut from a cured, extracted and dried
film and weighed (denoted 𝑚𝑖). Each sample was then swollen in excess
water for at least 24h, blotted dry and weighed again in the hydrated state
(denoted 𝑚𝑠). The water uptake was defined as:
𝑤𝑎𝑡𝑒𝑟 𝑢𝑝𝑡𝑎𝑘𝑒 =𝑚𝑠 − 𝑚𝑖
𝑚𝑖× 100%
Synthesis of an OH-functional PTMC polymeric diol
A PTMC polymeric diol was synthesized by the previously reported ring-
opening polymerization of trimethylene carbonate, using 1,6-hexandiol as
the initiator and Sn(II)Oct2 as the catalyst.8 Under dry conditions, 101.1g
(0.99 mol) of trimethylene carbonate was weighed into a dried and nitrogen-
secured 250mL 3-neck flask. Next, 2.93g (24.8 mmol) of 1,6-hexanediol was
added and the mixture was heated to 130°C. Next, 55mg of Sn(II)Oct2 was
added and the mixture was kept at 130°C for 40h. The resulting product was
collected as a viscous, transparent, colorless resin and analyzed by SEC and 1H-NMR (600 MHz, CDCl3,): δ = 4.00 (160H, brs), 3.39 (4H, brs), 1.64 (80H,
brs), 1.35 (4H, brs), 1.04 (4H, brs) ppm.
Synthesis of NCO-functionalized prepolymers
All polymeric diols were dried prior to use by azeotropic distillation in
toluene, which was subsequently removed under reduced pressure. In a
typical experiment, 0.25 g of dibutylphosphate was added to 100 g of dried
polymeric diol, which was kept in the melt using a heat gun where
applicable. This mixture was added dropwise to HDI kept at 100°C, using a
molar ratio of NCO:OH of 15:1, and the resulting mixture was subsequently
kept at 100°C for 3h. In the case of PPG polymeric diols, the procedure was
performed at 120°C and kept at 120°C for 6h. The resulting mixture was then
117
transferred dropwise into a cylindrical glass short-path thin-film evaporator
equipped with Teflon wipers with a volume of approximately 500mL, which
was operated at a reduced pressure of 1∙10-2 mbar at 140°C. All
functionalized prepolymers were collected as clear, colorless, viscous resins,
of which the PEG- and PCL-based resins crystallized upon cooling to room
temperature. Products were analyzed by SEC, GC and 1H-NMR (700 MHz,
C6D6, representative data for PEG-4k): δ = 4.60 (2H, brs), 4.26 (4H, t), 3.50
(410H, m), 2.97 (4H, q), 2.56 (4H, t), 1.12 (4H, t), 1.02 (4H, t), 0.88 (8H, m)
ppm.
Synthesis of combinatorial poly(urethane-isocyanurate) networks by the
trimerization of mixtures of NCO-functionalized prepolymers in solution
Combinatorial poly(urethane-isocyanurate) networks were synthesized in
analogy to a previously reported procedure,28 with slight modifications. In a
typical experiment, 4g of propylene carbonate was weighed into a
polypropylene cup. Next, the respective NCO-prepolymers were added in
the melt in the reported weight ratios, always totaling 4g. The mixture was
then kept under nitrogen under continuous stirring, until a macroscopically
homogenous solution was formed. Whenever crystallization was observed,
the mixture was gently heated to 50-60°C until the mixture was liquid again.
The mixture was subsequently degassed by gently applying a reduced
pressure of ca. 1-10mbar until no more bubbles were observed, whereby it
was observed that no residue was built up in the cooling trap of the vacuum
pump. Next, ca. 60 mg of Sn(II)Oct2 (corresponding to ca. 0.75wt% relative
to the solution) was added. After mixing for 15 seconds in a Hauschild
Speedmixer DAC150FVZ, the mixture was brought into a mold consisting of
a 0.5mm-thick polycarbonate frame clamped between two silanized glass
plates measuring 20 x 10cm and kept at 90°C for 48h. Unless reported
otherwise, all reaction mixtures were optically transparent within 15min
after heating to 90°C. Cured samples were collected as flexible, non-sticky
films, which were macroscopically homogeneous, unless reported
otherwise. The films were subsequently extracted by swelling in an excess
of acetone for 12-24h while periodically refreshing the solvent. Typically, 6-
8 films were extracted simultaneously using 3x 500mL acetone. Next, 5L of
118
n-hexane was slowly dripped into the 500mL of acetone over the course of
24h which was allowed to overflow, thereby slowly exchanging the acetone
with n-hexane. During this process, the films slowly contracted. The
resulting films were then dried, first in air and subsequently in vacuum, and
analyzed by FTIR, swelling, tensile testing and DSC experiments as described.
The remainder of each film was then swollen in an excess of water for at
least 24h, and analyzed in the water-swollen state by tensile testing and DSC
experiments.
119
6.3 Results and Discussion
Synthesis of NCO-functionalized prepolymers
All polymeric diols were functionalized using hexamethylene diisocyanate
(HDI), in a procedure analogous to the one described before.28 In the case of
the PPG polymeric diols, the reaction temperature and time were increased
to 120°C and 6h, respectively. This was done because the uncatalyzed
urethanization reaction typically occurs less readily for secondary alcohols
than for primary alcohols. Since as-received commercial polymeric diols may
contain traces of base, working under slightly acidic conditions was always
ensured by adding a small amount of dibutyl phosphate (DBP) to the
reaction mixture prior to reaction. This was done in order to prevent any
undesired side-reactions involving the isocyanates, such as isocyanurate
(trimer), uretdione (dimer) or allophanate formation.
Based on 1H-NMR, it was confirmed that the functionalization of all
polymeric diols with HDI succeeded selectively and quantitatively. No side
products were observed in the 1H-NMR spectra. From the 1H-NMR-spectra,
the average molecular weights of the functionalized prepolymers were
calculated. In order to synthesize functionalized prepolymers with a narrow
molecular weight distribution, a large molar excess of HDI of 15:1 was used.
Using SEC, it was shown that the polydispersity index (Đ = 𝑀𝑤/𝑀𝑛) of the
prepolymers after functionalization was relatively low and was always <1.5.
Finally, it was confirmed by GC that the unreacted excess of monomeric
diisocyanate was effectively removed by short-path thin-film evaporation.
This resulted in functionalized prepolymers with a residual HDI content
lower than 1wt% in all cases. The characteristics of all functionalized
prepolymers are reported in Table 6.1.
120
Table 6.1. Characterization of the NCO-functionalized prepolymers. Mn
represents the molecular weight of the NCO-functionalized prepolymer as
calculated from the respective 1H-NMR spectrum; Đ represents the
polydispersity index of the NCO-functionalized prepolymer; and Residual HDI
Content represents the mass percentage of residual free (unreacted) HDI in
the NCO-functionalized prepolymer after thin-film evaporation.
NCO-Functionalized
Prepolymer
Mn
[kg/ mol]
Đ
[-]
Residual HDI
Content
[wt%]
PEG-4k-diNCO 3.9 1.17 <0.01
PEG-10k-diNCO 9.4 1.20 0.09
PCL-4k-diNCO 3.9 1.30 0.21
PCL-8k-diNCO 7.8 1.43 0.17
PPG-4k-diNCO 4.1 1.30 0.02
PPG-8k-diNCO 8.0 1.13 <0.01
PTMC-4k-diNCO 4.1 1.44 0.98
Synthesis of combinatorial poly(urethane-isocyanurate) networks by the
trimerization of mixtures of NCO-functionalized prepolymers in solution
Various combinatorial poly(urethane-isocyanurate) (PUI) polymer networks
were synthesized by the trimerization of mixtures of NCO-functionalized
prepolymers. The synthesis approach was similar to the one reported
previously for single-network PUIs, with the main difference that here the
crosslinking reaction was performed in solution instead of in bulk.28 This was
done because typically NCO-prepolymers of different nature are not
homogeneously miscible as such. Propylene carbonate was selected as a
high-boiling aprotic inert solvent in which the combinatorial trimerization
reactions could be carried out at a concentration of 50wt%. In most cases,
the 50wt% solutions of combined NCO-functionalized prepolymers in
propylene carbonate were stable, homogeneous and transparent at room
temperature. In some cases, the prepolymer mixture was slightly turbid at
room temperature, indicating a certain degree of phase separation.
121
Typically, these prepolymer solutions turned transparent and homogeneous
within 15min after addition of the catalyst and heating to 90°C and remained
transparent throughout the trimerization reaction. In a few cases involving
PPG prepolymers, macroscopic phase separation into domains with a size of
ca. 0.1 – 1mm was observed during the trimerization reaction. The resulting
polymer films were heterogeneous after crosslinking. These films were not
characterized. Whenever this was the case, it is indicated in the table(s)
where the material properties are reported.
To obtain the networks in the dry state after the trimerization
reaction was completed, the high boiling propylene carbonate was extracted
from the polymer networks using acetone. In order to subsequently de-swell
the polymer films in a controlled manner, the acetone was slowly exchanged
for n-hexane over the course of 24h. Since n-hexane is a non-solvent for all
the network components used in this study, this solvent-exchange-process
resulted in the contraction of the networks. Typically, the films also turned
opaque during this process. This indicates that (micro)phase separation of
the different network components took place within the polymer networks
upon de-swelling.
After drying, the completion of the trimerization reaction was
confirmed in all cases by FTIR. This was done by confirming the
disappearance of the absorption peak in the range of 2270cm-1 associated
with the NCO groups. Simultaneously, the appearance of an extra carbonyl
peak in the range of 1690cm-1 associated with the aliphatic isocyanurate
(trimer) was confirmed, in analogy to what was reported previously.28 It was
observed that the polymer films remained opaque upon drying, indicating
phase separation.
Material properties of combinatorial poly(urethane-isocyanurate)
networks in the dry state
The thermal and tensile material properties of the polymer networks in the
dry state are reported in Table 6.2. In the tensile tests, the networks
displayed widely varying mechanical properties, ranging from relatively stiff
to more flexible and ductile elastomeric behavior. A selection of tensile
stress-strain diagrams is shown in Figure 6.1a. The wide range of mechanical
Table 6.2. Characterization of combinatorial PUI networks in the dry state. The prepolymers present in the networks (indicated
by a grey box) are present in equal weight percentages. Numbers in parentheses represent standard deviations.
# Network Component Thermal Tensile
Net
w. n
r.
PEG
-4k
PEG
-10
k
PLC
-4k
PC
L-8
k
PP
G-4
k
PP
G-8
k
PTM
C-4
k
Tg,dry
[°C]
Tm,p,dry
[°C]
ΔHm
[J·g-1]
Emod,dry
[MPa]
σb,dry
[MPa]
εb,dry
[%]
Wtensile,dry
[MJ·m-3]
1 -49.9 55.6 103.6 73.6
(-)a
17.5
(-)a
864.5
(-)a
72.79
(-)a
2 -58.8 39.5 62.4 65.0
(-)a
23.8
(-)a
1305
(-)a
150.4
(-)a
3 -58.4 1) 46.2
2) 50.6
1) 13.2
2) 57.3
184.7
(19.4)
19.5
(0.5)
871.2
(26.9)
103.0
(3.2)
4b -b -b -b -b -b -b -b
5b -b -b -b -b -b -b -b
6 -c - c - c 3.2
(0.8)
1.5
(0.2)
107.2
(23.9)
1.06
(0.28)
7 -55.3 54.5 76.4 124.9
(30.4)
20.1
(1.3)
1025
(40.5)
112.7
(3.3)
8 -58.4 54.6 81.6 221.1
(46.7)d
16.0
(0.7)d
555.4
(22.7)d
60.8
(0.3)d
9b -b -b -b -b -b -b -b
10b -b -b -b -b -b -b -b
11 -45.7 42.6 50.2 34.6
(7.4)
4.1
(0.6)
255.8
(65.0)
8.65
(1.91)
12 -57.8 53.5 53.6 105.4
(29.2)
25.1
(8.5)
586.5
(108.5)
78.2
(24.6)
13b -b -b -b -b -b -b -b
14 -67.1 36.2 24.2 15.6
(2.2)
5.2
(0.1)
339.9
(13.9)
11.03
(0.70)
15 -19.7 37 34 43.7
(2.1)
5.6
(0.4)
198.5
(18.4)
6.95
(0.65)
16 -64.3 41 26.7 21.6
(1.2)
4.7
(0.3)
170.9
(24.1)
5.84
(0.86)
17 -64.7 50.4 35 42.3
(1.8)
13.0
(0.8)
520.3
(36.9)
36.91
(3.23)
18 1) -56.1
2) -18.1 47.3 27.8
16.4
(0.8)
3.2
(0.1)
47.2
(2.6)
1.01
(0.08)
19b -b -b -b -b -b -b -b
20 1) -63.9
2) -19.5 - c - c
1.5
(0.2)
0.5
(0.1)
67.5
(12.8)
0.24
(0.06) an=1 bmacroscopic phase separation observed during crosslinking, resulting in a heterogeneous polymer film cnot observed
dn=2
124
properties is best indicated by the elastic moduli measured in the dry state
(Emod,dry), which covered a range of values from 1.5 MPa to 221.1 MPa. Also
the ultimate tensile strengths (σb,dry) and elongation at break values (εb,dry)
varied widely and ranged from 0.5 MPa to 25.1 MPa and from 47.2% to
1305%, respectively. Starting from only 4 different types of common
polymeric diols, this shows that the synthesis method allows for the
production of a wide variety of materials. The toughness values of the
networks (Wtensile,dry) were calculated from the work at break values and are
reported in Table 6.2 as well. They cover a range from a few MJ·m-3 for the
weaker networks up to 150 MJ·m-3 for the toughest network.
Analogous to the single-network PUIs reported previously, the
thermal properties of the combinatorial networks (reported in Table 6.2) are
dictated by the thermal properties of the respective polymer network
components present in the network. It was found that whenever a
crystallizable prepolymer was present in the combinatorial PUI network, a
corresponding crystallization peak was observed in the DSC curve of the
network. In fact, in the cases where two crystallizable prepolymer
components of a different nature were present in one combinatorial
network (networks 2, 3, 7 and 8), an overlapping double melting peak was
observed in the DSC measurements of the network. In the case of network
3, of which the DSC curves are shown as an example in Figure 6.1b, even two
separate melting- and crystallization peaks were observed. At this point it is
worthwhile to note that all solutions containing PEG and PCL prepolymers
were optically transparent and homogeneous throughout the crosslinking
reaction. It is therefore expected that the prepolymers were
homogeneously mixed at the molecular level during crosslinking. The fact
that separate melting peaks are then observed for the covalently crosslinked
combinatorial polymer network in the dry state is noteworthy. This indicates
that the two separate crystallizable network components of different nature
were able to crystallize within the same combinatorial polymer network.
125
Figure 6.1. Physical characterization of selected combinatorial
poly(urethane-isocyanurate) networks in the dry state: a) tensile stress-
strain curves of several PUI networks in the dry state. The numbers
correspond to the network numbers reported in Table 6.2. To illustrate the
wide range of elastic moduli, the initial part of the stress-strain diagrams is
shown in the inset; b) DSC curves of network 3 (a PEG-4k/PCL-8k
combinatorial PUI network) in the dry state, showing a double melting peak
and a double crystallization peak in both heating runs and in the cooling run,
respectively.
126
Mechanical properties of combinatorial poly(urethane-isocyanurate)
networks in the hydrated state
All combinatorial PUI networks were characterized by swelling in water. The
resulting water uptake values of the materials are reported in Table 6.3. As
expected, all networks containing at least one hydrophilic PEG component
(samples 1W-11W, marked in blue in Table 6.3) strongly swelled in water
and can therefore be classified as hydrogels. These materials all
demonstrated water uptakes >100wt%. Samples that did not contain any
hydrophilic PEG component (samples 12W-20W) were hydrophobic in
nature and therefore barely swelled in water. These networks all
demonstrated water uptakes <2.5wt% and are not considered hydrogels.
Generally, it was found that water uptake of the networks was dictated by
the molecular weight and nature of the prepolymers present in the network.
Networks based on lower molecular weight prepolymers, having a denser
network structure, typically demonstrated lower water uptake values than
their higher molecular weight counterparts.
The mechanical properties of all combinatorial networks in the
hydrated state are reported in Table 6.3. The mechanical properties of the
hydrophobic networks (networks 12W-20W) did not change considerably
compared to the dry state and will not be further discussed. The mechanical
properties of the networks classified as hydrogels (networks 1W-11W) on
the other hand changed considerably compared to the dry state. These
hydrogels demonstrated elastic moduli in the hydrated state (Emod,wet)
ranging from 0.5 MPa to 10.1 MPa. The ultimate tensile strengths of the
hydrogels in the hydrated state (σb,wet) ranged from 0.03 MPa to 7.7 MPa,
increasing up to 9.8 MPa for the best-performing sample of network 2W.
The elongation at break values in the hydrated state (εb,wet) ranged from
82.2% to 624%. The toughness values of the hydrogels in the hydrated state
(Wtensile,wet) were calculated from the work at break values and ranged from
0.02 MJ·m-3 to 33.2 MJ·m-3, increasing up to 53.4 MJ·m-3 for the best-
performing sample of network 2W. These are considered excellent
mechanical properties for hydrogels in the hydrated state for the given
amounts of water content. The tensile stress-strain diagrams of several PUI
hydrogels in the hydrated state are shown in Figure 6.2a.
127
Figure 6.2. Physical characterization of selected combinatorial PUI hydrogels
in the hydrated state; a) tensile stress-strain diagrams of several hydrogels
in the hydrated state. The numbers correspond to the network numbers
reported in Table 6.3. In the inset, a more detailed plot is shown of the curves
of hydrogels 3W, 6W, 8W and 11W; b) The DSC heating curve of hydrogel
2W (a PEG-4k/PCL-4k combinatorial PUI hydrogel) in the hydrated state,
showing a clear melting peak around 42.6°C; c) schematic representation of
the intermolecular interactions present within the hydrophobic phase of a
combinatorial hydrogel. These intermolecular interactions can be semi-
crystalline in nature (e.g. for hydrogel 2W, on the left) or amorphous in
nature (e.g. for hydrogel 6W, on the right).
Notably, ultimate tensile strength values of several MPa and
toughness values in the range of several tens of MJ·m-3 are comparable to
the toughest double network (DN) hydrogels that are well-known from the
literature.20,22,25,34,35 For comparison, the mechanical characteristics of a
selection of tough hydrogels reported in the literature are reported in Table
6.4. Both double network (DN) type hydrogels and combinatorial
photocrosslinked (CP) type hydrogels are included. The highest ultimate
tensile strengths and toughness values of the new PUI hydrogels lie in the
Table 6.3. Characterization of combinatorial PUI networks in the hydrated state. The prepolymers present in the networks
(indicated by a grey box) are present in equal weight percentages. Numbers in parentheses represent standard deviations.
Networks demonstrating a water uptake (WU) >100% are defined as hydrogels and are marked in blue.
# Network Components Swelling Thermal Tensile
Net
w. n
r.
PEG
-4k
PEG
-10
k
PLC
-4k
PC
L-8
k
PP
G-4
k
PP
G-8
k
PTM
C-4
k
WU
[wt%]
Tg,wet
[°C]
Tm,p,wet
[°C]
ΔHm
[J·g-1]
Emod,wet
[MPa]
σb,wet
[MPa]
εb,wet
[%]
Wtensile,wet
[MJ·m-3]
1W 1799.0
(128.4) -a -a -a
0.5
(0.3)
0.03
(0.01)
87.8
(32.0)
0.02
(0.01)
2W 107.5
(0.4) -a 42.6 14.0
10.1
(1.9)
7.7
(1.9)
624.0
(250.7)
33.2
(14.3)
3W 166.3
(1.3) -a 57.0 14.5
3.5
(0.01)
1.0
(0.1)
122.0
(32.5)
1.02
(0.28)
4Wb -b -b -b -b -b -b -b -b
5Wb -b -b -b -b -b -b -b -b
6W 109.4
(0.9) -a -a -a
2.9
(0.4)
1.1
(0.2)
82.2
(23.6)
0.58
(0.22)
7W 154.0
(4.4) -a 44.5 16.9
3.7
(1.7)
4.1
(0.5)
305.5
(51.6)
8.89
(2.23)
8W 287.9
(2.4) -a 56.7 10.8
0.6
(0.4)
0.6
(0.05)
208.1
(35.6)
0.96
(0.19)
9Wb -b -b -b -b -b -b -b -b
10Wb -b -b -b -b -b -b -b -b
11W 138.4
(8.8) -a -a -a
2.7
(0.1)d
0.9
(0.2)d
135.5
(49.9)d
0.86
(0.45)
12W 0.9
(0.3) -a 58.6 87.8
140.1
(113.2)d
25.8
(5.2)d
510.5
(82.3)d
76.3
(13.5)
13Wb -b -b -b -b -b -b -b -b
14W 1.1
(0.1) -a 49.9 32.8
37.9
(2.2)
7.2
(0.4)
295.4
(54.7)
14.7
(2.34)
15W 1.3
(0.1) -a 44.6 36.5
56.9
(2.1)
5.3
(0.3)
190.8
(19.0)
8.11
(0.81)
16W 1.3
(0.1) -a 55.1 36.2
38.2
(2.7)
5.9
(0.6)
162.9
(28.2)
7.57
(1.44)
17W 1.7
(0.4) -a 54.4 42.2
44.6
(5.4)
12.9
(0.6)
428.1
(19.5)
32.1
(1.58)
18W 1.6
(0.1) -a 54.2 35.3
19.5
(1.8)
3.7
(0.3)
55.9
(9.6)
1.42
(0.30)
19Wb -b -b -b -b -b -b -b -b
20W 2.4
(1.1) -a -a -a
1.6
(0.6)d
0.8
(0.1)d
134.7
(40.5)d
0.71
(0.25) anot observed bmacroscopic phase separation observed during crosslinking, resulting in a heterogeneous polymer film cn=1 ; dn=2
130
same range as the PUU-hydrogels reported by Yang et al, which are
presently the toughest hydrogels reported in the literature.12 Considering
the relative simplicity and especially the versatility of the synthesis method
reported here, combinatorial PUI hydrogels are considered a promising
candidate for the future development of tough hydrogels and related
biomaterials.
To investigate the effect of the swelling in water on the thermal
properties of the combinatorial PUI hydrogels, we also performed DSC
experiments in the water-swollen state (see Table 6.3). Interestingly, for all
combinatorial hydrogels containing a crystallizable PCL prepolymer as the
hydrophobic component (hydrogels 2W, 3W, 7W and 8W) a melting peak
was observed in the hydrated state. The melting peaks were all found in the
range of 42°C to 57°C, which corresponds to the typical melting range of PCL
domains within a PUI polymer network.28 Significant melting enthalpies of
10-20 J/g were measured for these hydrogels. As an example, the DSC
heating curve of hydrogel 2W in the hydrated state is shown in Figure 6.2b,
showing a clear semi-crystalline melting peak around 42.6°C.
The observed melting peaks suggest that the segregated
hydrophobic domains observed in the dry state remain intact when the
combinatorial network is swollen in water. The presence of such segregated
hydrophobic domains within a swollen hydrogel network has been reported
in the literature before and has been correlated to a significant
enhancement of the mechanical performance of a hydrogel.17,18,37 The high
toughness values and generally good mechanical properties of the
combinatorial PUI hydrogels in the hydrated state can likely be attributed to
this effect.
131
Table 6.4. Characteristics of several tough hydrogel materials reported in the
literature. Numbers in parentheses represent standard deviations.
Hydrogel Material
Swelling Tensile
Water Uptake
[wt%]
Emod,wet
[MPa]
σb,wet
[MPa]
εb,wet
[%]
Wtensile,wet
[MJ·m-3]
PDLLA-PCL-PEG CPa 132
(59)
0.43
(0.04)
1.08
(0.47)
320
(78)
1.46b
(0.71)
PTMC-PDLLA-PCL-
PEG CPc
181
(2)
1.49
(0.38)
1.81
(0.31)
283
(60) -d
PAMPS-PAAm DNe 900f -d 0.68 75 -d
PDMAEA-Q/PNaSS
DNg
73.9f
(7.4)
7.9
(0.6)
5.1
(0.6)
750
(80)
18.8
(1.9)
Alginate-PAAm DNh 614f 0.029 0.156 2200 -d
SA/PAMAAc JDNi 567f -d 1.8
(0.03)
655.7
(26.7) -d
PUU3-12j 186f 2.5 8.5 1000 45
Cartilagek 400 2.5 1.5 100 -d
aData on a PDLLA-10k/PCL-4k/PEG-10k combinatorial photocrosslinked
network (PDLLA-PCL-PEG CP)10 bRecalculated (and corrected) from original data cData on a PTMC-4k/PDLLA-4k/PCL-4k/PEG-4k combinatorial
photocrosslinked network (PTMC-PDLLA-PCL-PEG CP)9 dData not reported eData on a poly(2-acrylamido-2-methylpropanesulfonic acid)-
poly(acrylamide) double network (PAMPS-PAAm)22 fRecalculated from original data gData on an acryloyloxethyltrimethylammonium chloride-sodium p-
styrenesulfonate (PDMAEA-Q/PNaSS) double network25 hData on a calcium alginate/poly(acrylamide-acrylic acid) double network
(Alginate-PAAm DN)35 iData on a sodium alginate/poly(acrylamide-acrylic acid) joint double
network (SA/PAMAAc JDN)20 jData on a PEG-2k/IPDI Polyurethane-urea physical network (PUU3-12)12 kData on patella cartilage36
132
In the literature, hydrophobic phase segregation within swollen hydrogel
networks has mainly been limited to amorphous materials. In such cases,
the intermolecular interactions between the hydrophobic network
components are random and non-specific in nature, as sketched in Figure
6.2c on the right side. The current PUI hydrogels containing an amorphous
PPG- or PTMC prepolymer as the hydrophobic component (hydrogels 4W,
5W, 6W, 9W, 10W and 11W) are also examples of such hydrogels. The
mechanical performance of these hydrogels is better than that of single-
network hydrogels, but is not exceptional.
In the combinatorial PUI hydrogels containing a crystallizable PCL
prepolymer as the hydrophobic network component however, the
segregated hydrophobic domains are semi-crystalline in nature. In these
hydrogels (2W, 3W, 7W and 8W), the intermolecular interactions between
the hydrophobic network components are well-ordered and cooperative, as
sketched in Figure 6.2c on the left side. The presence of such crystalline
domains is generally associated with a strong increase in the toughness of a
material.38 In fact, the combinatorial hydrogels containing a semi-crystalline
PCL phase demonstrate the highest toughness values in the hydrated state
of all combinatorial PUI hydrogels studied. The high toughness values are
likely attributable to the semi-crystalline nature of the hydrophobic domains
of these hydrogels.
133
6.4 Conclusions Combinatorial poly(urethane-isocyanurate) (PUI) polymer networks can be
effectively synthesized by the trimerization of mixtures of NCO-
functionalized PEG-, PCL-, PPG- and PTMC prepolymers. The resulting
polymer networks demonstrate widely varying material properties in the dry
state, highlighting the wide applicability and versatility of the synthesis
method. Combinatorial networks consisting of two different crystallizable
prepolymers demonstrate a double melting peak in the DSC experiments in
the dry state, showing that two different network components are able to
crystallize simultaneously within one combinatorial PUI network.
Combinatorial PUI networks containing at least one hydrophilic PEG
component demonstrate high water uptake values of >100wt%. The
resulting combinatorial PUI hydrogels exhibit excellent mechanical
characteristics in the water-swollen state. This is indicated by elastic moduli,
ultimate tensile strengths, elongation at break values and toughness values
of up to 10.1 MPa, 9.8 MPa, 624.0% and 53.4 MJ·m-3, respectively. These
values are considered very high and show that combinatorial PUI hydrogels
are among the toughest hydrogels reported in the literature so far.
The presence of a hydrophobic network component in combinatorial PUI
hydrogel networks is shown to enhance the toughness and tensile strength
of the hydrogel in the hydrated state. This increase is the largest when the
hydrophobic network component is crystallizable in nature. The
combinatorial PUI hydrogels containing a crystallizable hydrophobic
network component were semi-crystalline in the water-swollen state.
6.5 Acknowledgements This project has received funding from the European Union’s Horizon 2020
research and innovation program under the Marie Skłodowska-Curie Grant
Agreement no. 642890 (http://thelink-project.eu/).
The support of the labs of Dr. Frank Richter and the analytical departments
of Covestro Deutschland AG Leverkusen is greatly acknowledged.
134
6.6 References
1. Hoffman, A. Hydrogels for biomedical applications. Adv. Drug Deliv. Rev. 64, 18–23 (2012).
2. Annabi, N. et al. 25th Anniversary Article: Rational Design and Applications of Hydrogels in Regenerative Medicine. Adv. Mater. 26, 85–124 (2014).
3. Liu, Y. et al. Recent Developments in Tough Hydrogels for Biomedical Applications. Gels 4, 46 (2018).
4. Creton, C. 50th Anniversary Perspective: Networks and Gels: Soft but Dynamic and Tough. Macromolecules 50, 8297–8316 (2017).
5. Gong, J. P. Materials both Tough and Soft. Science (80-. ). 344, 161–162 (2014).
6. Gong, J. P. Why are double network hydrogels so tough? Soft Matter 6, 2583–2590 (2010).
7. Chen, Q., Chen, H., Zhu, L. & Zheng, J. Fundamentals of double network hydrogels. J. Mater. Chem. B 3, 3654–3676 (2015).
8. Zant, E., Bosman, M. J. & Grijpma, D. W. Combinatorial synthesis of photo-crosslinked biodegradable networks. J. Appl. Biomater. Funct. Mater. 10, 197–202 (2012).
9. Zant, E. & Grijpma, D. W. Synthetic Biodegradable Hydrogels with Excellent Mechanical Properties and Good Cell Adhesion Characteristics Obtained by the Combinatorial Synthesis of Photo-Cross-Linked Networks. Biomacromolecules 17, 1582–1592 (2016).
10. Zant, E. & Grijpma, D. W. Tough biodegradable mixed-macromer networks and hydrogels by photo-crosslinking in solution. Acta Biomater. 31, 80–88 (2016).
11. Güney, A., Gardiner, C., McCormack, A., Malda, J. & Grijpma, D. Thermoplastic PCL-b-PEG-b-PCL and HDI Polyurethanes for Extrusion-Based 3D-Printing of Tough Hydrogels. Bioengineering 5, 99 (2018).
12. Yang, N., Yang, H., Shao, Z. & Guo, M. Ultrastrong and Tough Supramolecular Hydrogels from Multiurea Linkage Segmented Copolymers with Tractable Processablity and Recyclability. Macromol. Rapid Commun. 38, 1–6 (2017).
13. Drozdov, A. D. & Christiansen, J. D. Nanocomposite Gels with Permanent and Transient Junctions under Cyclic Loading. Macromolecules 51, 1462–1473 (2018).
14. Tang, X.-Z. et al. Tough and highly stretchable graphene oxide/polyacrylamide nanocomposite hydrogels. J. Mater. Chem. 22, 14160 (2012).
135
15. Aouada, F. A., de Moura, M. R., Orts, W. J. & Mattoso, L. H. C. Preparation and Characterization of Novel Micro- and Nanocomposite Hydrogels Containing Cellulosic Fibrils. J. Agric. Food Chem. 59, 9433–9442 (2011).
16. Okumura, Y. & Ito, K. The Polyrotaxane Gel : A Topological Gel by. Adv. Mater. 13, 485–487 (2001).
17. Argun, A., Tuncaboylu, D. C., Sari, M., Okay, O. & Sahin, M. Structure optimization of self-healing hydrogels formed via hydrophobic interactions. Polymer (Guildf). 53, 5513–5522 (2012).
18. Zhang, Y., Li, Y. & Liu, W. Dipole-dipole and h-bonding interactions significantly enhance the multifaceted mechanical properties of thermoresponsive shape memory hydrogels. Adv. Funct. Mater. 25, 471–480 (2015).
19. Yang, J., Li, Y., Zhu, L., Qin, G. & Chen, Q. Double Network Hydrogels with Controlled Shape Deformation : A Mini Review. J. Polym. Sci. PART B Polym. Phys. 1351–1362 (2018).
20. Zhang, M., Ren, X., Duan, L. & Gao, G. Joint double-network hydrogels with excellent mechanical performance. Polymer (Guildf). 153, 607–615 (2018).
21. Gong, J. P. Why are double network hydrogels so tough? Soft Matter 6, 2583 (2010).
22. Gong, J. P., Katsuyama, Y., Kurokawa, T. & Osada, Y. Double-Network Hydrogels with Extremely High Mechanical Strength. Adv. Mater. 15, 1155–1158 (2003).
23. Haque, M. A., Kurokawa, T. & Gong, J. P. Super tough double network hydrogels and their application as biomaterials. Polymer (Guildf). 53, 1805–1822 (2012).
24. Mai, T. T., Matsuda, T., Nakajima, T., Gong, J. P. & Urayama, K. Distinctive Characteristics of Internal Fracture in Tough Double Network Hydrogels Revealed by Various Modes of Stretching. Macromolecules 51, 5245–5257 (2018).
25. Luo, F. et al. Oppositely charged polyelectrolytes form tough, self-healing, and rebuildable hydrogels. Adv. Mater. 27, 2722–2727 (2015).
26. Dragan, E. S. Design and applications of interpenetrating polymer network hydrogels. A review. Chem. Eng. J. 243, 572–590 (2014).
27. Zant, E., Blokzijl, M. M. & Grijpma, D. W. A Combinatorial Photocrosslinking Method for the Preparation of Porous Structures with Widely Differing Properties. Macromol. Rapid Commun. 36, 1902–1909 (2015).
28. Driest, P. J., Dijkstra, D. J., Stamatialis, D. & Grijpma, D. W. The
136
Trimerization of Isocyanate-Functionalized Prepolymers : An Effective Method for Synthesizing Well-Defined Polymer Networks. Macromol. Rapid Commun. 40, 1800867 (2019).
29. Read, B. E. Mechanical relaxation in some oxide polymers. Polymer (Guildf). 3, 529–542 (1962).
30. Bailey, F. E. (Frederick E. & Koleske, J. V. Poly(ethylene oxide). (Academic Press, 1976).
31. P.Pêgo, A., Poot, A. A., Grijpma, D. W. & Feijen, J. Copolymers of trimethylene carbonate and ε-caprolactone for porous nerve guides: Synthesis and properties. J. Biomater. Sci. Polym. Ed. 12, 35–53 (2001).
32. Faucher, J. A. The dependence of glass transition temperature on molecular weight for poly(propylene oxide) and poly(butylene oxide). J. Polym. Sci. Part B Polym. Lett. 3, 143–145 (1965).
33. Pêgo, A. P., Grijpma, D. W. & Feijen, J. Enhanced mechanical properties of 1,3-trimethylene carbonate polymers and networks. Polymer (Guildf). 44, 6495–6504 (2003).
34. Li, J., Illeperuma, W. R. K., Suo, Z. & Vlassak, J. J. Hybrid hydrogels with extremely high stiffness and toughness. ACS Macro Lett. 3, 520–523 (2014).
35. Sun, J.-Y. et al. Highly stretchable and tough hydrogels. Nature 489, 133–136 (2012).
36. Chin-Purcell, M. V. & Lewis, J. L. Fracture of Articular Cartilage. J. Biomech. Eng. 118, 545 (1996).
37. Przeradzka, M. A., van Bochove, B., Bor, T. C. & Grijpma, D. W. Phase-separated mixed-macromer hydrogel networks and scaffolds prepared by stereolithography. Polym. Adv. Technol. 28, 1212–1218 (2017).
38. Galeski, A. Strength and toughness of crystalline polymer systems. Prog. Polym. Sci. 28, 1643–1699 (2003).
137
Chapter 7 - Structure-Property Relations
in Semi-crystalline Combinatorial PUI
Hydrogels
P.J. Driest1,2, D.J. Dijkstra1, D. Stamatialis2, D.W. Grijpma2
1Covestro Deutschland AG, CAS-Global R&D, 51373 Leverkusen, Germany
2Technical Medical Centre, Faculty of Science and Technology, Department
of Biomaterials Science and Technology, University of Twente P.O. Box 217,
7500AE Enschede, The Netherlands
To be submitted
Abstract
The development of tough hydrogel
materials for load-bearing
biomedical applications is a
challenging topic that has received
much attention over the past years.
For polymer materials in the dry
state, a correlation between
toughness and crystallinity of the
material is well-known. However,
such a correlation has not been
extensively studied for hydrogel materials in the water-swollen state.
In this work, we aim to get better insights into the relationship
between the crystallinity and toughness of poly(urethane-isocyanurate)
(PUI) hydrogels in the water-swollen state. For this, series of model
combinatorial PUI hydrogels with increasing crystallizable and amorphous
hydrophobic content are synthesized and characterized.
Using melting enthalpy as a measure for crystallinity, it is shown that
the toughness of combinatorial PUI hydrogels in the water-swollen state
strongly increases with increasing crystallinity of the hydrogel in the water-
swollen state. This correlation is an important structure-property-
relationship for the future development of tough hydrogel materials for
load-bearing biomedical applications.
140
7.1 Introduction The development of tough, mechanically resilient hydrogels for load-bearing
biomedical applications has received much attention over the past years.1–3
At present, the most widely adopted strategy for synthesizing tough
hydrogels is by an (interpenetrating) double network (DN) approach, as
sketched in Scheme 7.1a.1,4,5 Although DN hydrogels typically demonstrate
good mechanical performance in the water-swollen state, the ease and
versatility of their synthesis is limited. A cumbersome synthesis procedure is
needed to obtain the interpenetrating double network structure required to
achieve the desired material properties.6 A simpler and more widely
applicable synthesis method for hydrogels with a comparable or better
mechanical performance in the water-swollen state would be of great use.
Scheme 7.1. Methods for synthesizing tough hydrogels; a) a sequential
crosslinking-swelling-crosslinking approach used to synthesize an
interpenetrating double network structure; b) the combinatorial
trimerization of mixtures of NCO-functionalized prepolymers.
141
Recently, a synthesis method was developed by our group to
synthesize combinatorial poly(urethane-isocyanurate) (PUI) type polymer
networks and hydrogels by the trimerization of mixtures of NCO-
functionalized prepolymers in solution (Scheme 7.1b).7 It was shown that
these combinatorial PUI type hydrogels demonstrate excellent mechanical
performance in the water-swollen state, as illustrated by high ultimate
tensile strengths of up to 9.8 MPa and high toughness values of up to 53.4
MJ·m-3. Also, the simple two-step synthesis is applicable to a wide range of
suitable starting materials and it is more versatile than the DN approach
typically used to obtain tough hydrogels.
Our previous work showed that the presence of a hydrophobic
network component in a combinatorial PUI hydrogel significantly increases
the toughness of the hydrogel in the water-swollen state. It was shown that
this increase in toughness is the largest when the hydrophobic network
component is crystallizable in nature. In fact, the toughest PUI hydrogels
were shown to be semi-crystalline in the water-swollen state at room
temperature, suggesting a correlation between the crystallinity and
toughness of the hydrogel in the water-swollen state. Such a correlation is
well-known for polymer materials in the dry state,8 but has not been
extensively studied for hydrogel materials in the water-swollen state.
In this work, we aim to get better insights into the relationship
between crystallinity and toughness of PUI hydrogels in the water-swollen
state. For this, two series of combinatorial PUI hydrogels with increasing
hydrophobic contents are synthesized. In one series, the hydrophobic
component is crystallizable in nature, while in the other series the
hydrophobic component is amorphous in nature. The resulting hydrogels
are characterized for their thermal and mechanical properties, both in the
dry state and in the water-swollen state. The melting enthalpy, obtained
from dynamic scanning calorimetry (DSC) measurements, is used as a
measure for the crystallinity of the hydrogel and is correlated to the
toughness of the hydrogel in the hydrated state.
Poly(ethylene glycol) (PEG), poly(ε-caprolactone) (PCL) and
poly(propylene gylcol) (PPG) with number-average molecular weights (𝑀𝑛)
of 4 kg/mol are used as the polymeric diols. In the molecular weight used,
142
PEG is a crystallizable hydrophilic polymer with a glass transition
temperature (Tg) between -24°C and -37°C and a peak melting temperature
(Tm) between 58°C and 66°C,9,10 PCL is a crystallizable hydrophobic polymer
with a Tg and Tm of approximately -60°C and 59°C, respectively,11 and PPG is
an amorphous hydrophobic polymer with a Tg of approximately -60°C.12
143
7.2 Experimental Details
Materials and methods
Poly(ethylene glycol) (𝑀𝑛 4 kg/mol), stannous octoate (Sn(II)Oct2),
propylene carbonate and toluene were purchased from Sigma-Aldrich.
Poly(ε-caprolactone) (𝑀𝑛 4 kg/mol) was kindly supplied by Perstorp
Chemicals GmbH. Hexamethylene diisocyanate (HDI) and poly(propylene
glycol) (𝑀𝑛 4 kg/mol) were supplied by Covestro Deutschland AG. All other
chemicals were purchased from Sigma-Aldrich.
Differential scanning calorimetry (DSC) measurements were
performed under nitrogen on ca. 10 mg samples on a Perkin-Elmer
Calorimeter DSC-7. Samples in the dry state were measured in 2 heating runs
from -100°C to +150°C with a heating rate of 20 °C/min and a cooling rate of
20 °C/min. Samples in the hydrated state were measured in a high-pressure
DSC-pan in 1 heating run from +5°C to +80°C with a heating rate of 20
°C/min. The glass transition temperature(s) (Tg), peak melting
temperature(s) (Tm,p) and melting enthalpy value(s) (∆𝐻𝑚, obtained from
the area under the melting peak) were determined based on the last heating
run.
Tensile tests were performed in triplicate at room temperature
using a Zwick Z0.5 tensile tester equipped with a 500N load cell at
200mm/min. with a grip-to-grip separation of 35mm on strips of
approximately 60 x 4 x 0.5mm cut from a cured film. Elastic moduli and strain
values were derived from the initial grip-to-grip separation and are
therefore indicative only. The toughness values of the materials (Wb) were
calculated from the work at break value, normalized for the dimensions of
the specimen.
For swelling measurements, three samples measuring
approximately 10 x 20 x 0.5mm were cut from a cured, extracted and dried
film and weighed (denoted 𝑚𝑖). Each sample was then swollen in excess
water for at least 24h, blotted dry and weighed again in the hydrated state
(denoted 𝑚𝑠). The water uptake (WU) was defined as:
144
𝑤𝑎𝑡𝑒𝑟 𝑢𝑝𝑡𝑎𝑘𝑒 (𝑊𝑈) =𝑚𝑠 − 𝑚𝑖
𝑚𝑖× 100%
Synthesis of NCO-functionalized prepolymers
All reactions were conducted under a nitrogen atmosphere using standard
Schlenck-line techniques. All polymeric diols were dried prior to use by
azeotropic distillation in toluene, which was subsequently removed under
reduced pressure. In a typical experiment, 0.25 g of dibutylphosphate was
added to 100 g of dried polymeric diol, which was kept in the melt using a
heat gun where applicable. This mixture was added dropwise to HDI kept at
100°C, using a molar ratio of NCO:OH of 15:1, and the resulting mixture was
subsequently kept at 100°C for 3h. In the case of PPG polymeric diols, the
procedure was performed at 120°C and kept at 120°C for 6h. The resulting
mixture was then transferred dropwise into a cylindrical glass short-path
thin-film evaporator equipped with Teflon wipers with a volume of
approximately 500mL, which was operated at a reduced pressure of 1∙10-2
mbar at 140°C. All functionalized prepolymers were collected as clear,
colorless, viscous resins, of which the PEG- and PCL-based resins crystallized
upon cooling to room temperature. Products were analyzed by SEC, GC and 1H-NMR (700 MHz, C6D6, representative data for PEG-4k): δ = 4.60 (2H, brs),
4.26 (4H, t), 3.50 (410H, m), 2.97 (4H, q), 2.56 (4H, t), 1.12 (4H, t), 1.02 (4H,
t), 0.88 (8H, m) ppm.
Synthesis of combinatorial poly(urethane-isocyanurate) networks by the
trimerization of mixtures of NCO-functionalized prepolymers in solution
Combinatorial poly(urethane-isocyanurate) networks were synthesized
according to a previously reported procedure.7 In a typical experiment, 4g of
solvent (propylene carbonate in the case of PEG/PCL mixtures and toluene
in the case of PEG/PPG mixtures) was weighed into a polypropylene cup.
Next, the respective NCO-prepolymers were added in the melt in the
reported weight ratios, always totaling 4g. The mixture was then kept under
nitrogen under continuous stirring, until a macroscopically homogenous
solution was formed. Whenever crystallization was observed, the mixture
was gently heated to 50-60°C until the mixture was liquid again. The mixture
145
was subsequently degassed by gently applying a reduced pressure of
approx. 10 to 100mbar until no more bubbles were observed, whereby it
was observed that no residue was built up in the cooling trap of the vacuum
pump. Next, ca. 60 mg of Sn(II)Oct2 (corresponding to ca. 0.75wt% relative
to the solution) was added. After mixing for 15 seconds in a Hauschild
Speedmixer DAC150FVZ, the mixture was brought into a mold consisting of
a 0.5mm-thick polycarbonate frame clamped between two silanized glass
plates measuring 20 x 10cm and kept at 90°C for 48h. All reaction mixtures
were optically transparent within 15min after heating to 90°C. Cured
samples were collected as flexible, non-sticky films, which were
macroscopically homogeneous. The films were subsequently extracted by
swelling in an excess of acetone for 12-24h while periodically refreshing the
solvent. Typically, 6-8 films were extracted simultaneously using 3x 500mL
acetone. Next, 5L of n-hexane was slowly dripped into the 500mL of acetone
over the course of 24h which was allowed to overflow, thereby slowly
exchanging the acetone with n-hexane. During this process, the films slowly
contracted. The resulting films were then dried, first in air and subsequently
in vacuum, and analyzed by attenuated total reflection Fourier-transformed
infrared spectroscopy (ATR-FTIR), swelling, tensile testing and DSC
experiments as described. The remainder of each film was then swollen in
an excess of water for at least 24h, and analyzed in the water-swollen state
by tensile testing and DSC experiments.
146
7.3 Results and Discussion
Synthesis of NCO-functionalized prepolymers and PUI networks by the
selective trimerization of NCO-functionalized prepolymers
Polymeric diols were functionalized using hexamethylene diisocyanate (HDI)
and subsequently trimerized as reported previously.7 It was confirmed that
the functionalization of all polymeric diols with HDI succeeded selectively
and quantitatively in all cases. The characteristics of all functionalized
prepolymers are reported previously.7
Combinatorial poly(urethane-isocyanurate) (PUI) polymer networks
were synthesized by the trimerization of mixture of the NCO-functionalized
prepolymers in solution. In this case, to prevent phase separation, toluene
was used as the inert solvent for the PPG-containing prepolymer mixtures
instead of propylene carbonate. All reaction mixtures were transparent and
homogeneous throughout the trimerization reaction. The completion of the
trimerization reaction was confirmed using attenuated total reflection
Fourier-transformed infrared spectroscopy (ATR-FTIR) as reported
previously (data not shown).7
Material properties of combinatorial PUI networks in the dry state
The thermal and tensile mechanical properties of the PUI networks in the
dry state are reported in Table 7.1. Similar to the PUI networks on which we
reported previously, the mechanical properties varied based on the
molecular weight and nature of the prepolymers present in the network.7
The tensile stress-strain diagrams of all PUI networks in the dry state are
shown in Figure 7.1. In the case of PEG/PCL combinatorial PUI networks, all
networks were semi-crystalline in nature. These networks were generally
tough materials, which is consistent with the literature.8 In the case of
PEG/PPG combinatorial networks, a melting peak was only observed in the
dry state for the networks containing PEG. In these networks, a decreasing
melting enthalpy is observed for decreasing PEG content.
Table 7.1. Mechanical and thermal properties in the dry state of combinatorial PUI networks with increasing hydrophobic
content (HC). On the left, the hydrophobic component is crystallizable in nature (PEG/PCL), on the right, the hydrophobic
component is amorphous in nature (PEG/PPG).
PEG/
PCL Tensile Thermal
PEG/
PPG Tensile Thermal
HC
[wt%]
Emod,dry
[MPa]
σb,dry
[MPa]
εb,dry
[%]
Wb,dry
[MJ·m-3]
Tg
[°C]
Tm
[°C]
ΔHm
[J·g-1]
HC
[wt%]
Emod,dry
[MPa]
σb,dry
[MPa]
εb,dry
[%]
Wb,dry
[MJ·m-3]
Tg
[°C]
Tm
[°C]
ΔHm
[J·g-1]
0 208.9
(42.3)
15.4
(3.1)
403.8
(62.0)
42.34
(9.53) -50.3 46.7 98.4 0
208.9
(42.3)
15.4
(3.1)
403.8
(62.0)
42.34
(9.53) -50.3 46.7 98.4
20 85.8
(26.2)
10.4
(1.5)
401.4
(78.2)
28.33
(5.16) -54.7 42.9 76.8 20
83.8
(-)b
11.8
(-)b
291.0
(-)b
26.55
(-)b -50 44.9 90.3
40 55.7
(8.5)
7.5
(0.6)
322.6
(41.9)
16.34
(1.45) -56.6 35.3 62.7 40
19.5
(5.4)a
3.8
(0.6)a
32.0
(3.3)a
0.78
(0.05)
1) -62.5
2) -0.3 42.4 55.1
60 37.4
(9.4)
7.8
(1.3)
327.2
(68.3)
15.00
(2.98) -58.1 30.5 58.0 60
6.2
(2.0)
2.2
(0.5)
54.0
(7.7)
0.71
(0.19)
1) -61.7
2) -22.5 42.2 36.3
80 66.0
(3.5)
11.4
(1.0)
318.2
(22.2)
23.01
(1.69) -57.4 35.3 60.2 80
4.1
(0.2)
0.9
(0.2)
33.4
(5.6)
0.18
(0.06)
1) -61.7
2) -3.2 42 15
100 111.1
(1.0)a
12.5
(1.6)a
221.4
(11.2)a
16.42
(1.40) -53.40 33.6 42.9 100
3.6
(0.4)
0.7
(0.04)
21.6
(2.0)
0.09
(0.01) -61.7 -c -c
an=2 bn=1 cnot observed
148
Figure 7.1. Tensile stress-strain diagrams in the dry state of combinatorial
PUI networks with increasing hydrophobic content (HC); a) the hydrophobic
component is crystallizable in nature (PCL); b) the hydrophobic component is
amorphous in nature (PPG).
149
Material properties of combinatorial PUI networks in the water-swollen
state
The swelling properties of the combinatorial PUI networks and their thermal
and tensile mechanical properties in the water-swollen state are reported in
Table 7.2. The tensile stress-strain diagrams in the water-swollen state are
shown in Figure 7.2. Both for the crystallizable and amorphous PUI
hydrogels, the water uptake (WU) decreased with increasing hydrophobic
content. The water uptake values ranged from 380 wt% for a completely
hydrophilic PEG PUI network to <2wt% for both of the completely
hydrophobic networks.
The tensile mechanical properties of the PUI hydrogels in the water-
swollen state vary and strongly depend on the nature of the hydrophobic
network component. The crystallizable PEG/PCL combinatorial hydrogels
were generally tough and demonstrated ultimate tensile strengths and
toughness values of up to 335 MPa and 23.8 MJ·m-3, respectively. The
amorphous PEG/PPG hydrogels on the other hand were mechanically very
poor and demonstrated ultimate tensile strengths and toughness values <1
MPa and <0.3 MJ·m-3, respectively.
As reported in Table 7.2, the PEG/PCL combinatorial hydrogels
containing more than 40wt% of PCL were semi-crystalline in the water-
swollen state. This is in accordance with what we reported previously.7 The
combinatorial PEG/PCL hydrogels demonstrated increasing melting
enthalpies for increasing PCL content, reaching up to approximately 25 J·g-1
for a 100 wt% PCL network. Melting enthalpy values were generally lower
than the ones we reported previously: the PUI hydrogel containing 60 wt%
PCL demonstrated a melting enthalpy of approximately 2 J·g-1 whereas the
PUI hydrogel containing 50 wt% PCL on which we reported previously
demonstrated a melting enthalpy of approximately 14 J·g-1.7 Finally, it was
found that all hydrogels containing an amorphous hydrophobic network
component were amorphous in the water-swollen state. In other words, no
crystalline phase attributable to PEG domains was observed in any of the
hydrogels in the water-swollen state.
Table 7.2. Mechanical properties in the water-swollen state of semi-crystalline and amorphous combinatorial networks with an
increasing hydrophobic content (HC). On the left, the hydrophobic component is crystallizable in nature (PCL-4k), on the right,
the hydrophobic component is amorphous in nature (PPG-4k).
PEG/
PCL Swelling Tensile Thermal
PEG/
PPG Swelling Tensile Thermal
HC
[wt%]
WU
[wt%]
Emod,wet
[MPa]
σb,wet
[MPa]
εb,wet
[%]
Wb,wet
[MJ·m-3]
Tg
[°C]
Tm
[°C]
ΔHm
[Jg-1]
HC
[wt%]
WU
[wt%]
Emod,wet
[MPa]
σb,wet
[MPa]
εb,wet
[%]
Wb,wet
[MJ·m-3]
Tg
[°C]
Tm
[°C]
ΔHm
[Jg-1]
0 380.5
(3.8)
1.1
(0.5)
0.4
(0.05)
34.6
(5.0)
0.08
(0.02) -a -a -a 0
380.5
(3.8)
1.1
(0.5)
0.4
(0.05)
34.6
(5.0)
0.08
(0.02) -a -a -a
20 220.8
(1.2)
2.5
(0.2)
0.7
(0.1)
63.6
(8.9)
0.29
(0.06)
1) 40.9
2) 64.5 -a -a 20
313.4
(7.5)
1.7
(0.4)
0.2
(0.1)
9.4
(3.5)
0.01
(0.01) -a -a -a
40 116.9
(1.2)
3.9
(0.1)b
1.6
(0.6)b
96.7
(66.9)b
1.11
(0.76)b
1) 25.8
2) 41.2 -a -a 40
176.6
(2.9)
2.0
(0.3)b
0.5
(0.0)b
37.2
(11.6)b
0.12
(0.03)b
1) 32.8
2) 38.1 -a -a
60 51.3
(0.6)
12.0
(0.5)
3.7
(1.5)
158.2
(120.1)
4.86
(3.65) -a 43 2.0 60
120.5
(1.9)
2.5
(0.3)
0.7
(0.1)
56.6
(19.2)
0.27
(0.11)
1) 31.0
2) 57.6 -a -a
80 16.3
(0.1)
54.7
(2.3)
10.9
(1.1)
335.4
(40.8)
23.84
(3.04) -a 45.3 19.3 80
39.6
(8.8)
1.8
(0.4)
0.7
(0.04)
52.8
(7.4)
0.24
(0.04) -a -a -a
100 0.9
(0.1)
103.6c
(4.6)
6.6c
(0.1)
77.9c
(25.4)
4.74c
(1.28) -a 44.8 25.5 100
1.8
(0.1)
2.3
(0.7)
0.7
(0.03)
28.0
(1.0)
0.12
(0.01) -a -a -a
anot observed bn=2 cpremature failure
151
Figure 7.2. Tensile stress-strain diagrams in the water-swollen state of
combinatorial PUI networks with an increasing hydrophobic content (HC).
The water uptake (WU) of each network is reported next to the respective
curve; a) the hydrophobic component is crystallizable in nature (PEG/PCL); b)
the hydrophobic component is amorphous in nature (PEG/PPG). The 100wt%
PCL film (HC:100%, top) failed prematurely.
152
Of high interest are the combined mechanical and thermal
properties of the combinatorial PUI hydrogels. Most importantly, a
considerable increase in the toughness of PUI hydrogels is shown with
increasing melting enthalpy values. As mentioned previously, such a
correlation between crystallinity and toughness has been reported for
polymer materials in the dry state.8 The results presented here form a strong
indication that the same correlation is applicable to hydrogel materials in
the water-swollen state. In other words, the toughness of hydrogel materials
can be increased considerably by increasing the crystallinity of the
hydrophobic component of the hydrogel in the water-swollen state. This is
considered an important structure-property-relationship in the future
development of tough hydrogel materials. As shown here, an effective
method to incorporate crystallinity into hydrogel materials is by the
combinatorial trimerization of mixtures of NCO-functionalized prepolymers.
7.4 Conclusions
The combinatorial trimerization of mixtures of NCO-functionalized
prepolymers is an effective method for synthesizing tough hydrogels. For
semi-crystalline combinatorial PUI hydrogels, it is shown that he toughness
of the hydrogel in the water-swollen state increases significantly with
increasing melting enthalpy. This correlation between melting enthalpy and
toughness of the hydrogel in the water-swollen state is an important
structure-property-relationship for the future development of tough
hydrogels for load-bearing biomedical applications.
7.5 Acknowledgements This project has received funding from the European Union’s Horizon 2020
research and innovation program under the Marie Skłodowska-Curie Grant
Agreement no. 642890 (http://thelink-project.eu/).
The support of the labs of Dr. Frank Richter and the analytical departments
of Covestro Deutschland AG Leverkusen is greatly acknowledged.
153
7.6 References 1. Liu, Y. et al. Recent Developments in Tough Hydrogels for
Biomedical Applications. Gels 4, 46 (2018). 2. Creton, C. 50th Anniversary Perspective: Networks and Gels: Soft
but Dynamic and Tough. Macromolecules 50, 8297–8316 (2017). 3. Gong, J. P. Materials both Tough and Soft. Science (80-. ). 344, 161–
162 (2014). 4. Gong, J. P. Why are double network hydrogels so tough? Soft
Matter 6, 2583–2590 (2010). 5. Chen, Q., Chen, H., Zhu, L. & Zheng, J. Fundamentals of double
network hydrogels. J. Mater. Chem. B 3, 3654–3676 (2015). 6. Dragan, E. S. Design and applications of interpenetrating polymer
network hydrogels. A review. Chem. Eng. J. 243, 572–590 (2014). 7. Driest, P. J., Dijkstra, D. J., Stamatialis, D. & Grijpma, D. W. Chapter 6
of this Thesis. (2019). 8. Galeski, A. Strength and toughness of crystalline polymer systems.
Prog. Polym. Sci. 28, 1643–1699 (2003). 9. Read, B. E. Mechanical relaxation in some oxide polymers. Polymer
(Guildf). 3, 529–542 (1962). 10. Bailey, F. E. (Frederick E. & Koleske, J. V. Poly(ethylene oxide).
(Academic Press, 1976). 11. P.Pêgo, A., Poot, A. A., Grijpma, D. W. & Feijen, J. Copolymers of
trimethylene carbonate and ε-caprolactone for porous nerve guides: Synthesis and properties. J. Biomater. Sci. Polym. Ed. 12, 35–53 (2001).
12. Faucher, J. A. The dependence of glass transition temperature on molecular weight for poly(propylene oxide) and poly(butylene oxide). J. Polym. Sci. Part B Polym. Lett. 3, 143–145 (1965).
Chapter 8 - Poly(ethylene glycol)
Poly(urethane-isocyanurate) Hydrogels
for Contact Lens Applications
P.J. Driest1,2, I.E. Allijn2, D.J. Dijkstra1, D. Stamatialis2, D.W. Grijpma2
1Covestro Deutschland AG, CAS-Global R&D, 51373 Leverkusen, Germany
2Technical Medical Centre, Faculty of Science and Technology, Department
of Biomaterials Science and Technology, University of Twente P.O. Box 217,
7500AE Enschede, The Netherlands
Published in:
Polym. Int. (in press), DOI: 10.1002/pi.5938
Abstract
Over the past decades, the global
use and market of contact lenses
has expanded steadily. Due to the
many demands on material
properties (e.g. mechanical,
optical and biological) the
development of novel contact lens
materials is challenging.
Specifically, the ideal combination
of high equilibrium water content (EWC), high toughness in the hydrated
state and low protein adsorption is difficult to realize.
In this work, poly(ethylene glycol) poly(urethane-isocyanurate) (PEG
PUI) type hydrogels that combine the above important properties are
presented as a new class of materials for contact lens applications. It is
shown that these PEG PUI hydrogels demonstrate high toughness values in
the hydrated state ranging from 98 to 226 kJ·m-3 and elastic moduli ranging
from 0.8 to 17.2 MPa for networks with equilibrium water contents ranging
from 76.3 to 16.1wt%. These hydrogels also demonstrate transmittance
values >90% across the visible spectrum, clarities close to 100% in most
cases and refractive indices ranging from 1.48 to 1.36. Importantly, these
hydrogels are non-cytotoxic and demonstrate lower bovine serum albumin
adsorption values than several commercial contact lenses of 0.24 to 0.65
mg/g compared to 0.55 to 1.38 mg/g after 24h, respectively.
This combination of high EWC, high toughness in the hydrated state
and low protein adsorption is exceptional. These properties can be
attributed to the PEG PUI network structure: the use of a PEG polymeric
backbone provides hydrophilicity and chemical inertness while the PUI type
crosslinking units provide high toughness in the hydrated state.
156
8.1 Introduction Since the introduction of soft contact lenses in the 1960s, the development
of contact lens materials has received much attention over the years and is
still an actively discussed topic today.1,2 At present, the market for soft
contact lenses is estimated to exceed $8.5billion globally, with an estimated
150 million users of contact lenses worldwide in 2019.3,4 Historically, the
market for soft contact lenses is dominated by 2 main classes of materials:
poly(2-hydroxyethyl methacrylate) (pHEMA) type hydrogels and silicone
type hydrogels.1 An overview of common contact lens materials and their
general physical properties is shown in Table 8.1.2
Despite the long history of the field, novel (classes of) soft contact
lens materials are still actively sought after and reported on today.5 The
development of new contact lens materials is considered a challenging field
though. This is easily appreciated when looking at the impressive list of
requirements a material has to fulfil before being considered for contact lens
applications (see Table 8.2).2,5,6
Typically, modification of the various material characteristics is not
complementary and improving one property often deteriorates another. A
specifically challenging combination of contradictory material properties is
high hydrophilicity, high toughness and low biofouling.7,8 One of the reasons
for this is that the hydrophilicity of contact lens materials is generally
realized by the introduction of hydrogen-bonding moieties (e.g. -OH, -NH2
or -COOH groups) into the polymeric backbone. However, the presence of
such hydrogen bonding moieties has been shown to play a key role in the
biofouling process of hydrogel materials.9 By facilitating biofouling, contact
lenses can hold microbes in prolonged contact with the cornea of the eye,
leading to ocular infections and complications.8 In fact, even with good
contact lens care practice, up to 80% of contact lenses and contact lens cases
were found to harbor disease-causing microbes.10 This shows that there is
still need for a low-biofouling hydrophilic hydrogel material for contact lens
applications.6,8,11–13
157
Table 8.1. General properties of some common contact lens materials.
Adapted from Musgrave et al.2
Material EWC
[%]
Emod
[MPa]
O2-
perm.a
[Dk]
Wear
timeb
[days]
Polymer structure
PMMA 0 1000 0 <1
PMMA-
Silicone 0 -c 15 -c
Silicone-
HEMA (rigid) 0 10 10-100 -c
pHEMA
Hydrogel
30-
80 0.2-2 10-50 1-7
Silicone
(PDMS)
Hydrogel
20-
55 0.2-2 60-200 ~7-28
PVA 60-
70 -c 10-30 <1
aOxygen permeability, typically reported in Dk-units (ml O2/[ml⋅mmHg])
brecommended maximum wear time before lens disposal
cdata not reported
158
Table 8.2. List of typical requirements for contact lens materials.
Material Property Requirement / Comment
Physical and
Chemical
Hydrophilicity As high as possible. Typically indicated by the
equilibrium water content (EWC)
Oxygen permeability As high as possible
Mechanical
properties in the
hydrated state
Need to be suitable. Most notably, a suitable
elastic modulus and high toughness
Chemical stability Sufficient. With respect to both shelf-life and
common sterilization techniques such as UV-
exposure or autoclaving
Optical
Transmittance As high as possible across the visible spectrum
Clarity As high as possible. Clarity is typically considered
a measure for the see-through quality of the
material
Refractive index Needs to be suitable
Biological
Biocompatibility Required. Typically indicated by non-cytotoxicity
Biofouling As low as possible. Specifically, low protein
adsorption
159
In this context, poly(ethylene glycol) (PEG) is considered a promising
material.2,5 Specifically the combination of high hydrophilicity with chemical
inertness is attractive. The main drawback associated with PEG hydrogel
materials however, is the poor mechanical performance (i.e. the poor
toughness) in the hydrated state that is typically associated with crosslinked
PEG hydrogels.14–16 Consequently, the use of PEG has thus far been limited
to surface modification procedures of existing contact lens materials.8,11,17
Hydrogels based on crosslinked PEG have not been reported in this context.
Scheme 8.1. Schematic representation of the synthesis of a PEG PUI polymer
network. First, a PEG polymeric diol is functionalized using hexamethylene
diisocyanate (HDI); next the PUI network is formed by the selective
trimerization of the NCO-functionalized PEG prepolymer.
160
In this work, we investigate the use of a recently developed new
class of PEG-based poly(urethane-isocyanurate) (PUI) type polymer
networks for contact lens applications.18 The synthesis of these networks is
based on the functionalization and subsequent trimerization of PEG
polymeric diols using hexamethylene diisocyanate (HDI), as shown in
Scheme 8.1. The resulting PEG PUI networks were shown to exhibit high
water uptakes of up to 508wt% combined with high toughness in the
hydrated state.18 Since both PEG and PUI type polymer networks are also
known for their excellent optical properties19,20, high refractive indices21–23
and biocompatibility24,25, we hypothesize that these PEG PUI type hydrogels
would be a promising class of materials for contact lens applications. PEG
PUI hydrogels of varying crosslinking density are synthesized and
characterized for properties relevant to contact lens applications (see Table
8.2) and compared to commercially available contact lenses.
161
8.2 Experimental Details
Materials and methods
Materials
Poly(ethylene glycol)s (𝑀𝑛 = 0.4, 0.6, 1, 2, 4 and 10 kg/mol), stannous
octoate (Sn(II)Oct2) and resazurin sodium salt were purchased from Sigma-
Aldrich. Hexamethylene diisocyanate (HDI) was supplied by Covestro
Deutschland AG. A human retinal pigment epithelium cell line (Arpe-19) and
Dulbecco's Modified Eagle Medium/Nutrient Mixture F-12 (DMEM/f12) with
10% fetal bovine serum (FBS) cell culture medium were purchased from
ATCC. Commercial contact lenses (Midafilcon A day lenses manufactured by
Menicon and Riofilcon A day lenses, Comfilcon A month lenses and
Somofilcon A month lenses manufactured by Cooper Vision) were kindly
supplied by a local optometrist. A DC Protein Assay kit was purchased from
Bio-Rad. All other chemicals were purchased from Sigma-Aldrich.
Hydrogel synthesis and characterization
All chemical reactions were conducted under a nitrogen atmosphere using
standard Schlenck-line techniques. NMR-spectra were collected on a Bruker
Advance III-700 in C6D6. Size exclusion chromatography (SEC) was performed
according to DIN 55672-1:2016-03, on 4 PSS SDV Analytical columns (2x
100Å, 5µm; 2x 1000Å, 5µm) using an Agilent 1100 Series pump and an
Agilent 1200 Series UV Detector (230 nm) with tetrahydrofuran as the
elution solvent at 40°C and 1.00 mL/min. Residual monomeric diisocyanate
contents were determined by gas chromatography (GC) according to DIN EN
ISO 10283 on an Agilent Technologies 6890N system using a 15.0m DB17
column and tetradecane as an internal standard. Attenuated total reflection
Fourier-transformed infrared (ATR-FTIR) spectra were recorded on a Bruker
FTIR Spectrometer Tensor II with a Platinum-ATR-unit with diamond crystal.
Physical properties
Differential scanning calorimetry (DSC) measurements were performed
under nitrogen on ca. 10 mg samples on a Perkin-Elmer Calorimeter DSC-7
162
in 2 heating runs from -100°C to +150°C with a heating rate of 20 K/min and
a cooling rate of 20 K/min. The glass transition temperature (Tg) and peak
melting temperature (Tm,p) were determined based on the second heating
run. Tensile tests were performed in triplicate at room temperature using a
Zwick Z0.5 tensile tester equipped with a 500N load cell at 200mm/min. with
a grip-to-grip separation of 35mm on strips of ca 60 x 4 x 0.5mm cut from a
cured film. Elastic moduli and strain values were derived from the initial grip-
to-grip separation and are indicative. The toughness values of the materials
were calculated from the work at break value, normalized for the
dimensions of the specimen.
For swelling measurements, three samples measuring
approximately 10 x 20 x 0.5mm were cut from a cured film and weighed in
the dry state (values denoted 𝑚𝑑). Each sample was then swollen in excess
water for at least 24h, blotted dry and weighed again in the hydrated state
(values denoted 𝑚𝑤). The water uptake of the material was then estimated
as:
𝑤𝑎𝑡𝑒𝑟 𝑢𝑝𝑡𝑎𝑘𝑒 =𝑚𝑤−𝑚𝑑
𝑚𝑑× 100% (eq. 1)
The equilibrium water content (EWC) was estimated as:
𝐸𝑊𝐶 =𝑚𝑤−𝑚𝑑
𝑚𝑤× 100% (eq. 2)
Oxygen permeabilities were estimated according to the model proposed by
Morgan and Efron26 using the equation:
𝐷𝑘 = 1.67𝑒0.0397𝐸𝑊𝐶 (eq. 3)
163
where Dk is the oxygen permeability of the material in Dk-units (ml
O2/[ml⋅mmHg]) and EWC is the equilibrium water content.
Optical properties
Wavelength-dependent transmission experiments were performed using a
Konica Minolta CM-5 spectrophotometer. Scattering experiments were
performed using a BYK-Gardner Haze-Gard Plus haze meter after norm
ASTM D 1003_11. Scattering characteristics are divided into three
parameters: transmittance (T), haze (H) and clarity (C). Of these,
transmittance is defined as the ratio of transmitted light compared to the
incident light. Haze indicates the amount of light diffused under wide angle
(>2.5°) scattering. Clarity indicates the amount of light diffused under
narrow angle (<2.5°) scattering. Refractive index measurements were
performed at 20.0°C using a Rudolph Research J357 automatic
refractometer equipped with a temperature-controlled sample presser for
gels. Six independent measurements were taken for each sample, of which
the average value is reported.
Biological properties
Sterilization by UV-exposure was performed for 1h under four 9 Watt UV-
lamps at 365 nm and approx. 6 mW/cm2. Sterilization by autoclaving was
performed for 20min at 120°C in a PBI International pressure cooker from
Salm and Kipp BV. Fluorescence measurements were performed on a Perkin
Elmer Wallac Victor3 spectrophotometer.
Synthesis and characterization of NCO-functionalized PEG prepolymers
All polymeric diols were dried prior to use by azeotropic distillation in
toluene, which was subsequently removed under reduced pressure. In a
typical experiment, 0.25 g of dibutylphosphate was added to 100 g of dried
polymeric diol (where necessary, kept in the melt using a heat gun). This
mixture was added dropwise to HDI at 100°C, using a molar ratio of NCO:OH
of 15:1, and subsequently kept at 100°C for 3h. The resulting mixture was
then transferred dropwise into a short-path thin-film evaporator, which was
operated at a reduced pressure of 1∙10-2 mbar and 140°C. Products were
164
collected as transparent, colorless, viscous resins, of which the PEG-1k-,
PEG-2k-, PEG-4k- and PEG-10k-based resins crystallized upon cooling to
room temperature. Products were analyzed by SEC, GC and 1H-NMR (700
MHz, C6D6, representative data for PEG-4k): δ = 4.60 (2H, brs), 4.26 (4H, t),
3.50 (410H, m), 2.97 (4H, q), 2.56 (4H, t), 1.12 (4H, t), 1.02 (4H, t), 0.88 (8H,
m) ppm.
Synthesis and characterization of PEG PUI polymer networks by the
selective trimerization of NCO-functionalized PEG prepolymers
Poly(urethane-isocyanurate) (PUI) polymer networks were synthesized in
analogy to a previously reported procedure, with slight modifications.18 In a
typical experiment, 8g of NCO-prepolymer was weighed into a
polypropylene cup, heated to 90°C and subsequently degassed by gently
applying reduced pressure until no more gas formation was observed. Next,
ca. 60 mg of Sn(II)Oct2 (corresponding to ca. 0.75wt% relative to the
prepolymer) was added at 90°C. After mixing for 15 seconds in a Hauschild
Speedmixer DAC150FVZ, the reaction mixture was brought into a mold
consisting of a 0.5mm-thick polycarbonate frame clamped between two
silanized glass plates measuring approximately 20 x 10cm and kept at 90°C
for 24h. Products were collected as flexible transparent homogeneous films,
of which the films based on PEG-4k and PEG-10k prepolymers slowly turned
opaque upon cooling to room temperature. The films were analyzed in the
dry state by ATR-FTIR, water swelling measurements, tensile testing and DSC
experiments and subsequently swollen in excess water for at least 24h.
Finally, the films were analyzed in the hydrated state for physical properties
(tensile testing and mass-loss after sterilization by UV-exposure or
autoclaving), optical experiments (transmittance, scattering and refractive
index) and biological experiments (cytotoxicity and protein adsorption,
described below).
165
Indirect cytotoxicity measurements of PEG PUI hydrogels and commercial
contact lenses
Cytotoxicity measurements were performed on the PEG PUI hydrogels and
several commercial contact lenses as described previously, with minor
modifications.27 In short, PEG PUI hydrogels were swollen in milliQ water,
cut in discs with a diameter of 12mm (n=3), blotted dry and weighed. Next
the hydrogels were transferred to a phosphate buffered saline (PBS) solution
and autoclaved for 20min at 120℃ in a pressure cooker. The PEG PUI
hydrogels and the commercial contact lenses (n=3) were then incubated in
DMEM/f12 with 10% FBS (0.1 g/mL) at 37℃ in a humidified incubator
containing 5% CO2. Meanwhile, Arpe-19 cells were seeded on a polystyrene
tissue culture well plate with a concentration of 12.5·104cells/mL and left for
24h at 37℃ in a humidified incubator containing 5% CO2. After 24h, the
medium was removed from the Arpe-19 cells and replaced with conditioned
medium from the PEG PUI hydrogels or commercial contact lenses,
respectively. A control cell culture was kept in which the medium was
replaced with fresh medium that had not been in contact with hydrogel
materials. After 48h the cell viability was measured using an Alamar blue
solution (440µM resazurin sodium salt in phosphate buffered saline (PBS))
and normalized with respect to the control cell-culture. To measure the cell
viability, a 10:1 (v/v) mixture of cell-medium and Alamar blue solution was
added to the cells. After incubating for 2h at 37℃, the fluorescence emission
of the solution at 590nm was measured using an excitation wavelength of
560nm.
Protein adsorption measurements of PEG PUI hydrogels and commercial
contact lenses
Protein adsorption measurements were performed on the PEG PUI hydrogel
films and several commercial contact lenses as described previously, with
minor modifications.6 In short, the PEG PUI hydrogels were swollen in milliQ
water and cut in discs with a diameter of 12mm (n=3). Commercial contact
lenses were used as supplied. All samples were incubated for 24h in 2mL
PBS, containing 1.5mg bovine serum albumin (BSA) per mL PBS, under gentle
shaking at 37°C. Afterwards, the samples were extracted three times with
166
neat PBS for at least 1.5h per extraction step and subsequently dried. The
dry samples were then swollen in PBS containing 1% (v/v) sodium dodecyl
sulfate (SDS) and kept at 37°C under gentle shaking for three days. The
concentration of BSA in the supernatant was then determined using a Bio-
Rad DC Protein Assay kit according to the supplier’s instructions.
167
8.3 Results and Discussion
Synthesis of NCO-functionalized PEG prepolymers and PEG PUI polymer
networks by the selective trimerization of NCO-functionalized PEG
prepolymers
Poly(ethylene glycol) (PEG) polymeric diols were functionalized using
hexamethylene diisocyanate (HDI) and subsequently trimerized as reported
previously.18 Using 1H-NMR, SEC and GC, it was confirmed that the
functionalization of all PEG polymeric diols with HDI succeeded selectively
and quantitatively. The characteristics of all functionalized PEG prepolymers
are reported in the supporting information in Table S1. Next, the NCO-
functionalized prepolymers were crosslinked by the selective trimerization
of the terminal NCO groups. This resulted in the formation of PEG
poly(urethane-isocyanurate) (PUI) type polymer networks. The completion
of the trimerization reaction was confirmed using attenuated total reflection
Fourier-transformed infrared spectroscopy (ATR-FTIR) (supporting
information, Figure S1).
Physical properties of PEG PUI networks and hydrogels: equilibrium water
content, oxygen permeability, mechanical properties and stability
Water uptake, EWC and oxygen permeability
Due to the hydrophilicity of PEG, the water uptake and resulting equilibrium
water content of the PEG PUI networks were generally high. The shortest-
chain PEG PUI network, synthesized from a 0.4 kg/mol PEG prepolymer,
demonstrated a water uptake of 19.2wt%. This corresponds to an EWC of
16.1wt% in the hydrated state. The longest-chain PEG PUI hydrogel on the
other hand, synthesized from a 10 kg/mol PEG prepolymer, demonstrated a
water uptake of 321.6wt%, corresponding to an EWC of 76.3wt%. All water
uptakes and EWCs are reported in Table 8.3. Owing to the well-defined
network structure on which we reported previously, the water uptakes and
EWCs of the hydrogels could easily be controlled over a wide range.18
Notably, the range of EWCs measured for the PEG PUI hydrogels
168
encompasses almost the entire range of EWCs typically reported for contact
lens hydrogels (i.e. approx. 20 to 80wt%, see Table 8.1).2 This includes both
soft and rigid contact lens hydrogels.
From the EWCs, the oxygen permeabilities of the PEG PUI hydrogels
were approximated using the model proposed by Morgan and Efron (eq. 3)
and are reported in Table 8.3.26 This semi-empirical model correlates oxygen
permeability (the product of oxygen diffusivity and solubility, typically
referred to as “Dk” in contact lens literature) to EWC and is used for
conventional, non-silicon hydrogels.28 The model assumes that oxygen
transport only takes place through the water phase of the hydrogel while no
oxygen transport takes place through the solid phase of the hydrogel.
Consequently, the oxygen permeability values estimated using this model
should be considered as minimal values rather than the actual oxygen
permeability values. The difference between the predicted and actual
oxygen permeability values evidently depends on the specific oxygen
transport properties of the hydrogel´s solid phase. In this respect, it is
worthwhile to note that PEG polymer networks have been shown to exhibit
a non-zero oxygen permeability in the dry state and actually demonstrate a
certain level of oxygen affinity.29 Therefore, we expect that the PEG PUI
hydrogels have higher actual oxygen permeability values than the
approximated values reported in Table 8.3. However, an experimental
assessment will be needed to determine the exact oxygen permeabilities of
the PEG PUI hydrogels.
Table 8.3. Physical properties of the PEG PUI networks and hydrogels in the dry state and in the hydrated state. WU stands for
water uptake. Numbers in parentheses are standard deviations.
PEG PUI Hydrogel
Prepolymer
WU
[%]
EWC
[%]
Tg
[°C]
Tm,p
[°C]
Emod,dry
[MPa]
σb,dry
[MPa]
εb,dry
[%]
Emod,wet
[MPa]
σb,wet
[MPa]
εb,wet
[%]
Wb,wet
[kJ·m-3]
O2-perm.c
[Dk]
PEG-0.4k-diNCO 19.2
(0.1)
16.1
(0.1) -20.2 - a
17.1
(2.4)
2.7
(0.8)
13.9
(2.6)
17.0
(0.8)
2.0
(0.4)
9.8
(2.3)
110
(40) 3.2
PEG-0.6k-diNCO 26.1
(0.3)
20.7
(0.2) -23.2 - a
18.7
(1.3)
2.8
(0.4)
13.6
(1.6)
17.2
(0.5)
2.0
(0.3)
9.4
(1.6)
99
(27) 3.8
PEG-1k-diNCO 87.1
(0.6)
46.5
(0.2) -47.2 - a
10.9
(0.6)
1.4
(0.2)
11.5
(2.1)
9.6
(0.3)
1.6
(0.8)
14.1
(8.5)
142
(123) 10.6
PEG-2k-diNCO 150.6
(0.7)
60.1
(0.1) -42.9 21.6
6.6
(1.1)
1.6
(0.4)
25.1
(10.1)
5.9
(0.4)
1.1
(0.1)
17.0
(2.8)
98
(20) 18.2
PEG-4k-diNCO 222.6
(0.7)
69.0
(0.1) -49.5 39.3
171.5
(6.1)
19.3
(0.5)
250.7
(26.2)
4.0
(0.2)b
0.8
(0.1)b
21.7
(1.0)b
104
(5)b 25.9
PEG-10k-diNCO 321.6
(4.2)
76.3
(0.2) -49.4 53.8
309.2
(30.0)
26.4
(2.7)
517.0
(40.9)
0.8
(0.3)
0.4
(0.1)
88.9
(23.8)
226
(94) 34.5
anot observed bn=2 capproximated value calculated from the EWC assuming zero transport of oxygen through the solid phase of the hydrogel. Reported in Dk-
units (ml O2/[ml⋅mmHg])
170
Mechanical properties in the dry state and in the hydrated state
The mechanical properties of all polymer networks were assessed both in
the dry state and in the hydrated state and are reported in Table 8.3. It was
found that the dry polymer networks were generally tough materials, with
varying mechanical properties dependent on network density and
crystallinity. In the dry state, the elastic moduli (Emod,dry) of the networks
ranged from 6.6 to 309.2 MPa, while the stress at break (σb,dry) and
elongation at break (εb,dry) values ranged from 1.4 MPa to 26.4 MPa and from
11.5% to 517.0%, respectively. The PUI networks synthesized from PEG
prepolymers with a molecular weight of 2kg/mol and higher were semi-
crystalline in nature in the dry state, displaying melting peaks in the range of
20-60°C.
Table 8.3 also reports the mechanical properties of the hydrogels in
the hydrated state, which are important for their application as contact
lenses. The PEG PUI hydrogels were generally tough in the hydrated state,
covering a wide range of material properties. In the hydrated state, the
elastic moduli of the materials (Emod,wet) ranged from 0.8 to 17.2 MPa, the
stress at break (σb,wet) values ranged from 0.4 to 2.0 MPa and the elongation
at break (εb,dry) values ranged from 9.4% to 88.9%. The hydrogels
demonstrated toughness (Wb,wet) values in the hydrated state ranging from
98 to 226 kJ·m-3.
The tensile stress-strain diagrams of the PUI hydrogels in the
hydrated state are shown in Figure 8.1a, illustrating the high toughness of
the gels. Importantly, the stress at break values were >400 kPa for all
hydrogels and the toughness values were approximately 100 kJ·m-3 or more.
This illustrates the high mechanical resilience of these hydrogels in the
water-swollen state. In practice, it was observed that in the water-swollen
state all hydrogels could easily be handled without fracture or rupture. It
should be noted that this is non-trivial for crosslinked PEG hydrogels. As far
as we know, these PEG PUI hydrogels are the toughest crosslinked PEG
hydrogels reported in the literature, clearly outperforming other typical
crosslinked PEG hydrogels.14–16 The high toughness of the PEG PUI type
hydrogels can likely be attributed to the poly(urethane-isocyanurate) type
network structure on which we reported previously.18
171
To illustrate the wide range of mechanical properties accessible by
the synthesis method used, the tensile elastic moduli of the hydrogels are
plotted together with the EWCs in Figure 8.1b. The measured range of elastic
moduli covers most of the range of elastic moduli typically reported for soft
contact lens materials (0.2-2MPa) and also includes typical values reported
for rigid contact lens materials (ca. 10 MPa) (see Table 8.1).2 It should be
noted that the current span of elastic moduli is obtained simply by varying
the molecular weight of the starting material. Other strategies typically used
to manipulate the mechanical properties of hydrogel materials may also be
considered, such as the incorporation of mono-functional prepolymers
and/or prepolymers with a different chemical nature.
Hydrogel stability during sterilization
A first assessment of the stability of the PEG PUI hydrogels was done by
exposing the hydrogels to two common sterilization procedures. The
hydrogels were analyzed gravimetrically to detect any changes in mass and
visually by inspecting each sample for changes in color, size and/or opacity.
No mass loss was determined for any of the hydrogels after 1h of UV-
exposure nor after autoclaving at 120°C for 20min. Also, no changes in
color, size or opacity of the hydrogels were observed. Although changes in
other material properties, such as mechanical properties, degree of
swelling or optical properties, have not been investigated, these results
indicate that the PEG PUI hydrogels are stable under the sterilization
procedures used.
Figure 8.1. Mechanical properties of PEG PUI hydrogels in the water-swollen state; a) tensile stress-strain diagrams of PEG PUI
hydrogels in the hydrated state. The molecular weights of the respective prepolymers as well as the equilibrium water contents
are reported next to the curves; b) tensile elastic moduli (Emod; black squares; approximated as the tangent of the linear regime
of the tensile stress-strain dependence) and equilibrium water content (EWC; blue circles) of the PEG PUI hydrogels as a function
of the number average molecular weight between crosslinks Mc of the polymer network, approximated by the number average
molecular weight of the respective prepolymer Mn.
173
Optical properties of PEG PUI hydrogels: transmittance, haze, clarity and
refractive index
Transmittance, Haze and Clarity
The transparency of the PEG PUI hydrogels was investigated by two
methods. Firstly, the transmittance of the hydrogels was determined as a
function of wavelength. The transmission spectra of all hydrogels are shown
in Figure 8.2a. Typically, an average transmittance of 90% is considered a
minimum requirement for contact lens applications,2,30 although minimum
values of 95% are also reported.5 As seen in the spectra, the transmittance
of the hydrogels was generally >90% throughout the visible spectrum and
on average around 95% for most hydrogels. It should be noted that these
measurements were performed on hydrogel films with a thickness of 0.5-
1mm, whereas the typical center-point-thickness of a contact lens is around
50-250µm.31 In other words, even in non-optimized form the PEG PUI
hydrogels already meet the general transmittance requirements for contact
lens applications.
Secondly, the scattering characteristics of the transmitted light were
investigated (see results in Table 8.4). All hydrogels demonstrated high
transmittance values of >90%, with the PEG-1k, PEG-2k and PEG-4k
hydrogels demonstrating transmittance values of around 95%. This is
consistent with the wavelength dependent measurements discussed
previously. More importantly, the PEG PUI hydrogels generally showed very
high clarities, in most cases with values close to 100%. Clarity is also referred
to as the “see-through quality” of the material and describes how well fine
details and sharp images can be perceived when looking through the
material.32 Finally, the haze values, considered as a measure for the loss of
contrast32, were in most cases generally low and in the range of 1-4%.
Figure 8.2. Optical properties of PEG PUI hydrogels; a) wavelength-dependent transmission spectra of the PEG PUI hydrogels in
the UV-visible spectrum. A dotted grey line is included at the 90%-transmittance threshold as a guide for the eye; b) refractive
indices (nD20) and equilibrium water contents (EWC) of the PEG PUI hydrogels as a function of the molecular weight between
crosslinks of the PEG PUI networks. Marked by grey lines is the range of refractive indices described by Caló et al. as “ideal
hydrogels should have a refractive index value matching the range 1.372–1.381”.5
175
Table 8.4. Scattering characteristics (transmittance, haze and clarity) and
refractive indices (nD20) of PEG PUI hydrogels.
PEG PUI
Hydrogel
Prepolymer
Transmittance
[%]
Haze
[%]
Clarity
[%]
Refractive
Index (nD20)
[-]
PEG-0.4k-diNCO 93.1 3.4 96.2 1.4796
PEG-0.6k-diNCO 93.1 1.8 99.5 1.4707
PEG-1k-diNCO 95.2 1.9 99.4 1.4010
PEG-2k-diNCO 95.1 3.0 99.4 1.3768
PEG-4k-diNCO 95.7 2.2 98.7 1.3611
PEG-10k-diNCO 90.5 22.8 86.3 1.3562
Refractive index
The refractive indices of the PEG PUI hydrogels in the hydrated state
are reported in Table 8.4 and plotted in Figure 8.2b together with their
EWCs. The refractive indices ranged from ca. 1.48 for the shortest-chain
(PEG-0.4k) hydrogel to ca. 1.36 for the longest-chain (PEG-10k) hydrogel.
Indicated by the trends shown in Figure 8.2b, the refractive index seems to
be mostly dependent on the EWC of the hydrogel. This is consistent with
what is reported in the literature.33
The literature regarding the preferred value of the refractive index
of contact lens materials is inconsistent. On the one hand, it is reported that
the refractive index value of a contact lens material should preferably be as
high as possible, in order to allow the manufacture of the thinnest possible
contact lenses.34,35 On the other hand, it is also reported that the refractive
index of a contact lens material should preferably match that of the human
cornea as closely as possible, values ideally lying between 1.372 and
1.381.5,36 Typical values reported in the literature for commercial and novel
scientific contact lens materials range from approximately 1.38 to 1.45.33–35
The PEG PUI hydrogels span a wide range of refractive index values,
encompassing all values mentioned in the literature. In this regard, it is safe
to say that the range of refractive index values of the PEG PUI hydrogels is
176
suitable for contact lens applications. The intrinsically high refractive indices
of the materials comprising the hydrogel networks are considered beneficial
in this context. For example, a refractive index value of 1.48 is reported for
a pure HDI polyisocyanurate network22 and a value of 1.47 is provided by the
manufacturer for a liquid PEG-0.6k polymeric diol. Furthermore, the
refractive index of PEG polymeric diols has been shown to increase for
increasing molecular weight PEG.21 These high values allow for hydrogel
formulations with a higher EWC and/or thinner contact lenses, both of which
are associated with improved wearing comfort.34 Also the broad range and
easy control over the refractive index of the PEG PUI hydrogel materials is
considered beneficial for further optimization.
Biological properties of PEG PUI hydrogels: biocompatibility and protein
adsorption
Biocompatibility
The biocompatibility of the prepared and extracted PEG PUI hydrogels was
investigated in an indirect cytotoxicity study in analogy to a procedure
reported previously.27 A human retinal pigment epithelium cell line (Arpe-
19) was used as a model cell line for this. The viabilities of the cells after 48h
is reported in Table 8.5. As can be seen, all PEG PUI hydrogels were non-
cytotoxic under these conditions, indicating good biocompatibility. This is in
line with our expectations, considering that all the chemical moieties
present in the hydrogel network are known to be biocompatible.24,25 In most
cases, PEG PUI hydrogels had a slight stimulating effect on cell growth,
indicated by cell viabilities >100% compared to the control cell culture.
Table 8.5. Biological properties of the PEG PUI hydrogels and a selection of commercial contact lenses. Numbers in parentheses
are standard deviations.
PEG PUI Hydrogel
Prepolymer
EWC
[%]
Cell viability
after 48h
[%]
Adsorbed BSA
after 24ha
[mg/g]
Contact Lens Material
(manufacturer, wear time)
EWC
[%]
Cell
viability
after 48h
[%]
Adsorbed
BSA after
24ha
[mg/g]
PEG-0.4k-diNCO 16.1
(0.1)
111
(3)
0.24
(0.09)b
Midafilcon A
(Menicon, day) 56c
99
(4)
0.85
(0.07)b
PEG-0.6k-diNCO 20.7
(0.2)
110
(3)
0.45
(0.05)
Riofilcon A
(Cooper Vision, day) 58c
92
(3)
1.38
(0.33)
PEG-1k-diNCO 46.5
(0.2)
109
(2)
0.39
(0.18)
Comfilcon A
(Cooper Vision, month) 48c -d
0.55
(-)e
PEG-2k-diNCO 60.1
(0.1)
106
(1)
0.64
(0.19)
Somofilcon A
(Cooper Vision, month) -f -d
0.71
(-)e
PEG-4k-diNCO 69.0
(0.1)
102
(1)
0.65
(0.08)
PEG-10k-diNCO 76.3
(0.2)
104
(3)
0.64
(0.14)
aquantified as the total mass of adsorbed BSA in 24h, normalized for the dry weight of the hydrogel sample; bn=2; cas provided by the manufacturer; dnot determined; en=1; fnot provided
178
Protein Adsorption
To investigate the biofouling properties of the PEG PUI hydrogels, a protein
adsorption study was performed in analogy to a procedure reported
previously.6 Bovine serum albumin (BSA) was used as a model protein. The
amounts of adsorbed BSA after 24h, normalized for the dry weight of the
hydrogel, are reported in Table 8.5 for all PEG PUI hydrogels and several
commercial contact lenses. The PEG PUI hydrogels demonstrated BSA
adsorption values ranging from 0.24 mg/g to 0.65 mg/g. These values are
significantly lower than those of the commercial contact lenses tested,
which ranged from 0.55 mg/g to 1.38 mg/g. Especially the disposable
contact lenses intended for single-day use demonstrated high BSA
adsorption values of 0.85 mg/g and 1.38mg/g. This indicates that PEG PUI
hydrogels are suitable for the production of contact lenses with a wear time
of ≥1 month, at least as far as biofouling properties are concerned.
The low biofouling properties of the PEG PUI hydrogels can likely be
attributed to the inert chemical structure of the PEG PUI hydrogel networks.
The new PEG PUI hydrogels are unique in this context, in the sense that they
are hydrophilic but do not contain any free -OH, -NH2, -COOH or charged
moieties. This results in the desirable combination of high EWC hydrogels
with low biofouling properties, which is challenging to achieve using the
currently used contact lens materials. This is considered a significant
improvement, since biofouling is at present one of the main issues
associated with contact lens materials.6,8,11–13 In combination with the high
toughness in the hydrated state discussed previously, this makes PEG PUI
hydrogels an attractive new class of materials for contact lens applications.
179
8.4 Conclusions PEG PUI hydrogels comprise an attractive novel class of hydrogel materials
for contact lens applications, meeting all requirements on material
properties reported for the application (see Table 8.2). Owing to the specific
PEG PUI network structure, a desirable combination of high EWC, high
toughness in the hydrated state, excellent optical properties, good
biocompatibility and low biofouling properties is achieved. Also the oxygen
permeability and the chemical stability with respect to common sterilization
techniques of the PEG PUI hydrogels are suitable for the application in mind.
This particular combination of properties can be attributed to the
PEG PUI polymer network structure comprising the hydrogel networks. The
use of a PEG polymeric backbone provides the unusual combination of
hydrophilicity and chemical inertness. As a result, high EWC hydrogels that
demonstrate a low protein adsorption are obtained. The PUI type
crosslinking units provide the high toughness needed for contact lens
applications, which conventional crosslinked PEG hydrogels typically lack.
8.5 Acknowledgements This project has received funding from the European Union’s Horizon 2020
research and innovation program under the Marie Skłodowska-Curie Grant
Agreement no. 642890 (http://thelink-project.eu/) and from ZonMW under
Grant Number MKMD 114022507. The support of the analytical
departments of Covestro Deutschland AG and Currenta AG is greatly
acknowledged. Duco Schriemer and 20Med Therapeutics BV are kindly
acknowledged for supplying an Arpe-19 cell line and cell culture medium.
180
8.6 References
1. Nicolson, P. C. & Vogt, J. Soft contact lens polymers : an evolution. 22, 3273–3283 (2001).
2. Musgrave, C. S. A. & Fang, F. Contact Lens Materials: A Materials Science Perspective. Materials (Basel). 12, 261 (2019).
3. Nichols, J. J. & Fischer, D. Contact Lenses 2018. Contact Lens Spectr. 34, 18–23 (2019).
4. Moreddu, R., Vigolo, D. & Yetisen, A. K. Contact Lens Technology: From Fundamentals to Applications. Adv. Healthc. Mater. 1900368 (2019). doi:10.1002/adhm.201900368
5. Caló, E. & Khutoryanskiy, V. V. Biomedical applications of hydrogels : A review of patents and commercial products. Eur. Polym. J. 65, 252–267 (2015).
6. Zhang, W., Li, G., Lin, Y., Wang, L. & Wu, S. Preparation and characterization of protein-resistant hydrogels for soft contact lens applications via radical copolymerization involving a zwitterionic sulfobetaine comonomer. J. Biomater. Sci. Polym. Ed. 28, 1935–1949 (2017).
7. Annabi, N. et al. 25th Anniversary Article: Rational Design and Applications of Hydrogels in Regenerative Medicine. Adv. Mater. 26, 85–124 (2014).
8. Xiao, A. et al. Strategies to design antimicrobial contact lenses and contact lens cases. J. Mater. Chem. B 6, 2171–2186 (2018).
9. Elder, M. J., Stapleton, F., Evans, E. & Dart, J. K. G. Biofilm-related infections in ophthalmology. Eye 9, 102–109 (1995).
10. Dart, J. The inside story: why contact lens cases become contaminated. Contact Lens Anterior Eye 20, 113–118 (1997).
11. Thissen, H. et al. Clinical observations of biofouling on PEO coated silicone hydrogel contact lenses. Biomaterials 31, 5510–5519 (2010).
12. Lord, M. S., Stenzel, M. H., Simmons, A. & Milthorpe, B. K. The effect of charged groups on protein interactions with poly ( HEMA ) hydrogels. 27, 567–575 (2006).
13. Korogiannaki, M., Samsom, M., Schmidt, T. A. & Sheardown, H. Surface-Functionalized Model Contact Lenses with a Bioinspired Proteoglycan 4 ( PRG4 ) -Grafted Layer. ACS Appl. Mater. Interfaces 10, 30125–30136 (2018).
14. Zant, E., Bosman, M. J. & Grijpma, D. W. Combinatorial synthesis of photo-crosslinked biodegradable networks. J. Appl. Biomater. Funct. Mater. 10, 197–202 (2012).
181
15. Gong, J. P. Why are double network hydrogels so tough? Soft Matter 6, 2583–2590 (2010).
16. Sakai, T. et al. Design and fabrication of a high-strength hydrogel with ideally homogeneous network structure from tetrahedron-like macromonomers. Macromolecules 41, 5379–5384 (2008).
17. Moser, T. et al. Hydrophilic Organic Electrodes on Flexible Hydrogels. (2016). doi:10.1021/acsami.5b10831
18. Driest, P. J., Dijkstra, D. J., Stamatialis, D. & Grijpma, D. W. The Trimerization of Isocyanate-Functionalized Prepolymers : An Effective Method for Synthesizing Well-Defined Polymer Networks. Macromol. Rapid Commun. 40, 1800867 (2019).
19. Meier-Westhues, U., Danielmeier, K., Kruppa, P. & Squiller, E. P. Polyurethanes - Coatings, Adhesives and Sealants. (Vincentz-Network Verlag, 2019).
20. Driest, P. J. et al. Aliphatic isocyanurates and polyisocyanurate networks. Polym. Adv. Technol. 28, 1299–1304 (2017).
21. Kolská, Z., Valha, P., Slepička, P. & Švorčík, V. Refractometric study of systems water-poly(ethylene glycol) for preparation and characterization of Au nanoparticles dispersion. Arab. J. Chem. (2016). doi:10.1016/J.ARABJC.2016.11.006
22. Flipsen, T. A. C., Steendam, R., Pennings, A. J. & Hadziioannou, G. A novel thermoset polymer optical fiber. Adv. Mater. 8, 45–48 (1996).
23. Fabbri, P. et al. Highly stable plastic optical fibre amplifiers containing [Eu(btfa)3(MeOH)(bpeta)]: A luminophore able to drive the synthesis of polyisocyanates. Polymer (Guildf). 55, 488–494 (2014).
24. Liu, G. et al. Cytotoxicity study of polyethylene glycol derivatives. RSC Adv. 7, 18252–18259 (2017).
25. Burel, F., Poussard, L., Tabrizian, M., Merhi, Y. & Bunel, C. The influence of isocyanurate content on the bioperformance of hydrocarbon-based polyurethanes. J. Biomater. Sci. Polym. Ed. 19, 525–540 (2008).
26. Morgan, P. B. & Efron, N. The oxygen performance of contemporary hydrogel contact lenses. Contact Lens Anterior Eye 21, 2–6 (1998).
27. Ulu, A. et al. Poly(2-hydroxyethyl methacrylate)/boric acid composite hydrogel as soft contact lens material: Thermal, optical, rheological, and enhanced antibacterial properties. J. Appl. Polym. Sci. 135, 46575 (2018).
28. Seitz, M. E. et al. Influence of silicone distribution and mobility on the oxygen permeability of model silicone hydrogels. Polymer (Guildf). 118, 150–162 (2017).
182
29. Lin, H. & Freeman, B. D. Gas permeation and diffusion in cross-linked poly(ethylene glycol diacrylate). Macromolecules 39, 3568–3580 (2006).
30. Tummala, G. K., Rojas, R. & Mihranyan, A. Poly(vinyl alcohol) Hydrogels Reinforced with Nanocellulose for Ophthalmic Applications: General Characteristics and Optical Properties. J. Phys. Chem. B 120, 13094–13101 (2016).
31. Lira, M., Pereira, C., Real Oliveira, M. E. C. D. & Castanheira, E. M. S. Importance of contact lens power and thickness in oxygen transmissibility. Contact Lens Anterior Eye 38, 120–126 (2015).
32. Optical Properties of Transparent Materials: Transmission - Haze - Clarity. (BYK-Gardner Catalog 2010/2011).
33. Gonzalez-Meijome, J. M. et al. Refractive index and equilibrium water content of conventional and silicone hydrogel contact lenses. Ophthalmic and Physiological Optics 26, (Elsevier Science, 2006).
34. Childs, A. et al. Fabricating customized hydrogel contact lens. Sci. Rep. 6, 34905 (2016).
35. Varikooty, J., Keir, N., Woods, C. A. & Fonn, D. Measurement of the Refractive Index of Soft Contact Lenses During Wear. Eye Contact Lens Sci. Clin. Pract. 36, 2–5 (2010).
36. Patel, S., Marshall, J. & Fitzke, F. W. Refractive index of the human corneal epithelium and stroma. J. Refract. Surg. 11, 100–5
183
Summary and Outlook
185
Summary
This thesis describes our research on aliphatic poly(urethane-isocyanurate)
(PUI) networks, their synthesis, material properties and applications.
Generally speaking, PUI networks are considered promising materials for
biomedical applications. Specifically, the high toughness (both in the dry
state and in the water-swollen state), excellent optical properties and good
biocompatibility are highly attractive characteristics for biomedical
applications. In this thesis, first the viscosity characteristics of liquid aliphatic
isocyanurates are investigated. These are important industrial precursors for
the production of PUI networks. Next, a novel synthesis method for the
formation of well-defined PUI networks is presented. Using the newly-
developed synthesis method, various single and combinatorial PUI networks
are synthesized and characterized, demonstrating promising material
properties for biomedical use. Finally, the potential use of a specific type of
PUI networks based on poly(ethylene glycol) is explored for contact lens
applications.
Chapter 1 provides a brief description of the perspective in which the work
was performed. The scope, aims and outline of this thesis are presented.
In Chapter 2, a general introduction to isocyanate (NCO) and
polyurethane chemistry is provided. The unique reactivity of the isocyanate
functional group is discussed, as well as common reaction partners and
products. Next, various synthesis methods for the production of PUI
networks are reviewed. A general distinction can be made between the
conventional 2-component mixing approach and more recently-developed
1- and 1.5-component techniques. Finally, the general material properties
and typical applications of aliphatic PUI networks are discussed. It is shown
that PUI networks are generally appreciated for their versatility and high
toughness and durability. Owing to these properties, the largest application
field of aliphatic PUI networks is the coatings, adhesives and sealants
industry. At present, PUI networks are increasingly used for biomedical
applications and devices as well.
186
Isocyanate-functional aliphatic isocyanurates are important
industrial precursors for the 2-component synthesis of PUI networks. The
viscosity of these compounds is an important property for their production,
processing and application, but the underlying chemical interactions
dictating the viscosity of these liquids are poorly understood. In Chapter 3,
we report on the high-purity synthesis and isolation (>99%) of aliphatic
isocyanurate model compounds. The viscosities of these model compounds
are accurately quantified. By varying the carbon chain-length and comparing
isocyanate-functional isocyanurates to non-functional isocyanurates, the
effect of their molecular structure on viscosity is investigated. It is shown
that the viscosity of the liquid is significantly higher when NCO groups and
rings are simultaneously present and that viscosity increases for increasing
relative NCO- and ring content. Using density functional theory (DFT), the
presence of a strong bimolecular binding potential between the NCO group
and the isocyanurate-ring is identified for the first time. Using molecular
dynamics (MD) simulations, the viscosities of the model compounds were
accurately predicted, providing a more complete picture of the structure-
property-relations that dictate the viscosity of these liquids. The presence of
the newly-identified NCO-Ring interaction is shown to give rise to a strong
increase in viscosity.
In Chapter 4, we investigate the intermolecular interactions in
aliphatic isocyanurate liquids in more detail. We provide a first framework
of which specific intermolecular interactions are present in these liquids and
how they affect viscosity. Using DFT, we identify and quantify several
bimolecular interaction potentials involving NCO groups and isocyanurate
rings. It is shown that NCO-NCO interactions are relatively weak, with
interaction energies ranging from 5.8 to 7.9 kcal/mol. Ring-Ring interactions
are stronger, showing significant interaction energies of up to 13.8 kcal/mol.
NCO-Ring interactions are less predictable: interaction energies between 7
and 8 kcal/mol are found for single interactions while high interaction
energies increasing up to 17.6 kcal/mol are found when multiple,
cooperative interactions are possible. Using MD simulations, we show that
these multiple, cooperative NCO-Ring interactions have a strong increasing
effect on the viscosity of the liquid state. Thereby we provide a first
187
explanation for why the viscosity of aliphatic isocyanate-functional
isocyanurates is typically higher than that of non-functional isocyanurates or
other types of (poly)isocyanate oligomers.
In Chapter 5, a new synthesis method is presented for synthesizing
both hydrophilic and hydrophobic PUI polymer networks with a well-defined
network structure. The newly-developed synthesis method is based on the
functionalization of common polymeric diols using hexamethylene
diisocyanate (HDI) and their subsequent trimerization. It is shown that by
trimerizing NCO-functionalized prepolymers with a narrow molecular
weight distribution, PUI networks with a well-defined network structure can
be obtained. In contrast to conventional PUI networks, our PUI networks
demonstrate both a well-defined molecular weight between crosslinks (Mc)
and a well-defined functionality of the crosslinking units of exactly 3. These
PUI networks were shown to be flexible, tough materials both in the dry
state and in the water-swollen state.
In Chapter 6, we report on the synthesis of various combinatorial
PUI networks by the trimerization of mixtures of NCO-functionalized
prepolymers in solution. We show that combinatorial PUI networks
containing at least one hydrophilic PEG component show high water uptakes
of >100wt%. After swelling in water, the resulting hydrogels have elastic
moduli of up to 10.1 MPa, ultimate tensile strengths of up to 9.8 MPa,
elongation at break values of up to 624.0% and toughness values of up to
53.4 MJ·m-3. These values are exceptionally high and show that
combinatorial PUI hydrogels are among the toughest hydrogels reported in
the literature. Next, we show that the presence of a hydrophobic network
component significantly enhances the toughness and tensile strength of a
combinatorial PUI hydrogel in the hydrated state. This enhancement is the
largest when the hydrophobic network component is crystallizable in
nature. Finally, using dynamic scanning calorimetry, we show that PUI
hydrogels containing a crystallizable hydrophobic network component are
in fact semi-crystalline in the water-swollen state.
In Chapter 7, structure-property relations in semi-crystalline
combinatorial PUI hydrogels are investigated. Model combinatorial PUI
hydrogel networks, containing varying amounts of a crystallizable
188
hydrophobic network components, are synthesized and characterized.
These are compared to model combinatorial PUI hydrogel networks
containing varying amounts of an amorphous hydrophobic network
component. It is shown that the toughness of combinatorial PUI hydrogels
in the water-swollen state strongly increases with an increasing degree of
crystallinity of the hydrogel in the water-swollen state.
In Chapter 8, PUI networks based on poly(ethylene glycol) (PEG) are
presented as a new class of materials for contact lens applications. After
swelling in water, the PEG PUI hydrogels demonstrate high toughness values
ranging from 98 to 226 kJ·m-3 and elastic moduli ranging from 0.8 to 17.2
MPa for networks with equilibrium water contents (EWCs) ranging from 76.3
to 16.1wt%. The hydrogels demonstrate transmittance values >90% across
the visible spectrum, clarities close to 100% in most cases and refractive
indices ranging from 1.48 to 1.36. We show that these hydrogels are non-
cytotoxic, indicating good biocompatibility. Using bovine serum albumin as
a model protein, we show that PEG PUI hydrogel films demonstrate low
protein adsorption values, ranging from 0.24 mg/g to 0.65 mg/g after 24h.
These values are significantly lower than those of several commercial
contact lenses tested, which ranged from 0.55 mg/g to 1.38 mg/g after 24h.
The reported combination and range of material properties is considered
highly attractive for contact lens applications. Specifically, the contradictory
combination of high EWC, high toughness in the hydrated state and low
protein adsorption is exceptional and beneficial. These properties are
attributable to the PEG PUI network structure: the use of a PEG polymeric
backbone provides hydrophilicity and chemical inertness while the PUI type
crosslinking units provide high toughness in the hydrated state.
Outlook
Aliphatic PUI networks comprise a promising class of materials for
biomedical applications. The high toughness in both the dry state and the
water-swollen state associated with these materials is attractive for their
potential use as load-bearing biomaterials. Also, in combination with their
189
excellent optical properties, good biocompatibility and anti-biofouling
properties, PUI hydrogels are a promising class of materials for contact lens
applications. Before these materials can be applied in a biomedical context
though, further investigation and development is needed. This concerns
both industrial PUI precursors such as liquid aliphatic isocyanurates and PUI
networks and hydrogels in their final form.
Viscosity of liquid aliphatic isocyanurates
For a complete understanding of the structure-property-relations that
dictate the viscosity of liquid aliphatic isocyanurates a more detailed
experimental study is needed. Using specifically designed model
compounds, the molecular interactions and bimolecular complexes
proposed in Chapters 3 and 4 of this thesis could be experimentally
identified. For example, a detailed nuclear magnetic resonance (NMR) study
on high-purity model compounds can be considered to identify and quantify
the proposed NCO-NCO, NCO-Ring and Ring-Ring interactions
experimentally. Also, an X-ray diffraction study on crystalline small-
molecular NCO-Ring compounds can be considered to quantify the
intermolecular distances between the chemical moieties.
Regarding the computational part, the scope of the study could be
extended to include more (types of) (poly)isocyanates. Other industrially
relevant (poly)isocyanates such as methylenediphenyl diisocyanate (MDI),
toluene diisocyanate (TDI) and isophorone diisocyanate (IPDI) were not yet
included in this study and can be considered a good starting point. The
including of various other structure types is of high interest as well. For
example, allophanate, uretdione and iminooxadiazinedione type structures
are known to have significantly lower viscosities than the equivalent
isocyanurate type structures.12 A complete overview of all the relevant
intermolecular interactions present in these structures and their relative
strengths would be of great significance for the development of industrial
low-viscosity polyurethane precursors.
190
Network formation in aliphatic PUI networks
For an accurate study of PUI materials, having access to molecularly well-
defined PUI networks is important. Using the newly-developed synthesis
method reported in Chapter 5 of this thesis, the synthesis of PUI networks
with a variable, well-defined network structure is now possible. As this
synthesis method primarily provides a means for performing more accurate
structure-property-relations studies on PUI networks, an obvious next step
is to perform these studies.
An interesting example would be to investigate and quantify to what
extent the isocyanurate type crosslinking unit contributes to the mechanical
performance of PUI type networks.3 By comparing well-defined PUI
networks to PU networks containing other types of crosslinking units, the
specific contribution of the isocyanurate ring to the mechanical
performance of PUI networks can be targeted.
Aliphatic PUI networks for biomedical applications
At present, we consider biomedical materials and devices the most
promising novel application field for aliphatic PUI networks. Before these
materials can be applied in a biomedical context though, further
investigation and development of the materials is needed. In Chapter 8 of
this thesis, we already reported that a good biocompatibility of these
materials is indicated based on an indirect cytotoxicity test. This study can
be expanded to a more detailed cytotoxicity study using varying model cell-
lines and circumstances. Also, the biodegradation products should be
considered in such an experiment. Ultimately, the biocompatibility of these
materials should be investigated using an in-vivo assessment.
In Chapters 6 and 7 of this thesis, we have shown that the synthesis
of combinatorial PUI networks based on the trimerization of mixtures of
NCO-prepolymers can be used to produce exceptionally tough hydrogels
demonstrating toughness values in the range of several tens of kJ·m-3. Such
hydrogels can be applied as load-bearing biomaterials. For example, the use
of PUI type hydrogels for tissue engineering, artificial cartilage or artificial
implants can be considered. Before application in this context is possible, a
191
more detailed investigation of the physical properties of these combinatorial
PUI hydrogels is needed. An investigation of the compressive stress-strain
behaviour and the (repeated) load-bearing characteristics of these materials
can be considered. Also, the biodegradation process and rate of aliphatic PUI
networks is presently unknown, as are the biodegradation products and
their biocompatibility. A thorough in-vitro study, potentially followed by an
in-vivo and/or clinical study, should be performed before these materials
can be applied inside the human body.
Making use of the expertise and knowledge of PUI networks present
within the industrial coatings, adhesives and sealants industry, synergy can
be found with the field of biomedical materials and devices. For example,
PUI type materials can be optimized for use as (lubricious) coatings for
biomedical devices such as catheters and cannulas. The material properties
of the PUI networks reported in Chapters 6, 7 and 8 of this thesis are
promising characteristics in this respect. These include controllable water
uptake values, high toughness values (both in the dry state and in the
hydrated state), good biocompatibility, low biofouling and high stability with
respect to common sterilization techniques.
Finally, in Chapter 8 of this thesis we have demonstrated that PEG
PUI type hydrogels comprise a promising class of materials for contact lens
applications. Before commercialization in this context is possible though, the
detailed formulation of PEG PUI hydrogels should be optimized for this
specific application. This includes the optimization of the elastic modulus of
the hydrogel, optimization of the equilibrium water content and
optimization of the industrial production. Methods that are typically used to
manipulate the material properties of hydrogels may be considered, such as
the incorporation of mono-functional prepolymers and/or prepolymers with
a different chemical nature.
192
References 1. Ludewig, M., Weikard, J. & Stockel, N. Allophanate-containing
modified polyurethanes (US 2006/0205911 A1). (2006). 2. Widemann, M. et al. Structure-Property Relations in Oligomers of
Linear Aliphatic Diisocyanates. ACS Sustain. Chem. Eng. 6, 9753–9759 (2018).
3. Sasaki, N., Yokoyama, T. & Tanaka, T. Properties of isocyanurate-type crosslinked polyurethanes. J. Polym. Sci. Polym. Chem. Ed. 11, 1765–1779 (1973).
193
Acknowledgements
Dankwoord
Danksagung
195
Het tot stand komen van dit proefschrift was niet mogelijk geweest zonder
de hulp van vele mensen. Die wil ik hier graag bedanken voor hun rol, groot
of klein, in het proces.
Fistly, I would like to thank my promotors Dirk Grijpma and Dimitrios
Stamatialis. Your help and supervision are greatly acknowledged; without it
this thesis would not have existed.
Beste Dirk, heel erg bedankt voor al je hulp, advies, begeleiding,
sturing en geduld bij het tot stand komen van dit proefschrift. Als jij na het
eerste jaar niet vriendelijk doch dringend het onderzoek in de richting van
de materiaalontwikkeling had gestuurd was ik denk ik nu nog steeds
trimeren aan het synthetiseren. Ook wil ik je graag bedanken voor je
waardevolle adviezen omtrent het structureren van wetenschappelijk
onderzoek en wetenschappelijke teksten. Dat zijn inzichten om een leven
lang plezier van te hebben. Tot slot wil ik je nog een belofte doen. Ik. Zal.
Kortere. Zinnen. Maken. De niet-meer-dan-één-komma-per-zin-maatregel
heeft de leesbaarheid van dit boekje gered. Daar ben ik je dankbaar voor; en
de rest van de lezers ook.
Dear Dimitrios, thank you very much for all your help in the
structuring and fine-tuning of the manuscript. Your eye for detail and feeling
for organizing scientific text have been of great value. Also, your active
participation in conferences, colloquia and project meetings, including those
of TheLink, is much appreciated. I know that your family is not always spilling
with joy when you travel for work, but for us students your presence at these
events is very valuable. Please thank them as well for sharing you with us.
Secondly, I would like to thank the members of the graduation
committee for taking the time to read my thesis. Your effort to evaluate my
work are much appreciated.
Next I would like to thank my supervisors at Covestro, Leverkusen,
foremostly Dick Dijkstra and Frank Richter.
Beste Dick, heel erg bedankt voor de goede begeleiding en prettige
samenwerking gedurende mijn 4 jaar bij de Rheologie afdeling. Een
onuitputtelijke bron van kennis in de polymeerfysica aan het einde van de
196
gang is een luxe waar maar weinig PhD-studenten van kunnen genieten.
Bedankt voor het feit dat ik altijd kon aankloppen voor uitleg; of het nu ging
over de mechanische eigenschappen van polymeren, over wat de leukste
wijken en kroegen van Keulen zijn of over het doorgronden van de Duitse
cultuur vanuit een Nederlands perspectief.
Dear Frank, thank you very much for all your help and supervision,
both inside the lab and out. Your enthusiasm for chemistry (“Chemie muss
auch im freien Fall klappen” and “Never talk yourself out of an experiment”
are wisdoms I am not likely to forget soon) has been highly contagious and
has often motivated me to go the extra mile. Especially after having gelled
another HDI-trimerization or ruining one of your flasks again (sorry about
those)… But mostly I enjoyed our conversations that started with “Frank, do
you have any experience in working with …”, followed by some dodgy, dark
and forgotten corner of organic chemistry I had come across in an attempt
to realize one of my fantasy ideas. I was always amazed when you were able
to give me the exact state-of-the-art in that specific field from the top of
your head and direct me to the right papers and people I would need to find
my answer. Thank you for that; I am proud of the papers that we co-
authored on the chemistry of isocyanate-derivatives together.
Zunächst möchte ich mich gerne bei verschiedene Abteilungsleiter
bedanken für die persönliche Beratung und für die Unterstützung bei
meinem Projekt. Lieben Dank, dass eure Türe immer auf standen und ihr
euch immer bereit erklärt habt Zeit für mich zu finden.
Raul Pires, danke für deine immer gute Beratung und dein
Vertrauen in mich. Das hat mich stark motiviert und wird ich nicht schnell
vergessen. Dirk Achten, lieben Dank für deine positive Mentalität und gute,
kreative Ideen. Es hat mich immer Spaß gemacht bei dir vorbei zu kommen
um über Chemie (oder sonstiges ☺) zu quatschen. Péter Krüger, herzlichen
Dank für die Hilfe und Unterstützung in der Physikabteilung. Die gute
Atmosphäre innerhalb der Physik ist zum großen Teil zurückzuführen auf
deine Anerkennung des Personals. Ich (und vielen mit mich) habe mich
dementsprechend immer wohl gefühlt auf der 4. Etage in Q24. Karsten
Danielmeier, danke, für die Möglichkeit mal einen größeren internen
197
Projekt intensiv mitzumachen. Die Projektmanagement Skills die ich bei dir
abschauen konnte haben sich sehr nützlich erwiesen beim Fertigschreiben
der Doktorarbeit. ☺
Offensichtlich gibt es noch viele andere Covestro-kollegen die ich
erkenntlich bin. Lieben Dank an alle, für die super Atmosphäre, gute Ideen,
schöne Chemie, leckerer Kuchen, gute Witzen, schlechte Witzen,
Weihnachtsfeier, Sommerfeier, etc. etc. Insbesondere danke ich die
Kollegen aus der NCO-Runde, bzw. aus Q1, unter anderem: Dieter Mager,
Hans-Josef Laas, Christoph Eggert, Sebastian Doerr, Jan Suetterling, Florian
Golling, Rainer Trinks, Anne Kutz, Robert Maleika (Danke für al deine Hilfe
im Land des Patentrechts!), Heike Heckroth, Bernd Garska, Hans Grablowitz
und Serguei Kostromine. Natürlich gehören auch die Kollegen aus Q24 dazu,
unter anderem: Christoph Thiebes, Thomas Büsgen, Saskia Beuck, Andreas
Hecking, Yvonne Reimann, Daniel Raps, Walther Rauth, Mathias Matner,
Richard Meisenheimer und Joerg Tillack.
Auch vielen Dank an alle Laboranten in Q1, Q18 und Q24 die mich
die vergangene Jahren geholfen haben bei der Analyse (ich weiß, dass meine
Proben nicht immer die Einfachsten wahren… ☺). Insbesondere danke ich:
Steven Thiering für alle Hilfe im Labor Richter, Dirk Witschonke für die
vielen GPCs und GCs, Thomas Cüper, Edgar Eimermacher und Mischa
Simons für die rheologie, Brigitte Stechschulte, Kathrin Kemper und
Michaela Mallon für die Zugprüfungen, Dennis Turowski für deine
wunderschöne IR Messungen, Frank Ortmann und Sebastian Pinkert für die
thermische Messungen und die Kollegen bei Currenta in Q18 für die NMR
Messungen und Kristallstrukturanalyse. Auch war die administrative
Unterstützung von Marlis Beenen-Fuchs und Silke Köster von unschätzbare
Wert: vielen lieben Dank euch!
Of course I would also like to thank my colleagues at the University of
Twente for their input, ideas and help. Our conversations on the red couches
about anything but chemistry were always enjoyable and a nice distraction
from the scientific meetings. Ilaria Geremia and Kasja Skrzypek (I really
hope that that is spelled correctly… ☺), thanks for all the fun times that we
had together and for a great Christmas dinner. My visits to Twente were
198
always worth it when you guys were around. ☺ Also, thanks to Natalia
Chevtchik, Denys Pavlenko, Iris Allijn, Duco Schriemer and the rest of the
BST-group for the good times.
The same applies to my partners-in-crime within Covestro: the
group of “Covestro Doktoranden”. Thomas Prenveille, thanks for the good
talks and the bad jokes (yours mostly of course) and for several
(un)forgettable Covestro Weihnachtsfeier. Veronika Eilermann, es hat mich
sehr viel Spaß gemacht mit dir zusammen an dem PU-Buch zu Arbeiten.
Dank dafür und viel Erfolg in deinem Projekt. Anne Hansen, lieben Dank für
deine Interesse in meinem Projekt und deine Humor. Du hast gute Ideen
über Chemie und ich wünsche auch dich viel Erfolg bei deinem Projekt.
Also, many thanks to my fellow-PhD students in the consortium of
TheLink. I’ve really enjoyed our biannual meetings; both the project dinners
and the scientific progress meetings.
Wenn man sich mit ca. 20 verschwitzten deutsche Männer in eine Umkleide
antrifft, ist normalerweise irgendwas ganz schief gelaufen. Aber die 2.
Herren der RTHC Bayer Leverkusen sind darauf der Ausnahme: ich hätte
mich kein bessere verschwitzten deutsche Männer für sowas aussuchen
können! Ich danke euch für alle nette, sportliche, lustige, intensive,
lehrreiche, mit Kölsch angereicherte und immer angenehme
Hockeymomente, sowohl auf dem Platz als daneben. Es hat mich sehr viel
Spaß gemacht, euch die vergangene Jahren mal ein bisschen richtiges
Hockey beizubringen! ;)
Lieber Heike Heuser, Mischa Simons und Edgar Eimermacher, vielen Dank
für eure tägliche gute Laune und die angenehme Atmosphäre im Raum 424.
Ich hab mich vom Anfang bis Ende des Projekts bei euch im Büro zuhause
gefühlt und die vier Jahr sind vorbei geflogen. Es hat mich sehr viel Spaß
gemacht das Büro mit euch zu teilen und ich bedanke mich für die viele
angenehme Momente zusammen.
Das gleiche gilt natürlich für die PEAChT-bande: Tanja Hebestreit,
Andreas Hebestreit, Christoph Moethe und Edgar Eimermacher. Lieber
Tanja, vielen Dank für das hochhalten des Niveaus wenn es nötig war und
199
für die leckere Kaffee die wir immer bei dir genießen konnten (natürlich viel
leckerer als die in H4..!). Lieber Chris, vielen Dank für deine immer gute
Laune und für al deine Hilfe im Labor. Lieber Andreas, Tanja habe ich
bedankt für das Hochhalten des Niveaus wenn es nötig war; dich danke ich
für das Absenken wenn es nötig war! Das gemeinsame Essen in H4 und die
Kaffeekränzchen in Café Ratz haben mich immer viel Spaß gemacht. Ich wird
niemals mehr eine Zitronenkuche essen, ohne an dich zu denken.
Zum Schluss: für einen erfolgreichen Abschluss einer Doktorarbeit
ist neben der Doktorvater noch eine Person sehr wichtig: der Bürovater. Ich
bin sehr froh, dass ich es mit meine Bürovater nicht besser treffen haben
konnte. Lieber Edgar, vielen lieben Dank für deine Freundschaft und deine
Hilfe während die vergangene vier Jahren. Von alle Höhepunkte bei Covestro
waren die tägliche Höhepunkte in der Kantine von H4 doch immer noch die
Beste. Same table. Same time. Same jokes. Unter dem Genuss von den
typisch deutschen Nudeln mit Spinat und Fisch hast du mich mit viel Humor
die deutsche (zum Teil Eifeler?) Sprache und Kultur beigebracht und dafür
bin ich dich sehr dankbar. Insbesondere war dein Unterricht zum Thema
Flexibilität und Pünktlichkeit (Überpünktlichkeit?) beim Essen sehr lehrreich.
Auch das Bescheid Geben, dass man im Bayer-gefängnis landet, wenn man
zwei Schnitzels unter dem Reis versteckt, hat mich viele Monate Zeit
gespart. Also, mach es gut, viel Erfolg bei deinem letzten Jahr schuften (aber
lass bitte auch noch ein paar Proben für Mischa übrig, du hast mich selbst
gesagt: „es ist kein Kameradschaft, wenn nur den Kamerad schafft!“). Und
nachdem: genieß deine wohlverdiente Rente. Wenn du nochmal unterwegs
bist von oder nach Vrouwenpolder, schau dann gerne mal in Utrecht vorbei.
Für dich und deins steht der Tür immer auf und ein leckeres Hertog-Jan
bereit.
Ook al speelde het uitvoerende deel van dit onderzoek zich vooral in het
buitenland af, dit boekje had er niet kunnen liggen zonder de afleiding en
ondersteuning van vrienden vanuit Nederland. Velen van jullie hebben in de
afgelopen jaren de moeite genomen om mij (later ons) in Keulen op te
komen zoeken en dat waardeer ik enorm. De schnitzels en de braadworsten
kwamen me op een gegeven moment wel bijna de neus uit als we wéér een
200
weekendje Duitse tour van Brauhaus naar Brauhaus op de planning hadden.
Maar aangezien ze bij een enkeling van jullie letterlijk weer de neus uit
kwamen (ik zal geen namen noemen; die worst zal misschien niet helemaal
gaar geweest zijn denk ik. Of was het toch de Kölsch?) zal ik zeker niet klagen.
We hebben in ieder geval een hoop pret gemaakt in de weekenden dat we
bezoek hadden en ik bedank eenieder die langs is geweest voor het afleggen
van de trip.
Daarnaast dank aan mijn (oud)huisgenoten van DKP (o.a. Michiel
Marck, Sander Sinninghe Damsté; Marco van der Schoot; Harmen Booij,
Martijn Bekke; Jan-Wigbold Sluiter, Antonie van Eijk, Philip van Heemstra;
Jan-Willem de Bakker, Duco Klynstra; Bob ter Haar, Derk-Willem Duynstee;
Marijn Gelders) voor de geslaagde OHD’s, Oktoberdiners etc. Dank aan mijn
“vriendjes van thuis-thuis”: de C4 (Joris Wielders, Rogier Gundlach en Leon
Unk) voor de langlopende vriendschap. Dank aan mijn clubgenoten (Aus,
Arwin, Benni, Bob, Dré, Eijsvogel, Haas, Hans, Harmen, Jap, Jip, Joep, JW,
Langenberg, Michiel en Sloof) voor de vele huwelijken / feestjes /
housewarmings / Lumstrumthemakleurenbekendmakingsborrels en andere
belangrijke evenementen die de moeite waard waren om een weekend voor
op en neer te komen. Hetzelfde geldt natuurlijk voor vrienden van de ON
Hoogh / Mannendag (Ischa, Maurits, Pieter DJ, Kluis, Remmelt, Pang, Prick
en Ap). Bedankt voor alle activiteiten en mooie weekenden, met als kers op
de taart de Lumstrumeditie van de Mannendag op Chateau de la Noue de
Coq. Ik kijk nu al weer uit naar de volgende 10 jaar touwtrekken, vuur
maken, bijlen gooien, ver-plassen, ad-estafettes, …, …, etc., etc., etc.
Speciale dank gaat natuurlijk uit naar mijn paranimfen, Jeroen van Dorp en
Benjamin Thijs. Dit is echter het enige bedankje in dit dankwoord dat onder
voorbehoud geldt; eerst nog even zien hoe die verdediging straks loopt.
Maar dat komt vast goed, en natuurlijk bedank ik jullie wel graag nu al voor
jullie rol in de aanloop daar naar toe.
Jeroen, wij begonnen ons gezamenlijke studie-avontuur meteen
Samen op het Matje: we hadden in de eerste week van onze studie een of
ander natuurkunde-practicum gemist en mochten ons direct melden bij
Prof. Knoester voor uitleg. Al snel bleek dat wij met z’n tweeën ook de
201
voltallige intersectie vormden tussen onze studentenvereniging en de studie
scheikunde (wiskundig ook wel: VAP ∩ SK), wat toch een extra band schept.
Ik vind het erg leuk dat we door de jaren heen elkaars scheikundige
avonturen zijn blijven volgen en ik hoop dat we dat in de toekomst ook zo
blijven doen. Soortgenoten met dezelfde dubbele interesse voor sociale
interactie én chemie zijn per slot van rekening zeldzaam en ik ben blij dat we
elkaar daar in vinden.
Bennie, a.k.a. het Orakel van Groningen, de man die het antwoord
weet op alle vragen. Geen slechte eigenschap voor een paranimf. Maar dat
is natuurlijk niet de (enige) reden dat ik je aan mijn zijde gevraagd heb.
Tijdens onze vele gesprekken aan de bar in de Pintelier in Groningen was het
me een genoegen om samen met jou de ins & outs (en de tergend langzame
voortgang) van onze beide scripties te bespreken. Ik kijk met veel plezier
terug op onze discussies, waarin we samen de grenzen van de wetenschap
en van de bierkaart hebben verkend. Inmiddels liggen er nieuwe spannende
wetenschappelijke projecten in het verschiet en ik kijk er naar uit om ook
die samen te fileren onder het genot van een LaChouffe-Delirium-combo.
Lieve familie, ook al was ik de afgelopen jaren fysiek verder weg, jullie zijn
voor mij dichterbij geweest dan ooit. In mijn eentje in een vreemde stad met
een andere taal en een andere cultuur om me heen was het zeker niet altijd
makkelijk en op die momenten heb ik jullie steun gevoeld. Dankjewel dat
jullie zo lief hebben meegeleefd, meegedacht en meegevoeld.
Ik begon dit kopje aanvankelijk met “Lieve (schoon)familie”, maar ik
realiseerde me dat, zeker nu June er is, ik jullie met recht ook gewoon familie
mag noemen. Lieve Jacob en Sietske, dankjewel voor jullie gastvrijheid en
de warmte waarmee we altijd bij jullie thuis worden ontvangen. Ik voel me
inmiddels aan de Texellaan bijna net zo thuis als bij mijn eigen ouders. Dank
natuurlijk ook aan Pieter, Jelte, Haitske, Gesan en Dennis voor de vele
gezellige spelletjesmiddagen, Sinterklazen, kersten, oud-en-nieuws, etc.
Juul en Marthe, bedankt voor jullie gezelligheid, lekkere eten en alle
keren dat we bij jullie op de bank mochten slapen in Amsterdam als we
vanuit Keulen op bezoek waren! Mart, bovendien bedankt voor je gute
Laune en dito grappen (ook als de rest daar soms anders over denkt; niet
202
iedereen snapt wat humoor is). Bedankt, für die Blumen. Bedankt voor je
volledige vertrouwen in mij/ons, altijd en overal. Bedankt voor je interieur-
en mode-advies; gevraagd en ongevraagd. ☺ Bedankt voor al je support. Ik
zou me geen beter zusje kunnen wensen; en June zich geen betere tante.
Lieve Mars, dankjewel dat je altijd klaar staat. Je hebt altijd tijd om
even te bellen, ook wanneer het duidelijk is dat je het zelf druk hebt. Je bent
er altijd om even iets te overleggen. We kunnen altijd bij je terecht voor
advies, dat ook nog eens altijd goed is (jij zou consultant moeten worden ☺).
Dit boekje was nooit afgekomen zonder jouw suggesties over hoe ik het
project beter kon structureren in de beginfase. Dankjewel voor al het
meedenken, meekijken, nalezen en suggesties geven: altijd opbouwend, op
basis van en met ruimte voor eigen inbreng, maar altijd met verbetering tot
gevolg. The best kind.
Lieve pap, dankjewel dat je altijd de goede dingen weet te zeggen.
Wanneer ik worstel met het verwoorden van een ervaring of een gevoel,
weet jij soms in een paar zinnen te vangen wat ik bedoel. Dat heeft al een
hoop zaken in perspectief geplaatst en het helpt altijd bij het maken van
keuzes, of het reflecteren daarop. De gezellige (en lekkere!) etentjes waarbij
dit meestal plaatsvindt maken het plaatje compleet. Laten we er daar nog
veel van doen.
Lieve mam, dankjewel dat je altijd trots bent. En wat zou opa het
ook leuk gevonden hebben; die kijkt vast ergens aandachtig mee. Dankjewel
dat je altijd zo enthousiast bent over de dingen die wij doen en ons daarin
ondersteunt. De energie is aanstekelijk en motiveert om nog een stapje
extra te doen (daarin geef je natuurlijk zelf ook het goede voorbeeld). Lekker
ondernemen, geloven in en handelen vanuit eigen kracht, het ijzer smeden
wanneer het heet is en hoezo genoegen nemen met óf-óf zolang er
misschien ook nog én-én in het vat zit?? Het zijn credo’s om naar te leven en
ze zijn de output en kwaliteit van dit boekje zeker ten goede gekomen.
Lieve schat. Ik zou een tweede boek kunnen schrijven als ik alles zou
opschrijven dat ik je graag wil zeggen. Dat kan niet, maar ik geloof dat je het
meeste - de belangrijke dingen - ook onuitgesproken al wel weet. Toch zijn
er ook een paar dingen waar ik je graag nog expliciet voor wil bedanken.
203
Bedankt voor al je support de afgelopen jaren. Het was geen
makkelijke keuze om ineens voor vier jaar naar Duitsland te verhuizen en ik
heb me, zeker in het begin, regelmatig afgevraagd of ik er wel goed aan had
gedaan. Toen we na het eerste jaar merkten dat er door de afstand toch ook
wat druk op de relatie kwam te staan was het besluit heel dichtbij om ermee
te stoppen en terug te komen naar Nederland. Onze relatie was een offer
dat ik niet wilde maken, hoe leuk het werk ook, en daar sta ik nog steeds
volledig achter. Maar dat ik zou stoppen wilde jij niet, en dat jij toen besloten
hebt om jouw leven in Utrecht op pauze te zetten en voor mij naar Duitsland
te verhuizen is het liefste dat iemand ooit voor mij gedaan heeft. Zonder dat
besluit was dit boekje (dit hele project) er niet geweest. Dankjewel.
Vanaf dat jij naar Duitsland kwam werd alles leuker en ik heb enorm
genoten van al onze avonturen. Van het stapelbedje van Felix tot de Breite
Strasse (“I’m not famous around here..!”) tot ons fijne huis aan de
Rothgerberbach, wat hebben we een hoop pret gemaakt samen. Lekkere
avondjes borrelen thuis (meer oog voor de snacks dan voor de ring die ook
nog op tafel stond ☺), weekendjes weg, heel veel spellen spelen, lunchen bij
het graantentje, een kast bouwen, een prachtige vakantie naar Afrika, alles
is leuk zolang we het samen doen. Van alle mensen die ik ken kan ik met jou
het hardste lachen en ik ben superblij dat wij bij elkaar horen.
Het grote Duitsland-avontuur is inmiddels achter de rug, maar het
volgende is alweer begonnen. Lieve kleine June, wat fijn dat jij er bent; weet
dat er veel van je gehouden wordt.
Piet Driest
Januari 2020
204
List of Publications
• P.J. Driest, V. Lenzi, L.S.A. Marques, M.M.D. Ramos, D.J. Dijkstra, F.U.
Richter, D. Stamatialis and D. W. Grijpma; Aliphatic isocyanurates
and polyisocyanurate networks; Polym. Adv. Technol. 28 (2017),
1299-1304
• M. Widemann, P.J. Driest, P. Orecchia, F. Naline, F.E. Golling, A.
Hecking, C. Eggert, R. Pires, K. Danielmeier and F.U. Richter;
Structure−Property Relations in Oligomers of Linear Aliphatic
Diisocyanates; ACS Sustainable Chem. Eng. 6 (2018), 9753-9759
• C. Thiebes, (…) and P.J. Driest; Wasserfrei Härtende Klebstoffe auf
Polyisocyanatbasis, submitted (2018), EP18201522.2
• P.J. Driest, D.J. Dijkstra, D. Stamatialis and D.W. Grijpma; The
Trimerization of NCO-functionalized Prepolymers: an Effective
Method for synthesizing Well-defined Polymer Networks;
Macromol. Rapid Commun. 40 (2019), 1800867
• V. Lenzi, P.J. Driest, D.J. Dijkstra, M.M.D. Ramos and L.S.A. Marques;
GAFF-IC: Realistic Viscosities for Isocyanate Molecules with a GAFF-
Based Force Field; Mol. Simulat. 45 (2019), 207-214
• E. Avtomonov, K. Danielmeier, P.J. Driest et al.; Chemical Principles;
Chapter 3 in: Polyurethanes – Coatings, Adhesives and Sealants (U.
Meier-Westhues et al.); Vincentz-network, Hannover, 2019
• P.J. Driest, F.E. Golling, I.C. Latorre-Martinez, R. Pires and D. Achten;
Crosslinking technologies based on the polyaddition reactions of
polyisocyanates; Chapter 4.6 in: Polyurethanes – Coatings,
Adhesives and Sealants (U. Meier-Westhues et al.); Vincentz-
network, Hannover, 2019
205
• V. Lenzi, P.J. Driest, D.J. Dijkstra, M.M.D. Ramos and L.S.A. Marques;
Investigation on the Intermolecular Interactions in Aliphatic
Isocyanurate Liquids: Revealing the Importance of Dispersion; J.
Mol. Liq. 280 (2019), 25-33
• P.J. Driest, D. Kleinschmidt, R. Meisenheimer and D. Achten;
Storage-stable Thermolatent Catalysts for the Polymerization of
Isocyanates; “Erfindungsmeldung“ submitted (2019)
• P.J. Driest, V. Lenzi, L.S.A. Marques, M.M.D. Ramos, F.U. Richter and
D.J. Dijkstra; The Influence of Molecular Interactions on the Viscosity
of Aliphatic Polyisocyanate Precursors for Polyurethane Coatings;
ECS Conference Proceedings (2019)
• P.J. Driest, I.E. Allijn, D.J. Dijkstra, D. Stamatialis and D.W. Grijpma;
Poly(ethylene glycol) Poly(urethane-isocyanurate) Hydrogels for
Contact Lens Applications; Polym. Int. (in press), DOI:
10.1002/pi.5938
(To be) submitted
• P.J. Driest, D.J. Dijkstra, D. Stamatialis and D.W. Grijpma; Tough
Combinatorial Poly(urethane-isocyanurate) Polymer Networks and
Hydrogels Synthesized by the Trimerization of Mixtures of NCO-
Prepolymers; resubmitted
• P.J. Driest, D.J. Dijkstra, D. Stamatialis and D.W. Grijpma; Structure-
Property Relations in Semi-crystalline Combinatorial PUI Hydrogels;
to be submitted