The Effects of Silica Nanoparticles in Toughened Epoxy Resins and Fiber-Reinforced Composites

Sprenger
The Effects of Silica Nanoparticles in Toughened

Epoxy Resins and Fiber-Reinforced Composites
Stephan Sprenger
The Effects of Silica

Nanoparticles in Toughened
Epoxy Resins and Fiber-
Reinforced Composites
Hanser Publishers, Munich Hanser Publications, Cincinnati

The Author:
Dr. Dr. Stephan Sprenger, Evonik Hanse GmbH, Charlottenburger Straße 9, 21502 Geesthacht, Germany
Distributed in North and South America by:

Hanser Publications
6915 Valley Avenue, Cincinnati, Ohio 45244-3029, USA
Fax: (513) 527-8801
Phone: (513) 527-8977
www.hanserpublications.com
Distributed in all other countries by
Carl Hanser Verlag
Postfach 86 04 20, 81631 München, Germany
Fax: +49 (89) 98 48 09
www.hanser-fachbuch.de
The use of general descriptive names, trademarks, etc., in this publication, even if the former are not especially identified, is
not to be taken as a sign that such names, as understood by the Trade Marks and Merchandise Marks Act, may accordingly be
used freely by anyone. While the advice and information in this book are believed to be true and accurate at the date of going
to press, neither the authors nor the editors nor the publisher can accept any legal responsibility for any errors or omissions
that may be made. The publisher makes no warranty, express or implied, with respect to the material contained herein.
The final determination of the suitability of any information for the use contemplated for a given application remains the
sole responsibility of the user.
Cataloging-in-Publication Data is on file with the Library of Congress
ISBN 978-1-56990-627-9
E-Book ISBN 978-1-56990-628-6
All rights reserved. No part of this book may be reproduced or transmitted in any form or by any means, electronic or me-
chanical, including photocopying or by any information storage and retrieval system, without permission in writing from the
publisher.
© 2016 Carl Hanser Verlag, Munich

Coverdesign: Stephan Rönigk
Printed and bound by BoD - Books on Demand, Norderstedt
Printed in Germany
Table of contents:
Short summary 6
Kurzzusammenfassung 7
1. Introduction 8
1.1. Motivation 8
1.1.1. References 8
1.2. Objectives and structure 9
2. State-of-the-art of the science and technology 10
2.1. Manufacturing technologies for fiber-reinforced composites 10
2.1.1. Prepreg methods 11
2.1.2. Filament winding 12
2.1.3. Pultrusion 13
2.1.4. Injection processes via a mold 13
2.1.5. Infusion processes via a vacuum bag 15
2.1.6. References 16
2.2. Fiber reinforcements for epoxy resin systems 17
2.2.1. Woven fabrics 17
2.2.2. Unidirectional fabrics 18
2.2.3. Multiaxial fabrics 19
2.2.4. Preforms 19
2.2.5. Fiber sizing 20
1
2.3. Epoxy resin systems for fiber-reinforced composites 22
2.3.1. Epoxy resins 22
2.3.2. Curing agents for epoxy resins 24
2.3.2.1. Amine curing agents 24
2.3.2.2. Anhydride curing agents 24
2.3.3. Epoxy resin systems 26
2.3.4 References 27
2.4. Toughening of epoxy resins 28
2.4.1. Differentiation between toughening and increasing

flexibility 28
2.4.2. Reactive liquid rubbers 28
2.4.3. Core-shell rubber particles 30
2.4.4. Thermoplastic particles 31
2.4.5. Self-organizing block copolymers 32
2.5. Modification of epoxy resins with SiO2 nanoparticles 35
2.5.1. SiO2 nanoparticles 35
2.5.2. Property improvements of cured bulk resin

systems 38
2.5.3. Property improvements of cured bulk

hybrid resin systems 39
2.5.4. Property improvements of fiber-reinforced

composites 41
3. Results and Discussion 48
3.1. Property improvements of epoxy resins modified with SiO2

nanoparticles 48
2
3.1.1. Abstract 48
3.1.2. Introduction 48
3.1.3. Discussion 51
3.1.3.1. Amine-cured epoxy resins 51
3.1.3.2. Anhydride-cured epoxy resins 56
3.1.3.3. Comparison between different types of hardeners 60
3.1.4. Conclusions 62
3.2. Property improvements of epoxy resins modified with SiO2

nanoparticles and elastomers (hybrid systems) 66
3.2.1. Abstract 66
3.2.3.1. Epoxy resins modified with reactive liquid rubbers

(CTBNs) and silica nanoparticles, amine cured 69
3.2.3.2. Epoxy resins modified with styrene-butadiene rubber

(SBR) and silica nanoparticles, amine cured 72
3.2.3.3. Epoxy resins modified with reactive liquid rubbers

(CTBNs) and silica nanoparticles, anhydride cured 72
3.2.3.4. Epoxy resins modified with silica nanoparticles, cured

with amines and amino-functional reactive liquid
rubbers (ATBNs) 73
3.2.3.5. Epoxy resins modified with core-shell elastomers

(CSRs) and silica nanoparticles, amine cured 75
3.2.3.6. Epoxy resins modified with core-shell elastomers

(CSRs) and silica nanoparticles, anhydride cured 77
3.2.3.7. Short overview of improvements achieved 79
3.2.3.8. Synergy or no synergy? 80
3
3.3. Property improvements of fiber-reinforced composites based on

epoxy resins modified with SiO2 nanoparticles 86
3.3.1. Abstract 86
3.3.3.1. Glass fiber-reinforced epoxy resin composites 90
3.3.3.2. Carbon fiber-reinforced epoxy resin composites 94
3.3.3.3. Comparision between bulk and laminate property

improvements 98
3.3.3.4. Mechanisms of toughening 99
3.3.3.5. How much silica nanoparticles do you really

need – and where? 100
3.4. Property improvements of fiber-reinforced composites based on

epoxy resins modified with SiO2 nanoparticles and elastomers
(hybrid systems) 106
3.4.1. Abstract 106
3.4.3.1. Glass fiber-reinforced epoxy resin composites 111
3.4.3.2. Carbon fiber-reinforced epoxy resin composites 115
3.4.3.3. Mechanisms for property improvements of hybrid

epoxy resin systems 118
3.4.3.4. The transfer of improved bulk resin properties into

the fiber-reinforced composite 119
4
3.5. Carbon fiber-reinforced composites with epoxy resin modified with
reactive liquid rubber and SiO2 nanoparticles 128
3.5.1. Abstract 128
3.5.3. Experimental 132
3.5.3.1. Materials 132
3.5.3.2. Bulk sample preparation 133
3.5.3.3. Laminate preparation 133
3.5.3.4. Bulk sample testing 133
3.5.3.5. Laminate testing 134
3.5.4. Results and discussion 135
3.5.4.1. Bulk resin properties 135
3.5.4.2. Laminate properties 139
3.5.4.3. Microscopical investigations 146
3.5.4.4. Formation of nanosilica agglomerates 149
3.5.5. Conclusions and outlook 153
4. Summary and Outlook 157
4.1. Summary and outlook 157
4.2. Zusammenfassung und Ausblick 159
5. Bibliography 163
5
Short Summary
The properties of cured epoxy resins can be improved significantly by the

addition of surface-modified SiO2 nanoparticles. A linear relationship between the
increase of modulus respectively the fracture toughness and the addition level of
nanosilica exists for most epoxy curing agents. Compressive strength and fatigue
performance can be improved as well.
Combining this modification with the classic toughening concept using reactive
liquid rubbers or core-shell elastomers leads to so-called hybrid systems, which
are characterized by both high toughness and high stiffness. The fatigue
performance is improved further.
Laminates manufactured by using these modified resins exhibit improved

performance as well. Regardless if glass or carbon fibers are used as reinforcing
material, the relative property improvements of the laminates are much smaller.
A linear relationship between the percentual increase in fracture toughness (GIc)
of the cured bulk resin systems and the percentual increase of the GIc of the
laminates made thereof seems to exist; with a conversion factor of 0.18. If the
fracture toughness of a fiber-reinforced part shall be increased by 100 %, then
instead of the resin used a hybrid resin with a 555 % higher bulk GIc needs to be
used. For the fatigue performance of laminates made from hybrid resins a tenfold
increase in cyclic loadings upon failure can be achieved.
In own trials a fast curing epoxy resins system based on DGEBA/IPD/TMD was
employed to manufacture carbon fiber reinforced laminates; whose properties
were investigated. The cured bulk resin systems exhibit the expected property
improvements for the hybrid systems. However, the laminates based on hybrid
resin systems, modified with both reactive liquid rubber and SiO2 nanoparticles,
show no further improvements compared to the rubber-only modification but
rather slightly lower values for GIc and GIIc. The ILSS is comparable; the residual
strength after impact reduced.
The agglomerates of silica nanoparticles, which were discovered, might be a

potential cause. They form during the fast cure in presence of reactive liquid
rubber; probably caused by the forced rapid phase separation of the rubber upon
cure. Future research will provide further insights.
In summary it can be stated that laminates made from hybrid epoxy resins are
tough and stiff and exhibit improved compressive strength as well as excellent
fatigue performance. This makes them especially suitable for highly stressed
composites parts like in automotive applications.
6
Kurzzusammenfassung
Die Eigenschaften von gehärteten Epoxidharzen können durch den Zusatz von
oberflächenmodifizierten SiO2-Nanopartikeln deutlich verbessert werden. Es
besteht für die meisten Härter ein annähernd linearer Zusammenhang zwischen
der Steigerung des Moduls bzw. der Bruchzähigkeit und der Zusatzmenge an
Nanopartikeln. Druckbeständigkeit und Ermüdungsverhalten können ebenfalls
verbessert werden.
Kombiniert man diese Modifikation mit der klassischen Schlagzähmodifizierung

unter Verwendung von reaktiven Flüssigkautschuken oder Core-Shell-
Elastomeren, so erhält man sogenannte Hybridsysteme, welche sich durch eine
besonders hohe Zähigkeit bei gleichzeitiger Steifigkeit auszeichnen. Auch das
Ermüdungsverhalten wird weiter verbessert.
Beim Einsatz dieser modifizierten Harze zur Herstellung von Laminaten können
an diesen ebenfalls Eigenschaftsverbesserungen beobachtet werden. Unabhängig
davon, ob Glas- oder Carbonfasern als Verstärkungsmaterial eingesetzt werden,
fallen diese Verbesserungen prozentual deutlich geringer aus. Es scheint eine
lineare Beziehung zwischen den prozentualen Steigerungen der Bruchzähigkeit
(GIc) der gehärteten Reinharzhybridsysteme und den daraus hergestellten
Laminaten zu existieren; mit einem Übertragungsfaktor von 0,18. Soll ein
Faserverbundbauteil um 100 % in der Zähigkeit gesteigert werden, so muss statt
des bisher verwendeten Harzsystems ein Hybridsystem mit einem 555 %
höheren GIc eingesetzt werden. Beim Ermüdungsverhalten der Laminate auf
Hybridharzbasis kann eine zehnfache Steigerung der zyklischen Belastung bis
zum Versagen erzielt werden.
In den eigenen Versuchen wurden unter Verwendung eines schnellhärtenden

Epoxidharzsystems basierend auf DGEBA/IPD/TMD kohlefaserverstärkte
Laminate mittels RTM hergestellt und untersucht. Die gehärteten
Reinharzsysteme zeigen für die Hybride die erwarteten Eigenschafts-
verbesserungen. Die Laminate auf Basis Hybridharz, sowohl mit reaktivem
Flüssigkautschuk als auch mit SiO2-Nanopartikeln modifiziert, weisen jedoch
keine weitere Verbesserung der Bruchzähigkeit gegenüber der reinen
Kautschukmodifikation auf, sondern eher geringfügig niedrigere Werte für GIc
und GIIc. Die ILSS ist vergleichbar; die Restdruckfestigkeit nach Impact jedoch
verringert.
Eine mögliche Ursache sind die entdeckten Agglomerate der SiO2-Nanopartikel.

Sie entstehen während der raschen Vernetzung in Gegenwart von reaktivem
Flüssigkautschuk; wahrscheinlich verursacht durch die erzwungene schnelle
Phasentrennung des Kautschuks während der Härtung. Zukünftige Forschungs-
arbeiten werden hier weitere Erkenntnisse bringen.
Zusammenfassend kann festgestellt werden, daß Laminate auf Basis von

Hybridharzen gleichzeitig zäh und steif sind und eine verbesserte
Druckbeständigkeit sowie ein hervorragendes Ermüdungsverhalten aufweisen.
Sie sind daher besonders für hochbelastete Bauteile wie beispielsweise in
automobilen Anwendungen geeignet.
7
1. Introduction
Lightweight construction is one of the key technologies to master the challenges

of the 21st century. Be it to enable mankind to exploit renewable energy (just
think about rotor blades for wind energy generators) or to reduce consumption of
fossil energies in transportation by air, land or sea. In lightweight construction
fiber-reinforced composites (abbreviated to "composites" for short) play a very
prominent role. These new materials experienced, and are still experiencing, an
extremely rapid development. After the initial developments in aerospace
applications, and later rotor blade construction, it is now time to think about the
extensive use of such composites in automotive construction. The political
pressure on European car manufacturers to offer cars low in fuel consumption
and low in exhaust emissions is constantly increasing [1.1]. Thus, fiber-
reinforced composites offer here a tremendous potential for weight reduction.
The optimization of the manufacturing process of composites (see chapter 2.1.),
the identification of the best design and the types of fiber reinforcement are of
great importance. However, the optimization of the matrix materials is also a
very important topic of international research.
1.1. Motivation
Silica nanoparticles have been available as concentrates in epoxy resins in

industrial quantities for more than 10 years now. They are used to improve the
mechanical properties of epoxy resins and fiber-reinforced composites made from
these modified epoxy resins. They do increase the toughness of epoxy resins,
however not to the same extent as traditional tougheners like reactive liquid
rubbers or core shell rubber particles do. Nevertheless, when combining the
nanosilica modification with a toughening of the epoxy resin there seem often to
exist synergistic effects. After many own research work in this area, many papers
published, and common research projects with universities and institutes the
need arose to summarize and integrate the many results. Other research groups,
active in similar or related fields, have published often similar results but
sometimes contradictory results as well. Unfortunately very often either bulk
resin systems or fiber reinforced composites have been investigated - but not
both. Also, the many individual results based on totally different resin/fiber-
combinations made the preparation of an overview mandatory.
Of course the question needs to be asked, if fundamental laws or correlations

between the modifications and property improvements exist. Especially with
regard to a potential synergy between rubber-modification of the epoxy resin and
the addition of silica nanoparticles. The same question is also valid for the
transfer of property improvements of the bulk resin into the fiber reinforced
composite as a function of the different modifications and fiber reinforcements.
To have the possibility of predicting the composite performance from bulk resin
performance would be also extremely helpful for industrial formulators, who can
then validate the predicted results from their own test programs on the
composites.
1.1.1. References
[1.1] Regulation (EC) No 443/2009 OF THE EUROPEAN PARLIAMENT AND

OF THE COUNCIL; Official Journal of the EU, 05.06.2009, L 140/1
8
1.2. Objectives and structure
It is the intention of this work to provide an overview of the actual state-of-the-

art of the research regarding the mode of action of SiO2 nanoparticles in modified
epoxy resins and fiber-reinforced composites made thereof, to evaluate the
different results and, if possible, to identify fundamental correlations. One
example would be the potential existence of a relationship between the
properties of an optimized resin matrix and the properties of the resulting fiber-
reinforced composite.
First the modification of epoxy resins with SiO2 nanoparticles regarding the
achievable property improvements of cured bulk resins needs to be investigated
– this will be the subject of chapter 3.1. Secondly, the so-called hybrid epoxy
resins, that is epoxy resins modified with both a rubbery (i.e. elastomeric)
toughener and silica nanoparticles, will be evaluated with a focus on mechanical
properties in chapter 3.2. Thirdly, the use of nanosilica-modified epoxy resins in
fiber-reinforced composites based on glass or carbon fibers is the subject of
further investigations in chapter 3.3. One of the essential questions here is
whether property improvements found for modified bulk epoxy resins will be
found for fiber-reinforced composites as well. Fourthly, of course the same
question is even more important for fiber-reinforced composites made from
hybrid epoxy resins which will be investigated in chapter 3.4.
Finally, the insights obtained from the above studies will be verified by
investigating a bulk epoxy resin system and a composite made by using a
common resin system of major industrial importance in chapter 3.5. A carbon
fiber-reinforced system based on a fast-curing epoxy resin system, as typically
used for manufacturing automotive parts, was selected to perform this task.
To obtain a clear structure of this work it is divided in these five work packages
(chapters 3.1. - 3.5.). To ensure a correct interpretation of the many different
results from many individual sources as well as the results of my own work the
single work packages have been published up front in renowned scientific
journals.
As each work package has is a separate chapter and coherent in itself, the
necessity arose to use graphs or pictures several times. Examples are the SEM
picture of core shell elastomers used in chapters 2.4.3. and 3.2., or the SANS
curve for the particle size distribution presented in chapters 2.5.1. and 3.3., or
the TEM picture of the silica nanoparticles shown in chapters 2.5.1. and 3.2. This
recurrence is necessary to give a clear and readable structure. In the
bibliography the cross-references are given.
9
2. State-of-the-art of the science and technology
2.1. Manufacturing technologies for fiber-reinforced

composites
The manufacturing of fiber-reinforced composites is based on quite a few very

different processes [2.1.1], [2.1.2]. An excellent overview on actual technologies
especially used for epoxy resins as matrix materials is given by Constantino et al.
[2.1.3]. As each manufacturing technology implies specific requirements
regarding the processability of the resin/hardener system used, this affects the
formulations which potentially could be used. Thus the most important
manufacturing technologies for fiber-reinforced composites based on epoxy
resins will be introduced briefly in the present chapter.
General trends are the increasing automation of the manufacturing processes

and the reduction of the cycle times in order to reduce the process costs.
Therefore, much work has been performed on resin systems which enable cycle
times of 3-5 minutes to be achieved – which is one of the requirements of the
automotive industry [2.1.4]. In the near future even the fully automated
production of large rotor blades for wind energy generators will be possible
[2.1.5].
Sometimes single parts are manufactured by different methods and assembled to

form larger structures afterwards. If you have a close look at the chassis of the
Porsche 918 Spyder (Figure 2.1.1) you will discover that the monocoque is
manufactured using a resin transfer molding (RTM) process. The engine support
however, being subject to much higher operating temperatures, is made using
prepreg technology in combination with an autoclave cure [2.1.6].
Figure 2.1.1: Rolling chassis of the Porsche 918 Spyder [2.1.6]
10
2.1.1. Prepreg methods
Here woven or nonwoven fabrics (the latter often being called non-crimp fabrics -
see chapters 2.2.2. and 2.2.3.) are impregnated with a blend of epoxy resin and
curing agent and precured to a certain degree of crosslinking. Using a film
transfer process a matrix film is first formed using the resin/hardener blend. In a
second step this film is transferred from the release paper to the fiber
architecture (i.e. the textile) which has been selected for the composite (Figure
2.1.2). The classic solvent-based solution route uses a solution of the
resin/hardener blend in a volatile solvent, e.g. methylethylketone. After
impregnating the fibers the solvent is evaporated in a vertical or horizontal dryer
(see Figure 2.1.2). Another variation is the hot-melt process, where the
resin/hardener blend is applied as a melt. In all these three processes the resin
systems are partially polymerized, to a relatively low degree, at an elevated
temperature (so-called B-stageing). The final prepreg is coiled up and stored in a
refrigerator.
Figure 2.1.2: Prepreg manufacturing processes [2.1.7]
11
Later, very often in a different facility, the prepregs are cut by hand or by robot
and laid into the form of the end-product. Then the epoxy resin is fully cured;
e.g. at 135 °C or 190 °C, depending on the choice of hardener. This can be done
in a vacuum, at room pressure, in a press, using vacuum bags or, typically for
aerospace parts, in an autoclave.
The flow behaviour of the precured epoxy resin at higher temperatures is very
critical for this process. Fillers are used rarely. Nanofillers, of course, can be
used.
Prepregs are used to manufacture a wide variety of composite parts: for

aerospace (extensively in the Boeing 787 and the Airbus A 350 XWB), rotor
blades, sporting goods, shipbuilding, yacht construction or railway construction.
In the automotive industry prepregs are used as well, an example would be the
roof module of the Lamborghini Gallardo monocoque [2.1.8].
2.1.2. Filament Winding
The winding process is especially suitable for pipes, tubes used as ski poles or
masts of sailing boats as well as any other rotationally-symmetrical parts such as
tanks or pressure vessels for gases. The fiber rovings are passed through a
resin/hardener liquid blend and then wound around a mandrel; followed by being
cured at elevated temperatures.
Figure 2.1.3: Pipe manufacturing by filament winding [2.1.9]

12
In Figure 2.1.3 the winding of the impregnated glass fiber rovings can be seen
very well. Automotive applications are drive shafts, used for example in the Ford
Mustang or other rear wheel driven sports cars. It is important that the resin
systems used exhibit a relatively low degree of shrinkage upon cure, as
otherwise this could cause tremendous problems in removing the mandrel. Fillers
are used occasionally.
2.1.3. Pultrusion
The pultrusion process is designed for continuous manufacturing of standard

profiles, especially pipes. In Figure 2.1.4 a manufacturing line is shown. The fiber
rovings are passed through a bath containing the resin/hardener blend and
impregnated. They are then brought together in a die to give their final shape
(i.e. forming) and cured at elevated temperatures in a heating zone. It is
essential to use resin formulations which cure rapidly and have very low
shrinkage upon cure. In this process fillers, which could reduce the shrink, are
rarely ever used.
Figure 2.1.4: Pultrusion processing - picture courtesy of Strongwell [2.1.10]
2.1.4. Injection processes via a mold
Most important are injection processes such as resin transfer moulding (RTM) or
vacuum assisted resin transfer moulding (VARTM). They are very suitable for the
semiautomatic production of relatively complex composite parts. A fiber textile,
from simple layers of fabric to complex preforms or layups with core materials
such as foams or honeycomb structures, is placed in a mold. Then the
resin/hardener blend is injected and cured at an elevated temperature. Figure
2.1.5 shows schematically the process. Of great importance in the RTM process
is the need for low resin viscosities of 100-500 mPas at the injection temperature
to obtain good fiber wetting and reasonably long flow distances for the injected
resin/hardener blend.
13
Fillers which would improve mechanical properties like compressive strength
cannot be used, as they would be filtered out at the first layer of fibers.
Figure 2.1.5: Schematic illustration of the HD-RTM process [2.1.11]
RTM, VARTM and similar processes, such as the wet-press process, are suitable
for the high volume production of fiber-reinforced composite parts. Even complex
geometries can be realized – since then, instead of simple layers of fabrics,
three-dimensional layups or preforms are placed in the mold. Figure 2.1.6 shows
a small series automotive part (sports car bracket):
Figure 2.1.6: Composite part for automotive application manufactured by RTM -

picture courtesy of Kegelmann Technik [2.1.12]
14
Consequently injection processes are widespread in the composites industry. An
example of an automotive application is the convertible soft top cover of the VW
Eos. Also, the body of the electric BMW i3 is made entirely from composite parts,
which are made using RTM from preforms and assembled using structural
adhesives [2.1.13]. The use of preforms or very complex layups, which was
considered to be exotic a few years ago, is now an industrial standard practice.
2.1.5. Infusion processes via a vacuum bag
Infusion processes, such as the vacuum assisted resin infusion process (VARI)
which is used to manufacture very large rotor blades for wind energy generators,
are of also great importance in the industry. The fabrics, unidirectional or
multiaxial textile layups are placed in an open mold, as can be seen in Figure
2.1.7. The fibers are covered with a so-called vacuum bag, which is sealed
hermetically at the edges. Subsequently the resin/hardener blend is sucked at
different entry points using a vacuum line into the textile and the resin/hardener
blend then impregnates the fibers. Cure can be undertaken at room temperature
or moderately elevated temperatures. Relatively low resin viscosities are
required. Thus, as for with above injection processes, fillers cannot be used in
the resin formulations.
Figure 2.1.7: Boat hull made from carbon fibers using VARI - picture courtesy of
NORCO GRP Ltd [2.1.14]
15
2.1.6. References
[2.1.1] Ehrenstein, G.W. (2006) Faserverbundkunststoffe. Hanser Verlag

München, Germany, 148 - 209
[2.1.2] Campbell, F. (2004) Manufacturing processes for advanced

composites. Elsevier Advanced Technology, Oxford, UK, 104-179,
and 306-357
[2.1.3] Constantino, S., Waldvogel, U. (2010) Composite processing: State

of the art and future trends. In: Pascault J.P., Williams R.J.J. (eds)
Epoxy Polymers. Wiley, Weinheim, Germany, 271–287
[2.1.4] Hillermeier R., Hasson T., Friedrich L., Ball C. (2013) Advanced
thermosetting resin matrix technology for next generation high
volume manufacture of automotive composite structures SAE
Technical Paper 2013-01-1176. doi:10.4271/2013-01-1176
[2.1.5] Ohlendorf, J.H., Rolbiecki, M., Schmohl, T.; Franke, J. Thoben, K.D.,
Ischtschuk, L. (2013) Innovationen in der Handhabungs- und Textil-
technik zur Rotorblattfertigung, Lightweight Design 6 (5), 50-57
[2.1.6] Trender, L., Schromm, M. (2013) 918 Spyder. Proceedings of

Aachener Karosseriebautage, 24./25. September, RWTH, Aachen,
Germany
[2.1.7] Prepreg Technology. (2005) Company information No. FGU 017b,

Hexcel Ltd., Duxford, UK
[2.1.8] De Oto, L. (2012) Lightweight inspiration from the supersports car

segment – the Lamborghini Aventador. Proceedings of the
Lightweight Car Body Conference, 18./19. September, Neckarsulm,
Germany
[2.1.9] Company brochure. (2012) Magnus Venus Plastech, Clearwater, FL,

USA
[2.1.10] Pulsstar brochure. (2013) Strongwell, Bristol, VA, USA
[2.1.11] Mitzler, J., Renkl, J., Würtele, M. (2011) Hoch beanspruchte

Strukturbauteile in Serie. Kunststoffe 3, 36-40
[2.1.12] Product presentation. (2013) Kegelmann Technik, Rodgau-

Jügesheim, Germany
[2.1.13] Schmid, L., Schnaufer, T. (2013) The lightweight structure of the

BMW I3. Proceedings of Aachener Karosseriebautage, 24./25.
September, RWTH, Aachen, Germany
[2.1.14] Company information. (2013) Norco GRP Ltd., Holton Heath,UK
16
2.2. Fiber reinforcements for epoxy resin systems
A relevant overview regarding the different reinforcing fibers, their

manufacturing and their property profiles is given by Flemming et al. [2.2.1] and
Ehrenstein [2.2.2]. In combination with epoxy resins mainly glass fibers are used
(approx. 750.000 tons in 2013 [2.2.3]), followed by carbon fibers with approx.
43.000 tons in 2013 [2.2.4]. For specialties, such as ballistic applications, aramid
fibers and ultrahigh strength polyethylene fibers are employed. The selection of
the reinforcing fiber material is determined by the requirements of the final
composite part.
Unlike unsaturated polyester resins and thermoplastic matrices, short fiber

reinforcement is irrelevant for epoxy resins, with the exception of some
adhesives applications. When discussing fiber-reinforced composites based on
epoxy resins, the implication is always that long fiber reinforcement with
"endless fibers" will be involved.
For manufacturing processes such as filament winding or pultrusion in general

fiber strands, so-called rovings, are used.
For both injection and infusion processes (as well as in the manufacture of
prepregs) fabrics and non-crimp fabrics, such as unidirectional or multiaxial
fabrics, are used. In a few applications fiber mats with randomly oriented glass
or carbon fibers are processed using epoxy resin systems.
2.2.1. Woven fabrics
Figure 2.2.1: Glass fiber fabric with weave linen-type [2.2.5]
17
Fabrics with different weavings can be made from glass and carbon fibers. The
weavings can be made in the form of, for example, plain or linen-type, twill,
satin and many others. Figure 2.2.1 shows a typical example. The different
weavings not only lead to different mechanical properties but can be especially
important for esthetic aspects in composites parts, where the reinforcement is
shown intentionally for decorative purposes.
In woven fabrics, the fibers are always undulating as a consequence of the

weaving process. This can be a disadvantage when it comes to the load
transmission in the composite part. Therefore, so-called non-crimp fabrics have
been developed, in which the fibers are stretched out totally flat without any
undulation.
2.2.2. Unidirectional fabrics
In unidirectional fabrics (UD), the fibers are all oriented parallel to each other
and are fixed in place by an extremely thin sewing thread which is typically a
polyester yarn. In the example of a UD made from carbon fibers shown in Figure
2.2.2 this can be seen very well.
Figure 2.2.2: Unidirectional fabric or textile - picture courtesy of Spinteks Textil

[2.2.6]
A typical application for unidirectional fabrics made from glass fibers are the
horizontal beams of rotor blades for wind turbines, as they have to carry the
maximum load. An automotive application is the leaf spring for light trucks, such
as the Mercedes Sprinter.
18
2.2.3. Multiaxial fabrics
Besides unidirectional textiles, bidirectional fabrics have also been developed,

where two layers of unidirectional fibers are placed in a 45° or 90° angle to each
other and then sown together. This offers advantages for applications where the
application of force comes from different directions simultaneously.
The next step in the development of such textiles has been the multiaxial fabrics.
Here the orientation of the fiber layers to each other and the number of layers
can be chosen practically without any restriction. Figure 2.2.3 shows an example
of such a multiaxial textile layup. At the right side the stitching needles applying
the fixing thread can be seen.
Figure 2.2.3: Schematic layup of a multiaxial fabric or textile [2.2.7]
2.2.4. Preforms
Preforms are braided, woven, laid or sown three-dimensional fiber architectures

whose shape is already very close to the often complex shape of the final
composite part. An example of the braiding of a preform is given in Figure 2.2.4.
19
Figure 2.2.4: Braiding of a tubular preform using carbon fiber rovings [2.2.8]
An example for a composite part, manufactured from several different preforms

which are joined in the mold, by using a RTM process, can be seen in Figure
2.2.5.
Figure 2.2.5: Body in white of the BMW i3; manufactured by using preforms
which are impregnated subsequently in a HP-RTM process [2.2.9]
2.2.5. Fiber Sizing
All commercial glass and carbon fibers are pretreated with a coating, the so-
called fiber sizing by the manufacturer. This coating has several different
functions. It is supposed to protect the fiber during processing, e.g. during the
weaving process.
20
Furthermore it should significantly increase the adhesion between the fiber and
the resin matrix. Sizings for epoxy resins typically contain epoxy functional
silanes as adhesion promoters. Most fiber manufacturers formulate their sizings
themselves and consider them as part of their core technology. Hence it is
difficult to obtain detailed information regarding their composition. Indeed, a
further detailed consideration would go beyond the scope of this dissertation.
However, as an introduction to this topic, I would like to recommend the recently
published review by Thomason [2.2.10].
2.2.6. References
[2.2.1] Flemming, M., Ziegmann, G., Roth, S. (1995) Faserverbund-

Bauweisen. Springer Verlag Berlin, Germany, 6-179
[2.2.2] Ehrenstein, G.W. (2006) Faserverbundkunststoffe. Hanser Verlag,

München, Germany, 19-49
[2.2.3] Market Research Report (2012) Global Glass Fiber Market 2012-
-2017: Trend, Forecast and Opportunity Analysis, Lucintel, Las
Colinas, TX, USA
[2.2.4] Jahn, B., Witten, E. (2013) Composites Market Report 2013. Carbon
Composites e.V., September 2013, Augsburg, Germany
[2.2.5] Product presentation glas fiber fabric. (2013) Jiyuan Wuayang

Composite Materials Co. Ltd., He Nan, China
[2.2.6] Product presentation DU carbon textile (2014) Spinteks Tekstil

Insaat San. ve Tlc. A.S., Honaz, Turkey
[2.2.7] Baitinger, S. (2013) Möglichkeiten im Leichtbau durch multiaxiale

Gelege bei Auslegung und Fertigung. Proceedings of Neue
Technologien im textilbasierten Faserverbund-Leichtbau, 18. April,
DLR Stuttgart, Germany
[2.2.8] Product presentation. (2013) SGL Kümpers GmbH, Rheine/

Gellendorf, Germany
[2.2.9] BMW I3 Monocoque (2012), Composites Europe, 9.-11. October

2012, press information Reed Exhibitions
[2.2.10] Thomason, J.L. (2013) Glass Fiber Sizings: A Review of the Scientific
Literature. ISBN 978-0-9573814-0-7, Create Space, UK
21
2.3. Epoxy resin systems for fiber-reinforced composites
Epoxy resins are one of the most important class of thermosetting polymers,
along with polyurethanes (PU), unsaturated polyester resins (UPE), vinyl ester
resins and phenolic resins. Since their introduction in the 1950s they have been
used for many very different industrial applications. For example, they are
employed as adhesives, electrical encapsulants, floor coatings, fiber-reinforced
composites, heavy duty marine coatings, repair pastes, abrasive systems and
have been used in aircraft, train and car manufacturing, electronic assembling
and jewelry - indeed the abundance of applications is almost endless. A
comparison of various matrix resin systems and their usability for fiber-
reinforced composites is given by Hasson [2.3.1].
In the year 2009 1.8 million tons of epoxy resin were used worldwide; 52.000
tons in fiber-reinforced composites. This equals a quite small share of only 2.9 %
[2.3.2]. For the years 2009 to 2014 an annual increase in epoxy resin demand of
5.8 % is expected so that the consumption in 2014 will be approximatively 2.4
million metric tons. The consumption in composite applications however is
growing much faster by 8-10 % per year and will grow even faster if the
automotive manufacturers develop new, large volume, composite parts. The
epoxy resins and associated hardeners relevant for fiber-reinforced composites
will be introduced briefly below. For a more detailed study, the work of Ellis
[2.3.3] is recommended.
2.3.1. Epoxy resins
The working horse of the epoxy resin industry is the diglycidylether of bisphenol
A (DGEBA). In composite applications it is by far the most commonly used type
of epoxy resin. Figure 2.3.1 gives an overview of the chemical structures of the
epoxy resins relevant for fiber-reinforced composites. Higher molecular weight
homologues of DGEBA, so-called solid epoxy resins, are frequently used in
prepreg manufacturing. The diglycidylether of bisphenol F (DGEBF) is used in
formulations were low viscosities are necessary. The higher homologues of
DGEBF, epoxidized novolac resins, are used in applications were increased
thermal and chemical stabilities are required. Due to their higher epoxy
functionality they form closer meshed networks which consequently possess a
higher glass transition temperature (Tg). Short chain aliphatic epoxy resins, such
as hexandiol diglycidylether (HDDGE) or butanediol diglycidylether (BDGE) are
introduced into formulations when viscosities need to be lowered. Aliphatic epoxy
resins with longer molecular chains, such as epoxidized polyether polyols or
epoxidized ethoxylated derivatives of trimethylolpropane are occasionally added
to flexibilize resin systems. They lower the crosslink density and consequently
the Tg decreases. Cycloaliphatic epoxy resins, such as hydrogenated DGEBA or
3,4-epoxycyclohexane carboxylic acid -3,4- epoxycyclohexane methylester (EEC),
are currently not relevant for fiber-reinforced composites applications. However,
very important are the higher functional resins, such as the triglycidyl-p-
aminophenol (TGpAP) and tetraglycidyl methylendianiline (TGMDA), and without
these resins most aerospace applications would not be possible.
22
Diglycidylether of bisphenol A Solid DGEBA resin (higher homologue, n=1)
(DGEBA)
Diglycidylether of Epoxidized novolac resin

bisphenol F (DGEBF) (higher homologue of DGEBF)
Hexandiol diglycidylether (HDDGE)
Triglycidylether of p-aminophenole Tetraglycidylether of methylen-

(TGpAP) dianiline (TGMDA)
Figure 2.3.1: Chemical structures of epoxy resins relevant for composites

23
2.3.2. Curing agents for epoxy resins
The hardeners relevant for composites applications can be divided roughly into
amines and anhydrides. Catalytic curing agents such as BF3-complexes or UV
initiated cationic curing agents, are rarely used. Figure 2.3.2 gives an overview
of the chemical structures of the most important hardeners for fiber-reinforced
composites. For further reading the papers published by Hare [2.3.4], [2.3.5] are
to be recommended.
2.3.2.1. Amine curing agents
One of the most important amines is the cycloaliphatic isophorone diamine (IPD)
which is a major component for most infusion resin systems. It provides an
excellent balance between giving a relatively high crosslink density and a fast
reaction speed.
Linear polyether diamines are used as flexibilizers. Such amine formulations are
typically cured at relatively low temperatures – therefore these systems do not
exhibit very high glass transition temperatures (Tgs).
Especially in prepreg applications latent curing agents, such as dicyandiamide

(DICY) are of major importance. They start to cure the epoxy resin above a
starting temperature (approx. 145 °C) when dissolved in the epoxy resin. They
are often combined with accelerators, such as 4-methyl imidazole. Imidazoles
are also used as single curing agents. Ternary amines, such as benzyl
dimethylamine (BDMA), are used as accelerators for DICY as well in quite a few
formulations.
Aromatic amines are used for epoxy resin systems which require high glass
transition temperatures due to high service temperatures being experienced by
the composite parts, as in many aerospace applications. An important hardener
is the 4,4´-diaminodiphenylsulfone (DDS). It is used in prepregs as well as in
injection or infusion resin systems. Other important aromatic amines are the
4,4´-methylenbis-2,6-diethylaniline (M-DEA) and the 4,4´-Methylenbis-2,6-diiso-
propylaniline. A blend of both is used to formulate the benchmark for aerospace
injection resins, i.e. the famous RTM 6 resin/hardener blend.
2.3.2.2. Anhydride curing agents
Anhydride hardeners are characterized by their low viscosity. They need to be

cured at high temperatures and can achieve high glass transition temperatures.
Their main field of application is in resin systems for filament winding processes.
Methylhexahydrophthalic acid anhydride (MHHPA), an important material, is

under scrutiny by the European Commission and very recently was declared as a
critical substance. Methyltetrahydrophthalic acid anhydride (MTHPA) is not
considered as critical. Another common anhydride hardener is methyl-5-
norbornene-2,3-dicarboxylic acid anhydride (MNA).
New dianhydrides such as the 4,4´-carbonyldiphthalic acid anhydride (BTDA) can

achieve very high Tgs. However, these materials are solids and therefore more
difficult to use in formulations.
24
Isophorondiamine (IPD) Polyetherdiamine
Dicyandiamide (DICY)
Diaminodiphenylsulfone (DDS) Methylendiethyldianiline (M-DEA)
Methylendiisopropyldianiline (M-DIPA) Methylhexahydrophthalic acid

anhydride (MHHPA)
Methylnorbornendianhydride (MNA) Carbonyldiphthalic acid anhydride

(BTDA)
Figure 2.3.2: Chemical structures of the most relevant epoxy curing agents
25
2.3.3. Epoxy resin systems
An epoxy resin system is defined by the combination of resin(s) and hardener(s)

plus additives and other functional ingredients. The formulations are optimized
regarding the desired property profile of the fiber-reinforced composite and the
processing parameters defined by the manufacturing process. These areas are
the subject of very intensive research, which is very understandable when the
possible improvements shown later in this work are considered. In principle,
there are endless possibilities of resin/hardener combinations, but some are very
characteristic and will be mentioned below.
In standard pultrusion processes typically DGEBA or DGEBA/DGEBF blends with

anhydride hardeners are used, even more in combination with amine hardener
blends. These blends consist essentially of IPD, polyamidoamines and small
amounts of tertiary amines.
Typically for filament winding processes, e.g. for ski pole manufacturing, the
combination of DGEBA with an anhydride curing agent is common. If the
performance requirements of the composite part increase, other resin systems
are used. For example, pressure vessels made by filament winding are made
from epoxidized novolac resins and anhydrides or DDS.
In prepreg manufacturing a distinction is made regarding the cure temperature

of the prepreg: a typical 120 °C curing prepreg resin system is composed of
DGEBA and DICY as well as an accelerator. A 180 °C curing prepreg resin system
is based on DGEBA and DDS or TGMDA and DDS. (It will be recalled that during
the so-called B-stageing the resin system is prepolymerized onto the fiber
textile.) Prepregs, especially 180 °C systems for aerospace applications, used to
be always cured in a press or an autoclave. More recently "out-of-autoclave"
developments typically have led to the use of DGEBA and amine blends, e.g. with
imidazoles. For some fiber-reinforced composite parts, prepregs are wound
around mandrels as well.
In RTM processes very often DGEBA is formulated with IPD/amine blends, e.g.
for automotive parts. We have used a similar system in our processes (see 4.5.).
For aerospace applications TGpAO or TGMDA together with aromatic amines such
as DDS are typical. It should be recalled that a relatively low viscosity resin
system is very important for the processability via RTM methods.
In infusion processes especially low viscosity resin systems are required.

Therefore, DGEBA or DGEBF resins are blended with so-called reactive diluents.
A typical infusion resin system for the manufacture of rotor blades consists of
DGEBA with 20 wt% HDDGE and a hardener blend of IPD with 30 wt% of an
amino-functional polyether polyol. The relatively low curing temperatures of 60-
80 °C lead to a relatively low Tg.
26
2.3.4. References
[2.3.1] Hasson, T. (2013) Neue Matrixtechnologie für die CFK-Fertigung in

der Automobilindustrie. Proceedings of Neue Technologien im
textilbasierten Faserverbund-Leichtbau, 18. April, DLR Stuttgart,
Germany
[2.3.2] CEH Marketing Research Report (2010) Epoxy Resins. SRI

Consulting, Mento Park, CA, USA
[2.3.3] Ellis, B., Hrsg. (1993) Chemistry and Technology of Epoxy Resins.
Springer Science+Business Media, Dordrecht, Netherlands
[2.3.4] Hare, C.J. (1994) Amine curing agents for epoxies. J. Prot. Coat.
Lin., 9, 77-103
[2.3.5] Hare, C.J. (1994) Epoxy curing agents II. J. Prot. Coat. Lin.,
10, 197-213
27
2.4. Toughening of epoxy resins
A disadvantage of cured epoxy resins is their extreme brittleness. The tighter the
close-meshed three-dimensional networks become, e.g. when using tetra-
functional epoxy resins, the greater is the observed brittleness. Since their
introduction in the 1950s, increasing the toughness of cured epoxy resins has
been the subject of many ongoing developments.
2.4.1. Differentiation between toughening and increasing flexibility
One has to distinguish between toughening and simply incressing the flexibility of
the cured epoxy resin.
For an increase in the flexibility of the cured epoxy resin, the network density of
the epoxy resin is reduced on purpose by adding flexible, long chain molecules,
which often are monofunctional. Thus, the failure strain may be increased but
the modulus and glass transition temperature are consequently lowered. As a
result the toughness is often somewhat increased. However, the tensile strength
is also reduced in most cases. Examples of such molecules are epoxidized
polyethers such as polypropylene glycol diglycidylether, C12-C14-glycidylether,
epoxidized soja bean oils, epoxidized cashew nut shell oil or dimeric fatty acid
esters.
When a cured epoxy resin is significantly toughened, the associated aims are not
to greatly reduce the modulus, glass transition temperature or tensile strength.
Thus, the usual approach here is to deliberately create a two-phase structure.
Therefore, in the brittle epoxy resin matrix exists a second discontinuous phase
in the form of particles or agglomerated/aggregated molecules.
2.4.2. Reactive liquid rubbers
One of the first modifications to try and toughen cured epoxy resins, which still
today is state-of-the-art and the industrial benchmark, was introduced into the
market in the 1980s – i.e. the modification of epoxy resins with reactive liquid
rubbers. Figure 2.4.1 shows the chemical structure of such a carboxy-terminated
butadiene-acrylonitrile copolymer (CTBN rubber):
Figure 2.4.1: Chemical structure of a reactive liquid rubber [2.4.1.]
28
However, carboxy-functional nitrile-butadiene copolymers are incompatible with
epoxy resins and when blended into an epoxy resin they phase-separate within
hours. Nevertheless, if the carboxy groups are first reacted with an excess of
epoxy resin at elevated temperatures then epoxy-rubber-epoxy block copolymers
are formed, i.e. termed rubber adducts. These can now be blended with epoxy
resins and do not phase separate. However, they have impart a relatively high
viscosity. Upon formation of the three-dimensional network during cure, then
phase separation of the rubber adduct molecules occurs. Being distributed
homogenously before the cure, they now form particles or domains in which the
long rubber molecules are entangled with each other and are bound covalently to
the epoxy matrix via the former carboxy groups.
After the extensive research in the 1980s and 1990s the mechanisms of rubber-
toughening are well understood [2.4.1], [2.4.2], [2.4.3], [2.4.4]. Recently a
comprehensive review was published by Pearson et al. [2.4.5].
Kinloch et al., the pioneers of research in rubber-toughening, proposed a model

regarding the mechanisms of toughening [2.4.6] which has been consistently
confirmed and is still valid today. As can be seen in the illustration in Figure
2.4.2, the rubber particles can stop crack formation in the cured epoxy resin by
(a) rubber bridging, (b) local shear bands which can be formed between the
particles, and (c) the rubber particles can cavitate, followed by a plastic growth
of the epoxy matrix. Crack tips are blunted upon cavitation.
Figure 2.4.2: Mechanisms of toughening of epoxy resins with reactive liquid

rubbers [2.4.6]
29
Epoxy resins modified with reactive liquid rubbers are used in some composite
manufacturing processes, especially in prepregs and in filament winding.
However, since this type of modification considerably increases the resin
viscosity, such systems cannot be readily used in injection or infusion processes.
Another drawback of this type of modification is the fact that a small fraction of
the long, flexible rubber molecules do not participate in the phase separation
process but crosslink randomly into the three-dimensional molecular network of
the epoxy resin. Thus the network density is lowered, with an accompanying
decrease in the modulus and Tg of the cured epoxy resin.
2.4.3. Core-shell rubber particles
To overcome the above disadvantages, in the 1980/90s the so-called core-shell

rubber technology was developed. The morphology which is obtained using this
technology is very similar to that achieved with reactive liquid rubbers, as can be
seen in Figure 2.4.3.
(1)
(2)
Figure 2.4.3: Comparison of fracture surfaces of toughened epoxies using SEM:

(1) reactive liquid rubber (CTBN) [2.4.7]; (2) with core-shell-particles [2.4.13]
30
Well-defined elastomeric particles are synthesized, which have a shell compatible
with epoxy resins around the rubbery core. The shell possesses reactive groups,
e.g. epoxy groups, to ensure covalent bonding with the resin matrix. The core
might be a polybutadiene, a polyacrylate or a polysiloxane. Depending on the
manufacturer, the core-shell particles have different particle sizes from
approximatively 110 nm to 300-400 nm up to 500-700 nm [2.4.8], [2.4.9],
[2.4.10], [2.4.11], [2.4.12].
Creating a similar morphology in the cured bulk resin by using core-shell

particles implies that the mechanisms valid for rubber toughening are valid here
as well.
Dispersions of such particles in epoxy resins have significantly lower viscosities

and therefore are more suitable for use in fiber-reinforced composites. As the
modulus and Tg are slightly, or not affected, resin systems containing core-shell
particles are being increasingly used, even in injection processes.
2.4.4. Thermoplastic particles
Another possibility to increase toughness of cured epoxy resins is the

modification of the epoxy resin with a second phase of a thermoplastic; e.g.
polyethersulfones [2.4.14]. To obtain favourable properties, covalent bonding to
the epoxy resin matrix is indispensable. Pearson et al. [2.4.15] proposed several
mechanisms for thermoplastic toughening, as can be seen in Figure 2.4.4.
Figure 2.4.4: Schematic illustration of proposed toughening mechanisms with

thermoplastic particles: (1) crack bridging, (2) interruption of crack propagation,
(3) crack deflection, (4) formation of shear bands, (5) deformation of particles
[2.4.15]
A comparision of the proposed mechanisms with the model of Kinloch et al.

[2.4.6] is striking. The key to toughening is clearly a two phase system with the
particles covalently bonded to the epoxy resin matrix.
31
Thermoplastic modification is used in prepreg manufacturing for aerospace
applications, where for example the thermoplastic powder is sprinkled onto the
textile during the prepreg manufacturing process. A variation is to weave
thermoplastic fibers into the textile used for the prepreg which then dissolve
upon curing into the epoxy resin matrix. Due to the incompatibility of the
thermoplastic with the epoxy, they form particles and create the desired two-
phase morphology [2.4.16]. A leading epoxy resin system toughened with a
thermoplastic is commercially available under the trade name Cycom 977-2.
2.4.5. Self-organizing block copolymers
Another variation of the toughening of epoxy resins are self-organizing

thermoplastic block copolymers. The copolymers are blended into the epoxy
resin, in which they dissolve. Upon curing the epoxy resin, they phase separate
and form a secondary phase of core-shell particles, or another form of a phase-
separated structure. The principle is similar to that for the reactive liquid rubbers.
However, the epoxy resin systems employing such self-organizing thermoplastic
copolymers exhibit much lower viscosities. In Figure 2.4.5 the process of phase
separation is shown schematically.
Figure 2.4.5: The phase separation process of self-organizing thermoplastic

tougheners [2.4.17]
This relatively new technology is very rarely used in fiber-reinforced composites,

although commercial products from several manufacturers are offered in the
market. The reason for the lack of their use might be the losses in strength,
modulus and glass transition temperature seen for such modified epoxy resin
systems, which are similar in extent to those observed in reactive liquid rubber
modified systems.
32
2.4.6. References
[2.4.1] Okamoto, Y. (1983) Thermal Aging Study of Carboxyl-Terminated

Polybutadiene and Poly(Butadiene-Acrylonitrile)-Reactive Liquid
Polymers. Polymer Engineering and Science, Vol.23, No 4, 222-225
[2.4.2] Kinloch, A.J., Shaw, S.J., Tod, D.A., Hunston D.L. (1983)
Deformation and fracture behaviour of a rubber-toughened epoxy:
1. Microstructure and fracture studies. Polymer, 24, 1341-1354
2. Failure criteria. Polymer, 24, 1355-1363
[2.4.3] Siebert, A.R. (1984) Morphology and Dynamic Mechanical Behaviour

of Rubber-Toughened Epoxy Resins. ACS Adv. in Chem. Series, 208,
12, 179-191
[2.4.4] Kinloch, A.J. (2003) Toughening Epoxy Adhesives to Meet Today’s

Challenges. MRS Bull. June 2003, 445-448
[2.4.5] Bagheri, R., Marouf, B.T., Pearson,R.A. (2009) Rubber-Toughened

Epoxies: A Critical Review. J. Macromol. Sci. Part C Polym. Rev., 49,
201-225
[2.4.6] Huang, Y., Kinloch A.J. (1992) Modelling of the toughening

mechanisms in rubber-modified epoxy polymers. Part II A
quantitative description of the microstructure-fracture property
relationships. J. Mater. Sci. 27, 2763-2769
[2.4.7.] Sprenger, S., Weber, C., Pulliam, L. (1997) Elastomer-modified

epoxy prepolymers - the new generations. European Adhesives &
Sealants, September 1997, 9-12
[2.4.8] Giannakopoulos, G., Masania, K., Taylor, A.C. (2011) Toughening of

epoxy using core-shell particles. J. Mater. Sci. 46, 327-338
[2.4.9] Tsai, J.L., Chang, N.R. (2011) Investigating damping properties of

nanocomposites and sandwich structures with nanocomposites as
core materials. J. Compos. Mater. 45, 2157-2164
[2.4.10] Block, H., Pyrlik, M. (1988) Silicones are the Key: Modifying
Thermosetting Resins with Silicone Elastomers. Kunstst. Ger. Plast.
78, 1192-1196
[2.4.11] Lai, M., Friedrich, K., Botsis, J., Burkhart, T. (2010) Evaluation of
residual strains in epoxy with different nano/micro-fillers using
embedded fiber Bragg grating sensor. Compos. Sci. Technol. 70,
2168-2175
[2.4.12] Chen, J., Kinloch, A.J., Sprenger, S., Taylor, A.C. (2013) The
mechanical properties and toughening mechanisms of an epoxy
polymer modified with polysiloxane-based core-shell particles.
Polymer 54, 4276-4289
33
[2.4.13] Unpublished research report of nanoresins AG, Geesthacht,
Germany
[2.4.14] Yoon, T.H., Priddy, D.B., Lyle, G.D., McGrath, J.E. (1995)
Mechanical and morphological investigations of reactive polysulfone
toughened epoxy networks. Macromol. Symp. 98, 673-686
[2.4.15] Pearson, R.A., Yee, A.F. (1993) Toughening mechanisms in thermo-

plastic-modified epoxies: 1. Modification using poly-(phenylenoxid).
Polymer 34, 3658-3670
[2.4.16] WO 02/16481 A1 of 28.02.2002
[2.4.17] Product presentation (2007) Novel toughened epoxy resins. The

Dow Chemical Company, Freeport, TX, U.S.A.
34
2.5. Modification of epoxy resins with SiO2 nanoparticles
As mentioned already in chapter 2.1., in most processes to manufacture fiber-

reinforced composites no fillers can be used in the epoxy resin. At least no fillers
in the classic sense with particle sizes of several micrometers. Since the
commercial availability of nanoparticles this classic statement has changed
fundamentally.
2.5.1. SiO2 nanoparticles
In the years 2002 and 2003 the first commercial surface-modified silica
nanoparticles were introduced into the market. The particles are manufactured
using a modified sol-gel process, surface-modified and are supplied as
concentrates in a range of epoxy resins.
Figure 2.5.1 shows the molecular model of a 1.2 nm particle without surface
modification.
Figure 2.5.1: Model of a SiO2 nanoparticle [2.5.1]
The large number of hydroxy groups at the particle surface is remarkable. It

explains the importance of the surface modification for the particle production:
the OH groups are blocked by the reaction with organosilanes, hence a further
growth of the particle by a condensation reaction of the hydroxy groups is
prevented. Also, the formation of agglomerates is suppressed and the particles
remain monodisperse. Furthermore, the surface modification forms an organic
layer around the particles and makes them compatible with the organic matrix:
in this case the epoxy resin.
The commercially available particles are spherical, 20 nm in diameter and exhibit

a very narrow particle size distribution. Figure 2.5.2 shows the particle size
distribution determined by small angle neutron scattering [2.5.2]:
35
Figure 2.5.2: Particle size distribution of commercial SiO2 nanoparticles [2.5.2]
The particles are very well dispersed in the epoxy resin, as can be seen in Figure
2.5.3. Due to their small size they do not interact with the visible light; hence
cured bulk resins containing them seem to be transparent. This makes resins
modified with SiO2 nanoparticles suitable for use in visible composite parts
where, for instance, the appearance of the carbon fiber itself will be visible for
aesthetic reasons.
Figure 2.5.3: Transmission electron microscopic (TEM) image of a cured epoxy

resin with approx. 20 wt% SiO2 nanoparticles [2.5.3]
36
Another advantage of their small size is the prevention of sedimentation
occurring. This makes these nanoparticles especially suitable for prepreg
manufacturing using epoxy resin solutions (see 2.1.1.). Furthermore they
increase the viscosity of the epoxy resin only at relatively high loading levels.
This makes them very suitable for all injection and infusion processes, where
resin viscosity is in general very critical.
Figure 2.5.4 shows the viscosity increase of a DGEBA epoxy resin with increasing
nanosilica content [2.5.4]. The silica nanoparticles are introduced into the DGEBA
resin by blending with two different commercial nanosilica
concentrates.
Figure 2.5.4: Viscosity as function of silica nanoparticle content [2.5.4]
Nanopox F 400 is a concentrate of 40 wt% silica nanoparticles in DGEBA,

whereas Nanopox F 520 is based on DGEBF. Thus, the smaller viscosity increase
seen in Figure 2.5.4 for the latter epoxy resin system. In the range of 5-15 wt%
nanosilica, where most commercial formulations operate, the viscosity increase is
very small for both epoxy resin systems. Further, due to their small size the silica
nanoparticles can easily penetrate even close-meshed fabrics easiliy. Indeed,
long flow distances in injection and infusion processes do not result in a gradient
in nanosilica concentration over the distance of the infusion.
A TEM picture of a glass fiber-reinforced composite (GFRC) made with an epoxy

resin containing SiO2 nanoparticles (see Figure 2.5.5) shows the differences in
size of the glass fibers and the nanoscale filler particles. Though glass fibers
usually have diameters of 10-15 micron, they appear to have a diameter of 0.5
microns here. This is a result of the 80 nm thick microscopy sample preparation.
37
Figure 2.5.5: TEM picture of a GFRC with 15 wt% SiO2 nanoparticles [2.5.5]
2.5.2. Property improvements of cured bulk resin systems
An improvement of the properties of cured epoxy resins, such as their modulus,

by the addition of SiO2 nanoparticles is to be expected, as they are a rigid filler.
However, very intriguing was an observed improvement in toughness [2.5.6]. It
is well known that for fillers such as fumed silica the toughness of the epoxy
resin can be marginally improved. However they have complex three-
dimensional structures which are totally different from the isolated spherical
nanoparticles.
Rosso et al. [2.5.7] investigated this phenomenon with fumed silica in cured bulk
epoxy resin and found that with an addition level of 5 wt% an increase in
modulus of 20 %, with no change in the tensile strength, could be achieved. The
fracture toughness (KIc) was improved by 70 % and the fracture energy (GIc) by
140 %.
The mechanisms of the toughening of epoxies by a modification with silica

nanoparticles was investigated in detail by Taylor et al. [2.5.8]. Void formation at
the interface of the nanoparticle-resin matrix followed by plastic void growth was
shown to be responsible for most of the toughness increase. Local shear bands
also contributed a minor share to the increase in the measured toughness.
Zhang et al. [2.5.9] modified DGEBA with pyrogenic silica and compared this
system with the commercially available nanosilica. However, using more than 6
vol% of fumed silica was impossible due to the extreme increase in the viscosity
of the epoxy resin system.
38
At 6 vol% both fumed silica and nanosilica increased the modulus by 17 %. KIc
was increased by the pyrogenic silica by 49 % and by 29 % by using nanosilica.
At 3 vol% (approx. 5 wt%) the modulus was increased by 9 %, KIc by 24 % and
GIc by 43 %. Zhang et al. used an anhydride curing agent, whereas Rosso et al.
employed an amine curing agent. Although the epoxy resin molecules and
hardener molecules were much smaller than the 20 nm silica particles, being
approximately 2 nm for DGEBA [2.5.10] and 0.7 nm for DETDA [2.5.11], the
hardener used seems to have had an influence on the improvements found.
Different curing agents caused different crosslink densities of the cured bulk
resins, and hence different values of the toughness and the glass transition
temperature were recorded. The improvements upon nanosilica addition as a
function of the curing agent needs to be examined in detail (see chapter 3.1.),
since formulation guidelines would be of great help to the industrial formulator.
2.5.3. Property improvements of cured bulk hybrid resin systems
Relatively early the possibilities of the combination of the classic toughening

modifiers, namely reactive liquid rubbers, and the addition of SiO2 nanoparticles
were recognized and patented [2.5.12]. One of the first industrial application of
such hybrid epoxy resins was as adhesives and many improvements in their
mechanical properties have been found [2.5.13]. One the disadvantages of
rubber toughening is the reduction in modulus, as discussed above, and this loss
in stiffness can be more than compensated by the addition of silica nanoparticles
to the toughened epoxy resin. The formation of the rubber domains during curing
of the epoxy resin typically seems not to be affected by the presence of the
nanosilica [2.5.14]. Figure 2.5.6 shows a TEM picture of a cured hybrid epoxy
resin [2.5.15]:
Figure 2.5.6: TEM picture of a cured epoxy resin modified with CTBN and
nanosilica – unknown addition levels for both modifications [2.5.15]
39
Tsai et al. [2.5.16] investigated the system DGEBA/isophorone diamine. They
found that the highest value of GIc was for the rubber-toughened system and
that the hybrid had a lower toughness. However, the loss in modulus due to the
reactive liquid rubber was nearly compensated for by the nanosilica addition. No
synergy was found between the two modifications.
A totally different behaviour was reported for an anhydride-cured DGEBA

[2.5.17]: the loss in modulus by the rubber addition was now completely
compensated. Figure 2.5.7 shows the synergy between both modifications for the
values of the fracture energy. The hybrids exhibit higher values of GIc than seen
for the separate addition of rubber and nanosilica. These contradicting results
need to be looked at closely in chapter 3.2.
Figure 2.5.7: Bulk GIc of a DGEBA modified with CTBN and SiO2 nanoparticles
[2.5.17]
The combination of core-shell particles with nanosilica has also been subject of
several investigations. An extensive study conducted by researchers of the IVW
[2.5.18] revealed a homogenous distribution of both the micron-sized core-shell
and the 20 nm silica particles. They found that the modulus was lowered only
marginally by the core-shell particles. The addition of nanosilica up to 8 vol%
increased the modulus and KIc by 25-50 %. The improvements found for this
amine-cured epoxy resin system were, however, minor.
In contrast Liu et al. [2.5.19] reported an increase in GIc by nearly 800 % for a
combination of 10 wt% core-shell particles and 10 wt% nanosilica for their
amine-cured system.
40
2.5.4. Property improvements of fiber-reinforced composites
One of the most important questions for an industrial formulator is whether the
improvements achieved for cured bulk resin systems will be found in a fiber-
reinforced composite based upon the same epoxy resin system as a matrix.
Unfortunately most results of industrial research regarding nanosilica-containing

composites or hybrid systems have never been published. Instead the
composites parts developed by the industrial research have been introduced into
the market, but without any research publications.
Mahrholz et al. [2.5.20] reported very early improvements for a glass fiber-
reinforced laminate based on an anhydride cured DGEBA by the addition of
nanoparticles. The bidirectional layup (+45°/-45°) was used in an injection
process.
A carbon fiber-reinforced laminate based on a tetrafunctional epoxy resin system

and 8 plies linen-type fabric showed a decrease in GIc by the addition of
nanosilica, although the bulk resin properties were improved [2.5.21].
Caccavale et al. [2.5.22] reported on carbon fiber-reinforced laminates made by

RTM. The addition of 3.7 wt% nanosilica increased bulk KIc by 11 % but the value
of the laminate GIc remained more or less unchanged.
Glass fiber-reinforced composites made from anhydride cured epoxy and a

[(+45°/-45°, 90°/0°)8]2 layup by an infusion process were studied by Kinloch et
al. [2.5.23]. The addition of 10 wt% SiO2 nanoparticles to the epoxy resin matrix
increased GIc by 207 % to 1015 J/m2.
Manjunatha et al. [2.5.24] investigated the fatigue performance of composites

based on an anhydride cured epoxy resin. They found significant improvements
in the fatigue behaviour. For example, the addition of 10 wt% nanosilica enabled
the composite laminates to withstand five times the number of loading cycles
before failure occurred.
Tsai et al. [2.5.16], [2.5.25] made laminates from an amine cured epoxy and UD
glass fibers. The addition of 10 wt% SiO2 nanoparticles increased GIc by only 8
%. Although the GIc of the modified bulk resin was improved by 47 %, the tensile
strength and compressive strength of the laminates did not increase.
Tang et al. [2.5.26] used piperidine as the hardener and UD carbon fiber textiles
to make laminates by the VARTM process. The addition of 10 wt% of nanosilica
increased the bulk GIc of the epoxy resin matrix by 93 % and 20wt % of
nanosilica increased the bulk GIc by 180 %. For the composites, however, the
improvements in GIc were only 1 % and 21 %, respectively. Also, the
interlaminar shear strength (ILSS) was reduced by the nanosilica addition to the
epoxy matrix.
Falling dart tests on glass fiber composites showed no improvements due to

nanosilica addition [2.5.27]. Unfortunately the corresponding compression-after-
impact (CAI) data could not be published due to a secrecy agreement. In Figure
41
2.5.8 the delaminated areas from the falling dart test are shown and there is no
visible difference.
Figure 2.5.8: GFRC laminates after falling dart test with 30 J impact without and
with nanosilica [2.5.27]
To study these contradictions, and then to expand and complement them with
other research results and eventually to identify basic mechanisms is the subject
of chapter 3.3. Another aim is to undertake a detailed examination of the
influence of the curing agent used in the epoxy resin system in combination with
nanosilica on composite properties, as silica nanoparticles are slightly acidic,
which might have an influence on amine curing agents. Since, many actual
composite parts made in industry today contain SiO2 nanoparticles: be it
aerospace components, machine parts, automotive parts or sporting equipment.
From the above studies, apparently the mechanisms of toughening with

nanoparticles do not apparently work under fast impact conditions. Thus, it was
obvious to combine them with a classic toughening modifier like reactive liquid
rubber or core-shell particles; which are known to be good for impact
performance. It was expected to be able to compensate the disadvantages of
rubber-toughening, such as reduced strength and modulus, by the nanosilica.
Thus, tough and stiff composite laminates could be made and the best of both
worlds united in the hybrid epoxy resin. This was considered to be an achievable
aim, since for bulk hybrid resin systems such improvements were already known
(see 3.2.).
42
In the very first investigations promising results were found. Brandt et al.
[2.5.28] reported on carbon fiber-reinforced laminates based on a tetrafunctional
epoxy hybrid resin system with core-shell particles and SiO2 nanoparticles.
Adding nanosilica to the core-shell-toughened resin increased the laminate GIc
from 155 J/m2 to 250 J/m2. Further, the delamination area after a 30 J impact
was reduced significantly.
In another study carbon fiber-reinforced laminates were made from anhydride-

cured DGEBA [2.5.17]. Although the bulk resin properties were improved
significantly (see Figure 2.5.7), the improvements achieved in GIc for the
corresponding composites were much smaller (see Figure 2.5.9).
Figure 2.5.9: GIc of a CFRC based on DGEBA modified with CTBN and SiO2
nanoparticles [2.5.17]
The investigation of a glass fiber-reinforced laminate based on an amine-cured

DGEBA showed some improvements in fracture energy, but only for relatively
low nanosilica addition levels of approximatively 4 % [2.5.5].
Caccavale et al. [2.5.22], [2.5.29] tested carbon fiber-reinforced laminates made

using an amine-cured DGEBA. The KIc of the hybrid with 7.3 wt% CTBN and 3.7
wt% SiO2 nanoparticles was improved by 132 %.
Kinloch et al. [2.5.23] investigated glass fiber composites based on an

anhydride-cured DGEBA resin matrix.
43
A resin modification with 9 wt% CTBN and 10 wt% silica nanoparticles increased
the composite laminate GIc from 330 J/m2 to 860 J/m2; which is an improvement
of 160 %. The composite mode II fracture energy (GIIc) of the hybrid was 46 %
higher compared to the unmodified control.
Manjunatha et al. [2.5.24] tested the fatigue performance of glass fiber-

reinforced composites. They reported a tenfold increase in the number of cyclic
loadings before laminate failure occurred for a hybrid epoxy system based on an
anhydride-cured DGEBA modified with 9 wt% CTBN and 19 wt% nanosilica.
Tsai et al. [2.5.16], [2.5.26] formulated hybrid resin systems with CTBN as well
as with core-shell particles. The loss in modulus was in both cases reduced
signifcantly by the addition of 10 wt% nanosilica. The bulk GIc of the hybrid was
improved by modification with 10 wt% CTBN by 390 % and with 10 wt% core-
shell particles by 442 %. However, the composite GIc of a glass fiber-reinforced
laminate was improved only by 48 % for CTBN and by 82 % for the core-shell
particles system.
The very interesting question how much of a property improvement of the bulk
resin may be transferred into the composite laminate has already been studied in
the past for different rubber-toughened systems [2.5.30]. In Figure 2.5.10 the
relationship between bulk GIc and laminate GIc is given.
Figure 2.5.10: Bulk GIc of cured epoxy resins versus laminate GIc [2.5.30]
It is obvious that the increase in laminate fracture energy with increasing cured
bulk epoxy resin fracture energy is not proprtional.
44
First there is a significant increase in laminate GIc up to a neat resin fracture
energy of approximately 250 J/m2, then the slope of the curve flattens. It then
takes considerable increases in bulk GIc to improve the laminate fracture
toughness just by a small amount.
According to Altstädt these two different areas are the result of the constriction
of the plastic deformation zone in the epoxy resin matrix between the
reinforcement [2.5.30]. Tougher resins have a more extensive deformation zone
ahead of the crack tip than brittle resins.
Of course the question arises, will similar relationships be found to exist for the
hybrid epoxy resins with silica nanoparticles present, and this will be discussed
extensively in chapter 3.4.
It is to be expected that brittle, untoughened resins will behave similar -

regardless of the nature of the epoxy resin used and the curing agent employed.
However, there might be differences for the very tough hybrid epoxy resin
systems.
2.5.5. References
[2.5.1] Odegard, G.M., Clancy, T.C., Gates, T.S. (2005) Modeling of the
mechanical properties of nanoparticle/polymer composites. Polymer
46, 553-563
[2.5.2] Sprenger, S., Eger C., Kinloch A.J., Taylor, A.C. (2003) Nano-
toughening of Epoxies. Proceedings of Stick! Conference April 9th,
Nürnberg, Germany, Vincentz Verlag 2003
[2.5.3] Picture courtesy of IVW Kaiserslautern and Polymerservice

Merseburg, Germany
[2.5.4] Product information (2013) Evonik Hanse GmbH, Geesthacht,

Germany
[2.5.5] Sprenger S., Kinloch A.J., Taylor, A.C., Mohammed, R.D. (2005)
Rubber-toughened GFRCs optimized by nanoparticles. JEC Compos.
Magazine 21, 66-69
[2.5.6] Eger C., Schultz, P. (2005) Reinforcing epoxy resins with silica nano-
particles. In Proceedings of "High Performance Fillers 2005", March
8-9, Köln, Germany
[2.5.7] Rosso, P., Ye, L., Friedrich, K., Sprenger, S. (2006) A toughened
epoxy resin by silica nanoparticle reinforcement. J. Appl. Polym. Sci.
100, 1849-1855
[2.5.8] Johnsen, B.B., Kinloch, A.J., Mohammed, R.D., Taylor, A.C.,

Sprenger, S. (2007) Toughening mechanisms of nanoparticle-
modified epoxy polymers. Polymer 48, 530-541
45
[2.5.9] Liu, S., Zhang, H., Zhang, Z., Zhang, T., Sprenger, S. (2008)
Tailoring the mechanical performance of epoxy resin by various
nanoparticles. Polymers & Polymer Composites Vol. 16, No. 8, 471-
477
[2.5.10] Prasad, S., Grover T., Basu, S. (2010) Coarse-grained molecular

dynamics simulation of cross-linking of DGEBA epoxy resin and
estimation of adhesive strength. Int. J. Eng., Sci. Technol. 2, 17-30
[2.5.11] Gou, J., Fan, B., Song, G., Khan, A. (2006) Study of affinities
between single-walled nanotube and epoxy resin using molecular
dynamic simulation. Int. J. Nanosci. 5, 131-144
[2.5.12] Sprenger, S., Eger C. (2003) WO 2004081076 resp. EP 1 457 509
[2.5.13] Sprenger, S., Eger, C. Kinloch, A.J., Lee, J.H., Taylor, A.C., Egan, D.
(2003) Nanoadhesives: Toughness and high strength. Adhäsion,
Kleben & Dichten 03/2003, 24-30
[2.5.14] Army Research Laboratory Technical Report 4084 (2007)

Aberdeen Proving Ground, MD, U.S.A. April 2007
[2.5.15] Picture courtesy of J. Robinette, Army Research Laboratory,

MD, U.S.A.
[2.5.16] Tsai, J.-L., Huang, B.-H. Cheng, Y.-L. (2009) Enhancing Fracture
Toughness of Glass/Epoxy Composites by Using Rubber Particles
Together with Silica Nanoparticles. J. Comp. Mater. 43, 3107-3123
[2.5.17] Sprenger, S., Kinloch, A.J., Taylor, A.C., Mohammed, R.D., Eger, C.
(2005) Rubber-toughened FRCs optimized by nanoparticles. JEC
Comp. Mag. No 19, 73–76
[2.5.18] Final Report Stiftung Industrieforschung Projekt S 657 (2005),

Institut f. Verbundwerkstoffe, Kaiserslautern, Germany
[2.5.19] Liu, H.Y., Wang, G.T., Mai, Y.W. Zeng, Y. (2011) On fracture
toughness of nanoparticle modified epoxy. Composites: Part B 42,
2170-2175
[2.5.20] Mahrholz, T., Herbeck, L., Riedel U. (2004) New high-performance

fibre-reinforced nanocomposites. JEC Comp. Mag. No 9, 71-75
[2.5.21] Sprenger, S., Kinloch, A.J., Taylor, A.C., Mohammed, R. (2007)

Rubber-toughened CFRCs optimized by nanoparticles Part III. JEC
Compos. Mag. No 30, 54-57
[2.5.22] Caccavale V., Wichmann, M., Quaresimin, M., Schulte, K. (2007)

Nanoparticle/Rubber Modified Epoxy Matrix Systems: Mechanical
Performance in CFRPs. Proceedings of AIAS XXXVI Convegno
Nazionale, 4-8 September, Ischia, Neapel, Italien
46
[2.5.23] Kinloch, A.J., Masania, K., Taylor, A.C., Agarwal, R., Sprenger, S.,
Egan, D. (2008) The fracture of glass-fibre-reinforced epoxy
composites using nanoparticle-modified matrices. J. Mater. Sci.
Letters 43, 1151-1154
[2.5.24] Manjunatha, C.M., Taylor, A.C., Kinloch, A.J., Sprenger, S. (2009)

The effect of rubber micro-particles and silica nanoparticles on the
tensile fatigue behaviour of a glass-fibre epoxy composite. J. Mater.
Sci. Letters 44, 342-345
[2.5.25] Tsai, J.L., Huang, B-H., Cheng, Y.L. (2011) Enhancing Fracture
Toughness of Glass/Epoxy Composites for Wind Blades Using Silica
Nanoparticles and Rubber Particles. Procedia Engineering 14, 1982-
1987
[2.5.26] Tang, Y., Ye, L., Zhang, D., Deng, S. (2011) Characterization of
transverse tensile, interlaminar shear and interlaminate fracture in
CF/EP laminates with 10 wt% and 20 wt% silica nanoparticles in
matrix resins. Composites: Part A, 42, 1943-1950
[2.5.27] Sprenger, S., Eger, C., Kinloch, A., Mohammed, R., Taylor, A.
(2005) Rubber-toughening and Nanoparticles in Epoxies: Synergies
in FRC. Proceedings SAMPE, Paris, France, April 5 – 7, 2005
[2.5.28] Brandt, J., Drechsler, K., Schmidtke, K. (2004) Composites für den
Flugzeugbau. Kunststoffe 10, 290-294
[2.5.29] Caccavale, V. (2007) Nanoparticle/rubber modified epoxy matrix

systems: mechanical performance in CFRPs. Master of Science
Thesis 2007, University of Padua, Italy
[2.5.30] Altstädt, V. (1991) Effect of the polymer matrix on the properties of

advanced composites. Makromolekulare Chemie, Macromolecular
Symposia 50, 137-145
47
3. Results and Discussion
3.1. Property improvements of epoxy resins modified with

SiO2 nanoparticles
In this chapter first all property improvements of cured epoxy resins achievable
by the addition of surface-modified SiO2 nanoparticles will be investigated. The
influence of the particle modification on the cure performance of epoxy resins is
of interest as well. Furthermore it will be investigated if there are differences or
commonalities for the different resin/hardener combinations. The mechanisms
relevant for property improvements will be highlighted and the question
regarding the optimum addition level will be answered. With the gained
knowledge of the behavior of silica nanoparticles in bulk epoxy resin systems the
behavior in fiber-reinforced laminates can be studied later on (see chapter 3.3.).
3.1.1. Abstract
Surface-modified silica nanoparticles, 20 nm in size and with a very narrow

particle size distribution, have been available as concentrates in epoxy resins in
industrial quantities for the last 10 years. They can be used in epoxy resin
formulations to improve many different properties including the strength,
modulus, toughness and fatigue performance. In this review, I examine the
literature published in the last decade, compare results with a focus on
mechanical properties, and discuss the mechanisms responsible for property
improvements.
3.1.2. Introduction
Epoxy resins are very versatile raw materials for industrial products, from
windmill blades to highly sophisticated aerospace parts such as wings or fuselage
to coatings and adhesives for construction. They are used in large volumes in
generator encapsulations and in microlectronics and UV-cured electronic
adhesives. Structural adhesives for automotive or aerospace applications,
shipbuilding or windmill blade construction are mainly based on epoxy resins.
Therefore, quite a range of different epoxy resins is available: from low viscous
short-chain aliphatics such as the diglycidyl ether of hexanediole to high
performance, multifunctional aromatic resins such as the triglycidyl ether of
aminophenole or tetraglycidyl ether of methyl dianiline (TGMDA). Of course, the
diglycidyl ether of bisphenol F (DGEBF) and its higher molecular weight
variations, epoxidized novolac resins, are of great importance. However, by far,
the biggest volume of epoxy resins is produced as diglycidyl ether of bisphenol A
(DGEBA), the workhorse of the epoxy industry.
With regard to the many different applications and their substantially different
property profiles for the materials used, a big variety of hardeners is used in
industrial applications as well. An excellent and very comprehensive overview of
the different hardeners used in the industry, their chemical natures and their
network formation was published by Hare some years ago [3.1.1], [3.1.2.].
48
Sterically hindered aromatic amines are especially suitable for densely
crosslinked, high-glass-transition-temperature (Tg) aerospace formulations; they
are typically used in combination with trifunctional or tetrafunctional epoxy
resins.
In construction applications, low viscosity, fast-curing amines are preferred.

When they are slowed down to a certain extent, they are the hardener of choice
for room-temperature (RT) curing adhesives. Nonstochiometric hardeners such
as dicyandiamide are used in large quantities by the industry for one-part heat-
curing structural adhesives.
Hardeners appropriate for composites manufactured by injection methods are

mainly amine-based as well, most commonly on isophorone diamine or a
combination with short-chain aliphatic poly(ether amines) which tend to have
lower crosslink densities but somewhat tougher networks.
Acid anhydrides, being part of the formulation in almost equal amounts to epoxy
resins and exhibiting a very low viscosity, are very useful in highly filled
encapsulation systems and composites made by filament winding as they lower
the viscosities significantly. On curing, they form medium crosslinked resin
systems.
In quite a few epoxy resin formulations, fillers are necessary. Fillers improve
mechanical properties such as strength, stiffness and modulus. However, they
have a negative impact on the viscosity of the resin, which forbids their use in
some applications. Furthermore, they are filtered out by the fabric when the
resin formulation containing the filler is subjected to injection manufacturing
methods for fiber-reinforced composites. Thus, many applications whose
performance could be improved by the use of a filler do not permit the use of
classical micrometer-sized fillers, and of course, fillers cannot be used in
transparent applications.
49
Figure 3.1.1: Transmission electron microscopy picture of cured epoxy resin with
5 wt % nanosilica
In the years 2002 and 2003, the first commercial-grade surface-modified silica
nanoparticles were introduced into the market. They were manufactured in situ
directly in the epoxy resin by a modified sol-gel process and had an average
particle size of 20 nm and a very narrow particle size distribution. Odegard et al.
[3.1.3.] showed in their article molecular models of such particles and the huge
amount of hydroxyl groups on the particle surface. The industrially manufactured
particles were surface coated. The hydroxyl groups were reacted with silanes to
prevent agglomeration and to compatibilize the particles with the resin.
Nevertheless, there were still some remaining free hydroxyl groups, and thus,
the particles were slightly acidic. One needs to keep this in mind when looking at
the different additives for epoxy resins. The industrial material is very close to
the model: isolated spherical particles as concentrates in epoxy resins with an
average size of approximately 20 nm. Figure 3.1.1 shows a cured epoxy resin
with such silica nanoparticles (5 wt %).
They offer several advantages: being 20 nm small and completely monodisperse,

they do increase the resin viscosity only slightly at higher concentrations. In
contrast to fumed silica, they exhibit no thixotropic properties but behave like a
Newtonian liquid. Because of their size, they are transparent and can easily
penetrate even close-meshed fabrics in composite manufacturing when they are
injected.
Consequently, they are a very attractive raw material for epoxy resin
formulators.
50
Today, 10 years later, they are used in many industrial formulations, including
encapsulating resins, adhesives, and composites such as automotive parts and
machine parts. They improve various properties, including strength, modulus,
stiffness, toughness, and scratch resistance. Significant improvements in the
fatigue performance were reported when the epoxy resin was modified with
nanosilica.
Nevertheless, it is difficult to determine the optimum addition level on the

function of a resin and hardener system of choice. Different and sometimes even
contradicting results have been published. The aim of this review is to give a
comprehensive overview of the actual state of research with a focus on
mechanical properties and to provide formulating guidelines.
3.1.3. Discussion
Unless mentioned otherwise, the researchers cited used commercial 40 wt %

concentrated masterbatches of surface-modified nanosilica in DGEBA with an
average particle size of 20 nm and a very narrow particle size distribution. These
were then diluted down with commercial epoxy resins to vary the nanosilica
concentrations.
The dispersion of the nanosilica was investigated by all researchers and was
always found to be homogenous. Agglomerates or areas with different silica
nanoparticle concentrations were not observed.
This behaviour was mainly due to the surface coating of the particles as uncoated
particles tend to agglomerate. In some rare cases, when nanosilica-containing
epoxy resins were cured with amine-functional reactive liquid rubbers, some
agglomeration was found. Such exceptions are described in another article
currently under preparation.
Tg was not influenced in most cases and sometimes decreased by 1-2 °C at very
high addition levels of nanosilica. Sanctuary at al. [3.1.4] investigated the
complex specific heat capacity and reported that a blend of DGEBA and surface-
modified nanosilica behaved in a neutral manner with regard to the glass
transition dynamics of the resin matrix, just like a mixture.
3.1.3.1. Amine-cured epoxy resins
Aliphatic and Cycloaliphatic Amines as Hardeners
Rosso et al. [3.1.5] investigated the property improvements of piperidine-cured

DGEBA by the modification with 5 wt % nanosilica. Although the tensile strength
remained unchanged, the tensile modulus was increased by more than 20%. The
fracture toughness (KIc) was improved by 70%, and GIc was improved by more
than 140%.
In a continuation of this work, Wetzel et al. [3.1.6] explored the fracture and
toughening mechanisms using Al2O3 and TiO2 nanoparticles of similar sizes (ca.
20 nm) in 4,4'-methylene bis(2-methylcyclohexyl-amine) cured DGEBA.
51
They identified crack deflection processes, crack pinning, and energy dissipation
rather than debonding at the particle-resin interface as reasons for the
toughness improvements. A very detailed description is given in Wetzel´s Ph.D.
thesis [3.1.7].
The fracture behaviour of piperidine-cured DGEBA with various nanosilica

concentrations at high and low temperatures was reported by Deng et al. [3.1.8].
They found that the toughness increased significantly at RT and 50°C with a
maximum at approximately 5 wt % nanosilica. At 70°C, they found no increase
in modulus or toughness. At 0 and -50°C, the improvements were much smaller;
this indicated different mechanisms at different temperatures.
The curing kinetics of a modified DGEBA cured with piperidine were studied by
Rosso and Ye [3.1.9]. The addition of 1-5 vol % nanosilica led to a higher
reactivity in curing and an alteration in crosslinking. They suggested the
formation of an amino-rich interphase region around the silica nanoparticles,
which could have been responsible for the property improvements.
Haupert et al. [3.1.10] looked into the tribological properties of DGEBA cured
with an aliphatic amine. They found improvements at nanosilica addition levels
above 2 vol % and a maximum at 5.5 vol %. The wear resistance was improved
by 30%.
DGEBA and DGEBF cured with an accelerated aminoethyl piperazine were the
subject of studies by Dittanet [3.1.11]. She reported an increase in the modulus
of 108% for DGEBA and 90% for DGEBF at a 30 wt % addition level of
nanosilica. A significant reduction in coefficient of thermal expansion (CTE) was
reported as well.
The cyclic fatigue properties of a piperidine-cured DGEBA was studied by Mai et

al. [3.1.12]. They found a fatigue life improvement of 145% with 2 wt %
nanosilica, an even slightly higher improvement at 6 wt % nanosilica, but only a
56% improvement at 10 wt % nanosilica. Apparently, there is no linear
relationship between the silica content and increased fatigue performance, but a
maximum seems to exist at a certain addition level.
Liang and Pearson [3.1.13] investigated the toughening mechanisms of

piperidine-cured DGEBA. In addition to 20-nm silica nanoparticles they used 80-
nm particles from a small-scale manufacturer, which might have had a different
surface modification. Neither particle sizes influenced Tg (≤ 24.6 wt % of silica).
The modulus increased by approximately 20% for both particle sizes. The
compressive property and toughness increase were nearly identical as well. The
authors concluded that the influence of particle size was negligible in the range
of 20-80 nm.
The interactions between silica nanoparticles and diethylene triamine cured

DGEBA before and during network formation was the subject of research of
Baller et al. [3.1.14]. In the first stage of isothermal curing, there was no
difference between epoxy resins with different nanosilica contents, whereas later
in curing, the reaction rate was reduced, probably due to the reduced mobility of
the matrix with increased nanoparticle content.
52
This study was continued by Philipp et al. [3.1.15] and the generalized Cauchy
relation was investigated. It seems that the cured epoxy resins with different
amounts of nanosilica incorporated behaved similarly to porous silica glasses,
and this indicated a perfect distribution of monodisperse silica nanoparticles.
Tsai et al. [3.1.16] investigated nanosilica-containing DGEBA cured with a

modified isophorone diamine. The modulus was increased up to 19% and the KIc
by 81%. with 40 wt % nanosilica. Again, the improvements increased with
increasing nanoparticle addition.
Furthermore, Tsai and Chang [3.1.17] explored the damping properties of

isophorone-diamine-cured DGEBA and reported slightly improved damping
properties (+ 3.24%) at 10 wt % nanosilica addition.
Ye et al. [3.1.18] reported increases in the modulus from 2.9 to 3.3 GPa (with 10
wt % nanosilica) and 3.6 GPa (with 20 wt % nanosilica) for a DGEBA resin cured
with piperidine. GIc was increased from 238 to 458 and 666 J/m2 (improvements
of 92 and 180%, respectively).
Liu et al. [3.1.19] looked further into KIc of piperidine-cured DGEBA. They found
an increase in the modulus and KIc with increasing loading level: 22% increase in
the modulus and 304% increase in GIc at 20 wt % nanosilica.
In another study, Liu et al. [3.1.20] examined cyclic fatigue crack propagation
and reported significant improvements in the fatigue lifetimes for 6 and 12 wt %
nanosilica. They discussed extensively the contribution of the different
toughening mechanisms identified at high and low loading levels.
In continuation of earlier work, Dittanet and Pearson [3.1.21] tried to identify the
influence of the nanoparticle size on the toughening of a piperidine-cured DGEBA
epoxy resin. They used particles with average sizes of 23, 74 and 170 nm from a
small-scale manufacturer and reported improved properties with increasing
addition levels of nanosilica. The modulus was increased by approximately 60%
at 30 vol % nanosilica addition regardless of the particle size. GIc was improved
by 221% for the 23 nm particles, 239% by the 170 nm particles and 317% by
the 74 nm particles. Interesting was the reduction of the CTE size dependence as
well; the 23 nm particles performed best.
Mechanisms for the toughening effect of nanosilica were discussed as well, and
the model from Kinloch et al. [3.1.22] was confirmed; see the next two sections
in this article. Matrix shear banding was the dominant mechanism, matrix void
growth was secondary, and the debonding of silica nanoparticles had only a
minor effect.
Table 3.1.1 gives an overview of the increases in modulus and KIc (at RT) versus
addition levels of nanosilica. The same particles were used together with
piperidine as a hardener, and identical cure conditions were used.
As a short summary, I concluded that the tensile strength remains more or less
unchanged by the addition of silica nanoparticles. At very high addition levels,
there might by a slight increase [3.1.19]. The modulus increases with increasing
concentrations of silica nanoparticles.
53
However, toughness and fatigue improvements have been either reported to
increase steadily or have a maximum at 5-6% loading level.
SiO2 Modulus (GPa) KIc (MPam1/2)

content
(wt%)
0 2.80 2.86 2.86 0.967 0.89 0.95
± 0.03 ± 0.11 ± 0.08 ± 0.07 ± 0.03
2 2.89 2.90 2.88 1.01
± 0.07 ± 0.06 ± 0.03 ± 0.04
4 2.98 2.93 1.14
± 0.15 ± 0.03 ± 0.06
5 1.66
± 0.11
6 2.94 2.98 2.98 1.26
± 0.07 ± 0.08 ± 0.10 ± 0.04
8 3.18 3.10 1.43 1.39
± 0.12 ± 0.15 ± 0.07
10 3.14 3.14 1.57
± 0.14 ± 0.14 ± 0.02
12 3.20 1.70
± 0.05 ± 0.05
20 3.48 2.11
± 0.14 ± 0.01
Reference [3.1.8] [3.1.12] [3.1.19] [3.1.5] [3.1.8] [3.1.19]
Table 3.1.1: Properties of piperidine cured epoxy resins with various nanosilica
contents
Poly(ether amines) as Hardeners
Ma et al. [3.1.23] reported for DGEBA cured with a difunctional short-chain

poly(ether amine) an increase in the modulus by 32% at a 10 wt % addition
level of nanosilica. At a 20 wt % addition level, the modulus increased by 40%.
GIc increased by 110 and 274%, respectively. By extensive microscopic work, the
initiation and development of a thin dilatation zone and nanovoid formation were
identified as the dominant toughening mechanisms.
Kinloch et al. [3.1.22] investigated DGEBA and a DGEBA/DGEBF blend cured with
a difunctional short-chain poly(ether amine). They reported only very small
increases in the modulus (17 and 10%, respectively) with 20 wt % nanosilica.
Toughness by means of GIc was improved in both cases by approximately 280%.
A linear increase with increasing addition level was found. The toughening
mechanisms were investigated and compared with theoretical predictions.
Localized plastic shear bands initiated by the stress concentrations around the
periphery of the silica nanoparticles were the main contributor to the increase in
toughness. The debonding of the nanoparticles seemed to be less important, as
only approximately 15% of the nanoparticles were found to debond.
54
However, the plastic void growth following the debonding contributed to the
toughness increase.
DGEBA modified with various amounts of nanosilica and cured with the
difunctional short-chain poly(ether amine) was studied by Tsai et al. [3.1.16] as
well. They found exactly the same 17% improvement in the modulus at 20 wt %
nanosilica like Kinloch et al. [3.1.22] and a 40% improvement at a 40 wt %
loading level. The strength was slightly improved at the 40 wt % level. Three-
point bending tests showed an improvement in the flexural strength with
increasing addition of silica nanoparticles up to 16%. The toughness increase was
found to be very small because of the fact that KIc of the unmodified resin was
quite high. Improvements reached a plateau at approximately 10 wt %
nanosilica. One has to take into account the fact that the curing conditions were
different.
The work of Jajam and Tippur [3.1.24] focused on a DGEBA blended with 15% n-
butyl glycidyl ether cured with a commercial hardener formulation consisting of
poly(ether amine), trimethyl hexane diamine, benzene-1,3-dimethane amine,
nonyl phenol, and substituted phenol. In addition to nanosilica, they tested
micrometer-sized spherical glass particles with a mean diameter of 35 µm. They
found a linear increase of KIc with increasing addition level for both particles. At
10 vol %, the nanosilica provided a 78% enhancement relative to the 35 µm
glass particles. In dynamic fracture tests, both materials showed improved
dynamic KIc values with increasing loading levels. However, the nanosilica
showed only minor improvement of 34% at a 10 vol % addition level. In another
study [3.1.25], it was confirmed that the addition of nanosilica did not
necessarily improve the toughness when a fast impact occured. Nevertheless, it
was shown that quite significant improvements could be achieved when
commercial resin systems were used, with the hardeners typically being complex
amine blends. The effects found for nanosilica modification of epoxy resins have
not been limited to model systems.
Aromatic Amines as Hardeners
Kinloch et al. [3.1.22] also investigated a high performance, high Tg epoxy resin
system similar to the industrial benchmark RTM6. Nanosilica filled TGMDA was
cured with a blend of 4,4´-methylenbis(2,6-diethylaniline) and 4,4´-
methylenebis(2,6-diisopropylaniline). At 10 wt % nanosilica loading level, the
modulus was increased by 26%; and GIc was increased by 146%, although it was
still at a very low level of 172 J/m2.
DGEBA cured with 3,3´-diaminodiphenyl sulfone (3,3´-DDS), tested by Rhoney

et al. [3.1.26] et al showed a reduction in gel time with increasing silica levels
without much change in the cure profile. The Tg, determined by
thermomechanical analysis, was lowered from 163 to 146°C at approximately 33
wt % nanosilica. The CTE at 80° (below Tg) and 200°C (above Tg) were
measured, and a reduction of approximately 20% was found for approximately
33 wt % of nanosilica.
Ma et al. [3.1.23] studied DGEBA cured with 4,4´-DDS. The modulus was
increased by 18% at a 10 wt % addition level of nanosilica. Doubling the
addition level to 20% increased the modulus by 40% compared to the neat
epoxy resin. GIc was improved by 49 and 81%, respectively.
55
Transmission electron microscopy showed some dilatation in the propagated
crack-tip area and some nanovoid formation.
The research of Gurung [3.1.27] was based on a DGEBA cured with 4,4´-DDS as
well. An industrially available nanosilica epoxy masterbatch was compared to
silica nanopowder modified with aminopropyl triethoxysilane. Gurung reported an
acceleration of curing at the beginning of curing caused by the industrial
nanosilica as well. A significant drop in Tg from 173 to 130°C at approximately
33 wt % nanosilica was found in both thermomechanical analysis and differential
scanning calorimetry studies. The modulus was improved by 59%, but the stress
at break was reduced by 12%. The industrial material performed better than the
"homemade" nanosilica, and this was attributed to a better particle dispersion.
It is interesting to see the effects of the different network densities deriving from
the two different DDS molecules when they were modified with the same
nanoparticles with regard to the reduction of Tg.
3.1.3.2. Anhydride-cured epoxy resins
By far, most researchers have worked with anhydride curing agents for various
reasons, including their low viscosity and easily controlled curing cycle. The side
reactions can also be controlled and can be suppressed with a well-defined curing
schedule. Thus, a tremendous amount of test results is available in this field and
are described hereafter. All of the anhydride curing agents described were
accelerated with very small amounts (e.g., 1%) of ternary amines.
Methylhexahydrophthalic Acid Anhydride as a Hardener
- DGEBA epoxy resin
Eger and Schultz [3.1.28] reported a significant reduction in the viscosity when
nanosilica was used in combination with a DGEBA epoxy resin in place of fumed
silica. An amount of 25 wt % increased the modulus by 37% and KIc by 72%. A
slight increase in tensile strength was reported as well.
The toughening mechanisms involved were investigated thoroughly by Taylor et

al. [3.1.29]. The Tg, determined by differential scanning calorimetry and DMTA,
was not affected by the addition of silica nanoparticles. The modulus was
increased by 30% at a 20.2 wt % loading level. KIc was improved by 141%.
Crack pinning and crack deflection, found often as toughening mechanisms when
larger particles are used to toughen epoxy resins, were ruled out. Localized shear
banding might have delivered a minor contribution, but debonding of the
nanoparticles and subsequent plastic void growth were mainly responsible for the
increase in toughness.
Kinloch et al. [3.1.30] looked into the fatigue performance of such systems. The
modulus was increased by 30% and KIc by 73% at 20.2 wt % nanosilica.
Cyclic fatigue testing was applied to the compact tension test specimen, and it
was found that the addition of nanosilica clearly and significantly improved the
fatigue performance: the more, the better.
56
Zhang et al. [3.1.31] reported an almost linear increase in the modulus with
increasing silica nanoparticle content, from 2.75 to 3.95 GPa at a 14 vol %
loading level (an improvement of 44%). KIc was improved by more than 50%.
The tensile strength was improved only slightly. They reported an observation of
a polymer shell around the inorganic particle, as suggested by Wetzel et al.
[3.1.6] for other nanoparticles and Zhang et al. [3.1.32] for SiO2-nanoparticles.
In another article, Zhang et al. [3.1.33] compared industrial silica nanoparticles

manufactured by the sol-gel process to fumed nanosilica. The fumed nanosilica
increased the viscosity; the maximum loading was 6 vol %. At this
concentration, the modulus was increased by 17%, and KIc was increased by
49%. Microscopy revealed particle clusters of 100-200 nm in size. A similar
loading level of monodisperse nanosilica yielded the same improvement for the
modulus but only a 29% increase of KIc. A further increase in the nanosilica
concentration increased the modulus and KIc further.
Kinloch and coworkers [3.1.22], [3.1.34] observed for a 20 wt % silica

nanoparticle modification an increase in the modulus of 30% and an increase in
KIc of 73%. A slight decrease in Tg was noted. The observed toughening
mechanisms were a debonding of the epoxy polymer from the silica nanoparticles
followed by plastic void growth of the epoxy. Localised plastic shear banding was
observed as well. They proposed a model to predict the toughening by nanosilica
that correlated well with the experimental data.
In another investigation, Taylor et al. [3.1.35] looked into the combination of

nanosilica and carbon nanotubes (MWCNTs). An addition of 6 wt % nanosilica
increased the modulus by 4% and KIc by 9%. Additionally, a concentration of
0.18 wt % of multiwalled carbon nanotubes increased the modulus by another 1
% and the KIc by another 40%. Several different toughening mechanisms were
identified.
Zhang et al. [3.1.36] found the modulus to increase by 31% at a nanosilica

addition level of 12 vol %. The strength was increased considerably, by 45%. KIc
was improved by 93%, and Tg was only slightly reduced.
Table 3.1.2 gives an overview of the modulus and KIc (at RT) values versus the
addition level of nanosilica. The same particles were used in DGEBA as epoxy
resin and were cured with methyl hexahaydrophthalic acid anhydride. The curing
conditions varied slightly.
57
SiO2 Modulus (GPa) KIc (MPam1/2)
(wt
%)
0 2.96 2.75 2.96 3.01 0.59 0.55 0.51 0.46
± 0.05 ± 0.05 ± 0.12 ± 0.04
2 2.79 0.57
± 0.09 ± 0.11
4 3.20 3.20 1.03 0.64
5 3.00 3.39 0.68 0.68

± 0.07 ± 0.19 ± 0.05 ± 0.02
8 3.42 3.42 1.17 0.79
10 3.24 3.42 0.71 0.74

± 0.04 ± 0.34 ± 0.05 ± 0.04
14.5 3.56 3,63 0.75 0.80
± 0.05 ± 0.11 ± 0.05 ± 0.03
15 3.60 3.60 1.29 0.83
20 3.85 3.85 3.95 1.42 0.88 0.89

± 0.20 ± 0.05
22.8 3.95 0.83
± 0.12 ± 0.05
Ref. [3.1.29] [3.1.33]* [3.1.34] [3.1.36]* [3.1.2] [3.1.33]* [3.1.34] [3.1.36]*

[3.1.30] [3.1.30]
*Recalculated from the volume percentage
Table 3.1.2: Properties of anhydride-cured epoxy resins with various contents of

nanosilica
In short, I concluded that the tensile strength was slightly improved by the
addition of silica nanoparticles at high loading levels. The modulus and toughness
increased with increasing concentrations of silica nanoparticles. There seems to
be no maximum addition level.
- DGEBF epoxy resin
The effects of the modification of DGEBF with nanosilica were investigated by

Zhang et al. [3.1.37]. With increasing nanoparticle content, property
improvements were found. At 15 vol %, the modulus was increased by 48%, the
strength was increases by 8%, the impact energy was increased by 30%, and KIc
was increased by 77 %. When testing was done at 80°C instead of RT, the
improvements in the strength and modulus were in the same range; the KIc,
however, was increased by 125%. Zhang et al. reported different fracture
behaviours for the different temperatures. At RT, many dimples were found,
whereas at 80°C, a larger smooth zone on the fracture surface was observed. As
dominant toughening mechanism, an enhanced local deformability around the
crack tip induced by the silica nanoparticles was identified.
58
Gu et al. [3.1.38] examined the mechanical and tribological aspects of modified
DGEBF. The hardness and modulus increased nearly linearly with increasing
nanosilica content. When 15 vol % nanoparticles were used, the modulus was
improved by 40% and the hardness was improved by 33%. Plasticity index
decreased first, showed a minimum at 8 vol % nanosilica, and then increased.
The friction coefficient showed similar behaviour.
- Cycloaliphatic epoxy resin
Eger and Schultz [3.1.28] studied the properties of 3,4-epoxy cyclohexylmethyl-

3,4-epoxycyclohexane carboxylate (EEC) modified with silica nanoparticles and
fumed silica. At a 40 wt % loading level, the viscosity increased from 0.23 to 2.1
Pas for the silica nanoparticles and to 43 Pas for the fumed silica. At 22 wt %,
the strength remained unchanged, and the modulus was increased by
approximately 40% for both modifications. Fumed silica increased KIc by 45%;
silica nanoparticles increased it by 53%. A reduced water absorption was
reported.
Zhang et al. [3.1.32] investigated EEC as well. With a 22.7 wt % silica content,
the modulus was found to increase from 3.05 to 4.18 GPa or by 37%. KIc was
improved by 76%. Interestingly, the impact strength (by Charpy impact testing)
went from 25.6 to 31.4 J/m2 at 5.3 wt% nanosilica content to 23.7 J/m2 at 22.7
wt % loading. Apparently, a maximum existed. The strength was not affected by
the modifications. As a main toughening mechanism, the energy dissipation
caused by nanoparticle-induced dimples was claimed.
Bai et al. [3.1.39] applied dynamic nanoindentation to investigate the triboelastic

properties of modified EEC. The storage modulus increased with increasing
nanosilica content. The increase was not linear, with a rapid increase first and
then a moderate increase at higher loading levels. The absolute values were
higher than those found when DMTA or three-point bending tests were used.
Methyltetrahydrophthalic Acid Anhydride as a Hardener
The industrial curing agent used contained a certain amount of tetra-

hydrophthalic acid anhydride and was accelerated with 1-methyl imidazole.
Mahrholz et al. [3.1.40] explored achievable property improvements by
nanosilica modification regarding the use of modified resins for injection
techniques where low viscosities are indispensable.
They found a slight increase in tensile strength and an increase in the modulus
with increasing nanosilica content. At a 25 wt % loading level, an improvement
of 36% was found. The impact strength, determined by a Charpy test, showed a
maximum improvement at a content of 15 wt % nanoparticles and decreased
upon further addition of nanosilica. Shrinkage was reduced by 19% and the
thermal conductivity increased with increasing nanosilica content.
Duwe et al. [3.1.41] described a strong increase in the modulus of approximately

50% at 25 wt % nanosilica at RT and at 50°C. Further, they reported an increase
in thermal conductivity. In comparison to the SiO2, they investigated AlN and
boehmite modifications as well.
59
Methylnadic Acid Anhydride as a Hardener
Hodzic et al. [3.1.42] used benzyldimethylamine as an accelerator for an

anhydride curing agent in their study of static uniaxial compression. In contrast
to that of classic micrometer-sized fillers, the addition of nanoparticles enhanced
the compressive stress-strain behaviour of the cured epoxy resin. A
concentration of 13.6 wt % nanosilica increased the compressive modulus by
19%, the compressive strength by 33%, and the strain at break by 76%. Further
increases in the nanoparticle concentration caused further property
improvements.
In another study, Hodzic et al. [3.1.43] employed both cylindric and prismatic
test specimens. In both cases, the test specimen with the highest addition level
of nanosilica showed the best performance. Enhanced shear deformation of the
matrix and the formation of shear bands that influenced the crack propagation
were identified as mechanisms of action of the nanoparticles.
3.1.3.3. Comparision between different types of hardeners
Flemming et al. [3.1.44] summarized the common understanding that a higher

crosslink density leads to a higher Tg and typically a higher modulus as well as a
reduced elongation at break and an increased brittleness.
The Tg´s reported for the DGEBA/isophorone diamine system were around 80°C.
Cured DGEBA/piperidine systems exhibited Tg´s between 80 and 100°C. Short-
chain aliphatic poly(ether amine)s as curing agents achieved Tg´s around 80°C
as well. For DGEBA crosslinked with methylhexahydrophthalic acid anhydride, the
Tg´s were found to range between 150 and 164°C. A 4,4´-DDS hardener yielded
in combination with DGEBA Tg´s between 163 and 173°C. DDS or other aromatic
amines used in combination with tetrafunctional epoxy resin (TGMDA) can
achieve Tg´s of up to 260°C.
In Figure 3.1.2, the modulus data from Tables 3.1.1 and 3.1.2 and from some
literature are given [3.1.16], [3.1.22], [3.1.23], and a linear interpolation is
applied. The epoxy resin used for all systems was DGEBA. The increase in
modulus with increasing nanosilica content was larger for hardeners that form a
close meshed network. Piperidine led to polymers with the lowest network
density. This was followed by the difunctional poly(ether amine) (here called D
230), methylhexahydrophthalic acid anhydride (MHHPA), and finally 4,4´-DDS.
60
5
D 230
4,5 Piperidine
Modulus (GPa)
MHHPSA
4
DDS
Linear (DDS)
3,5
Linear (MHHPSA)
3 Linear (D 230)
Linear (Piperidine)
2,5
0 5 10 15 20 25
wt% nanosilica
Figure 3.1.2: Moduli of epoxy resins with various contents of nanosilica cured
with different hardeners
Figure 3.1.3 shows the KIc as function of the nanosilica content for these different
hardeners. It is evident that hardeners forming a closer meshed network and
therefore yielding cured resins with a higher modulus tend to be more brittle and
thus having a lower toughness. Again, the toughness increases more or less
linearly with increased addition level of silica nanoparticles. The improvements at
the same loading level are bigger for systems that already exhibit a higher
toughness. This is consistent with the fact that toughening mechanisms are more
efficient for polymer matrices with a higher ductility.
2,5
Fracture toughness K1c (MPam0,5)
D 230
2 Piperidine
MHHPSA
DDS
1,5
Linear (DDS)
Linear (MHHPSA)
1
Linear (D 230)
Linear (Piperidine)
0,5
0 5 10 15 20 25
wt% nanosilica
Figure 3.1.3: KIc of epoxy resins with various contents of nanosilica cured with
different hardeners
As shown in Figure 3.1.3, the linear approach did not work for the rather flexible,
short-chain poly(ether amine) (D 230) as a curing agent.
61
Here, there seemed to be a maximum; this means that an optimal addition level
of nanosilica existed, apparently at lower addition levels. This was consistent
with other studies where such a curing agent was used in combination with
flexible amine-functional reactive liquid rubbers [3.1.45].
3.1.4. Conclusions
Taking into account the data gathered and published over the last 10 years, I
drew the following conclusions:
1. Nanosilica particles are monodisperse, and even at very high

concentrations, no agglomerates are found. The rheological properties of
the resins are not affected.
2. The particles behave like a filler and do not significantly change the curing
characteristics of epoxy resin/hardener blend. Toward the end of the
curing cycle, curing is slowed down to a certain extent.
3. Although the topology of the three-dimensional network formed by the

crosslinking reaction between an epoxy resin and a hardener is different in
the close vicinity of the silica nanoparticles, the general network structure
does not seem to change significantly. This could be deducted from the
fact that Tg remains the same as for the systems without nanosilica for
most curing agents. Furthermore, in most cases, the tensile strength of
the cured epoxy resin (determined by lap shear testing) does not change.
4. The modulus increases with increasing addition level of silica nanoparticles

in a nearly linear function. At 10 wt % nanosilica, an increase in the
modulus of 30-50% can be expected.
5. The compressive strength and the compressive modulus increase as well.

Improvements of 10-30% can be expected at a 10 wt % addition level of
silica nanoparticles.
6. KIc increases with increasing level of silica nanoparticles but not always in
a linear function. At 10 wt % nanosilica, an increase of approximately 50%
can be expected.
7. The toughening mechanisms are the debonding of the epoxy polymer from
the silica nanoparticles followed by plastic void growth. Localised plastic
shear banding contributes as well. Crack deflection does not seem to play
a significant role.
8. Fatigue performance is improved as well, but there seems to be no linear

relationship between the nanosilica content and the level of improvement.
Very probably, a maximum exists. At a 10 wt % addition level, an
improvement in the fatigue performance of 50-60% can be expected.
62
3.1.5. References
[3.1.1] Hare, C. J. Prot. Coat. Lin. (1994), 09, 77-103.
[3.1.2] Hare, C. J. Prot. Coat. Lin. (1994), 10, 197-213.
[3.1.3] Odegard, G.M.; Clancy, T.; Gates, T.S. Polymer (2005), 46, 553–
563.
[3.1.4] Sanctuary, R.; Baller, J.; Krüger, J.-K.; Schäfer, D; Wetzel, B.;
Possart, W.; Alnot, P. Thermochimica Acta (2006), 445, No. 2, 111
–115.
[3.1.5] Rosso, P.; Ye, L.; Friedrich, K.; Sprenger, S. J. Appl. Polym. Sci.
(2006), 100, 1849-1855.
[3.1.6] Wetzel, B.; Rosso, P.; Haupert, F.; Friedrich, K. Engineering

Fracture Mechanics (2006), 73, 2375-2398.
[3.1.7] Wetzel, B. "Mechanische Eigenschaften von Nanoverbundwerk-

stoffen aus Epoxydharz und keramischen Nanopartikeln"; PhD
Thesis. Technical University Kaiserslautern, Kaiserslautern, Germany
(2006)
[3.1.8] Deng, S.; Ye, L.; Friedrich, K. J. Mater. Sci. (2007), 42, No. 8, 2766
-2774.
[3.1.9] Rosso, P.; Ye, L. Macromol. Rapid Commun. (2007), 28, 121-126.
[3.1.10] Walter, R.; Haupert, F.; Schlarb, A. Tribologie + Schmierungs-

technik (2008), 2/55, 26-30.
[3.1.11] Dittanet, P. "The use of nanosilica in epoxy resins"; Master Thesis,

Lehigh University, Bethlehem, Pennsylvania, U.S.A. (2008)
[3.1.12] Wang, G.-T.; Liu, H.-Y.; Saintier, N.; Mai, Y.-W. Eng. Failure
Analysis (2009), 16, 2635-2645.
[3.1.13] Liang, Y.L.; Pearson, R.A. Polymer (2009), 50, 4895-4905.
[3.1.14] Baller, J.; Becker, N.; Ziehmer, M.; Thomassey, M.; Zielinski, B.;
Müller, U.; Sanctuary, R. Polymer (2009), 50, 3211-3219.
[3.1.15] Philipp, M.; Müller, U.; Jiménez Riobóo, R.J.; Baller, J.; Sanctuary,
R.; Possart, W.; Krüger, J.K. New Journal of Physics (2009), 11,
023015.
[3.1.16] Tsai, J.-L.; Huang, B.-H.; Cheng, Y.-L. J. Comp. Materials (2010),
4/44, 505-524.
[3.1.17] Tsai, J.-L.; Chang, N.-R. J. Comp. Materials (2011), 2/45, 2157-
2164.
63
[3.1.18] Tang, Y.; Ye, L.; Zhang, D.; Deng, S. Composites: Part A (2011),
42, 1943-1950.
[3.1.19] Liu, H.Y.; Wang, G.T.; Mai, Y.W.; Zeng, Y. Composites: Part B
(2011), 42, 2170-2175.
[3.1.20] Liu, H.Y.; Wang, G.T.; Mai, Y.W. Comp. Sci. Tech. (2012), 72, 1530
-1538.
[3.1.21] Dittanet, P.; Pearson, R. Polymer (2012), 53, 1890-1905.
[3.1.22] Hsieh, T.H.; Kinloch, A.J.; Masania, K.; Taylor, A.C.; Sprenger, S.
Polymer (2010), 51, 6284-6294.
[3.1.23] Ma, J.; Mo, M.-S.; Du, X.-S.; Rosso, P.; Friedrich, K.; Kuan, H.-C.
Polymer (2008), 49, 3510-3523.
[3.1.24] Jajam, K.C.; Tippur, H.V. Composites: Part B (2012), in press,

doi:10.1016 / j.compositesb.2012.01.042.
[3.1.25] Sprenger, S.; Kinloch, A.J.; Taylor, A.C.; Mohammed, R.D. JEC
Composites Magazine (2005), No 21, 66–69.
[3.1.26] Pethrick, R.; Miller, C.; Rhoney, I. Polym. Int. (2010), 59, 236–
241.
[3.1.27] Gurung, R. "Effects of nanosilica filler on the thermal and

mechanical properties of an epoxy/amine resin system"; Master
Thesis, Wichita State University, Wichita, Kansas, U.S.A. (2011)
[3.1.28] Eger C.; Schultz, P. Proceedings of "High Performance Fillers 2005",

March 8-9, Cologne, Germany (2005)
[3.1.29] Johnsen, B.B.; Kinloch, A.J.; Mohammed, R.D.; Taylor, A.C.;

Sprenger, S. Polymer (2007), 48, 530-541.
[3.1.30] Blackman, B.R.K.; Kinloch, A.J.; Sohn Lee, J.; Taylor, A.C.; Agarwal,
R.; Schueneman, G.; Sprenger, S. J. Mater. Sci. Letters (2007), 42,
7049–7051.
[3.1.31] Liu, S.; Zhang, H.; Zhang, Z.; Sprenger, S. J. Nanosci. Nanotechnol.
(2008), Vol.8, 1-6.
[3.1.32] Zjang, H.; Zhang, Z.; Friedrich, K.; Eger, C. Acta Materialica (2006),
54, 1833-1842.
[3.1.33] Liu, S.; Zhang, H.; Zhang, Z.; Zhang, T.; Sprenger, S. Polymers &
Polymer Composites (2008), Vol. 16, No. 8, 471-477.
[3.1.34] Hsieh, T.H.; Kinloch, A.J.; Masania, K.; Sohn Lee, J.; Taylor, A.C.;
Sprenger, S. J. Mater. Sci. (2010), 45, 1193-1210.
64
[3.1.35] Hsieh, T.H.; Kinloch, A.J.; Taylor, A.C.; Sprenger, S. J. Applied
Polym. Sci. (2011), 119, 2135-2142.
[3.1.36] Tang, L.-C.; Zhang, H.; Sprenger, S.; Ye, L.; Zhang, Z. Composites
Science and Technology (2012), 72, issue 5, 558-565.
[3.1.37] Zhang, H.; Tang, L.-C.; Zhang, Z.; Friedrich, K.; Sprenger, S.
Polymer (2008), 49, 3816-3825.
[3.1.38] Wang, Z.,Z.; Gu, P.; Zhang, Z.; Gu, L.; Xu, Y.Z. Tribol. Lett. (2011),
42, 185-191.
[3.1.39] Zhang, Y.-F.; Bai, S.-L.; Li, X.-K.; Zhang, Z. J. Polym. Sci. Part B:
Polym. Phys. (2009), 47, 1030-1038.
[3.1.40] Mahrholz, T.; Stängle, J.; Sinapius, M. Composites: Part A (2009),

40, 235-243.
[3.1.41] Duwe, S.; Arlt, C.; Aranda, S.; Riedel, U.; Ziegmann, G. Comp. Sci.
Tech. (2012), 72; 1324-1330.
[3.1.42] Jumahat, A.; Soutis, C.; Jones, F.; Hodzic, A. "The effects of
nanosilica contents on thermal and mechanical properties of epoxy
polymers"; Proceedings of 10th SAMPE Europe conference, 12-14
April, Paris, France (2010)
[3.1.43] Jumahat, A.; Soutis, C.; Jones, F.; Hodzic, A. J Mater. Sci. (2010),
45, No. 21, 5973-5983.
[3.1.44] Flemming, M. ; Ziegmann, G. ; Roth, S. Faserverbundbauweisen:

Fasern und Matrices, ISBN 3-540-58645-8 Springer-Verlag Berlin
Heidelberg Germany (1995), 212–215
[3.1.45] Sprenger, S.; Kinloch, A.J.; Taylor, A.C.; Hsieh, T.-H. Adhesion,
Adhesives & Sealants (2009), 10, 8–11
65
3.2. Property improvements of epoxy resins modified with
SiO2 nanoparticles and elastomers (hybrid systems)
3.2.1. Abstract
Epoxy resins are inherently brittle. Thus they are toughened with reactive liquid
rubbers or core-shell elastomer. Surface-modified silica nanoparticles, 20 nm in
diameter and with a very narrow particle size distribution, are available as
concentrates in epoxy resins in industrial quantities since 10 years. Some of the
drawbacks of toughening, like lower modulus or a loss in strength can be
overcompensated when using nanosilica together with these tougheners.
Apparently there exists a synergy as toughness and fatigue performance are
increased significantly. In this article the literature published in the last decade is
studied with a focus on mechanical properties. Results are compared and the
mechanisms responsible for the property improvements are discussed.
3.2.2. Introduction
Epoxy resins are used for many years in a multitude of industrial products, like
structural automotive adhesives, high performance fibre reinforced composites,
electrical and electronic applications, heavy duty protective coatings and many
more. However, they are very brittle and therefore in most commercial
formulations tougheners are used.
Since the seventies and eighties of last century the use of reactive liquid rubbers
as tougheners for epoxy resins became industrial standard. Carboxy terminated
butadiene acrylonitrile copolymers (CTBNs) are reacted with an excess of epoxy
resin to form a so-called adduct, an epoxy-rubber-epoxy terpolymer. These are
soluble in epoxy resins, whereas the pure rubber is not. Upon cure and
subsequent formation of the three-dimensional network, the rubber molecules
become insoluble again, phase separate and form small rubber domains or
particles within the cured polymer matrix. These rubber particles are chemically
linked to the polymer. Since the very early work of Kinloch and his team [3.2.1],
[3.2.2] the mechanisms of rubber toughening have been the subject of intensive
research and are well understood. An excellent review was published recently
[3.2.3].
However, the phase separation and domain formation depend on the cure speed,
cure temperature and the curing agent itself. The acrylonitrile content of the
copolymer has an influence on the particle size as well.
Furthermore not all long-chain rubber molecules participate in the phase

separation, some of them are crosslinked randomly into the epoxy polymer
matrix. Consequently the network density is lowered, which results in a lower
strength and a lower modulus, and, of course, in a lower glass transition
temperature (Tg). Another issue is the relatively high viscosity of epoxy resins
containing reactive liquid rubbers which prohibit some applications where low
viscosities are required.
66
To overcome these disadvantages core-shell elastomers (CSRs) have been
developed in the 1980s-1990s. Instead of forming a second phase upon cure the
rubber particles were added from the beginning. They consist of an elastomeric,
rubber-like core of approx. 90 nm; typically a butadiene homopolymer or a
butadiene-styrene copolymer with a random copolymer shell of 10-20 nm which
is compatible with the epoxy resin [3.2.4], [3.2.5]. They will be referred to as
CSR Type I. Others are based on polyacrylate cores and have a diameter in the
range of 300- 400 nm [3.2.6]. They will be referred to as CSR Type II. If these
core-shell particles are dispersed in epoxy resins, the viscosities of modified
resins are much lower compared to epoxy resins modified with reactive liquid
rubbers. The toughening effects are independent from the curing agent and the
cure schedule. Sometimes strength and modulus are lowered, but not to the
same extent as with reactive liquid rubbers. The use in high temperature
applications however is limited, as the shell tends to soften at higher
temperatures followed by a drastic loss in strength and modulus of the cured
polymer.
Another CSR development in the mid-1980s created a material which can be

used at elevated temperatures as well [3.2.7], [3.2.8]. This was achieved by
reducing the thickness of the shell to a molecular monolayer - the result is rather
a core-skin than a core-shell material. These epoxy-functional CSR with a
polysiloxane core and an average diameter around 500 - 700 nm are very
efficient tougheners over a very broad range of temperatures [3.2.9]. They will
be referred to as CSR Type III. Figure 3.2.1 shows the unstained SEM picture of
the fracture surface of an anhydride cured epoxy resin containing 5.5 wt% of
CSR Type III. The morphology looks very similar to the pictures taken from
polymers toughened with reactive liquid rubbers (CTBNs). The rubber domains
are very uniform.
In the years 2002/2003 the first commercial grades of surface modified silica
nanoparticles were introduced into the market. They are manufactured in situ
directly in the epoxy resin by a modified sol-gel process and have an average
diameter of 20 nm as well as a very narrow particle diameter distribution.
These particles are completely monodisperse and do increase the resin viscosity
only slightly at higher concentrations. In contrast to fumed silica they exhibit no
thixotropic properties but behave like a Newtonian liquid. Due to their size they
are transparent and can easily penetrate even close meshed fabrics in composite
manufacturing when being injected. The property improvements which can be
achieved by modifying epoxy resins with these silica nanoparticles, like modulus,
toughness and fatigue performance, have been the subject of intensive research
in the last decade [3.2.10].
The mechanisms how nanoscaled spherical fillers can improve epoxy polymer
properties have been identified; however the contribution of each one might be
of a different proportion depending on the hardeners used to form the three-
dimensional network upon cure. Figure 3.2.2 shows the excellent dispersion and
the very narrow particle diameter distribution of the spherical nanosilica.
67
Figure 3.2.1: SEM image of an epoxy polymer with 5.5 wt% CSR Type III
Figure 3.2.2: TEM image of an epoxy polymer with approx. 20 wt% silica
nanoparticles [3.2.11]
The combination of reactive liquid rubbers or core-shell elastomers and silica

nanoparticles as additional modifier in epoxy resin systems yields additive and
sometimes synergistic property improvements. It becomes possible to formulate
tough and stiff materials. Therefore such hybrid systems are used in many
industrial epoxy formulations today.
68
3.2.3. Discussion
If not mentioned otherwise, the researchers cited used commercial 40% (by
weight) concentrated masterbatches of surface-modified nanosilica in DGEBA
from one supplier. These particles have an average diameter of 20 nm and a
very narrow particle diameter distribution. They were then diluted down using
commercial epoxy resins to vary the nanosilica concentrations.
3.2.3.1. Epoxy resins modified with reactive liquid rubbers (CTBNs)

and silica nanoparticles, amine cured
The diglycidyl ether of bisphenole A (DGEBA) is the most commonly industrially

used epoxy resin. Thus most of the research work was performed using DGEBA.
At first the silica nanoparticles, after being commercially available, had been
added to epoxy formulations containing CTBNs to reduce the loss in strength and
modulus caused by the rubber modification without increasing the viscosity. Very
soon in some applications a synergy between elastomeric tougheners and silica
nanoparticles was discovered and patented consequently [3.2.12].
However, as will be shown in this article, the synergy is not necessarily related to
morphology and sometimes only found for one polymer property or not at all.
One of the first industrial applications where nanosilica was used together with
reactive liquid rubbers were structural epoxy adhesives. We found an increase in
adhesive lap shear strength at low addition levels of nanosilica (<2 wt%) of a
one-component, heat-curing adhesive [3.2.13]. The toughness seemed not to be
increased further compared to the formulation without nanosilica. This might be
due to the fact that the curing agent of this adhesive, dicyandiamide, forms very
close-meshed molecular networks. The CTBN rubber in the formulation had 26 %
acrylonitrile in the copolymer.
In another study we used a rubber with 18 % of acrylonitrile and a commercial

amine curing agent based on 2,2´-dimethyl-4,4´-methylene-bis(cyclohexyl-
diamine) and isophorone-diamine. We found that the loss in modulus caused by
the rubber modification could be compensated by the addition of nanosilica for
rubber concentrations up to approx. 7 wt% [3.2.14]. The GIc of the unmodified
polymer was increased from 609 J/m2 to 1223 J/m2 by the addition of 4.6 wt%
CTBN and increased further to 2059 J/m2 for a system containing 4.1 wt% CTBN
and 2 wt% of nanosilica. This indicates the existence of an optimum nanoparticle
content which might be different for each different polymer system defined by
resin and hardener.
Caccavale [3.2.15] studied epoxy resin systems modified with the same rubber
and cured with another commercial polyamine hardener. The loss in modulus due
to a 7.3 wt% rubber modification was only partially compensated by the addition
of 3.7 wt% nanosilica and still approx. 6% lower than the control. Tg was lowered
significantly by 19 °C. KIc of the hybrid however was increased by 136% (dry
conditions) respectively 154% (wet conditions). Microscopical investigations
revealed a good dispersion of both silica nanoparticles and rubber domains
formed upon cure. Moisture absorption of the hybrid system was higher than for
the unmodified epoxy and this could be assigned to the CTBN modification.
69
Tsai et al. [3.2.16] based their research upon the same rubber with 18 %
acrylonitrile in the copolymer and isophorone diamine as a hardener. They
reported a fair dispersion of both rubber domains formed upon cure and silica
nanoparticles. The modulus of the unmodified system was lowered by 10 wt% of
CTBN from 3.25 GPa to 2.63 GPa compared to the control and brought back to
3.18 GPa by the addition of 10 wt% nanosilica. Similar behaviour was found for
the tensile strength - a loss of approx. 20% for the rubber-modified polymer and
approx. 3% for the hybrid. Fracture toughness by means of GIc was increased by
516% to 1170 J/m2 by the addition of the rubber. The hybrid system achieved
only 930 J/m2. This indicates a tough and stiff system, but no synergistic effects.
In another study they tested the damping properties at room temperature of

these systems [3.2.6]. The glass transition temperature (Tg) of the rubber is
below – 30 °C; the Tg of the cured epoxy polymer is estimated to be above +
100 °C. The loss factors reported were 2.95% for the unmodified system, 3.12%
for 10 wt % of rubber, 3.51% for 10 wt% of nanosilica and 3.88% for the hybrid
with 10 wt% of both. Modulus was reduced by 17.3% by the CTBN, increased by
8.74% by the nanosilica and only 5.3% lower for the hybrid. As damping
properties were best for the hybrid there might be a synergy.
Pearson et al. [3.2.17] reviewed the improvements obtained by the addition of

spherical glass beads with an average diameter of 2 and 50 µm to rubber
toughened epoxies, and compared them to their results with the spherical silica
nanoparticles of 20 nm diameter. Additionally they investigated silica
nanoparticles of 80 nm in diameter; those were supplied by a different
manufacturer. A significant increase in fracture toughness and fracture energy
with a maximum at approx. 5 vol% of nanosilica was found. Variations of the
rubber content revealed larger improvements at lower rubber levels than for high
rubber concentrations. This might eventually be caused by the observed
nanosilica agglomeration at very high rubber concentrations.
In the basic study they published further details [3.2.18]. The rubber used
contained 18% acrylonitrile in the copolymer. Piperidine was the curing agent of
choice. Compressive moduli are improved only slightly compared to the
unmodified control when nanosilica is added; with the values being lower for the
20 nm particles than for the 80 nm particles. The 80 nm particles seem to have a
smaller effect on the toughness; the addition level of the rubber dominates. The
20 nm particles, however, do increase the fracture toughness up to a maximum
at approx. 6 vol% nanosilica. Looking at the fracture energy of 3250 J/m2 for the
unmodified system containing approx. 21 vol% of CTBN, the 80 nm particles
increase the fracture energy up to 4310 J/m2 at 1.5 vol% addition level. The 20
nm particles show with 5720 J/m2 a maximum at 3.1% addition level. A similar
toughness was found with 12 and 17 vol% rubber. Pearson observed
agglomerate formation (clustering) of the nanosilica particles at rubber
concentrations above 12 vol%.
Robinette and his group [3.2.19] used two commercial epoxy resin formulations
which contain a CTBN adduct in the resin part. They modified the resin part with
nanosilica and adapted the amount of amine-based hardener accordingly. The
tougher, low Tg system called SC-15 showed an increase in modulus from 2.37
GPa to 2.96 GPa at 15 wt% addition level of nanosilica.
70
This equals an improvement of 25%. Tg seemed to be nearly unaffected. GIc
however, increased from 900 J/m2 to 1020 J/m2 at 1 wt% nanosilica.
For higher loading levels the GIc was more or less at 800 J/m2. In contrast, the
high Tg system SC-79 showed no increase in modulus at first, then a somewhat
higher modulus at 7.5 and 10 wt% addition level, then a drop in modulus again.
The glass transition temperature was lowered with increasing nanosilica
concentration from 180 °C to 162 °C at 15 wt% nanoparticles. GIc was increased
from 180 J/m2 to 380 J/m2 at 10 wt%, then drops again to 250 J/m2 at 15 wt%.
Apparently both systems have an optimum nanosilica addition level, however at
totally different concentrations. Figure 3.2.3 shows the good distribution of the
nanosilica and the typical rubber domains formed upon cure. It is interesting to
see that no silica nanoparticles are trapped in the rubber domains upon their
formation during cure.
Figure 3.2.3: TEM picture of epoxy polymer modified with reactive liquid rubber
(CTBN) and nanosilica [3.2.20]
Sun et al. [3.2.21] worked with the SC-79 epoxy resin system as well. Modulus
increased with increasing nanosilica addition level up to approx. +40% for a 15
wt% loading level. Flexural strength and strain at break showed a similar
behaviour. Additionally they investigated combinations with other nanoparticles
like alumina or carbon nanofibers but could not find further improvements.
Another team investigating the SC-79 two part epoxy system was the group of
Renukappa et al [3.2.22]. They focused on tribological and electrical properties
and reported a maximum for strength at 10 wt% nanosilica. Wear resistance was
increased significantly and showed a maximum at 10 wt% as well.
71
Dielectric strength, arc and track resistance were improved first with increasing
the nanosilica content, showed a maximum at 15 wt% and decreased with 20
wt% below the level of the unmodified control.
In continuation of this work Ranganathaiah et al. [3.2.23] reported a linear

increase in modulus with increasing nanosilica content from 2.97 GPa to 3.55
GPa at 20 wt%. Strength exhibited a maximum at 10 % again. They concluded
that 10 wt% is the optimum loading level as for 15 wt% and 20 wt%
agglomerates had been observed (by TEM) which could explain the lower
performance. Volume resistivity and surface resistivity had been investigated as
well.
Tetrafunctional epoxy resins, cured with sterically hindered aromatic amines are
used in high performance aerospace applications. We examined such a system
and observed a further increase in toughness by the addition of 10 wt%
nanosilica to the reactive liquid rubber modified system [3.2.24]. The loss in
modulus due to the rubber modification was overcompensated by the nanosilica.
Thus a tough and stiff resin system could be formulated. Dry Tg was not affected
by the modifications; however the wet Tg (after two weeks at 70 °C and 100 %
relative humidity) was the lowest for the hybrid system.
3.2.3.2. Epoxy resins modified with styrene-butadiene rubber (SBR) and

silica nanoparticles, amine cured
Gope et al. [3.2.25] studied an epoxy resin cured with triethylenetetramine,

which was modified with a styrene-butadiene rubber up to 1.5 wt% and silica
nanoparticles at up to 2 wt%. The silica nanoparticles with an average diameter
of 130 nm - it is unclear if they were surface modified or not. They reported a
very good dispersion of both rubber and nanosilica as well as improved
mechanical properties. Wear properties have been improved significantly by the
nanosilica addition (up to 4 times lower).
3.2.3.3. Epoxy resins modified with reactive liquid rubbers (CTBNs) and silica
nanoparticles, anhydride cured
In an early investigation we used the reactive liquid rubber with 18% acrylonitrile
in the copolymer and methyl hexahydrophthalic acid anhydride (MHHPA)
accelerated with a small amount of a ternary amine [3.2.26]. Modulus was
lowered by the rubber modification and increased again by the nanosilica
addition to the level of the unmodified system. This system had a fracture energy
(GIc) of 103 J/m2, which was increased to 406 J/m2 by the addition of 9 wt%
CTBN. Adding 4.5 wt% nanosilica pushed the GIc to 917 J/m2; 9 wt% nanosilica
to 973 J/m2. Compared to the control this is an increase of approx. 850%. The
glass transition temperature dropped by the rubber modification from 143 °C to
133 °C. The nanosilica addition had no influence on Tg.
Manjunatha et al. [3.2.27], working with the same system, investigated the cyclic
fatigue behaviour. Tensile strength and modulus of the polymer were lowered by
the rubber addition and increased by the nanosilica addition. The system with 9
wt% CTBN and 10 wt% nanosilica had approx. 10 % lower strength and
modulus.
72
The cyclic loading tests at different stress levels showed a clear shift towards
much more cycles before failure for both the rubber and the nanosilica modified
polymer. However, the hybrid performed best and gave a 6-10 times
enhancement of fatigue life.
Kinloch, Taylor et al. [3.2.28] used the same system and reported significant
improvements in toughness. Though the modulus of the hybrid system was
slightly lower than for the unmodified system (-4%), the fracture energy GIc was
significantly increased by the addition of 9 wt% CTBN and further increased by
the nanosilica addition. A maximum was found at 15 wt% nanosilica: 965 J/m2
compared to 77 J/m2 for the control which equals an improvement by 1150%. At
20 wt% nanosilica the toughness shifts to lower values. Interestingly the
formation of small silica agglomerates was observed, but did not seem to affect
the rubber domain formation nor the performance of the cured resin.
Zhang, Tang et al. [3.2.29] used a powdered CTBN rubber for their study,
together with accelerated methyl hexahydrophthalic acid anhydride. They
reported well dispersed rubber particles and well dispersed silica nanoparticles.
No agglomerate formation could be observed. The elastic modulus of all hybrid
formulations was lower than for the unmodified system with a modulus of 3.01
GPa; except for the formulation with 7 vol% nanosilica and 2 vol% rubber which
achieved a modulus of 3.25 GPa. GIc of the control was 62 J/m2 and increased to
186 J/m2 for 7 vol% nanosilica and 2 vol% rubber. 2 vol% nanosilica and 7 vol%
rubber gave a further increase to 371 J/m2. However, the best overall
performance was shown for the system with 4.5 vol% silica and 4.5 vol% rubber
with a GIc of 306 J/m2 (equals an improvement of approx. 394%), a slightly
lower modulus (-4%) and the highest impact energy of all systems tested. 31.29
kJ/m2 compared to 16.2 equals an improvement of approx. 93%. The glass
transition temperature was reported to be unaffected by the modification which
indicates a particulate nature of the rubber similar to core shell materials.
In another study they evaluated a carboxy-terminated polyurethane-co-polyether

block copolymer as toughener together with silica nanoparticles [3.2.30]. As
curing agent accelerated methylhexahydrophthalic acid anhydride was used. TEM
microscopy revealed a good dispersion of both rubber domains formed upon cure
as well as for the silica nanoparticles. The tensile strength was slightly lower for
the elastomer modified system and increased for the hybrid systems compared
to the unmodified control. A reduction in modulus from 3.01 GPa to 2.50 GPa at
9 vol% elastomer was observed. With 9 % nanosilica this loss was
overcompensated to 3.20 GPa; at 12 % nanosilica even to 3.40 GPa. KIc was
more than doubled from 0.46 MPam1/2 to 1.00 MPam1/2 by the toughener and
further increased with increasing concentration of silica nanoparticles. The hybrid
system with both 9 vol% exhibited a KIc of 1.2 MPam1/2, the system with 9 vol%
elastomer and 12 vol% nanosilica achieved 1.33 MPam1/2. This equals an
improvement of 189%.
3.2.3.4. Epoxy resins modified with silica nanoparticles, cured with amines
and amino-functional reactive liquid rubbers (ATBNs)
Totally different systems can be achieved when amine functional reactive liquid
rubbers are used to modify an epoxy resin. They are used as part of the
hardener blend in 2 part epoxy formulations, e.g. structural adhesives.
73
Compared to epoxy resins modified with CTBN adducts a much smaller
percentage of the rubber molecules phase separate and form rubber domains. A
large proportion is crosslinked randomly into the polymer and lowers the
crosslink density significantly - with the known effects of a much lower Tg, lower
modulus, increased elongation at break, increased toughness etc.
Kinloch et al. [3.2.31] used a combination of an amine terminated reactive liquid

rubber with 17% acrylonitrile in the backbone together with a commercial
polyamidoamine as a curing system. The lap shear strength of adhesive joints of
untreated aluminium alloy was increased significantly at low nanosilica addition
levels of approx. 1 wt%. Roller peel tests showed significant improvements as
well with a maximum around 2 wt% nanosilica. GIc was increased from 1200
J/m2 for the rubber toughened system to a maximum of 2300 J/m2 at 4.6 wt%
In continuation of this work we varied the rubber level of the formulations and
found similar behaviour: there was always a significant property improvement at
low addition levels of nanosilica (1-4 wt%) [3.2.32]. The same was found for
roller peel tests at different temperatures (substrate chromic acid etched
aluminium) as well as for wedge impact test with adhesive joints of degreased
steel. Most striking were the results of the adhesive fracture energy test where
the GIc was doubled from 1200 J/m2 to 2400 J/m2 by the addition of approx. 1
wt% nanosilica.
Further work with a different hardener composition (amine functional reactive

liquid rubber, amine functional polyether, polyamidoamine and isophorone
diamine) showed similar results [3.2.33]. The mode I fracture energy was
increased from 2256 J/m2 to a maximum 4550 J/m2 at 6 wt% nanosilica addition
level. Lap shear tests of adhesively bonded glass-fiber reinforced composites
gave comparable behaviour.
Atomic Force Microscopy (AFM) investigations revealed a different morphology

than for the CTBN modified resins. When an epoxy resin system containing both
the CTBN adduct and the nanosilica is cured, the rubber domain formation upon
cure seems not to be affected by the presence of the nanosilica. Agglomeration
of the nanosilica does either not occur or only small agglomerates are found
[3.2.21], [3.2.28], which seems not to affect the performance of such systems.
An aspect which will be discussed when looking into the synergy between CTBN
and silica nanoparticles.
If an epoxy resin containing nanosilica is cured with an amine blend containing

amine functional reactive liquid rubber, the phase separation and rubber domain
formation become quite irregular at higher nanosilica concentrations.
Furthermore it seems that partial agglomeration of the silica nanoparticles is

induced, with large agglomerates formed; nevertheless superior mechanical
properties were found [3.2.33]. These irregular morphologies need further
investigation in the future.
74
3.2.3.5. Epoxy resins modified with core-shell elastomers (CSRs) and silica
nanoparticles, amine cured
CSR Type I
Mai et al. [3.2.34] studied the cyclic fatigue behaviour of an epoxy resin system
modified with a core-shell elastomer characterized by a particle diameter of
approx. 110 nm. Piperidine was used as a curing agent. They reported a
reduction in strength by the addition of the core-shell rubber, which was not
improved by the addition of the 20 nm silica nanoparticles. The loss in modulus
however was compensated; it dropped from 2.86 GPa to 2.25 GPa at 10 wt%
elastomer and was brought back to 2.81 GPa with 10 wt% nanosilica. The silica
nanoparticles increased the fatigue life whereas the core-shell particles
decreased it. Thus the hybrid system did not perform better than the neat
polymer. Toughness was increased from 277 J/m2 to 1250 J/m2.
This work was continued by Liu et al. [3.2.35]. They looked further into
toughness and toughening mechanisms. Core-shell particles and silica
nanoparticles were well distributed, as TEM microscopy revealed. GIc for the
unmodified resin was 277 J/m2. The elastomer addition increased GIc with raising
addition level. At 10 wt% core-shell the GIc was 1930 J/m2. An increase in
toughness with the addition of nanosilica was found as well. At 10 wt% nanosilica
GIc was 690 J/m2. The hybrid system with 10 wt% of both modifications had the
highest toughness of 2480 J/m2. This seems to imply an additive behaviour, no
synergistic effects. However a good balance between toughness and modulus
could be achieved. As major toughness contributors nano-silica debonding and
bridging before pullout and core-shell rubber cavitation and matrix plastic
shearing were identified. They claimed that nanosilica and core-shell elastomer
act independently during crack growth.
Furthermore they examinated cyclic fatigue crack propagation of this hybrid

system and observed a synergistic effect on the fatigue crack growth threshold
[3.2.36]. The unmodified epoxy polymer has a ΔGth of 44 J/m2; 6 wt% core shell
increases it to 47 J/m2. The modification with 6 wt% nanosilica results in 65
J/m2, whereas the hybrid with 6 wt% of both achieves 102 J/m2; an
improvement of 132% compared to the control.
CSR Type II
Tsai et al. [3.2.16] used the larger Type II particles and isophorone diamine as a
hardener. TEM microscopy showed a good dispersion of both silica nanoparticles
and elastomer particles as well. The modulus was reduced from 3.25 GPa to 2.73
GPa by the addition of 10 wt% core shell particles and brought back to 2.97 GPa
by further addition of 10 wt% silica nanoparticles. The tensile strength was
reduced by approx. 27 % by the elastomer and this reduction was not changed
by the nanosilica addition - though the system modified only with nanosilica
exhibited no loss in strength. The fracture toughness of the epoxy resin was
determined to be 190 J/m2. It was increased by a 10 wt% core-shell modification
to 1420 J/m2. A modification with 10 wt% of nanosilica achieved 280 J/m2. The
hybrid system with 10 wt% of both modifications achieved only 1030 J/m2. A
good balance between toughness and modulus, but no synergistic effects.
75
In continuation of this work Tsai et al. [3.2.6] investigated the damping
properties at room temperatures and found a 42% improvement of the loss
factor for the hybrid system.
CSR Type III
Palinsky [3.2.37] reported about the modification of a commercial resin transfer

molding (RTM) resin based on tetraglycidyl ether of methyl dianiline (TGMDA)
and cured with sterically hindered aromatic amines. By modifying this resin
system with the core-shell elastomer strength and modulus were not affected,
however the (dry) Tg was lowered by 12 °C. KIc was increased from 0.69 MPam1/2
to 0.90 MPam1/2 which equals an improvement in toughness by 30%. Further
modification with silica nanoparticles did not affect strength but increased the
modulus by 12%. KIc was not increased further, however the glass transition
temperature was lowered by another 16 °C. The addition level of both
modifications was not communicated.
In a very detailed study Walter et al. [3.2.38] investigated an epoxy resin

system cured with a commercial aliphatic amine hardener.
Figure 3.2.4: TEM picture of epoxy polymer modified with 8 vol% Type III core-
shell elastomer and 5.5 vol% silica nanoparticles [3.2.11]
They found significant improvements in KIc by adding 3-8 vol% of core-shell

elastomer with only small losses in modulus. Nanosilica addition up to 8 vol%
increased both modulus and KIc by 25-50%. The combination of both
modifications showed further small increases in KIc. The morphologie of the
polymer containing 8 vol% core-shell rubber and 5.5 vol% nanosilica can be seen
in Figure 3.2.4, respectively 3 vol% core-shell rubber and 5.5 vol% nanosilica
can be seen in Figure 3.2.5. As the different magnifications show, both
nanoparticles and elastomer particles are well dispersed.
76
Figure 3.2.5: TEM picture of epoxy polymer modified with 3 vol% Type III core-
shell elastomer and 5.5 vol% silica nanoparticles [3.2.11]
In continuation of this work Schlarb et al. [3.2.39] studied the abrasive

properties of the cured epoxy resin systems. The addition of core-shell elastomer
increases the abrasion whereas the nanosilica lowers the abrasion up to 30%.
The hybrid performed like the unmodified epoxy resin system.
Botsis et al. [3.2.5] investigated cycloaliphatic polyamine cured epoxy resin.

They observed a good dispersion of both rubber and nanoparticles as well. A
small decrease in the coefficient of thermal expansion (CTE) was found for the
nanosilica addition; the core-shell particles slightly increased the CTE. The hybrid
behaves pretty much like the unmodified resin.
3.2.3.6. Epoxy resins modified with core-shell elastomers (CSRs) and silica
CSR Type I
Taylor et al. [3.2.40] used DGEBA cured with accelerated methylhexahydro-

phthalic acid anhydride for their studies. The strength was lowered from 83 MPa
to 69.5 MPa by the addition of 9 wt% core-shell elastomer. Further modification
with 9 wt% silica nanoparticles increased the strength slightly to 73 MPa. The
rubber addition reduced the modulus from 2.05 GPa to 1.7 GPa.
77
The addition of nanosilica brought the modulus back to 1.9 GPa. The toughness
of the core-shell modified polymer was increased from 1.18 MPam1/2 to 1.58
MPam1/2. This equals an improvement of approx. 34%.
TEM investigations showed again uniform dispersion of both the 20 nm silica

nanoparticles and the 110 nm core-shell rubber particles, as can be seen in
Figures 3.2.6 and 3.2.7 for different magnifications.
Figure 3.2.6: TEM picture of epoxy polymer modified with 9 wt% Type I core-
shell rubber and 6 wt% silica nanoparticles
78
Figure 3.2.7: TEM picture of epoxy polymer modified with 9 wt% Type I core-
shell rubber and 6 wt% silica nanoparticles
3.2.3.7. Short overview of improvements achieved
Having discussed all research results in detail, Table 3.2.1 gives an overview of
the results of some selected references; regardless of the resin/hardener
combination.
As can be seen, the modulus of the hybrids is a few percent lower than for the
unmodified systems; especially for higher rubber or elastomer levels.
Toughness is improved significantly; especially for higher elastomer

concentrations very high values can be obtained. Of course the absolute value is
depending on the resin/hardener combination of the formulation.
All systems are tough and stiff, which means that the best of both modifications
is found in the hybrid systems.
Additionally the fatigue performance of such formulations is outstanding.
79
Resin / Modifications Modulus Improv. GIc Improv. Ref.
Hardener (GPa) versus (J/m2) versus
control control
DGEBA 4.6 wt% CTBN* 2.89 -4% 241 + 289 % [3.2.29]
MHHPA 8.5 wt%
nanosilica*
DGEBA 9 wt% CTBN 2.77 -6% 683 + 787 % [3.2.28]
MHHPA 9 wt%
nanosilica
DGEBA 9 wt% CTBN 2.85 -4% 965 + 1153 % [3.2.28]
MHHPA 15 wt%
nanosilica
DGEBA unknown, but 3.25 - 21 % 380 + 111 % [3.2.19]
commercial low % CTBN
amine I 10 wt%
nanosilica
DGEBA 11.3 wt% CTBN 2.68 - 22 % 1843 + 203 % [3.2.14]
commercial 5.7 wt%
amine II nanosilica
TGMDA 8 wt% CTBN 3.37 +4% 708 + 302 % [3.2.24]
aromatic 10 wt%
amines nanosilica
DGEBA 10 wt% CSR I 2.78 -3% 2480 + 795 % [3.2.35]
Piperidine 10 wt%
nanosilica
DGEBA 10 wt% CSR II 2.97 -9% 1030 + 442 % [3.2.16]
Isophorone 10 wt%
diamine nanosilica
DGEBA 5.1 wt% CSR 3.50 +9% 1851 + 300 % [3.2.38]
commercial III*
amine 11 wt%
nanosilica*
*recalculated from vol%
Table 3.2.1: Overview of property improvements of epoxy polymers achieved by

hybrid modification
3.2.3.8. Synergy or no synergy?
If Figures 3.2.3 to 3.2.7 are compared, the morphology of the fracture surfaces
looks identical, though the elastomer particles have different sizes. However,
what looks the same, behaves differently regarding micro-mechanics.
We found in our investigations with the Type III core-shell rubber no synergy
between core-shell particles and silica nanoparticles. Toughness increases
compared to the control were additive. Similar findings have been reported by
Liu et al. [3.2.35], [3.2.36] for Type I core-shell rubbers except for fatigue
performance where a synergy seems to exist. Tsai et al. [3.2.16] could not
confirm the existence of a synergy for Type II core shell particles either.
Apparently the well-known mechanisms of rubber toughening with particulate

rubbers and the mechanisms of nanosilica toughening do not significantly
interact with each other, though interactions at the tip of a crack forming should
be expected.
80
No agglomeration was reported for all systems containing both core-shell rubber
particles and silica nanoparticles.
The toughness of a hybrid resin based on core-shell rubbers is the added

toughness of the nanosilica modification plus the core-shell elastomer
modification; regardless of the rubber particle size or chemistry.
In contrast we always found a synergy in toughening between reactive liquid

rubbers and nanosilica. Others, like Zhang confirmed the existence of a synergy.
Though the morphology looks the same, here the mechanisms of rubber
toughening with reactive liquid rubbers and nanosilica toughening are interacting
somehow.
Some researchers noticed the presence of small agglomerates or clusters of

nanosilica in their cured resin systems containing reactive liquid rubber [3.2.18],
[3.2.21], [3.2.27], [3.2.28].
This raises two questions. First, do agglomerates have an influence on the

performance of the cured epoxy polymer? Second, do the agglomerates have any
influence on the synergy found between CTBN adducts and silica nanoparticles?
Could they be the reason?
Pearson et al. [3.2.41] investigated two different sizes of nanosilica (20 nm and
80 nm) in piperidine cured epoxy resin. They reported a neglible effect of the
particle size on fracture toughness and compressive properties. Thus it seems to
be a fair assumption to consider these small agglomerates or clusters as
somewhat larger nanoparticles which do not or nearly not affect the performance
of the cured epoxy polymer.
Recently Pearson [3.2.18] studied the hybrids based on the two different particle
sizes and reactive liquid rubber. He found no considerable effect of particle size
on the toughening of the hybrid nor could he confirm a synergy.
A correlation between very high, synergistic values of GIc and the observation of
small nanosilica agglommerations or clusters could not be established.
It is well known that the GIc of rubber toughened epoxy polymers increases
dramatically with a lower crosslink density [3.2.42]. This is the reason why many
industrial formulations contain the CTBN rubber adduct along with a small
amount of a solid DGEBA with a higher molecular weight. The long chain
molecules of the solid epoxy resin lower the crosslink density to a certain extent
and make the resin matrix more ductile as well as the rubber toughening more
efficient.
When silica nanoparticles are added to an epoxy resin modified with CTBN, the
proportion of rubber molecules not participating in the phase separation upon
cure increases with increasing addition level of nanosilica [3.2.28]. If more long
flexible rubber molecules are crosslinked randomly into the polymer matrix, the
crosslink density is lowered and the ductility of the matrix is increased - thus
making the rubber toughening more efficient. This could be the explanation of
the synergistic improvement of GIc and fatigue performance for hybrid systems
based upon the combination CTBNs and nanosilica.
81
As indication for the lower crosslink density can be seen the development of the
glass transition temperature [3.2.28]: The addition of nanosilica does not lower
the Tg compared to the unmodified epoxy polymer system. This is different for
the CTBN modified system. At a 9 wt% rubber addition level the Tg is 150 °C.
Further addition of silica nanoparticles lowers the Tg to 145°C at 10 wt% and to
140 °C at 20 wt%.
Though the difference in Tg might seem small, this is not unusual as minor
changes in crosslink density can have a big impact on toughness. Pearson
[3.2.42] reported for two rubber-toughened epoxies with a small difference in
crosslink density and a difference in Tg of only 5 °C an increase in fracture
toughness by factor three.
Still there is a need for additional research to explain the synergy, like the nature
of interactions at the tip of a forming crack.
3.2.4. Conclusions
Taking into account all the data accumulated, a couple of conclusions can be
drawn:
1. The combination of rubber toughening and silica nanoparticles offers the

possibility to formulate tough and stiff epoxy resin systems; regardless of
the hardener used.
2. Losses in modulus due to a modification with reactive liquid rubbers or

core-shell rubbers can be compensated to a large extent, sometimes even
completely by the addition of nanosilica.
3. Losses in strength caused by core-shell elastomer modification cannot

always be compensated by the silica nanoparticles.
4. Lower glass transition temperatures caused by the toughener are not

improved by the nanosilica addition - no increases in Tg can be achieved.
5. Toughness of hybrid systems is always higher as for only rubber-

toughened epoxy resin systems.
6. Toughness increases of epoxy resins containing reactive liquid rubbers are

synergistic. Improvements of 80%-300% can be achieved for amine cured
epoxies and of 300%-1200% for anhydride cured epoxies.
7. Similar improvements can be achieved for the cyclic fatigue performance.
8. Typical addition levels providing best performance, high toughness and

stiffness as well as very high fatigue performance are 5-10 wt% core-shell
elastomer or 5-15 wt% reactive liquid rubber combined with 5-10 wt% of
82
3.2.5. References
[3.2.1] Kinloch, A.J.; Shaw, S.J.; Tod, D.A. Polymer (1983), 24, 1341–
1354.
[3.2.2] Kinloch, A.J. MRS Bulletin (2003), June 2003, 445-448.
[3.2.3] Bagheri, R.; Marouf, B.T.; Pearson, R.A. J. Macromol. Sci. Part C:
Polymer Reviews (2009), 49, 201-225.
[3.2.4] Giannakopoulos, G.; Masania, K.; Taylor, A.C. J. Mater. Sci. (2011),
46, 327-338.
[3.2.5] Lai, M.; Friedrich, K.; Botsis, J.; Burkhart, T. Composites Sci. Tech.
(2010), 70, 2168-2175.
[3.2.6] Tsai, J.-L.; Chang, N.-R. Journal of Composite Materials (2011), 45,
10/2011, 2157–2164.
[3.2.7] EP 0 407 834 (1989)
[3.2.8] Block, H.; Pyrlik, M. Kunststoffe German Plastics (1988), 78, 29.
[3.2.9] Chen, J.; Kinloch, A.J.; Sprenger, S.; Taylor, A.C. Polymer accepted
for publication May (2013)
[3.2.10] Sprenger, S. J. Appl. Polym. Sci. (2013) DOI: 10.1002/APP.39208
[3.2.11] Pictures courtesy of IVW Kaiserslautern and Polymerservice

Merseburg (see [3.2.38] as well)
[3.2.12] WO 2004081076 respectively EP 1 457 509 (2003)
[3.2.13] Sprenger, S.; Eger, C.; Kinloch, A.J.; Lee, J.H.; Taylor, A.C.; Egan,
D. Adhäsion, Kleben & Dichten (2003), 03/2003, 24-30.
Composites Magazine (2005), No 21, 66–69.
[3.2.15] Caccavale, V. Master of Science Thesis (2007) University of Padova,

Italy
[3.2.16] Tsai, J.-L.; Huang, B.-H.; Cheng, Y.-L. Journal of Composite

Materials (2009), 43, 3107-3123.
[3.2.17] Pearson, R.; Liang, Y. Polymer Nanocomposites (2010), Woodhead

Publishing Ltd, Chapter 23, 773-786.
[3.2.18] Liang, Y.L.; Pearson, R.A. Polymer (2010), 51, 4880-4890.
[3.2.19] Army Research Laboratory Technical Report 4084 2007, Aberdeen

Proving Ground, MD, U.S.A. April (2007)
83
[3.2.20] Picture with permission of J. Robinette, Army Research Laboratory
(see [3.2.19] as well)
[3.2.21] Uddin, M.F.; Sun, C.T. Composites Sci. and Tech. (2010), 70, 223-
230.
[3.2.22] Veena, M.G.; Renukappa, N.M.; Suresha, B.; Shivakumar, K.N.

Polymer Composites (2011), 32, 2038–2050.
[3.2.23] Veena, M.G.; Renukappa, N.M.; Raj, J.M.; Ranganathaiah, C.;

Shivakumar, K.N. J. Appl. Polym. Sci. (2011), Vol. 121, 2752–
2760.
Composites Magazine (2007), No 30 January-February, 54–57.
[3.2.25] Singh, V.K.; Gope, P.C. J. of Reinforced Plastics and Composites

(2010), 29, 2450-2468.
[3.2.26] Sprenger, S.; Kinloch, A.J.; Taylor, A.C.; Mohammed, R.D.; Eger, C.
JEC Composites Magazine (2005), No 19 August/September, 73–
76.
[3.2.27] Manjunatha, C.M.; Taylor, A.C.; Kinloch, A.J.; Sprenger, S. J. Mater.

Sci. Letters (2009), 44, 4487-4490.
[3.2.28] Hsieh, T.H.; Kinloch, A.J.; Masania, K.; Sohn Lee, J.; Taylor, A.C.;
Sprenger, S. J. Mater. Sci. (2010), 45, 1193-1210.
[3.2.29] Tang, L.-C.; Zhang, H.; Pei, Y.-B.; Chen, L.-M.; Song, P.; Zhang, Z.
Composites: Part A (2013), in press
[3.2.30] Tang, L.-C.; Zhang, H.; Sprenger, S.; Ye, L.; Zhang, Z. Composites
Science and Technology (2012), 72, issue 5, 558-565.
[3.2.31] Sprenger, S.; Eger, C.; Kinloch, A.J; Lee, J.H.; Taylor, A.C.; Egan,
D. Journal of Adhesion (2003), 79, 867-873.
[3.2.32] Sprenger, S.; Eger, C.; Kinloch, A.J.; Lee, J.H.; Taylor, A.C.; Egan,
D. Adhäsion, Kleben & Dichten (2004), 03/2004, 17-21.
[3.2.33] Sprenger, S.; Kinloch, A.J.; Taylor, A.C.; Hsieh, T.-H. Adhesion,
Adhesives & Sealants (2009), 10/2009, 8-11.
[3.2.34] Wang, G.-T.; Liu, H.-Y.; Saintier, N.; Mai, Y.-W. Eng. Failure
Analysis (2009), 16, 2635-2645.
[3.2.35] Liu, H.Y.; Wang, G.T.; Mai, Y.W.; Zeng, Y. Composites: Part B
(2011), 42, 2170-2175.
[3.2.36] Liu, H.-Y.; Wang, G.; Mai, Y.-W. Composites Sci Tech. (2012), 72,
1530-1538.
84
[3.2.37] Palinsky, A. Composites Processing (2006), 27 April 2006,
Merseyside, UK
[3.2.38] Final report Stiftung Industrieforschung Projekt S 657, (2005),

Institut f. Verbundwerkstoffe, Kaiserslautern, Germany
[3.2.39] Walter, R.; Haupert, F.; Schlarb, A. Tribologie + Schmierungs-

technik (2008), 55, 2/2008, 26–300.
[3.2.40] Taylor, A. personal communication (2012), Imperial College,

London, UK
[3.2.41] Pearson, R.A.; Liang, Y.L. Polymer (2009), 50, 4895-4905.
[3.2.42] Pearson, R.A.; Yee, A.F. J. Mat. Sci. (1989), 24, 2571–2580.
85
3.3. Property improvements of fiber-reinforced composites
based on epoxy resins modified with SiO2 nanoparticles
3.3.1. Abstract
Surface-modified silica nanoparticles, 20 nm in diameter and with a very narrow

particle size distribution, are available as concentrates in epoxy resins in
industrial quantities since 10 years. They improve many different properties like
strength, modulus, toughness and fatigue performance. Some of these
improvements can be found for fiber-reinforced composites as well. In this
review, the research results obtained with commercial silica nanoparticles
published in the last decade are studied, results are compared with a focus on
mechanical properties and the mechanisms responsible for the property
improvements are discussed. Silica nanoparticles present only in the interface
between fibers and resin matrix might be a very promising approach for future
cost-sensitive applications.
3.3.2. Introduction
The use of fiber-reinforced composites is the key to lightweight construction,

which is indispensable for modern life. Aerospace, automotive, shipbuilding or
railway, rotor blades for wind energy generators or sports equipment - the
applications for fiber-reinforced composites are manifold. Glass or carbon fibers,
woven, unidirectional or multiaxial, are the most common reinforcements.
Thermoplastic or thermosetting materials are used as matrices.
In many high performance applications epoxy resins are used as matrix

materials. All known manufacturing technologies for fiber-reinforced composites
are used in the industry in combination with epoxy resins. Ski poles are made by
using filament winding of carbon fibers in combination with anhydride cured
epoxy resins. Machine parts are manufactured by using amine cured glass fiber
prepregs. Profiles are produced by pultrusion using liquid resin systems.
Aerospace parts are made by using carbon fiber prepregs based on
tetrafunctional epoxy resins cured with aromatic amines in an autoclave. The
most popular manufacturing technologies are injection technologies like resin
transfer moulding (RTM), single line injection (SLI), vacuum-assisted resin
transfer moulding (VARTM) etc. due to the relatively short manufacturing cycles.
Rotor blades for wind energy generators are manufactured mainly by using
vacuum-assisted infusion techniques. However, the viscosities of the resin
systems used for injection technologies need to be low. In engineering composite
parts, e.g. a crank shaft for a sports car, the reinforcing fiber or fabric or
nonwoven and their orientation towards the force along the part are of utmost
importance. However, the choice of resin and hardener is important as well - to
match performance as well as production cycle needs. Of course any
improvement of the resin performance is welcome to help to improve the
performance of the composite parts.
Mineral fillers like calcium carbonate or quartz powder can improve mechanical
properties of epoxy resins like strength, stiffness and modulus. However, they do
increase the viscosity of the resin significantly.
86
Due to their size, typically microns, they cannot be used in injection
manufacturing methods because they are filtered out by the fabric upon
injection. In prepreg manufacturing they tend to sediment in the process. Thus in
most composites applications no mineral fillers are used.
In the years 2002/2003 the first commercial grades of surface-modified silica

nanoparticles were introduced into the market. They are manufactured in situ
directly in the epoxy resin by a modified sol gel process. The spherical particles
have an average diameter size of 20 nm and a very narrow particle size
distribution [3.3.1]. Figure 3.3.1 shows the particle size distribution determined
by small angle neutron scattering (SANS).
Figure 3.3.1: Particle size distribution of silica nanoparticles determined by SANS
In a recent, extensive review the property improvements of epoxy resins

modified with silica nanoparticles have been investigated [3.3.2]. Incorporated
into cured bulk epoxy resins, regardless of the hardener choice, they do improve
several properties like strength and modulus (as expected for mineral particles).
Furthermore, they improve toughness which is less obvious, but a known effect
when using glass microspheres in epoxy formulations. Cyclic fatigue performance
can be improved significantly, by several hundred percent.
Used in resin systems for composite applications, they offer several advantages:
being 20 nm small and completely monodisperse, they do increase the resin
viscosity only slightly at higher addition levels. In contrast to fumed silica they do
not exhibit thixotropic properties but behave like a Newtonian liquid. Due to their
size they can easily penetrate even close meshed fabrics in composites
manufacturing when being injected.
Figures 3.3.2 and 3.3.3 show TEM pictures of glass fiber reinforced composites
(GFRC) made by using an epoxy resin containing silica nanoparticles [3.3.3]. The
laminates have been made by using VARTM.
87
The uniform dispersion of the particles is evident and the difference in size
between fiber and particle is obvious - thus it is self-explaining why there is no
gradient in nanosilica distribution even after quite a distance from the point of
resin injection.
Further the silica nanoparticles can be used in prepreg manufacturing as they do

not sediment - in contrast to larger particles. Being completely transparent, they
can even be used in composite parts where the surface appearance matters, e.g.
in sports car parts like roofs or car body side panels as well as interior surfaces
or in sporting equipment like golf clubs, helmets or ski poles.
As silica nanoparticles are a very attractive raw material for epoxy resin
formulators, they are used today, 10 years later, in many industrial composites
applications. They improve various properties like tensile strength, tensile
modulus, flexural stiffness, toughness and scratch resistance.
Figure 3.3.2: TEM picture of GFRC with 15 wt% silica nanoparticles
Very important for many applications is the significant improvement of fatigue

performance that the nanosilica modification of the epoxy resin provides to the
fiber-reinforced composite part. Thus many fiber-reinforced composites in
automotive, machine building or recreational equipment do contain silica
nanoparticles. Aerospace applications will follow soon.
88
Nevertheless, it is a difficult task to determine the optimum addition level in
function of resin and hardener used. Very often the improvements found for a
modified bulk resin [3.3.2] are not translated into the fiber-reinforced composite
made from this resin.
Mechanical properties of fiber-reinforced materials like tensile strength and

tensile modulus or compressive strength and compressive modulus are
considered to be fiber-dominated. Other properties like fracture toughness (KIc)
or fracture energy (GIc) are considered to be matrix-dominated, as the crack is
inserted into the test specimen in between the layers of fibers. Compressive
properties (before and after impact) are influenced by both fiber and matrix.The
fiber or fabric or textile used as reinforcement is therefore a further variable
which needs to be taken into consideration. Of course many research results
have never been published but converted into industrial formulations.
The aim of this review is to give a comprehensive overview regarding the actual
state of research, to understand the mechanisms of property improvements of
bulk resins as well as laminates and to provide formulating guidelines. First
results for laminates with nanosilica at the fiber/matrix interface are presented.
Figure 3.3.3: TEM picture of GFRC with 4 wt% silica nanoparticles
3.3.3. Discussion
If not mentioned otherwise, the researchers cited used commercial 40 % (by

weight) concentrated masterbatches of surface-modified nanosilica in diglycidyl
89
ether of bisphenol A (DGEBA) with an average diameter of 20 nm and a very
narrow particle size distribution. These were then diluted using commercial epoxy
resins to vary the nanosilica concentrations.
All investigations revealed a very good dispersion of the nanosilica particles;

typically determined by transmission electron microscopy (TEM).
3.3.3.1. Glass fiber-reinforced epoxy resin composites
Amine cured epoxy resin systems
Tsai et al. [3.3.4], [3.3.5] prepared laminates by impregnating one layer of

unidirectional (UD) glass fiber (E-glass) and subsequent addition of further layers
until 12 layers have been impregnated. Cure was performed in a press under
vacuum. The DGEBA epoxy resin was cured with an industrial
isophoronediamine. They reported an increase in fracture energy (GIc) of the bulk
resin by 47% at a loading level of 10 wt% nanosilica and by 84% at 20 wt%
nanosilica. GIc of the laminate was increased from 0.83 kJ/m2 to 0.90 kJ/m2 at 10
wt% silica nanoparticles in the resin, which equals an improvement of only 8%.
At 20 wt% an increase in GIc of 15% was reported.
In continuation of this work they investigated in-plane shear strength and off-
axis compressive strength [3.3.6]. DGEBA and a short-chain polyether diamine
as curing agent were blended to prepare thin laminates (1.45 mm, 5 layers of
UD, 10° off axis). Again hand lay-up and cure under pressure and vacuum were
used. The in-plane shear strength was increased from 23.7 MPa to 27.5 MPa at 5
wt% silica nanoparticles to 30.1 MPa at 10 wt% and 29.6 MPa at 20 wt% loading
level. This equals improvements of 16%, 27% and 25%, respectively.
To test the off-axis compressive strength thick laminates (6.15 mm, 22 layers of
UD, 0°, 5°, 10°, 15° and 90° off axis) were prepared. The unmodified control
was compared to a system with 20% of silica nanoparticles. The improvements
were different for the different fiber orientations: + 7% at 0°, + 17% at 5°, +
6% at 10°, +6 % at 15° and + 2% at 90°. For 0° the failure mechanism was
dominated by fiber splitting. At 90° the out-of-plane shear failure mechanism
was observed. For 5°, 10° and 15° the main failure mode was microbuckling
where improvements can be expected to be found as the silica nanoparticles do
increase the modulus of the epoxy resin matrix.
Tate et al. [3.3.7] used a commercial epoxy resin system, aliphatic amine cured,
which they modified with silica nanoparticles. They used a -45°/90°/+45°
bidiagonal stitched bonded fabric (E-Glass) and VARTM to prepare the laminates
(5.95 mm, 8 layers of fabric). Tensile strength was improved from 92.1 MPa to
122.4 MPa at 6 wt% silica nanoparticles, which equals an improvement of 33%.
At 7 wt% 109.5 MPa and at 8 wt% 101.4 MPa could be achieved. A similar
behaviour was found for the modulus which increased from 10.3 GPa to 13.2 GPa
at 6 wt% nanosilica, again an improvement of 29%.
7 wt% and 8 wt% achieved only 12.8 GPa and 11.7 GPa, respectively; which
might indicate the existence of a plateau in performance. Or, there might be
further improvements found at 10 wt% or 15 wt% addition levels – which have
not been investigated.
90
The interlaminar shear strength (ILSS: ASTM D 2344) showed the same
tendency: it went up from 24.5 MPa to 31.0 MPa (+27%) with 6 wt% silica
nanoparticle modification. With 7 wt% and 8 wt% 30.1 MPa and 29.8 MPa were
obtained. A slightly different picture was found for flexural strength and flexural
modulus. Strength increased from 146 MPa to 177 MPa to 181 MPa and 180 MPa.
Modulus was improved from 25.0 GPa to 29.5 GPa to 30.5 GPa to 30.7 GPa for 6
wt%, 7 wt% and 8 wt% nanosilica.
In a different investigation we found in three-point-bending tests with laminates

made by an anhydride cured system a plateau in performance improvements as
well; at about 8 wt% of silica nanoparticles.
Anhydride cured epoxy resin systems
If not mentioned otherwise, the systems investigated are all based on DGEBA,
cured with methyl hexahydrophthalic acid anhydride; accelerated with a very
small amount of a ternary amine.
Mahrholz et al. [3.3.8], [3.3.9] published very early promising results. They
made laminates with a bidirectional glass fiber at +45°/-45° orientation using
the SLI technique. At a 20 wt% addition level of silica nanoparticles, the tensile
strength was improved by 25%, tensile modulus was increased by 64% and the
shear modulus was increased by 75%.
Kinloch et al. [3.3.10] prepared 7-mm laminates consisting of 16 plies of

unidirectional glass fibres at 0°. Resin infusion under flexible tooling (RIFT) was
used to manufacture the laminates, with an infusion temperature of 50 °C. At 10
wt% silica nanoparticles in the matrix glass transition temperature (Tg) was
lowered by 2°C, flexural modulus was slightly lowered from 39.7 GPa to 38.5
GPa (equals - 3%). The interlaminar fracture energy GIc was increased from 330
J/m2 to 1015 J/m2, which equals an improvement of 207%. GIIc was improved
slightly from 1300 J/m2 to 1380 J/m2; which is equal to + 6%.
In continuation of this work they manufactured quasi-isotropic plates, 4-mm

thick with 8 plies of glass fibers with a 0°/0° interface and performed ballistic
impact testing [3.3.11]. The delaminated area of the nanosilica-containing
laminates was always larger than for the unmodified control - for all impact
energies tested. It seemed like the presence of nanosilica did not improve
laminate properties regarding fast impact.
Sprenger et al. [3.3.12] tested similar laminates with a falling dart impact test
(30 J) and could not detect any improvements. The delaminated area had exactly
the same size with or without silica nanoparticles present in the resin - see
Figure 3.3.4. Unfortunately, the compression after impact (CAI) test results have
not been released for publication.
Kuehn et al. [3.3.13] found slightly improved CAI properties for laminates with
nanosilica as will be shown in the chapter carbon fiber-reinforced composites.
The question which mechanisms of toughness improvements by silica
nanoparticles apply to fiber-reinforced composites and wether they work at high
speed impact will be discussed later.
91
Manjunatha et al. [3.3.14] investigated the tensile fatigue behaviour. They used
a fabric with a +45°/-45° pattern, laid up 16 ply in a quasi-isotropic sequence
(+45°/-45° alternating with 90°/0°). Infusion temperature was 50 °C. The
tensile strength was increased from 365 MPa to 381 MPa (+4%); modulus was
increased from 17.5 GPa to 18.8 GPa (+7%). The 10 wt% silica nanoparticles
improved significantly the fatigue performance of the laminate. At a very high
stress (225 MPa) the number of cycles until failure was more than doubled.
Figure 3.3.4: Photograph of GFRC after impact testing without and with
nanosilica [3.3.12]
Further investigations compared the improvements in strength and modulus to

the changes of the bulk resin system [3.3.15]. Bulk strength increased by 19%,
laminate strength by 4%. Bulk modulus was increased by 17%; laminate
modulus by 7%. The fatigue life of the bulk epoxy is three to four times higher
for the resin with 10 wt% silica nanoparticles than for the unmodified control.
The same improvement was found for the laminate, as can be seen in Figure
3.3.5. From investigations using transmission optical microscopy, it was found
that the crack density after a given number of cyclic loadings was much lower.
The nanosilica apparently suppresses the matrix cracking and reduces the crack
growth rate which results in an enhanced fatigue life.
92
300
Neat matrix
(MPa)
Modified matrix
250
max
Maximum Stress, s
200
150
GFRP composite
R = 0.1
n = 1 to 3 Hz
RT, Lab air
100
50
1.E+02 1.E+03 1.E+04 1.E+05 1.E+06
No. of cycles, N f
Figure 3.3.5: Stress versus lifetime for a GFRC with epoxy resin matrix;
unmodified and with 10 wt% nanosilica [3.3.15]
Kinloch, Taylor and their team prepared UD GFRCs, 12 ply, 6-mm thick and
investigated the toughness [3.3.16]. 10 wt% silica nanoparticles increased the
GIc from 330 J/m2 to 1015 J/m2, this equals an improvement of more than 200%.
The bulk resin GIc was improved by 130% at 11 wt% nanosilica. They reported
small agglomerates of silica nanoparticles for the bulk resin, which seemed to
have no negative influence on performance. Quasi-isotropic laminates, 8 ply, 4-
mm thick by using a biaxial stitched fabric laid up to give a 0°/0° interface
across the fracture plane were prepared. Interestingly the GIc was lowered by the
nanosilica modification from 718 J/m2 to 626 J/m2 (-13%). This is very probably
due to the different fibre architecture.
Figure 3.3.6: Scanning electron micrographs of fracture surfaces of quasi-

isotropic GFRC with epoxy resin matrix: (a) unmodified , (b) with 10 wt%
nanosilica [3.3.16]
93
Figure 3.3.6 compares the fracture surfaces of laminates without and with silica
nanoparticles. Crack propagation was in the direction of the fibers; from left to
right. Disregarding the different magnifications, there seems to be no big
difference for the fracture surface; both exhibit brittle failure with little plastic
failure.
3.3.3.2. Carbon fiber-reinforced epoxy resin composites
Amine cured epoxy resin systems
Caccavale et al. [3.3.17] made laminates by using VARTM to infiltrate 18 layers

of T300 carbon fiber in a 90°/0° lay-up. As curing agent a commercial polyamine
hardener was used. The dispersion of the nanosilica in the epoxy resin was found
to be very well. Fracture toughness (KIc) for the bulk resin was increased by 11%
at an addition level of 3.7 wt% silica nanoparticles to 1.01 MPam1/2. No
improvement was found in the laminate, however. The initiation value of GIc was
decreased by 9%; the propagation value increased by 4%, which is within the
standard deviation. Impact testing showed a slight decrease of energy
absorption. ILSS was increased slightly. The tests were performed for wet
conditions as well, but the changes by the nanosilica addition were similar.
In her thesis more details were given [3.3.18]. Cyclic fatigue performance was
not improved either by the silica nanoparticles addition. Improvements of the
matrix properties could not be transformed into improvements of the laminate.
The conclusion drawn was that these properties are fiber-dominated and not
improved by the nanosilica addition. However, it should rather be said that the
nanosilica had no influence on toughness and fatigue performance in this case.
Or, eventually higher loading levels of nanosilica are necessary to see an effect.
Tang et al. [3.3.19] used piperidine as curing agent and vacuum-assisted resin
infusion (VARI) to prepare laminates. The unidirectional carbon fiber fabric used
was UT70-20 (24 layers for ILSS testing and 20 layers for mode I and mode II
testing) and the resulting fiber volume fraction was 65 %. The tensile strength of
the bulk epoxy was increased from 64.8 MPa to 69.4 MPa at 10 wt% nanosilica
and to 72.4 MPa at 20 wt% loading. Tensile modulus increased from 2.9 GPa to
3.3 GPa to 3.6 GPa. GIc increased from 237.9 J/m2 to 457.9 J/m2 at 10 wt%
nanosilica to 666.4 J/m2 at 20 wt% nanosilica. These are improvements of 93%
and 180%, respectively.
The laminates with the modified resins were tested extensively. The G Ic of the
laminate (at crack propagation) was improved from 995 J/m2 to 1007 J/m2 at 10
wt% nanosilica and to 1203 J/m2 at 20 wt%; an improvement of 1% and 21%
only. The ILSS was lowered by the nanosilica addition by 3% and 12%. GIIc was
lowered as well – from 969 J/m2 to 929 J/m2 and to 750 J/m2 (-23 %). The
conclusion drawn was that especially for the high concentration of 20 wt% of
silica nanoparticles the shear strength of the matrix was reduced.
Liu et al. [3.3.20] investigated laminates made from plain woven carbon fibers,
18 ply, with a fiber volume fraction of 60%. Piperidine was used as curing agent,
VARI as manufacturing method.
94
GIc of the laminate was improved by approx. 24 % for 4 wt%, 6 wt%, 8 wt% and
12 wt% silica nanoparticles, whereas the GIc for the bulk resin is increased
linearly with increasing nanosilica level for more than 100%. This could indicate a
plateau in performance improvements already at 4 wt% nanosilica. Of course it
has to be taken into account that woven fabrics have a big influence on GIc;
which might “hide” some effects of the silica nanoparticles.
A completely different test specimen and totally different properties were studied
by Zhang and his team [3.3.21]. They prepared short carbon fiber reinforced
epoxy resins; cured with a cycloaliphatic amine. Tenax A-385 short carbon fibers
with an average diameter of 7 µm and a length of 40 -70 µm were added at 10
vol% as well as 8 vol% graphite flakes. Silica nanoparticles were introduced in
addition levels of 1–5 vol%. The distribution of the nanosilica was uniform;
strength and modulus were increased. However, the main topic of the
investigation was related to the wear properties on both rough and mirror-
polished surfaces. The coefficient of friction was lowered significantly by the
spherical nanosilica particles. The wear rate was improved by approximately
50%. A maximum in performance improvements was found for 3 vol% silica
nanoparticles which equals approx. 5 wt%.
Sprenger et al. [3.3.22] used tetraglycidyl ether of methylene dianiline epoxy

resin (TGMDA), cured with sterically hindered aromatic amines and 8 plies of 439
T fabric to prepare laminates by RTM. Fibre content was 60%. The bulk resin
fracture energy was increased from 180 J/m2 to 320 J/m2 at 10 wt% nanosilica
and lowered again to 160 J/m2 at 19.5 wt%. The fracture energies of the
laminates however were reduced significantly by more than 60% for 5 wt% and
10 wt% of silica nanoparticle addition; the laminate with 19.5 wt% was only
slightly inferior to the control.
Falling dart test with 30 J impact showed no significant differences in

delaminated areas for the laminates with different concentrations of nanosilica
and without nanosilica. This confirms other findings for fast impact testing of
GFRCs. Apparently the mechanisms of toughening found for cured epoxy resins
modified with nanosilica in bulk does not apply for fiber-reinforced composites
made from nanosilica containing epoxy resins; at least not for fast impact
[3.3.2]. This will be discussed later.
Hackett et al. [3.3.23] studied spherical functionalized silica nanoparticles from a

different source with a particle size of 84 nm. They were manufactured as well by
a sol-gel method and dispersed in an epoxydized phenol novolac resin.
They prepared prepregs using dicyandiamide as curing agent together with a

urea accelerator and TR50S carbon fiber; 12 plies with a 0° orientation were
used to prepare the laminate for compression testing; 24 plies for the flexural
test specimens.
The modulus of the bulk resin was increased with increasing nanosilica addition
level by more than 100% at 45 wt% nanosilica. GIc was increased by approx.
50%. Bulk strength remained unchanged up to 35 wt% and then increased
slightly. The compressive strength of the laminate was increased from 1.8 GPa to
2.0 GPa at 45 wt% nanosilica which equals an improvement of 11%. At 15 wt%
1.8 GPa or an improvement of 3% were found.
95
Flexural modulus was found unchanged even at very high addition levels of silica
nanoparticles and is apparently fiber-dominated. The flexural strength was
increased from 1.5 GPa to 1.9 GPa at 45 wt% (an improvement of 26%). At 15
wt% the improvement found was approximately 10%.
In continuation of this work they investigated TGMDA cured with 4,4´-diaminodi-

phenylsulfone (DDS) and modified with 45 wt% of silica nanoparticles from a
different source with an average particle size of 154 nm [3.3.24]. The bulk resin
modulus was increased from 3.8 GPa to 7.8 GPa at 45 wt% nanosilica; fracture
toughness from 0.57 MPam1/2 to 0.76 MPam1/2. The coefficient of thermal
expansion was reduced from 40 to 25 µm/m/°C due to the fact that 45% of
organic resin and hardener had been replaced by inorganic filler. Prepregs and
then laminates were made with 38 wt% silica nanoparticles, but no reference
data without silica were provided. The laminate was reported to have a high
strength, high compressive strength and a high modulus.
Furthermore a DGEBA-based system was evaluated. At 43 wt% of nanosilica the

tensile modulus was doubled from 3.1 GPa to 6.3 GPa, the strength was
increased moderately from 56.8 MPa to 67.2 MPa and the fracture toughness was
increased from 0.51 MPam1/2 to 0.72 MPam1/2; an improvement of 41%.
Prepregs and laminates were made and tested: at 43 wt% silica nanoparticles
the shear modulus was increased from 3.7 GPa to 5.1 GPa. Shear strength was
improved from 53.7 MPa to 61.6 MPa and the compressive strength from 0.7 GPa
to 1.0 GPa.
In recent papers they revealed more details [3.3.25], [3.3.26]: imidazole was
used as hardener. The prepregs were prepared using T300-3k and T700-12k 2x2
twill carbon fabrics. Four types of laminates were made, three with T300: 8 ply
0° for compression tests, 14 ply 0° for flexural testing, 36 ply +-45° for in-plane
shear. The fourth was made from 1 layer T300 followed by 6 respectively 8
layers of T700 and finally a last layer of T300. Nanoindentation tests showed an
increased strength and hardness. Vickers hardness tests revealed similar
improvements.
Chiu et al. [3.3.27] investigated the property improvements of CFRC by

polyacrylonitrile (PAN) nanofibers. As second control they used a system
containing 5 wt% of the commercially available nanosilica. The modification with
the PAN-nanomats did not really show any improvements, whereas the addition
of silica nanoparticles did improve the GIc by 25–41 % (depending on the fiber
reinforcement and the resin/hardener system).
Anhydride cured epoxy resin systems
Soutis at al. [3.3.28] worked with DGEBA cured with 1-methyl-5-norbornene-

2,3-dicarboxylic acid anhydride (NMA) accelerated with a small amount of
ternary amine. Eight layers of HTS40 12K carbon fibres were wound dry on a
frame, impregnated and cured using vacuum bag technology. The compressive
properties for different addition levels of silica nanoparticles were investigated.
The compressive modulus increased with increasing nanosilica content from 85.7
GPa to 132.4 GPa at 11.7 vol% nanosilica; which equals an improvement of
54%.
96
The compressive strength increased quickly from 825.9 MPa to 1150.6 MPa at
2.1 vol%, then to 1268 MPa at 5.5 vol% (54% improvement) and finally dropped
to 1172 MPa at 11.7 vol % silica nanoparticles. Strain at break did not change
significantly. It is clear that the presence of nanosilica in the resin improves the
compressive properties of a laminate significantly. This finding is used in quite a
few industrial applications, like composite leaf springs.
Thunhorst et al. [3.3.29] used methyltetrahydrophthalic acid anhydride (MTHPA)

and DGEBA to formulate a resin/hardener blend designed for pultrusion. It
contained 32.6 wt% of nanosilica from a different manufacturer with a particle
size of probably 86 nm [3.3.23]. Standard 12 K tows were used to prepare
pultruded rods with an average fiber volume of approx. 64% and an average
diameter of 1.27 cm. They reported significant processing improvements through
reductions in pull force, enabling slight increases in fiber volume and line speed
increase. Improvements of up to 10% were found for modulus and flexural
failure stress, however this could be the result of higher fiber volumes rather
than the nanosilica addition.
Sprenger et al. [3.3.30] had chosen accelerated methylhexahydrophthalic acid

anhydride to cure DGEBA. The GIc of the bulk resin was increased from 100 J/m2
to 205 J/m2 at 4 wt% nanosilica to a plateau around 400 J/m2 for 7.5 wt% and
11.2 wt% - an improvement of 300%. 6 plies of carbon-fiber fabric were used to
prepare 3 mm laminates with a 60% fiber content using VARTM technology. The
flexural modulus of the laminate was improved by 2.5% from 24.5 GPa to 25.1
GPa at 12 wt% nanosilica. GIc was increased from 430 J/m2 to 550 J/m2, which
equals an improvement of 28%.
Hsieh et al. [3.3.16] investigated the same system. The GIc of the bulk resin was
increased from 77 J/m2 with increasing silica addition level to 212 J/m2 at 20
wt% nanosilica which represents an improvement of 175%. Modulus was
improved from 2.96 GPa to 3.85 GPa at 20 wt% nanosilica. Laminates were
prepared by stacking a linen woven carbon fiber fabric (0°/90° weave) and using
VARTM. The modified system contained 12 wt% silica nanoparticles. GIc was
increased from 439 J/m2 to 489 J/m2, a minor improvement of only 11%. Figure
4.3.7 shows the fracture surfaces of the laminates with the unmodified and
modified epoxy resin matrix. Apparently the presence of silica nanoparticles has
no significant influence on the fracture behaviour in this fiber-reinforced system.
The research of Liu and his team [3.3.31] was based on DGEBA as well and used
a blend of methylhexahydrophthalic acid anhydride and sebacic anhydride as
hardener. The tensile strength of the bulk resin was increased from 74.0 MPa to
79.7 MPa at 2 vol% nanosilica and to 86.7 MPa at 14 vol% nanosilica.
The elastic modulus was increased from 2.6 GPa to 2.8 GPa (at 2 vol% SiO2)
respectively 3.8 GPa at 14 vol% SiO2. Single T-300 carbon fibers were embedded
into the center of the test specimen mold. Single-fiber fragmentation tests were
performed.
97
Figure 3.3.7: Scanning electron micrographs of fracture surfaces of CFRC with an
epoxy matrix: (a) unmodified; (b) with 12 wt% nanosilica [3.3.16]
The critical break density (number of breaks / mm at failure) of the fiber was
improved by 19% at 2 vol% and by 27% at 14 vol% nanosilica. The interfacial
shear strength was calculated and showed improvements of 21% and 30%
respectively. Taking into account the results of micro-Raman spectroscopy and
SEM microscopy, they concluded, two effects to be responsible: The stress
transfer from matrix to fiber is made more efficient by the nanosilica particles
and the nanoparticles increase the interfacial frictional stress between fiber and
matrix.
Kuehn et al. [3.3.13] prepared quasi-isotropic test specimen with 4 mm

thickness using 16 ply of carbon fiber fabric AUW 797-1. DGEBA containing 25
wt% silica nanoparticles was cured with MTHPA. They reported identical
delamination areas for both control and modified epoxy resin at different impact
energies of 20 J, 30 J and 40 J. However, the compressive strength was approx.
7% higher for the nanosilica containing laminate after 20 J impact and approx.
5% higher after 40 J impact.
3.3.3.3. Comparision between bulk and laminate property improvements
By the addition of silica nanoparticles strength and modulus of cured bulk epoxy
resins were increased, of course, depending on the epoxy resin/hardener
combination, by different magnitudes.
Summing up the many details provided in the paragraphs before, it can be

concluded that regardless of the fiber or fabric or nonwoven used as
reinforcement and their orientations in the laminate the composite modulus is
not or only slightly improved by the presence of silica nanoparticles in the epoxy
resin. A similar trend was observed for the tensile strength of laminates made
from woven textiles or UD laminates tested in fiber direction. These properties
are fiber-dominant. Tensile tests with UD laminates in other directions than the
fiber direction could reveal an effect of the nanosilica added.
Unfortunately, very often either bulk or laminate properties were studied by the
researchers, but not both. Few report GIIc data.
98
Thus Table 3.3.1 gives an overview of the GIc data provided by the researchers
who investigated both bulk and laminate properties. It becomes evident, that the
significant improvements found for the modified bulk resins are not reflected in
the fiber-reinforced laminates.
Reference [3.3.4] [3.3.4] [3.3.19] [3.3.19] [3.3.30] [3.3.16] [3.3.16] [3.3.16]

[3.3.5] [3.3.5]
Resin / DGEBA DGEBA DGEBA DGEBA DGEBA DGEBA DGEBA DGEBA
hardener IPD IPD Piperi- Piperi- An- An- An- An-
dine dine hydride hydride hydride hydride
Nanosilica 10 20 10 20 11 11 12 20
(wt%)
Bulk GIc 280 350 458 666 400 183 - 212
(J/m2)
Improve- +47 +84 +92 +180 +300 +138 - +175
ment (%)
Fiber rein- Glass Glass Carbon Carbon Carbon - Carbon -
forcement
Composite 900 950 1007 1203 550 - 489 -
GIc (J/m2)
Improve- +8 +15 +1 +21 +28 - +11 -
ment (%)
DGEBA: diglycidyl ether of bisphenol A
Table 3.3.1: GIc of various bulk resins and laminates
This raises the question regarding the toughening mechanisms which apply for
the fiber-reinforced materials and how they differ from the toughening
mechanisms of cured bulk resins.
3.3.3.4. Mechanisms of toughening
The mechanisms for the property improvements of bulk epoxy resins have been
well investigated and are fully understood [3.3.2]. Two main toughening
mechanism have been identified for silica nanoparticles: Localised shear-bands
initiated by the stress concentrations around the periphery of the nanoparticles
and the debonding of the silica nanoparticles followed by subsequent plastic void
growth.
In a fiber-reinforced system, additionally mechanisms like fiber bridging and fiber

debonding may play a certain role; fiber pullout probably a minor one.
Microscopical studies of fracture surfaces revealed some differences between the

laminates with nanosilica and without nanosilica at typical addition levels around
10 wt%. The fracture failure mode changes from interface failure between glass
or carbon fiber and epoxy matrix to a combination of interface failure and matrix
failure for the laminates with nanosilica [3.3.4], [3.3.6], [3.3.16], [3.3.19].
This might explain why the improvements in toughness found are relatively small
compared to bulk resins, as the major failure “source” is still the failure at the
interface between fiber and resin matrix.
99
Another issue to take into account is the effect of strain rate – GIc testing versus
falling dart test. There is abundant literature available which to discuss would go
beyond the scope of this paper. Jacob et al. [3.3.32] published a comprehensive
overview for various fiber reinforced composites.
As the modification of the matrix resin apparently has a rather small effect on
fiber matrix adhesion, this could explain why we found no difference in
delaminated area between control and modified system for the falling dart tests.
3.3.3.5. How much silica nanoparticles do you really need – and where?
A very interesting and promising aspect was researched by Gao et al. [3.3.33].
They investigated changes in laminate properties when the silica nanoparticles
are only present in the interphase between fiber and resin. Fiber sizings
containing 1% of 22 nm silica nanoparticles without surface modification from a
different source were prepared and applied to the unsized glass fibers. An
increase in fiber surface roughness was observed. They reported superior
interphase performance compared to the unsized control: strength, debonding
and total energy absorption were improved. Failure modes were studied by SEM
and it was found that the crack was deflected which apparently increased energy
consumption.
Similar to Gao et al., Yang et al. [3.3.34] investigated changes in laminate

properties with silica nanoparticles only present in the interphase between fiber
and resin. They made composites based on tetrafunctional epoxy resin (TGMDA)
cured with DDS. Unsized carbon fibers were prepared and treated with a sizing
agent based on an epoxy resin which contained 3 wt% silica nanoparticles with
an average diameter of 20 nm and a surface treatment with a silane coupling
agent. Then the fibers were impregnated, prepregs made and cured in a mould.
An increase in ILSS was reported: +14% compared to unsized fibers and +5%
compared to a sizing without nanosilica. They concluded that the interfacial
adhesion between the fiber and the matrix is increased. The crack propagation
path in the sizing layer was found to be changed by the presence of the
nanosilica.
We started research activities in this area as well: composites plates of

unidirectional ([0/0]) and quasiisotropic ([+45/-45/0/90/-45/+45/90/0]5) layups
were manufactured by resin infusion using flexible tooling method. Stitched
+45/-45 and unidirectional glass fiber fabrics were supplied by Saertex. A two
part epoxy resin system (Epikote MGS RIMR 135 + Epikure MGS RIMH 137 from
Momentive) was used as matrix material. Before infusion the glass fibers were
dipped in an aqueous emulsion of a silica nanoparticle containing epoxy resin and
dried subsequently to obtain a defined amount of epoxy resin on the fabric.
The first tests performed were fracture toughness tests. Figure 3.3.8 shows the
improvements in GIc of the laminates made from pretreated UD. An addition level
of 1.5 wt% aqueous emulsion equals 0.36 wt% silica nanoparticles on fiber
weight; 4 wt% equal 1.12 wt% silica nanoparticles on fiber weight.
100
Additionally, an increase in the adhesion between fibers and matrix was indicated
by the presence of a significant amount of resin bonded on the fibers when the
fracture surfaces of the modified samples were observed, a feature that was not
observed in the case of control fracture surfaces.
Next property to be investigated will be the fatigue performance. And the

question, if some or most of the SiO2 nanoparticles are washed off the fabric and
dispersed into the matrix during the resin infusion process or if the stay at the
interface fiber-matrix needs to be addressed.
Figure 3.3.8: GIc of pretreated glass fiber reinforced composites made from
commercial unmodified epoxy resin
If more data will confirm that for substantial improvements of the laminates it
will be sufficient to apply thin layers on the fabric, this will be of huge
commercial interest. Calculated on total it would mean that far less than 1 wt%
silica nanoparticles would be enough, which is negligible from a cost aspect –
compared to modify the whole resin matrix.
Of course this field needs further future research to fully understand such
systems.
3.3.4. Conclusions
1. Significant improvements in strength, modulus and toughness were found

for bulk epoxy resins modified with silica nanoparticles. When these resins
are used as matrix for fiber reinforced composites much smaller
improvements were found.
101
2. Not taking into account the reinforcing fibers, their volume fraction and
orientation respectively the layup used and disregarding the hardeners
used a general outline can be given: Typical improvements found for fiber
reinforced composites with an epoxy resin matrix modified with 10 wt%-
20 wt% of silica nanoparticles are increases of 5–25% in fracture
toughness.
3. Most significant improvements are found for the compressive properties
and for the fatigue performance.
4. Consequently many industrial composite parts today are based on epoxy
resins modified with silica nanoparticles – glass fiber-reinforced as well as
carbon fiber-reinforced.
5. Improving laminate properties by using silica nanoparticles in the sizing
seems to be a very cost-efficient route, which needs further evaluation.
6. To push the performance levels of fiber-reinforced systems further,
especially regarding toughness and fast impact, the combination of
classical toughening methods with the silica nanoparticle modification is
the next logical step in technology development [3.3.35]. This will be the
subject of future investigation.
3.3.5. References
[3.3.1] Sprenger S, Eger C, Kinloch AJ, et al. Nanotoughening of epoxies.

Proceedings of Stick! Conference (2003), April 9, Nuernberg,
Germany, Vincentz Verlag, 9 April 2003.
[3.3.2] Sprenger S. Epoxy resin composites with surface-modified silicon

dioxide nanoparticles. J. Applied Polym. Sci. (2013), 130, 1421.
[3.3.3] Sprenger S, Kinloch AJ, Taylor AC, et al. Rubber-toughened GFRCs

optimized by nanoparticles. JEC Compos. Magazine (2005), 21, 66.
[3.3.4] Tsai JL, Huang BH and Cheng YL. Enhancing fracture toughness of
glass/epoxy composites by using rubber particles together with
silica. J. of Compos. Mater. (2009) 43, 25, 3107.
[3.3.5] Tsai JL, Huang BH and Cheng YL. Enhancing fracture toughness of
glass/epoxy composites for wind blades using silica nanoparticles
and rubber particles. Procedia Eng. (2011), 14, 1982.
[3.3.6] Tsai JL, Huang BH and Cheng YL. Investigating mechanical

behaviours of silica nanoparticle reinforced composites. J. of
Compos. Mater. (2010) 44, 4, 505.
[3.3.7] Tate JS, Akinola AT and Sprenger, S. Mechanical performance of

nanomodified epoxy/glass composites for wind turbine applications.
In: Proceedings of the SAMPE Conference (2010), Seattle, WA, USA,
17–21 May 2010
[3.3.8] Mahrholz T, Herbeck L and Riedel U. New high-performance fibre-

reinforced nanopomposites. JEC Compos. Magazine (2004), 9, 71.
102
[3.3.9] Herbeck L, Mahrholz T, Mosch J, et al. Faserverstärkte Nano-
composites – Stand der Technik und Perspektiven. GAK (2005), 58,
161.
[3.3.10] Kinloch AJ, Masania K, Taylor AC, et al. The fracture of glass-fibre-
reinforced epoxy composites using nanoparticle-modified matrices.
J. Mater. Sci. Letters (2008), 43, 1151.
[3.3.11] Kinloch AJ, Masania K, Taylor AC, et al. The fracture of nanosilica
and rubber toughened epoxy fibre composites. In: Proceedings of
the Composites & Polycon (2009), American Composites
Manufacturers Association, Tampa, Florida, USA, 15-17 Jan. 2009
[3.3.12] Sprenger S, Eger C, Kinloch A, et al. Rubber-toughening and

nanoparticles in epoxies: synergies in FRC. Proceedings SAMPE
(2005), Paris, France, April 5 - 7, 2005
[3.3.13] Kuehn A, Mahrholz T and Mosch J. Matrix optimization of CFRP parts

concerning fire and impact properties with process acceleration.
CEAS Aeronaut. J. (2013), 4, 191.
[3.3.14] Manjunatha CM, Taylor AC, Kinloch AJ, et al. The effect of rubber
microparticles and silica nanoparticles on the tensile fatigue
behavior of a glass-fibre epoxy composite. J. Mater. Sci. Lett.
(2009), 44, 342.
[3.3.15] Manjunatha CM, Taylor AC, Kinloch AJ, et al. The tensile fatigue
behaviour of a silica nanoparticle-modified glass fibre reinforced
epoxy composite. Compos. Sci. Technol. (2010), 70, 193.
[3.3.16] Hsieh TH, Kinloch AJ, Masania K, et al. The toughness of epoxy
polymers and fibre composites modified with rubber microparticles
and silica nanoparticles.; J. Mater. Sci. (2010), 45, 1193.
[3.3.17] Caccavale V, Wichmann M, Quaresimin M, et al. Nanoparticle/rubber

modified epoxy matrix systems: mechanical performance in CFRPs.
In: Proceedings of AIAS XXXVI Convegno Nazionale, (2007), Ischia,
Naples, Italy, 4-8 September 2007.
[3.3.18] Caccavale V. Nanoparticle/rubber-modified epoxy matrix systems:

mechanical performance in CFRPs. Masters Thesis University of
Padova, Italy (2007)
[3.3.19] Tang Y, Ye L, Zhang D, et al. Characterization of transverse tensile,

interlaminar shear and interlaminate fracture in CF/EP laminates
with 10 wt% and 20 wt% silica nanoparticles in matrix resin.
Compos. Part A (2011), 42, 1943.
[3.3.20] Zeng Y, Liu HY, Mai YW, et al. Improving interlaminar fracture
toughness of carbon fibre/epoxy laminates by incorporation of
nanoparticles. Compos. Part B (2012), 43, 90.
103
[3.3.21] Zhang G, Sebastian R, Burkhart T, et al. Role of monodispersed
nanoparticles on the tribological behaviour of conventional epoxy
composites filled with carbon fibers and graphite lubricants, Wear
(2012), 292-293, 176.
[3.3.22] Sprenger S, Kinloch AJ, Taylor AC, et al. Rubber-toughened CFRCs

optimised by nanoparticles – Part III. JEC Compos. Magazine
(2007), 30, 54.
[3.3.23] Hackett SH, Nelson JM, Hine AM, et al. The effect of nanosilica
concentration on the enhancement of epoxy matrix resins for
prepreg composites. In: Proceedings SAMPE (2010), Salt Lake City,
UT, U.S.A. 11–14 Oct., 2010
[3.3.24] Nelson JM, Hackett SC, Goetz DP, et al. Development of nanosilica-
thermoset matrix resins for prerpreg composites. NSTI-Nanotech
(2012), ISBN 978-1-4665-6274-5 Vol. 1
[3.3.25] Nelson JM, Hine AM, Goetz DP, et al. Nanosilica-modified epoxy
matrix resin for prepreg composite tooling applications. In:
Proceedings SAMPE (2012), Baltimore, MD, U.S.A. 21–24 May,
2012
[3.3.26] Nelson M, Hine AM, Goetz DP, et al. Properties and applications of
nanosilica-modified tooling prepregs. SAMPE J. (2013), 49, No. 1, 7.
[3.3.27] Chiu KR, Duenas T, Dzenis Y, et al. Comparative study of nano-

materials for interlaminar reinforcement of fiber-composite panels.
Proc. of SPIE (2013), 8689, 86891D1
[3.3.28] Jumahat A, Soutis C, Jones F, et al. Improved compressive

properties of a unidirectional CFRP laminate using nanosilica
particles. Adv. Compos. Lett. (2010), 19 (6), 218.
[3.3.29] Thunhorst K, Goetz DP, Hine AM, et al. The effect of nanosilica
matrix modification on the improvement of the pultrusion process
and mechanical properties of pultruded epoxy carbon fiber
composites. In: Proceedings Composites (2011), Ft. Lauderdale,
FL, U.S.A., 2–4 Feb., 2011
[3.3.30] Sprenger S, Kinloch AJ, Taylor AC, et al. Rubber-toughened FRCs

optimised by nanoparticles. JEC Compos. Magazine (2005), 19, 73.
[3.3.31] Liu LQ, Li L, Gao Y, et al. Single carbon fiber fracture embedded in
an epoxy matrix modified by nanoparticles. Compos. Sci.Technol.
(2013), 77, 101.
[3.3.32] Jacob GC, Starbuck JM, Fellers JF, et al. Strain rate effects on the
mechanical properties of polymer composite materials. J. Appl.
Polym. Sci. (2004), 94, 296.
104
[3.3.33] Gao X, Jensen RE, McKnight SH, et al. Effect of colloidal silica on the
strength and energy absorption of glass fiber/epoxy interphases.
Composites: Part A (2011), 42, 1738.
[3.3.34] Yang Y, Lu CX, Su XL, et al. Effect of nano-SiO2 modified emulsion

sizing on the interfacial adhesion of carbon fibers reinforced
composites. Mater. Lett. (2007), 61,3601.
[3.3.35] Sprenger S. Epoxy resins modified with elastomers and surface-

modified silica nanoparticles. Polymer (2013), 54, 4790.
105
3.4. Property improvements of fiber-reinforced composites
based on epoxy resins modified with SiO2 nanoparticles
and elastomers (hybrid systems)
3.4.1. Abstract
The properties of fiber-reinforced composites made by using epoxy resin

formulations can be improved by using modified epoxy resins. As epoxies are
inherently brittle, they are toughened with reactive liquid rubbers or core-shell
elastomers. Surface-modified silica nanoparticles, 20 nm in diameter and with a
very narrow particle size distribution, are available as concentrates in epoxy
resins in industrial quantities for the past 10 years. Some of the drawbacks of
toughening like lower modulus or a loss in strength can be compensated when
using nanosilica together with these tougheners. Apparently, there exists a
synergy as toughness and fatigue performance are increased significantly. Some
of these improvements in bulk resin properties can be found for fiber-reinforced
composites as well. In this article, the literature published in the last decade is
studied with a focus on mechanical properties. Results are compared, and the
mechanisms responsible for the property improvements are discussed. A
relationship between the improvements of the fracture energy of the cured bulk
epoxy resins and the fracture energy of the fiber-reinforced composites could be
established.
3.4.2. Introduction
Fiber-reinforced composites are a key technology of the twenty-first century as

they enable lightweight construction in many different areas of modern life. Not
only transportation applications like aerospace, automotive, railway, yacht- or
shipbuilding are depending on composites components to be fuel-efficient and
long-lasting. Wind energy installations, crude oil pipes for offshore platforms,
energy recovery flywheels, compressurized gas tanks, and many other
applications in the energy market are impossible without fiber-reinforced
composites either. Construction, sports equipment, machine parts, protective
equipment like helmets – fiber-reinforced composites are everywhere.
In most of them, glass fibers are used as reinforcement, woven or nonwoven,

unidirectional or multiaxial, or as single filaments. Carbon fibers gain importance,
other fibers like aramides or high-strength polyethylenes are used as well.
Thermoplastic or thermosetting materials are used as matrices [3.4.1], [3.4.2].
In many composite applications, epoxy resins are used as matrix materials, and
the diglycidyl ether of bisphenol A (DGEBA) is the most important resin used by
the industry. The diglycidyl ether of bisphenol F (DGEBF) and the higher
molecular weight variations, epoxidized novolac resins, are popular raw materials
as well. For high-performance applications like aerospace parts, multifunctional
epoxy resins are used, mainly the tetraglycidyl ether of methyldianiline (TGMDA).
The combination of epoxy resin and hardener defines the matrix material and the
various possible combinations allow to formulate according to all possible
property requirements [3.4.1], [3.4.2].
106
The variety of epoxy curing agents used in the industry and their network
formation was described by Hare quite a few years ago [3.4.3], [3.4.4]. In low
viscosity infusion resins typically for rotor blade manufacturing, mainly DGEBA is
used as base resin, blended with low viscosity aliphatic epoxy resins like
hexanediole diglycidylether (HDDGE). As hardeners cycloaliphatic amines like
isophorone diamine (IPD) are the material of choice; often in combination with
low molecular weight polyetheramines. Medium performance injection resins are
based on DGEBA with aromatic amines like diaminodiphenylsulfone (DDS) as
curing agents.
High performance aerospace resin systems are normally based on tri- or

tetrafunctional epoxy resins and sterically hindered aromatic amines like 4,4'-
Methylenebis(2,6-diethylaniline) (M-DEA) and 4,4'-Methylenebis (2,6-
diisopropylaniline) (M-DIPA) [3.4.2]. RTM 6, manufactured by Hexcel, is still
considered to be the aerospace industry benchmark material [3.4.2], [3.4.5].
Standard prepreg resins systems are often formulated with DGEBA and
dicyandiamide (DICY) as curing agent. In filament winding applications acid
anhydrides like methyl hexahydrophthalic acid anhydride (MHHPA) are used. Due
to new EU regulations, MHHPA will probably be phased out, but new high
performance dianhydrides like 4,4´carbonyl diphthalic acid anhydride (BTDA) are
available and enable to achieve extremely high glass transition temperatures (Tg).
All common manufacturing technologies are used in combination with epoxy

resin formulations [3.4.5]. Prepregs made by wet or dry impregnation, cured in
an autoclave or outside, filament winding, pultrusion – just to name a few.
Injection and infusion methods are very popular, like resin transfer molding
(RTM), vacuum assisted resin transfer molding (VARTM), or vacuum assisted
resin infusion (VARI). Of course, each method requires adapted formulations
regarding processability (e.g., viscosity) and reactivity.
Composite parts manufactured for the automotive industry have to match the
production cycles of the industry of only a few minutes. This limits of course the
manufacturing methods and the epoxy resin/hardener combinations which are
suitable [3.4.6].
However, epoxy resins are inherently brittle. Therefore many epoxy resin
formulations are toughened. Very often, rubber adducts are used. Introduced in
the 1970s, these carboxyl terminated nitrile butadienes (CTBNs) are immiscible
with epoxies but become miscible after a pre-reaction with an excess of epoxy
resin. They form soluble adducts which phase separate again upon cure and form
rubber domains in the epoxy resin. The mechanisms of rubber toughening are
well known [3.4.7], [3.4.8], [3.4.9].
Rubber toughening has several drawbacks like lowering the modulus and
reducing the glass transition temperature. This is caused by some rubber
molecules not participating in the phase separation, but crosslinking randomly
into the matrix. Another issue is the relatively high viscosity of epoxy resins
containing reactive liquid rubbers. Nevertheless they are used in composites
manufacturing.
107
Figure 3.4.1 shows the scanning electron microscopy (SEM) picture of a carbon
fiber-reinforced composite (CFRC) made by RTM, the epoxy resin matrix
toughened with 6.9 wt% of a reactive liquid rubber (CTBN). The spherical rubber
domains, which have formed upon cure, are clearly visible in the resin matrix
and are well dispersed. Thus rubber toughened epoxy resins can be used for
infusion-manufacturing methods.
Figure 3.4.1: SEM picture of a CFRC with CTBN-toughened matrix [3.4.10]
In the 1980s, core-shell elastomers (CSRs) have been developed. Rubber

particles are dispersed in the epoxy resin, and the viscosity is only increased
slightly. CSR Type I is characterized by a butadiene homopolymer or butadiene
styrene copolymer core of approx. 90 nm with a random copolymer shell 10-20
nm thick [3.4.11]. This shell is compatible with epoxy resins. CSR Type II are
based on a polyacrylate core and are larger; approx. 300–400 nm [3.4.12].
Strength, modulus and Tg of modified epoxy resins are lowered, but not to the
same extent as by reactive liquid rubbers. A major drawback is the softening of
the shell at higher temperatures which prohibits their use in applications for
elevated temperatures.
In the 1990s CSR Type III have been developed [3.4.13], [3.4.14]. They affect
strength, modulus, or glass transition temperature less. This was achieved by
reducing the thickness of the shell which consequently cannot soften anymore.
These epoxy-functional particles with a polysiloxane core have an average
diameter of 500-700 nm.
In 2002/2003 surface-modified silica nanoparticles became available in industrial

quantities. They are manufactured in situ directly in the epoxy resin by a
modified sol-gel process and have an average diameter of 20 nm as well as a
very narrow particle size distribution; as shown in Figure 3.4.2 [3.4.15]. The
particle size distribution is usually determined by small angle neutron scattering
(SANS).
108
These spherical particles appear transparent, are monodisperse and do not
increase resin viscosity up to quite high addition levels [3.4.15]. They do not
exhibit thixotropic properties like fumed silicas but behave like a Newtonian liquid.
particle density
0 20 40 60 80 100
particle size [nm]
Figure 3.4.2: Particle size distribution of the commercial silica nanoparticles

determined by SANS
Figure 3.4.3: TEM picture of a cured hybrid epoxy resin [3.4.19]
Properties like strength, modulus, toughness and especially fatigue performance

of cured epoxy resins can be improved by a modification with silica nanoparticles.
109
This has been subject of intensive research in the last ten years [3.4.16]. The
mechanisms responsible for the property improvements are rather well
understood today.
Due to their size and the unchanged viscosity of the resin, they can be used in
injection or infusion methods for composite manufacturing where they can
penetrate even close meshed fabrics easily.
The combination of reactive liquid rubbers or CSRs and the use of silica
nanoparticles as additional modifier in epoxy resin systems yields additive and
sometimes synergistic property improvements [3.4.17]. It is possible to
formulate tough and stiff materials. These resin systems and their synergies in
various applications, including fiber-reinforced composites were patented
consequently [3.4.18].
Today such hybrid systems are used in many industrial epoxy formulations.
Figure 3.4.3 shows the transmission electron microscopy (TEM) picture of a
cured hybrid epoxy resin. Both the core shell particles Type III (8 vol%) and the
silica nanoparticles (5.5 vol%) are well distributed.
The property improvements found for bulk epoxy resins modified with silica
nanoparticles were not found in fiber-reinforced composites made with such
resins or the improvements were much smaller in scale [3.4.16], [3.4.20].
Thus one of the questions to be answered is if the significant property

improvements reported for hybrid resins will be transferred into the fiber-
reinforced composites [3.4.17].
The aim of this article with a focus on mechanical properties is to provide a

comprehensive overview of the actual state of research regarding the various
different aspects and to provide formulation guidelines.
3.4.3. Discussion
All the researchers cited used commercial 40 % (by weight) concentrated

masterbatches of surface-modified nanosilica in DGEBA with an average diameter
of 20 nm and a very narrow particle size distribution.
These were then diluted down using commercial epoxy resins to vary the
nanosilica concentrations. All investigations revealed good dispersion of the
nanosilica particles, sometimes small agglomerates; typically determined using
TEM.
Though, quite some research was done with hybrid epoxy resins, only a part of
the researchers used the commercial nanosilica. Furthermore were the results of
several research projects transferred directly into industrial products and not
published at all.
110
3.4.3.1. Glass fiber-reinforced epoxy resin composites
Epoxy resins modified with reactive liquid rubbers (CTBNs) and silica
Sprenger et al. [3.4.21] prepared their samples from a linen-type fabric made
from 200 g/m2 glass fibers; eight layers of fabric had been placed in a closed
mold and impregnated by using VARTM. Standard DGEBA and a commercial
amine curing agent were used as resin matrix. They reported a decrease in
Youngs modulus of the bulk epoxy resin upon the CTBN modification and an
increase for the nanosilica addition. The hybrid resin exhibited a modulus of the
bulk resin slightly below the control.
An increase in toughness for the cured bulk resin was observed with increasing
addition levels of both rubber and nanosilica. However, the corresponding
laminates showed a different behaviour at 4.1 wt% rubber and 2 wt% nanosilica;
the fracture toughness fracture energy (GIc) of the bulk resin was more than
doubled from 609 J/m2 to 1576 J/m2; an improvement of 159 %, the GIc of the
laminate made using this system did not change much from 1175 J/m2 to 1205
J/m2. At 6.9 wt% rubber and 3.5 wt% nanosilica, the increase for GIc of the bulk
resin was + 174 %; for the laminate + 30 %. The GIc for the laminate showed a
maximum at this level and decreased on further addition of both rubber and
nanosilica, whereas the GIc of the bulk resin was further increased.
A commercial rubber-toughened epoxy resin based on DGEBA and cured with a

cycloaliphatic amine was used by Uddin and Sun [3.4.22] for their investigations.
Unidirectional (UD) E-glass fiber cloth was arranged in 6 layers for the tensile
test specimen, respectively, in 16 layers for the other test specimen to
manufacture the laminates by VARTM. Silica nanoparticle concentration was 15
wt%; the rubber concentration of the commercial resin is unknown. The
longitudinal compressive strength was tested for fiber volume fractions of 42
vol% and 50 vol%. The addition of the nanosilica improved the longitudinal
compressive strength by 81 and 61 % for the hybrid. Longitudinal tensile
strength was increased by 11 %; modulus was unchanged. In transverse
direction, the tensile strength was raised by 32 % and the modulus by 41 %.
Off-axis compressive strength was increased as well for the hybrid by 12 %, 20

% and 12 % for 5°, 10° and 15° at a fiber volume fraction of 42 vol%. At 50
vol% the improvements were even larger. Clearly the hybrid resin improved the
composite performance versus the rubber toughened resin. Unfortunately, a
comparison to a totally unmodified resin is not possible as the commercial
material is already rubber-toughened. In a report regarding the development of
toughened and multi-functional nanocomposites for ship structures which was
prepared by Sun [3.4.23] more details regarding the trials were given.
Tsai et al. [3.4.24], [3.4.25] modified a DGEBA resin with CTBN and silica
nanoparticles and cured it with a modified isophorone diamine. 12 layers of UD
E-glass were stacked in a modified hand lay-up process to prepare the laminates.
At 10 wt% rubber and 10 wt% nanosilica the GIc of the bulk resin was increased
from 190 J/m2 to 930 J/m2 – this represents an improvement of 390 %. The
laminate based on the unmodified resin had a GIc of 840 J/m2 and the laminate
made from the hybrid resin 1230 J/m2.
111
This equals an improvement of only 48 %. The Youngs modulus of the bulk
hybrid resin was reported to be only 2 % lower than for the control.
A commercial DGEBA resin formulation, modified with a CTBN adduct and silica
nanoparticles, cured with a commercial aliphatic amine blend were investigated
by Schultz et al. [3.4.26]. The hybrid epoxy resin formulation contained 6.5 wt%
CTBN and 8.1 wt% nanosilica. 8-ply of a bidiagonal ±45° stitch bonded glass
fiber fabric were processed by VARTM. The fiber volume fraction of the control
laminate was 0.51 and of the hybrid 0.54. Low-velocity impact tests were
performed by a drop test with different heights and weights corresponding to
different impact energies. For each energy level the hybrid samples showed
visibly more matrix cracking and ply delamination than the laminates made with
the control resin. It was concluded that the hybrid resin system has a better
impact energy dissipation mechanism than the unmodified epoxy resin.
In continuation of this study, compression after impact was tested [3.4.27]. It

was found that the laminates made by using the hybrid resin system showing
more delamination after impact had a reduced compressive strength as well. For
all energy levels, the compression strength after impact was approx. 6–8 %
lower for the hybrid laminates. Microscopical investigations revealed that the
rubber domains were uniformely distributed, but small agglomerates of the
nanoparticles had formed during the manufacturing process. This has probably
no influence on mechanical properties as the agglomerates are still nanoscaled.
Methylhexahydrophthalic acid anhydride (MHHPA), accelerated with a small

amount of a ternary amine was the hardener used in the following studies.
Kinloch et al. [3.4.28] investigated DGEBA, modified with 9 wt% CTBN and 10
wt% nanosilica. They prepared the laminates by using 16 plies of unidirectional
glass fiber fabric. The fiber volume fraction was determined to be approximately
57 %. The modulus of the laminate made with the hybrid resin was 40.6 GPa, 2
% higher than for the control; the Tg was slightly lower. Fracture toughness tests
were performed with promising results. The GIc of the control panel was 330
J/m2, and the hybrid was improved by 160 % to 860 J/m2. GIIc of the control was
found to be 1300 J/m2 and 1895 J/m2 for the hybrid, again an improvement of
46 %.
In continuation of this work Charpy impact tests have been performed and
showed improvements of approx. 40 % in impact energy [3.4.29]. Ballistic
impact tests with different impact energies showed slightly higher values for the
laminates made with the hybrid epoxy system.
Non-crimp glass fiber fabric arranged in ±45° pattern was laid up in a quasi-
isotropic sequence [(+45/-45,90/0)8]2 and infused with DGEBA epoxy resin
modified with 9 wt% rubber and 10 wt% silica nanoparticles by Manjunatha et al.
[3.4.30]. Fatigue testing showed improvements for laminates for both rubber-
modified and nanosilica-modified epoxy matrices in the same range.
112
The hybrid system with both modifications clearly performed best, as can be
seen in Figure 3.4.4. Regardless of the stress applied during the cyclic testing,
the number of cycles before failure was increased by up to 10 times.
Figure 3.4.4: Stress versus lifetime curves of a glass fiber-reinforced composite

[3.4.30] NR control, NRR rubber modified, NRS nanosilica modified, NRRS
rubber, and nanosilica modified
Kinloch, Taylor et al. [3.4.31] continued this study and had a closer look at the
mechanisms of toughening involved. They observed minor agglomerations of the
silica nanoparticles, which apparently had no negative effect on mechanical
properties. The mechanisms of toughening for the bulk resin were debonding of
the epoxy polymer from the silica nanoparticles, followed by plastic void growth
of the epoxy. Localised shear-banding in the polymer was observed as well. The
synergy between rubber and nanosilica (as described by Sprenger [3.4.17] for
bulk epoxy resins) was confirmed. The increase in toughness of the bulk hybrid
resin was transferred into the fiber composite. Typical toughening mechanisms
for fiber reinforced materials like fiber bridging, fiber debonding, and fiber pullout
came into play as well. They extended an existing model to predict laminate
toughness and found good agreement with the experimental data.
Further research results regarding the fatigue performance of laminates made

with hybrid epoxy resin systems were published by Manjunatha et al. [3.4.32].
The hybrid resin system contained 9 wt% CTBN rubber and 10 wt% silica
nanoparticles. Bulk modulus was lowered by 10 % compared to the control; the
same reduction of 10 % was found for the laminate modulus. The fatigue
strength coefficient of the control laminate was 463 MPa and increased by 16 %
to 535 MPa. For the fatigue strength coefficient of the bulk resin an improvement
of 19 % was found. The fatigue strength exponent was increased by 7 % for the
bulk resin and by 4 % for the laminate. Optical measurements revealed the
drastic increase in crack density after only a few test cycles with the laminates
made from the unmodified epoxy resin compared to the hybrid system.
113
The modification obviously suppresses matrix cracking and reduces the crack
growth rate. The cyclic fatigue life is 6–10 times higher for the hybrid than for
the control; which is quite a substantial improvement regarding the lifespan of
composite parts used in high stress applications (e.g., rotor blades for wind
energy installations).
Manjunatha et al. [3.4.33] then investigated the fatigue performance under

variable-amplitude loading. Again, the significant better performance of fiber-
reinforced composites made with hybrid epoxy resin systems was confirmed –
fatigue lifes were enhanced 3-4 times for the hybrid. Using spectrum load
sequences again improvements in the same magnitude were found [3.4.34].
Epoxy resins modified with CSRs and silica nanoparticles, amine

cured
Tsai et al. [3.4.24], [3.4.25] modified a DGEBA epoxy resin with CSR Type II
particles with an average particle size of 300–400 nm and silica nanoparticles. A
modified isophorone diamine was used as curing agent. The laminates were
made of 12 layers of UD E-glass. Mode I fracture tests with the laminates from
the unmodified epoxy resin system resulted in a GIc of 830 J/m2. The fiber
reinforced composite made from a resin modified with 10 wt% CSR and 10 wt%
silica nanoparticles showed a GIc of 1510 J/m2; which equals an improvement of
82 %. However, they reported a reduced modulus of the bulk resin by 8.6 %.
When comparing 10 wt% CSR Type II to 10 wt% CTBN, they concluded that the
property profile of the CTBN/nanosilica hybrid is superior. Keeping in mind that
for bulk resins, CSR and nanosilica are not synergistic compared to CTBN and
nanosilica this could be expected for fiber reinforced composites too [3.4.17].
A large research project investigated the properties of glass fiber-reinforced

composites based on DGEBA epoxy resin modified with core shell particles Type
III and nanosilica, cured with a commercial amine [3.4.35]. 26 and 32 layers of
bidirectional glass fiber fabric (linen type) were processed by VARTM into
laminates with 42 vol% fibers and 50 vol% fibers. Microscopical investigations
showed a uniform dispersion of both core shell particles and silica nanoparticles.
Cured bulk resin modulus did not change by the modification. The interlaminar
fracture energy was increased from 640 J/m2 to 1020 J/m2 at 7 vol% nanosilica
and 5.5 vol% core shell rubber which equals an improvement of 59 %. For the
bulk resin a KIc of 0.75 MPam0.5 and for the hybrid 2.9 MPam0.5 – an improvement
of 287 %. Unfortunately no GIc data was provided.
As the core-shell technology is younger than the epoxy resin modification with
reactive liquid rubbers and less established in composites manufacturing, less
research was performed in this field. As an example, no data is published
regarding the combination of CSR Type I and nanosilica in amine cured systems.
Thorough literature search did not yield any data regarding glass fiber-reinforced
composites made from epoxy resins modified with CSRs and industrial 20 nm
silica nanoparticles, which were anhydride cured.
However, the recent increases in using core-shell toughening for fiber reinforced
composites started to push research projects with these materials.
114
3.4.3.2. Carbon fiber-reinforced epoxy resin composites
Caccavale et al. [3.4.36], [3.4.37] investigated DGEBA modified with 7.3 wt%
CTBN and 3.7 wt% silica nanoparticles cured with a commercial polyamine
hardener. The tensile testing of the cured bulk resin showed a 10 % reduction in
modulus for the hybrid resin system. 18 layers of T300 NCF carbon fiber were
laid up in [90°/0°]9s and VARTM was used to manufacture the test laminates with
a fiber volume content of 53 %. The fracture toughness (KIc) of the control was
increased from 0.91 MPam0.5 to 2.11 MPam0.5 which equals an improvement of
132 %. Fatigue performance was reported to be increased significantly.
Sprenger et al. [3.4.38] used 8 layers of a 439 T fabric to prepare laminates by

RTM. The resin was TGMDA, with sterically hindered aromatic amines as curing
agents. The formulation was similar to RTM6, the aerospace industry benchmark
RTM resin. As expected, the nanosilica addition to the resin increased modulus
whereas the CTBN modification lowered the modulus considerably. The hybrid
with 8 wt% CTBN and 10 wt% silica nanoparticles had a slightly higher modulus
than the control for the bulk resin. Laminate GIc was the highest for the rubber-
only modification and lowest for nanosilica-only.
In falling dart tests the laminates were impacted with 30 J and the laminate
based on the hybrid resin system performed best (lowest delamination after
impact). In the pictures given in Figure 3.4.5, it is obvious that the dart nearly
went through the control, whereas the hybrid shows less damage. This is
confirmed by the corresponding C-Scans where there can be seen less
delamination for the hybrid system.
115
Figure 3.4.5: Pictures of CFRC laminates after 30 J impact testing and
corresponding C-Scans [3.4.38]
The toughness of laminate made from the hybrid epoxy resin was 1017 J/m2,
which equals an improvement of 33 %. Dry Tg were identical, however, when
stored at 100 % relative humidity and 70 °C for two weeks and wet Tg is
determined, there are differences. With increasing levels of nanosilica, the wet Tg
is lowered as well as for the CTBN modification. Thus, the hybrid system has the
lowest wet Tg (which is still above 190 °C).
In a project under the Clean Sky program CFRCs made by using hybrid epoxy
resins have been investigated as well. Kowalik et al. [3.4.39] infused HexForce
G0926 D1304 TCT. The GIc of a DDS-cured DGEBA was increased from 724 to
1000 J/m2; an improvement of 38 %. A second system similar to RTM 6 was
increased from 662 to 980 J/m2; which equals an improvement of 48 %. Both
hybrid formulations contained approx. 5 wt% nanosilica and approx. 10 wt%
CTBN.
Microscopical studies revealed well dispersed rubber domains and well dispersed
silica nanoparticles. In some formulations, small agglomerates were found which
were evenly dispersed and seemed not to affect the performance of the
composites. In Figure 3.4.6, the TEM microscopical image of the carbon fiber-
reinforced laminate is given. The rubber domains, formed upon cure are clearly
visible between two carbon fibers. At this magnification, however, the nanosilica
particles cannot be seen.
Figure 3.4.6: TEM picture of a carbon fiber-reinforced laminate made from hybrid
epoxy resin [3.4.40]
116
Figure 3.4.7 shows the same specimen at higher magnification. The uniform
distribution of the nanosilica is confirmed again; though the particle
concentration at the interface rubber particle/epoxy resin seems somewhat
higher. This seems to have no negative effect on performance.
Figure 3.4.7: TEM picture of a carbon fiber-reinforced laminate made from hybrid
epoxy resin [3.4.40]
Sprenger et al. [3.4.41] stacked six plies carbon fiber fabric (linen type) to
manufacture 3-mm thick laminates by VARTM. The GIc of the laminate made with
the bulk resin was 439 J/m2. The hybrid laminate, modified with 9 wt% CTBN
and 5 wt% silica nanoparticles showed increased toughness with 1080 J/m2. The
system containing 9 wt% CTBN and 8 wt% nanosilica achieved a further increase
in GIc of 1110 J/m2. The hybrid laminate modified with 9 wt% CTBN and 11 wt%
nanosilica performed best; the GIc was found to be 1320 J/m2 which equals an
improvement of 200 % compared to the control. The modulus of the bulk resin
was found to be identical for control and hybrid.
117
Hsieh et al. [3.4.31] continued this investigation. They reported the presence of
small agglomerates which seemed not to affect the performance of the hybrid
resin systems. They looked into the toughening mechanisms and identified two
types of plastic deformation. First, localized shear-bands initiated by the stress
concentrations around the periphery of the silica nanoparticles. Second, the
debonding of the silica nanoparticles followed by subsequent plastic void growth
of the epoxy polymer. Of course the well-known mechanisms of the “classic”
rubber toughening apply to the hybrid systems as well.
Epoxy resins modified with CSRs and silica nanoparticles, amine

cured
18 layers of plain woven carbon fibers (203 g/m2) were used to prepare
laminates by the hand lay-up method by Zeng et al. [3.4.42]. Piperidine-cured
DGEBA was used as resin matrix. The elastomer chosen for the hybrid systems
was a CSR Type I. The addition of various amounts of silica nanoparticles
increased the GIc of the laminate by 20-30 %. The modification with the core
shell particles provided the biggest improvement of 150 % at 10 wt% CSR.
However, bulk resin modulus was lowered by approx. 21 %. The hybrid with 6
wt% silica nanoparticles and 6 wt% core shell elastomer showed an
improvement in GIc of approx. 125 % with a modulus only 6 % lower. Thus a
good balance between toughness and stiffness of the fiber-reinforced composite
could be achieved.
Brandt et al. [3.4.43] reported of a resin system similar to RTM 6, modified with
a core shell elastomer Type III and silica nanoparticles regarding aerospace
applications. Details regarding the fibers or fabrics used were not given. The
toughness of the laminate, already being improved by the presence of the core
shell particles to 155 J/m2 was further increased to 250 J/m2 by the addition of
nanosilica. Delamination after impact with 30 J was reduced significantly. Details
regarding addition levels were not given.
Similarly, as for the glass fiber-reinforced composites, the literature search did
not yield any data regarding CFRCs made from epoxy resins modified with CSRs
and industrial 20 nm silica nanoparticles, which were anhydride cured.
It may be due to the competitive nature of the composites industry, that though
such resin systems are used in industrial products, the research results were not
published.
3.4.3.3. Mechanisms for property improvements of hybrid epoxy resin

systems
The reactive liquid rubber domains formed upon cure or the core shell particles
toughen the epoxy by the well-known mechanisms of classical rubber
toughening. The silica nanoparticles increase toughness further by debonding
and subsequent plastic void growth. Localized plastic shear banding contributes
as well. Furthermore they do increase modulus (as a filler) and improve fatigue
performance [3.4.16].
118
The synergy between reactive liquid rubbers and silica nanoparticles as well as
the rather additive performance improvements of hybrids based on core shell
elastomers and nanosilica have been discussed earlier for bulk resins [3.4.17].
In case of the combination reactive liquid rubber/nanosilica more rubber

molecules crosslink randomly due to the presence of the nanosilica; thus
increasing toughness further. This conclusion is based on Tg measurements and
damping property measurements giving information regarding lower crosslink
densities of the cured bulk epoxy resins [3.4.13], [3.4.31]. A lower crosslink
density equals a more ductile polymer which makes rubber toughening more
efficient [3.4.2]. This was achieved in the past by using solid epoxy resins with a
longer molecular chain [3.4.2].
Of course, fiber reinforcement can increase toughness as well with the known
mechanisms of fiber bridging, fiber debonding, or fiber pullout. However, when
laminates made from unmodified resins, from resins modified with either
elastomer or nanosilica and from hybrid resin systems are compared, the fiber
(or fabric) related effects should be the same for all samples.
Sprenger et al. [3.4.41] reported a synergy for a carbon fiber-reinforced

laminate modified with 9 wt% CTBN and 10.5 wt% silica nanoparticles. The GIc
was 1317 J/m2 compared to 1120 J/m2, when the improvements of rubber-only
and nanosilica-only modification are added.
Tsai et al. [3.4.24], [3.4.25] found a synergy for a glass fiber-reinforced laminate
modified with 10 wt% CTBN and 10 wt% silica nanoparticles. The GIc was 1230
J/m2 compared to 1080 J/m2, when the improvements of rubber-only and
nanosilica-only modification are added. A modification with 10 wt% CSR Type II
and 10 wt% nanosilica showed no synergy.
Zeng et al. [3.4.42] found no synergy in carbon fiber-reinforced laminates

modified with CSR Type I and nanosilica.
It cannot be said which hybrid combination is more favourable, reactive liquid

rubbers and silica nanoparticles or core shell rubbers and silica nanoparticles.
Both have their advantages, synergistic GIc for CTBN/nanosilica versus lower
viscosity for CSR/nanosilica, which makes processing easier; especially for
infusion methods. Or toughness is needed at temperatures around – 100 °C,
which can be achieved by using a CSR Type III. Therefore, this decision has to
be made for every intended application of the fiber-reinforced composite and the
anticipated manufacturing method.
3.4.3.4. The transfer of improved bulk resin properties into the fiber-
reinforced composite
The question of the transfer of the property improvements of the bulk hybrid
resin systems is of great interest. As the fiber properties usually dominate;
especially for high fiber volume fractions which for most composites are between
50 and 70 %, the toughness and the fatigue performance are of bigger interest
than resin properties like modulus.
119
Table 3.4.1 gives an overview of the GIc data from the different sources for the
bulk cured epoxy resins and the laminates made from these resins as well as the
improvement in %.
Ref. Resin / Hybrid GIc bulk GIc bulk Impro- Fiber / GIc composite GIc composite Impro-
Hardener modification resin hybrid vement Fabric from bulk from hybrid vement
(J/m2) (J/m2) (%) resin (J/m2) resin (J/m2) (%)
[3.4 DGEBA / 10 wt% 77 906 1008 Glass 330 860 160
.26] MHHPA nanosilica,
9 wt% CTBN
[3.4 DGEBA / 10 wt% 190 930 390 Glass 840 1230 48
.19] IPD nanosilica,
10 wt% CTBN
[3.4 DGEBA / 2 wt% 609 1576 159 Glass 1175 1205 3
.16] Amine nanosilica
4.1 wt% CTBN
[3.4 DGEBA / 4.1 wt% 609 1344 121 Glass 1175 1349 15
.16] Amine nanosilica,
4.1 wt% CTBN
[3.4 DGEBA / 3.5 wt% 609 1671 174 Glass 1175 1525 30
.16] Amine nanosilica,
6.9 wt% CTBN
[3.4 DGEBA / 10 wt% 190 1030 442 Glass 830 1510 82
.19] IPD nanosilica
10 wt% CSR
Type I
[3.4 DGEBA / 10 wt % 77 906 1008 Carbon 439 1316 200
9 wt% CTBN
[3.4 TGMDA / 10 wt% 176 708 302 Carbon 766 1017 33
.33] aromat. nanosilica
Amines 8 wt% CTBN
[3.4 DGEBA / 4.7 wt% 103 917 790 Carbon 439 1078 146
[3.4 9 wt% CTBN
.36]
[3.4 DGEBA / 7.2 wt% 103 973 845 Carbon 439 1106 152
[3.4 9 wt% CTBN
.36]
[3.4 DGEBA / 11 wt% 103 1250 1114 Carbon 439 1320 200
[3.4 9 wt% CTBN
.36]
Table 3.4.1: GIc data for cured bulk epoxy resins and laminates made thereof
In Figure 3.4.8 the GIc values of the cured bulk resin systems versus the
laminates are given. There seems to be a linear relationship at low-fracture
toughness up to approx. 500 J/m2. Above that value, the relationship is no more
linear and exhibits a much smaller gradient.
Hunston et al. [3.4.44] reported a very similar behaviour for the fracture
toughness of CFRCs. They distinguished an area of low GIc below 200 J/m2 where
a toughness increase of the cured bulk resin was transferred into the laminate.
At higher GIc values, the bulk fracture behaviour was not fully transferred into
the fiber-reinforced composite, the data points more scattered (eventually no
linear relationship anymore).
120
Flemming et al. [3.4.2] showed for a system based on T-300 carbon fibers a
linear increase at first, and then a slower increase of the GIc of the laminates
versus the bulk resins – very similar to Figure 3.4.8.
Jordan and Bradley [3.4.45] investigated CFRCs made from rubber toughened
epoxy resins. They observed a sharp drop in fracture toughness improvements
for the laminates above 700 J/m2 – again very similar to Figure 3.4.8. They
concluded that delamination fracture is often dominated by interfacial failure and
that more ductile resins will not necessarily enhance delamination fracture
toughness. This statement can be confirmed for the subject of this review,
toughening by the combination of elastomers and silica nanoparticles as well.
Figure 3.4.8: GIc of cured bulk resin versus GIc laminate
In Figure 3.4.9, the relative improvement of the GIc of the cured bulk resins
versus the relative improvement of the GIc of the laminates when using a hybrid
resin instead of the unmodified control is shown. The nature of the epoxy resin
and curing agent as well as the nature of the reinforcement are not taken into
account. A linear slope can be drawn for the data points. As the line does not go
through the origin, this means a threshold of minimal improvement of fracture
toughness exists. In other words, small improvements of the GIc of the cured
bulk resin like 30 % will not yield an improvement in GIc of the laminates.
The gradient could be defined as transfer factor and would be 0.18 for the data
used in Figure 3.4.9. (accuracy range ± 0.02).
121
This means that an improvement in GIc of 100 % for the cured bulk resin
achieved by a resin modification improves the fracture toughness of the laminate
by only 18 %.
Figure 3.4.9: Relative improvements of GIc of cured bulk resins versus laminates
This could be very useful for fiber-reinforced composites manufacturers. If they

desire to increase the toughness of their existing product without changing the
design, they just have to replace the resin system used by a hybrid resin. Using
the linear relationship as given in Figure 3.4.9, they can select the modification
necessary.
If, example given, an improvement of the laminate GIc by 100 % is desired, then
the composite manufacturer will have to replace his resin system by a hybrid
which has a 555 % higher GIc. By simply measuring the GIc of the cured bulk
resin a suitable hybrid can be chosen without the necessity of making many
tedious trials with laminates from various resins.
Of course, further research will be necessary to gather more data to support and
enhance this theory. Eventually, it makes more sense to differentiate between
glass fiber reinforced and CFRCs. It is very likely that they will have different
transfer factors. Eventually, different epoxy resins used as matrix could be
distinguished as well.
122
Especially for fiber-reinforced composites modified with core shell particles and
nanosilica significantly more research is necessary – so far we have only one
data point for those. Though a similar linear relationship can be expected; it
could be well a different gradient than for hybrid resins with CTBN.
3.4.4. Conclusions
1. Bulk resin modulus of cured hybrid resins was found either unchanged or
up to 6 % lower as for the control at 6 wt% CSR Type I and 6 %
nanosilica. The CSR Type I seems to affect modulus, a resin containing 10
wt% rubber exhibits a 21 % lower modulus of the cured bulk system.
2. The modulus of both glass fiber reinforced and CFRCs made with resins
containing both elastomer particles and silica nanoparticles was reported
to be identical for UD and linen type reinforcements. A quasi-isotropic fiber
layup was reported to have a 10 % lower modulus [3.4.32].
3. The significant improvements in fracture toughness found for bulk epoxy
resins modified with both elastomers and silica nanoparticles were found
for the fiber-reinforced systems on a much smaller scale.
4. A nearly linear relationship exists between the relative GIc increase of the
cured bulk resin systems containing both CTBN and silica nanoparticles
and the laminates made thereof. The fiber reinforcement, be it glass or
carbon fiber, regardless of the three-dimensional orientation of the fibers,
seems to have little or no influence on this relationship.
5. Remarkable improvements of fatigue properties of fiber-reinforced
composites made from hybrid resins were found – up to 10 times of cyclic
loadings before laminate failure.
3.4.5. References
[3.4.1] Ehrenstein G.W. Faserverbundkunststoffe. Hanser Verlag München,

Germany (2006), pp 63 – 68.
[3.4.2] Flemming M., Ziegmann G., Roth S. Faserverbundbauweisen.

Springer Berlin (1995), pp 199 – 225.
[3.4.3] Hare C.J. Amine curing agents for epoxies. J. Prot. Coat. Lin.
(1994), 9, 77 – 103.
[3.4.4] Hare C.J. Epoxy curing agents II. J. Prot. Coat. Lin. (1994),10, 197 -
213.
[3.4.5] Constantino S., Waldvogel U. Composite processing: state of the art

and future trends. In: Pascault J.P., Willimas R.J..J (eds) Epoxy
Polymers. Wiley,Weinheim, Germany (2010), pp 271 – 287.
[3.4.6] Hillermeier R., Hasson T., Friedrich L., Ball C. Advanced

thermosetting resin matrix technology for next generation high
volume manufacture of automotive composite structures. SAE
Technical Paper (2013)-01-1176, 2013, doi:10.4271/2013-01-1176.
123
[3.4.7] Kinloch A.J., Shaw S.J., Tod D.A. Deformation and fracture behavior
of a rubber-toughened epoxy: 1. Microstructure and fracture
studies. Polymer (1983), 24, 1341 – 1354.
[3.4.8] Kinloch A.J. Toughening epoxy adhesives to meet today’s

challenges. MRS Bulletin (2003), 28, 445 – 448.
[3.4.9] Bagheri R., Marouf B.T., Pearson R.J. Rubber-toughened epoxies: a

critical review. J. Macromol. Sci. (2009), 49, 201 – 225.
[3.4.10] Sprenger S., Kinloch A.J., Taylor A.C., Mohammed R.D. Rubber-
toughened FRCs optimised by nano-particles IV. JEC Compos. Mag.
(2008), 38, 34 – 37.
[3.4.11] Giannakopoulos G., Masania K., Taylor A.C. Toughening of epoxy

using core–shell particles J. Mater. Sci. (2011), 46, 327 - 338.
[3.4.12] Tsai J.L., Chang N.R. Investigating damping properties of

nanocomposites and sandwich structures with nanocomposites as
core materials. J. Compos. Mater. (2011), 45, 2157 – 2164.
[3.4.13] Lai M., Friedrich K., Botsis J., Burkhart T. Evaluation of residual
strains in epoxy with different nano/micro-fillers using embedded
fiber Bragg grating sensor. Composites Sci. Technol. (2010), 70,
2168 – 2175.
[3.4.14] EP 0 407 834 (1989).
[3.4.15] Sprenger S., Eger C., Kinloch A.J., Taylor A.C. Nanotoughening of
Epoxies. Proceedings of Stick! conference, Vincentz Verlag,
Nürnberg, Germany, 9 April (2003).
[3.4.16] Sprenger S. Epoxy resin composites with surface-modified silicon

dioxide nanoparticles: a review. J. Applied Polym. Sci. (2013), 130,
1421 – 1428.
[3.4.17] Sprenger S. Epoxy resins modified with elastomers and surface-

modified silica nanoparticles. Polymer (2013), 54, 4790 – 4797.
[3.4.18] EP 1457509 and WO 2004081076 (2003).
[3.4.19] Pictures courtesy of IVW Kaiserslautern and Polymerservice

Merseburg (see [4.4.35] as well).
[3.4.20] Sprenger S. Improving mechanical properties of fiber-reinforced

composites based on epoxy resins containing industrial surface-
modified silica nanoparticles: review and outlook. J. Comp. Mat.
(2015), 49, 53-63.
124
[3.4.21] Sprenger S., Kinloch A.J., Taylor A.C., Mohammed R.D. Rubber-
toughened GFRCs optimised by nanoparticles. JEC Compos.
Mag. (2005), 21, 66 – 69.
[3.4.22] Uddin M.F., Sun C.T. Strength of unidirectional glass/epoxy

composite with silica nanoparticle-enhanced matrix. Compos. Sci.
and Technol. (2008), 68, 1637 – 1643.
[3.4.23] Sun C.T. Development of toughened and multifunctional nano-

composites for ship structures. Final Report for Office of Naval
Research, School of Aeronautics and Astronautics, Purdue
University, West Lafayette, IN 47907, U.S.A. (2012)
[3.4.24] Tsai J.L., Huang B.H., Cheng Y.L. Enhancing fracture toughness of
glass/epoxy composites by using rubber particles together with silica
nanoparticles. J. Compos. Mater. (2009), 43 (25), 3107 – 3123.
[3.4.25] Tsai J.L., Huang B.H., Cheng Y.L. Enhancing fracture toughness of
glass/epoxy composites for wind blades using silica nanoparticles
and rubber particles. Procedia Eng. (2011), 14, 1982 – 1987.
[3.4.26] Schultz R., Tate J.S., Gaikwad S., Trevino E., Jacobs C., Sprenger S.
Low velocity impact on epoxy glass composites modified with rubber
microparticles and silica nanoparticles. In: Proceedings of the SAMPE
TECH conference, Fort Worth, TX, USA SAMPE Covina, CA, U.S.A.
18-19 October (2011).
[3.4.27] Tate J., Trevino E., Gaikward S., Sprenger S., Rosas I., Andrews
M.J. Low velocity impact on epoxy glass composites modified with
rubber microparticles and silica nanoparticles. J. Nanosci., Nanoeng.
Appl. (2013), 3 (1), ISSN: 2231-1777.
[3.4.28] Kinloch A.J., Masania K., Taylor A.C., Agarwal R., Sprenger S., Egan
D. The fracture of glass-fibre-reinforced epoxy composites using
nanoparticle-modified matrices. J. Mater. Sci. Lett. (2008), 43, 1151
– 1154.
[3.4.29] Kinloch A.J., Masania K., Taylor A.C., Sprenger S. The fracture of
nanosilica and rubber toughened epoxy fibre composites. In:
Proceedings of the Composites & Polycon, American Composites
Manufacturers Association, Tampa, Florida, USA, 15-17 January
(2009).
[3.4.30] Manjunatha C.M., Taylor A.C., Kinloch A.J., Sprenger S. The effect of
rubber micro-particles and silica nano-particles on the tensile fatigue
behaviour of a glass-fibre epoxy composite. J. Mater. Sci. Lett.
(2009), 44, 342 – 345.
[3.4.31] Hsieh T.H., Kinloch A.J., Masania K., Sohn Lee J., Taylor A.C.,
Sprenger S. The toughness of epoxy polymers and fibre composites
modified with rubber microparticles and silica nanoparticles. J.
Mater. Sci. (2010), 45, 1193 – 1210.
125
[3.4.32] Manjunatha C.M., Taylor A.C., Kinloch A.J., Sprenger S. The tensile
fatigue behavior of a glass-fiber reinforced plastic composite using a
hybrid-toughened epoxy matrix. J. Compos. Mater. (2010), 44 (17),
2095 – 2109.
[3.4.33] Manjunatha C.M., Jagannathan N., Padmalatha K., Kinloch A.J.,

Taylor A.C. Improved variable-amplitude fatigue behavior of a glass-
fiber-reinforced hybrid-toughened epoxy composite. J. Reinf. Plast.
Compos. (2011), 30 (21), 1783 – 1793.
[3.4.34] Manjunatha C.M., Bojja R., Jagannathan N., Kinloch A.J., Taylor A.C.
Enhanced fatigue behavior of a glass fiber reinforced hybrid particles
modified epoxy nanocomposite under WISPERX spectrum load
sequence. Int. J. Fatigue (2013), 54, 25 – 31.
[3.4.35] Entwicklung von hochverschleißfesten Kunststoffen für tribologische

Anwendungen und mechanisch beanspruchte Bauteile. Final Report
Stiftung Industrieforschung Projekt S 657 2005, Institut für
Verbundwerkstoffe, Kaiserslautern, Germany, (2005).
[3.4.36] Caccavale V., Wichmann M., Quaresimin M., Schulte K. Nanoparticle

/rubber modified epoxy matrix systems: mechanical performance in
CFRPs. In: Proceedings of AIAS XXXVI Convegno Nazionale, Ischia,
Naples, Italy, 4-8 September (2007).
[3.4.37] Caccavale V. Nanoparticle/rubber modified epoxy matrix systems:

mechanical performance in CFRPs. Master of Science Thesis,
University of Padova, Italy (2007).
[3.4.38] Sprenger S., Kinloch A.J., Taylor A.C., Mohammed R.D. (2007)
Rubber-toughened CFRCs optimized by nanoparticles – Part III. JEC
Compos. Mag. 2007, 30, 54 – 57.
[3.4.39] Kowalik T., Hesebeck O., Flothmeier K., Koesling S. Zähelastisch

modifizierte CFK Harze zur Herstellung von Strukturbauteilen und im
Reparatureinsatz. MicroCar (2013), Leipzig, Germany, 25-26
February 2013.
[3.4.40] Picture courtesy of Fraunhofer IFAM, Bremen, Germany (see

[3.4.39] as well).
[3.4.41] Sprenger S., Kinloch A.J., Taylor A.C., Mohammed R.D., Eger C.
Rubber Rubber-toughened FRCs optimised by nanoparticles. JEC
Compos. Mag. (2005), 19, 73 – 76.
[3.4.42] Zeng Y., Liu H.Y., Mai Y.W., Du, X.S. Improving interlaminar
fracture toughness of carbon fibre/epoxy laminates by incorporation
of nanoparticles. Compos. B (2012), 43, 90 – 94.
[3.4.43] Brandt J., Drechsler K., Schmidtke K. Composites für den Flugzeug-
bau. Kunststoffe (2004), 10, 290 – 294.
126
[3.4.44] Hunston D.L., Moulton R.J., Johnston N.J., Bascom W.D. Tough
composites, STP 937, ASTM, Philadelphia, U.S.A. (1987), pp 74 –
94.
[3.4.45] Jordan W.M., Bradley W.L. The relationship between resin

mechanical properties and Mode I delamination fracture toughness.
J. Mat. Sci. Lett. (1988), 7, 1362 – 1364.
127
3.5. Carbon fiber-reinforced composites with epoxy resin
modified with reactive liquid rubber and SiO2
nanoparticles
3.5.1. Abstract
Carbon fiber-reinforced composites are gaining importance and are about to play
a very prominent role in automotive applications. In structural applications
mainly epoxy resins are used as matrix materials. The properties of epoxy resins
and laminates made thereof can be improved by tougheners like reactive liquid
rubbers. Further improvements can be achieved by adding surface-modified silica
nanoparticles, with 20 nm in size and a very narrow particle size distribution.
In this study carbon fiber-reinforced laminates made from epoxy resins modified
with reactive liquid rubber and silica nanoparticles have been prepared and
investigated. A very fast amine cure of 15 minutes has been chosen to match
industrial needs. Mechanical properties for bulk resins and laminates are
compared and the mechanisms responsible for the property improvements are
discussed. Structure property relationships between the neat resin fracture
toughness and the interlaminar GIc and GIIc of the reactive liquid rubber and silica
nanoparticles modified resins were established. Tough laminates could be
prepared.
However, CAI performance of the hybrid laminates was slightly inferior as

compared to the rubber-toughened laminates, most probably due to
agglomerates of nanoparticles found in the cured resin systems. The reason for
the nanoparticle aggregation was detected and their influence on laminate
performance discussed.
3.5.2. Introduction
Environmental requirements like reduced fuel consumption, lower carbon dioxide

emission and an improved carbon footprint are changing the automotive
industry. Especially in Europe, where carbon dioxide emissions are subject to
state regulations [3.5.1].
Lightweight construction becomes the key technology to solve these issues which
currently drives a trend towards the increased use of fiber-reinforced composites
in automotive construction [3.5.2].
Automotive could be the third large industrialization of composites technologies,

after aerospace and wind energy. In other areas of application like railway or
shipbuilding more and more fiber-reinforced materials are used as well. Of
course automotive manufacturing has its own rules, one is a very short part
production cycle, like the stamping of metal parts. Thus composites
manufacturing technologies and the materials used have to be adapted
accordingly [3.5.3].
Glass or carbon fibers, woven fabrics and unidirectional or multiaxial nonwovens

are used as reinforcements; natural fibers are under investigation. Thermoplastic
or thermosetting materials are used as matrices.
128
Besides unsaturated polyester resins mainly epoxy resins are used as
thermosetting matrices for automotive components. The diglycidylether of
bisphenol A (DGEBA) is the most commonly used epoxy resin in automotive
industry. The diglycidylether of bisphenol F (DGEBF) and the higher molecular
weight variations, epoxidized novolac resins, are only used for special
applications. So-called reactive diluents, short chain aliphatic epoxy resins like
hexanediole diglycidylether (HDDGE), are very often blended with the DGEBA in
order to reduce the viscosity. A low viscosity of the resin is very important as the
preferred manufacturing methods for automotive parts currently are injection or
infusion processes – which enable short cycle times [3.5.3].
The combination of epoxy resin and hardener defines the matrix material and the
various possible combinations enable to formulate according to all possible
property requirements [3.5.4]. Hare described the variety of epoxy curing agents
a few years ago [3.5.5], [3.5.6]. Most popular for automotive part manufacturing
are aliphatic or cycloaliphatic amines like isophorone diamine (IPD) due to their
low viscosity and fast cure at elevated temperatures. Various accelerators and
additives are typically part of the hardener formulation.
However, epoxy resins are inherently brittle. Therefore many epoxy resin
formulations are toughened. In the seventies of last century carboxy-terminated
nitrile butadienes (CTBNs) were introduced as tougheners. They are immiscible
with epoxies but become miscible after a pre-reaction with an excess of epoxy
resin. These soluble adducts phase separate again upon cure and form rubber
domains in the epoxy resin. The mechanisms of rubber toughening are well
known [3.5.7], [3.5.8]. Rubber toughening has several drawbacks like lowering
the modulus and reducing the glass transition temperature, due to some rubber
molecules not participating in the phase separation, but crosslinking randomly
into the matrix. Another issue is the relatively high viscosity of epoxy resins
containing reactive liquid rubbers, which can be critical for infusion processes.
Nevertheless they are used in composites manufacturing.
In 2002/2003 surface modified silica nanoparticles became available in industrial

quantities. They are manufactured in situ directly in the epoxy resin by a
modified sol-gel process and have an average particle diameter of 20 nm as well
as a very narrow particle size distribution [3.5.9].
These spherical particles are transparent, monodisperse and just slightly increase
the resin viscosity up to quite high addition levels. Resins containing such
nanoparticles exhibit no thixotropic properties but behave like Newtonian liquids.
Figure 3.5.1 shows the uniform particle distribution in a cured bulk epoxy resin
by transmission electron microscopy (TEM).
129
Figure 3.5.1: Transmission electron microscopy (TEM) image of amine-cured
DGEBA with 12.2 vol% silica nanoparticles at different magnifications [3.5.10]
Properties like strength, modulus, toughness and especially fatigue performance

of cured bulk epoxy resins can be improved by a modification with silica
nanoparticles. This has been subject of intensive research in the last ten years
[3.5.11]. The mechanisms responsible for the property improvements are rather
well understood today [3.5.12].
Due to their size and the only slight increase in viscosity of the resin they can be
used in injection or infusion methods for composite manufacturing where they
can penetrate even close meshed fabrics easily. It is well known that properties
like modulus, fracture toughness, compressive strength and especially fatigue
performance upon cyclic loading can be improved by modifying the epoxy resin
matrix with silica nanoparticles. Some properties, e.g. high speed impact
performance, were not improved. An extensive review regarding the potential of
silica nanoparticles to toughen fiber-reinforced composites was published
recently [3.5.13].
The combination of reactive liquid rubbers or core shell elastomers and the use
of silica nanoparticles as additional modifier in epoxy resin systems yields
synergistic property improvements in cured bulk epoxy resins [3.5.14]. It is
possible to formulate tough and stiff materials. These resin systems and their
synergies in various applications, including fiber-reinforced composites were
patented consequently at a very early stage [3.5.15].
Today such hybrid systems are used in many industrial epoxy formulations.
Figure 3.5.2 shows the transmission electron microscopy (TEM) picture of a
cured bulk hybrid epoxy resin. Obviously both the rubber particles which have
formed upon cure and the silica nanoparticles are well dispersed.
130
Figure 3.5.2: Transmission electron microscopy (TEM) image of amine-cured
hybrid epoxy resin [3.5.16]
The toughening mechanisms of such hybrid systems are well understood [3.5.14]
and described recently in a review article [3.5.17]. In summary, significant
property improvements reported for hybrid neat resins are transferred only
partially into improvements in the corresponding composites.
In this experimental study epoxy resin formulations similar to industrial resin

systems and modifications by reactive liquid rubber and silica nanoparticles were
studied on carbon fiber-reinforced composite made by resin transfer moulding
(RTM). Besides neat resin data we intended to obtain an extended property
profile based on interlaminar fracture toughness (GIc and GIIc) data, compression
after impact (CAI) data and interlaminar shear strength (ILSS) data as well. Neat
resin and corresponding laminates were investigated to achieve relationships
between neat resin and composite properties and to probe the conversion factor
as proposed recently by Sprenger [3.5.17].
The results published so far have been obtained with resin systems cured for
very long times. Amine cured systems with cure cycles of 4-6 hours; sometimes
followed by a postcure.
Anhydride cured systems cured for 3-12 hours at temperatures above 100 °C.
Thus a hardener system with a relatively fast cure of 15 minutes, more realistic
for industrial applications, was selected.
131
3.5.3. Experimental
3.5.3.1. Materials
A diglyciylether of bisphenol A (DGEBA) (Araldite® GY 250, Huntsman, Switzer-

land) was used. To lower the viscosity the triglycidylether of trimethylolpropane
(TGETMP) (Grilonit® V51-3, EMS, Switzerland) was added as reactive diluent. A
reactive liquid rubber was added as a 40 wt% concentrated adduct with DGEBA
(Albipox® 1000, Evonik Hanse GmbH, Germany). The silica nanoparticles were
introduced by using a 40 % concentrate in DGEBA (Nanopox® F 400, Evonik
Hanse GmbH, Germany).
As hardener a blend of 90 wt% isophorone diamine (Vestamin IPD) and 10 wt%

trimethyl hexamethylene diamine (Vestamin TMD) was used (Evonik Industries,
Germany). The active hydrogen of the hardener blend was calculated to be 42.3.
The epoxy equivalent weight of the different epoxy resin systems was calculated
and is given in Table 3.5.1 as well, which gives an overview of the different resin
system compositions used to prepare the test specimen.
Table 3.5.1: Resin system formulations
Nomenclature: system 10R contains 10 wt% of reactive liquid rubber, and

system 10R10N contains 10 wt% reactive liquid rubber as well as 10 wt% silica
nanoparticles.
132
Carbon fiber Tenax-E HTS40 F13 12 K was arranged into a -45°/45° nonwoven
with 254 g/m2. This material was kindly provided by Saertex GmbH, Saerbeck,
Germany.
3.5.3.2. Bulk sample preparation
The resin systems were blended according to the compositions given in Table
3.5.1 and cured for 15 minutes at 120 °C followed by 2 hours postcuring at 160
°C. The aluminum mold was coated with a release agent (Loctite Frekote 770-
NC, Henkel, Germany). Plates with a defined thickness of 4 mm respectively 2
mm were obtained.
3.5.3.3. Laminate preparation
Carbon fiber-reinforced laminates were produced using a resin transfer molding

(RTM) machine (Isojet 200915, Isojet, France) designed for 1 and 2- part
vacuum-assisted RTM. Stainless steel frames were put in the aluminum mold,
cavity size of approx. 40 cm x 40 cm, to obtain laminates with a defined
thickness. Before placing the dry carbon fibers, the aluminum mold and the steel
frame were preconditioned with a release agent (Loctite Frekote 770-NC, Henkel,
Germany) and thoroughly dried. For the determination of the interlaminar
fracture toughness Mode I and II (GIc, GIIC), 3 mm thick laminates with
[(90/0)]5s layup and 48 vol% fibers were prepared. A teflon foil was inserted in
the mirror plane to create the preinduced crack.
Laminates for compression after impact (CAI) testing were produced with a
[(90/0)/(-45/+45)]4s layup and 54 vol% fibers at a laminate thickness of 4.2
mm. The closed aluminum mold was put into a hot press (LZT 110 L,
Maschinenfabrik Langzauner GmbH, Austria) and evacuated (10 mbar). An
injection temperature of 70 °C was chosen for the resin systems in order to have
a reasonably fast filling of the mould. The injection pressure was set to 8 bars.
After 15 minutes curing at 120 °C the laminates were demoulded and postcured
in an oven for 2 hours at 160 °C.
3.5.3.4. Bulk sample testing
The viscosities of the uncured resin systems were determined using a rotational
rheometer (Physica MCR 301, Anton Paar GmbH, Austria) equipped with a
disposable aluminium plate tool and an aluminium bowl. Testing parameters
were rotational mode with a shear rate of 10 s-1 at 70 °C and a gap size of 1
mm.
All test specimens were prepared using a circular saw (Diadisc 6200, MUTRONIC
Präzisionsgerätebau GmbH & Co. KG, Germany) equipped with a diamond saw
blade.
Glass transition temperature of the cured bulk resin was determined by dynamic
mechanical thermal analysis (DMTA) according to DIN EN ISO 6721-7 using an
Advanced Rheometric Expansion System (ARES I, Rheometric Scientific,
Germany) in torsional mode. A sinusoidal deformation of 0.1 % with a frequency
of 1 Hz was applied from 25 °C to 200 °C with a heating rate of 3 K/min. The
glass transition temperature (Tg) was determined as maximum of the loss factor,
tan δ.
133
The critical stress intensity factor (KIc) and the critical energy release rate (GIc)
were determined according to ISO 13586 using compact tension (CT) test
specimens. The test specimen width was w = 33 mm, the thickness d = 4 mm.
For each sample, a sharp crack was generated in the V-notch by tapping a new
razor blade. The tests were performed using a universal tensile testing machine
(Zwick BZ 2.5/TN1S, Zwick Roell, Germany) with a test speed of 10 mm/min.
The crack opening displacement was measured using a clip extensometer
(632.29F-30, MTS, Germany). The modulus of the bulk resin samples was
calculated according to the method of Saxena and Hudak [3.5.18].
Transmission electron microscopy (TEM) characterizations were carried out using

a Zeiss LEO 902 A EFTEM (Carl Zeiss AG, Germany) applying an acceleration
voltage of 200 kV. Thin sections of 50 nm were cut on a Leica Ultracut
microtome (Leica Biosystems GmbH, Germany) equipped with a glass knife. The
Cryo-TEM was carried out by freezing the resin/hardener blend after mixing in
liquid nitrogen followed by specimen cutting over a bath of liquid nitrogen.
3.5.3.5. Laminate testing
The glass transition temperature (Tg) of the laminates was measured as

described in the chapter before.
Interlaminar fracture toughness Mode I (GIc) was determined in accordance to

AITM1-0005, whereas aluminum blocks were used instead of hinges – as
described in ASTM D 5528-94a, applying a test speed of 10 mm/min.
Interlaminar fracture toughness Mode II (GIIc) was determined using the tested
GIc-specimen as described in DIN EN 6034 applying a test speed of 1 mm/min.
Both tests were performed on a universal tensile testing machine (Zwick BZ
2.5/TN1S, Zwick Roell, Germany).
The interlaminar shear strength (ILSS) was measured according to DIN EN 2563
using test specimen with the dimension 25x10x3 mm prepared from the
laminates made for the GIc testing. The tests were performed on a universal
tensile testing machine (Zwick 1475, Zwick Roell, Germany) with a test speed of
1 mm/min.
Compression after impact (CAI) data was gathered in accordance to DIN 65561
using 3 samples with dimensions of 150x100 mm. The specimens were impacted
with an energy of 30 J using a drop-weight impactor. Subsequently the
delaminated area was measured using an ultrasonic inspection method (AirTech
HFUS 2400, Ingenieurbüro Dr. Hilger Ultraschall-Prüftechnik, Germany). The
compression after impact strength was measured on a universal tensile testing
machine (Zwick 1485, Zwick Roell, Germany) with a test speed of 0.5 mm/min.
Scanning electron microscopy (SEM) analysis was carried out using a Zeiss 1530
(Carl Zeiss AG, Germany) posessing a field emission cathode for high-resolution
micrographs (acceleration voltage 1.5 kV).
134
3.5.4. Results and discussion
All other studies cited used the same commercially available epoxy resins with 40
wt% silica nanoparticles with a diameter of 20 nm. The reactive liquid rubber
used was always introduced into the formulations as epoxy resin adduct;
typically 40 wt% CTBN rubber with an acrylonitrile content of 18 wt% had been
prereacted with 60 wt% DGEBA.
3.5.4.1. Bulk resin properties
First the viscosities of the formulations given in Table 3.5.1 were measured. At
70 °C they ranged from 200 mPas for the control to 8.000 mPas for 10R10N. It
has to be noted that processing resin systems with viscosities above 1.000 mPas
in an RTM setup is anything but trivial.
In Figures 3.5.3 and 3.5.4 the Tg´s and the moduli for the cured bulk resins are
given. As expected, the modification with 10 wt% reactive liquid rubber (system
10R) lowered both modulus and Tg by approx. 10 %. When epoxy resins
containing reactive liquid rubbers are cured, the rubber phase separates and
forms a secondary phase, the rubber domains.
Figure 3.5.3: Glass transition temperature (Tg) for cured bulk epoxy resin
systems
It is well known that some of the long rubber molecules do not participate in the
phase separation upon cure but crosslink randomly into the matrix. This lowers
the network density, which affects of course modulus and Tg.
135
Figure 3.5.4: E-modulus of the cured bulk epoxy resin systems
The addition of silica nanoparticles as a filler reinforces the matrix and thus
increases the modulus again. The epoxy resin system with 10 wt% rubber and
10 wt% nanosilica has the same modulus as the unmodified control. An increase
in modulus of 0.35 GPa for 10 wt% silica nanoparticle addition is similar to what
was found for other amine-cured DGEBA systems [3.5.11].
The glass transition temperature, however, is not increased by the addition of

nanosilica – as the silica nanoparticles have no influence on the network density.
Instead, the Tg is lowered slightly at addition levels above 7.5 wt% nanosilica.
Most probably higher addition levels of nanosilica hinder the formation of rubber
domains upon cure. Thus the crosslink density is further reduced by the higher
number of randomly crosslinked rubber molecules [3.5.14].
Xu et al. studied piperidine-cured DGEBA [3.5.19]. They reported a huge increase

in both KIc and GIc for the addition of CTBN which reaches a plateau at 10 wt%
addition level. Additional modification of a 5 wt% rubber containing system with
silica nanoparticles did only slightly increase KIc. GIc was not further improved.
Though they used an 18 hours cure cycle, our system cured for only 15 minutes
performs quite similar.
136
Figure 3.5.5: KIc of the cured bulk epoxy resin systems
Figure 3.5.6: GIc of the cured bulk epoxy resin systems
137
As can be seen in Figures 3.5.5 and 3.5.6, the increase in toughness by the
rubber addition is significant; the further addition of silica nanoparticles does not
contribute to toughening markedly. At 10 wt% nanosilica however, the
toughness is further increased – beyond the increase deriving from the rubber
modification. An improvement in GIc of approx. 250 % is not untypical. In a
recently published review improvements in the range of 111-302 % were
reported for amine-cured, CTBN-modified DGEBA resins [3.5.14].
Figure 3.5.7 shows that the hybrid system with 10R10N is a tough and stiff
matrix resin, clearly outperforming the unmodified control. The Tg with 140 °C is
still acceptable for structural automotive applications. In Table 3.5.2 an overview
of all bulk resin properties is given.
Figure 3.5.7: Modulus versus GIc for cured bulk epoxy resin systems
138
Table 3.5.2: Overview of cured bulk epoxy resin properties
3.5.4.2. Laminate properties
It is well known that very often property improvements of matrix resins do not
translate into improvements of fiber-reinforced composites made thereof.
Especially carbon fibers can be a very dominating reinforcement. The recently
published review from Ye et al. gives an excellent overview regarding the
improvements of interlaminar fracture toughness and compression after impact
strength by using nanoparticles in epoxy resin matrices for fiber-reinforced
composites [3.5.20]. They distinguished two different relationships between GIc
of the matrix and GIc of the laminate - depending if the matrix is rather brittle or
tough. Furthermore they concluded that a positive correlation between the GIc of
the matrix and the GIIc of the composite or the CAI strength is not clear.
Therefore they voiced the opinion that the research “… is at an immature
stage…”. Adding an additional component, elastomeric tougheners like reactive
liquid rubbers or core-shell elastomers makes the understanding of such systems
even more complex.
In a review regarding composites made from hybrid epoxy resins, Sprenger

recently suggested the existence of a conversion factor for hybrid epoxy resins
[3.5.17]. He found for both glass and carbon fiber reinforced materials that the
percentual improvement of the GIc of the cured bulk epoxy resin converted into
0.18 times the percentual improvement of the GIc of a laminate made from this
matrix resin.
The first laminate properties investigated in this study were GIc and GIIc. The
results are given in Figures 3.5.8 and 3.5.9.
139
Figure 3.5.8: GIc of carbon fiber-reinforced laminates based on various epoxy
resins systems
As can be seen in Figure 3.5.8, despite the higher modulus of the hybrid matrix
systems, the laminate toughness was improved by approx. 50 % for the hybrids
and even more by the rubber-only modified system.
For carbon fiber-reinforced materials much less literature data are available than
for glass fiber-reinforced composites made from hybrid epoxy resins containing
both the reactive liquid rubber and the silica nanoparticles [3.5.17]. Nevertheless
in most studies with CFRC published until now, the fiber-reinforced composites
based on a hybrid system were found to have the highest GIc. Kinloch et al.
reported about a CFRC made from anhydride cured DGEBA [3.5.21], [3.5.22].
The control with 439 J/m2 was increased to 1050 J/m2 by a modification with 9
wt% reactive liquid rubber. Further addition of 10.5 wt% silica nanoparticles
increased the GIc of the laminate further to 1320 J/m2. Ye et al. investigated
piperidine-cured DGEBA [3.5.23]. The GIc could be increased from 741 J/m2 to
1256 J/m2 by a modification with 10 wt% rubber. The addition of 10 wt%
nanosilica further increased the GIc to 1323 J/m2.
However, other findings have been published as well: Sprenger et al. reported
about a CFRC based on tetrafunctional epoxy resin cured with aromatic amines;
a system similar to the aerospace industry benchmark Hexflow® RTM6 [3.5.26].
Though the cured hybrid bulk system had the highest GIc, the hybrid laminate
was lower in GIc than the system containing only reactive liquid rubber.
140
Nevertheless the modulus of the hybrid was higher than for the rubber-only
modification.
Figure 3.5.9: Relative improvement of bulk GIc versus relative improvement of

laminate GIc
Figure 3.5.9 compares the improvement of the bulk GIc with the improvement of
the laminate GIc (in %). The average of laminate improvement versus bulk
improvement was 0.2 ± 0.07 - which is very close to the conversion factor of
0.18 ±0.02 recently proposed by Sprenger [3.5.17].
141
Figure 3.5.10: GIIc of carbon fiber-reinforced laminates based on various epoxy
resins systems
Not much GIIc data has been published for hybrid systems so far. The data shown
in Figure 3.5.10 indicates that once a crack has formed, the delamination under
mode II loading continues at a sligthly lower force. How this is related to the
hybrid modification is unclear. Further clarification of the fracture mechanisms is
expected from the scanning electron micrographs of the fracture surfaces of GIc
and GIIc test specimen.
The ILSS was tested and the results are given in Figure 3.5.11. As can be seen,
it is slightly lower for all modified systems, but still on a very high overall level.
Having a look at the error bars, there cannot be seen real differences between
the rubber-only modification and the hybrids with different nanosilica
concentrations. The very small differences in ILSS for all samples tested indicate
an excellent laminate quality.
Quaresimin et al. studied CFRC made from amine-cured DGEBA [3.5.25]. The
modification with 7.3 wt% reactive liquid rubber and 3.7 wt% silica nanoparticles
increased the bulk resin KIc from 0.91 MPam1/2 to 2.11 MPam1/2 or by 130 %. The
mode I interlaminar fracture toughness was increased by 74 %. However the
interlaminar shear strength was reduced from 55 MPa to 50 MPa - very similar to
our results.
142
Figure 3.5.11: Interlaminar shear strength of carbon fiber-reinforced laminates
based on various epoxy resins systems
Falling dart test and the CAI strength were investigated next. The results are
given in Figures 3.5.12 and 3.5.13. As can be seen, for higher nanosilica
contents the hybrid systems show more delamination and the compressive
strength is reduced. This is unexpected, as for a laminate made from a tougher
epoxy resin matrix one would expect an increase in CAI.
Kuehn et al. investigated carbon fiber reinforced laminates made from anhydride
cured DGEBA [3.5.26]. A modification with 10 wt% reactive liquid rubber reduced
the damage area (at 30 J) from 38 to 24 cm2 – the compressive strength after
impact was increased from 165 MPa to 205 MPa. A modification with 25 wt%
nanosilica did not reduce the delaminated area at all but increased the CAI to
175 MPa. Obviously the higher modulus of the systems modified with nanosilica
is improving the residual compressive strength. Unfortunately they did not
investigate hybrid systems.
Tate et al. reported results of a glass fiber-reinforced laminate based on DGEBA

and cured with a commercial amine blend [3.5.27]. They compared the
unmodified control with a hybrid containing 6.5 wt% reactive liquid rubber and
8.1 wt% of nanosilica. For all different impact energies tested the delamination
was always larger for the laminate made from the hybrid epoxy resin system and
the post impact compressive strength was lower. They concluded that the hybrid
is capable to dissipate the impact energy over a larger area, which would explain
the larger delaminated area and the lower CAI values.
143
They noticed that small agglomerates had formed which seemed not to affect the
performance of the nanosilica modification.
Impact 30 J
Figure 3.5.12: Delaminated area after impact of carbon fiber-reinforced

laminates based on various epoxy resins systems (after 30 J impact)
Impact 30 J
Figure 3.5.13: CAI strength of carbon fiber-reinforced laminates based on

various epoxy resins systems (after 30 J impact)
144
The comparision of the delaminated area between the rubber-modified system
10R and the hybrid 10R5N which contains additionally 5 % of nanosilica is very
interesting: they exhibit the same delaminated area, but the hybrid shows a
lower compression after impact strength.
GIc and GIIc of this hybrid are lower as well. This indicates a more brittle matrix
behaviour, which is confirmed by the bulk resin results for 10R and 10R5N.
Nevertheless, if the damage tolerance, given by the compression after impact

strength and the delaminated area after impact, is higher for 10R5N and very
similar for 10R7.5N to the unmodified control, as can be seen in Figure 3.5.14.
Impact 30 J
Figure 3.5.14: Damage tolerance as combination of CAI strength and

delaminated area (after 30 J impact)
Table 3.5.3 gives an overview of all laminate properties:
145
Table 3.5.3: Overview of carbon fiber-reinforced laminate properties
In a future paper the fatigue properties will be studied, as they have been
reported best for hybrid systems [3.5.17].
3.5.4.3. Microscopical investigations
Laminate fracture surfaces
In order to understand the mechanisms of failure the fracture surfaces were

investigated. Figure 3.5.15 shows the fracture surfaces from the GIc test
specimen and Figure 3.5.16 gives the fracture surfaces from GIIc test specimen.
Figure 3.5.15 reveals some information regarding the fracture mechanism. The
fracture surface of the unmodified control shows less residual epoxy resin than
the rubber-modified epoxy and the hybrids, which indicates a more interfacial
failure between fiber layer and matrix. A similar effect was found by Ye et al.
[3.5.23].
The difference becomes more evident on the fracture surfaces from GIIc test
specimen at a higher magnification in Figure 3.5.16. The matrix resin between
the carbon fibers of the control shows the typical angular structures of brittle
materials; designated "hackles".
146
Figure 3.5.15: SEM images of fracture surfaces from GIc test specimens. The
white arrows indicate the direction of crack propagation.
Figure 3.5.16: SEM images of fracture surfaces from GIIc test specimens. The
white arrows indicate the direction of crack propagation.
147
In the picture of the matrix containing 10 wt% reactive liquid rubber the rubber
domains, which have formed upon cure, are clearly visible. The matrix shows no
angular structures any more - it looks rather peeled or teared off. This indicates
a more ductile behavior. The hybrid matrix with 10 wt% rubber and 5 wt%
nanosilica looks similar; however at 10 wt% nanosilica the matrix does not look
ductile anymore. Apparently the nanosilica addition reduces the ductility gained
by the rubber modification. This would explain the increase in delaminated area
and subsequent decrease in compressive strength after impact with increasing
silica nanoparticle content.
Figure 3.5.17: SEM image of hybrid 10R5N (fracture surface of GIIc test
specimen).
The white arrow indicates the direction of crack propagation. A major aspect
comes into focus when looking closer. In Figure 3.5.17 the matrix resin of the
hybrid 10R5N with 10 wt% rubber and 5 wt% nanosilica can be seen between
the surfaces of the carbon fibers. The rubber domains with their typical size of
0.5 - 0.7 micron are clearly visible, and, unexpected, agglomerates of silica
nanoparticles are found. This triggered an additional microscopical investigation
of the cured bulk resins in order to understand these systems.
Bulk resin cross sections
Despite the rather fast cure the rubber domains have formed perfectly and are
evenly distributed, as can be seen in Figure 3.5.18 for the system 10R. All
hybrids show agglomerates of silica nanoparticles. They increase in size with
increasing addition level in nanosilica. At 5 wt% the agglomerates seem to be
approximately 100 nm large; at 7.5 wt% 200 - 300 nm and at 10 wt% they
seem to exceed 300 nm in size.
148
Eventually the agglomerates have formed due to an incompatibility of the
surface-modified silica nanoparticles with the hardener system. Another
possibility would be the very fast cure. This will be discussed in the next chapter.
Figure 3.5.18: TEM images of cross sections of cured bulk epoxy resins systems
3.5.4.4. Formation of nanosilica agglomerates
In most studies regarding hybrid epoxy systems, no agglomerates were detected

[3.5.14]. Very few researchers reported the presence of agglomerates in their
cured resin systems. Manjunatha et al. investigated the fatigue performance of
glass fiber reinforced laminates based on anhydride cured DGEBA [3.5.28]. They
observed agglomerates of 400-800 nm in the hybrid with 9 wt% reactive liquid
rubber and 10 wt% nanosilica. The significant improvement in fatigue
performance seemed not to be affected by their presence.
Hsieh et al. prepared laminates with carbon fibers and glass fibers using different
hybrid systems [3.5.29]. Scanning electron microscopy of the anhydride-cured
bulk DGEBA revealed the formation of agglomerates as well. The nanoparticles
appeared to cluster in necklace-like structures and increased in size with
increasing nanosilica content, however the GIc of the bulk resin and the laminates
was increased significantly – and was always higher for the hybrid than for the
rubber-only modified epoxy resin.
149
For the 10R10N system they reported agglomerate sizes of up to 2000 nm in
length and 100 nm in width. Therefore it was concluded that the formation of
agglomerates does not affect performance.
We started to examine the different components of our epoxy resin system and
found no incompatibility between silica nanoparticles and the curing system
consisting of IPD and TMD. If our system is cured without the presence of
reactive liquid rubber but with the same hardener/accelerator and the same fast
cure, no agglomeration occurs - as can be seen in Figure 3.5.19.
Figure 3.5.19: TEM image of cross section of cured epoxy resin with 5 wt%
nanosilica (0R5N) at different magnifications
The SiO2-nanoparticles are well dispersed in such a cured epoxy resin.
Next we investigated the blending process with reactive liquid rubber adducts.
When epoxy resins containing reactive liquid rubber adducts and epoxy resins
containing silica nanoparticles are blended, these blends are transparent and no
agglomeration takes place, even if such a resin blend is stored over weeks and
months. Now being mixed with the hardener and accelerator at room
temperature, the resin system remains transparent - still no agglomeration
occurs. We froze the system 10R5N immediately after blending with the hardener
using liquid nitrogen and examined using TEM under cryogenic conditions (Figure
3.5.20).
Despite the presence of hardener and accelerator and their intensive mixing with
the resin part no agglomeration has occurred – the silica nanoparticles are well
dispersed. As the system is uncured, the rubber phase separation has not
occurred and no domains have formed.
150
Figure 3.5.20: Cryo-TEM image of uncured epoxy resin system with 10 wt %
CTBN and 5 wt% nanosilica (10R5N)
Eventually the relatively fast cure cycle of 15 minutes, inducing a rather quick
rubber phase separation of the CTBN, is the cause and forces the nanoparticles
to agglomerate. This would explain why other researchers using unaccelerated
isophoronediamine curing did not report the formation of agglomerates [3.5.14].
In order to confirm this assumption the hybrid system 10R5N was cured at room
temperature for 48 hours. The microscopical investigation was evident, as can be
seen in in Figure 3.5.21.
Though the room temperature cure was almost completed after 2 hours we
prepared the microscopy test specimen after 48 hours of RT cure. As can be
seen, even during this somewhat slower cure agglomerates were formed. They
seem to be slightly smaller, but this is only a subjective impression.
151
Figure 3.5.21: TEM image of cross section of RT-cured epoxy resin with 10 wt%
CTBN and 5 wt% nanosilica (10R5N) cured 48 hours at RT
Curing 10R5N without the accelerator TMD did not impede nanosilica
agglomeration either.
It can be concluded that the agglomeration of the silica nanoparticles is linked to

the relatively fast rubber phase separation of the reactive liquid rubber due to
the fast increase in crosslink density defined by the fast curing system of
IPD/TMD. Eventually the rubber still present in the epoxy network and randomly
crosslinked into the matrix causes the nanosilica to agglomerate upon cure.
If the agglomerates affect the composite performance when increasing in size

could not be clarified finally. It seems not to be the case for properties like GIc
(bulk and laminate), GIIc (laminate) or ILSS. However the compressive strength
after impact (laminate) is the lowest for the system with the largest
agglomerates (10R10N).
In the literature Pearson et al. reported about bulk hybrid systems containing
large agglomerates as well [3.5.30]. They claimed that the formation of
agglomerates occurs for higher concentrations of reactive liquid rubber; probably
above 12 vol%. The improvements in fracture toughness by nanosilica addition
became smaller for systems with agglomerates present. This would explain our
findings for bulk GIc.
152
The question if the performance of our laminates would even be better with no
agglomerates present could not be answered as the rubber-phase separation of
our system, probably due to the relatively fast curing system, leads to the
formulation of agglomerates. Eventually another hardener system might behave
different. However, this will be the subject of future work.
Another approach for future investigations will be to use core shell rubber
particles instead of reactive liquid rubber; therefore no phase separation during
cure will occur. Thus, theoretically no agglomerates should be formed during
cure.
3.5.5. Conclusions and outlook
Using a fast curing epoxy resin system based on DGEBA/IPD/TMD carbon fiber-
reinforced laminates suitable for automotive applications were made in a
reasonable short manufacturing cycle. The modification of the matrix with both
reactive liquid rubber and silica nanoparticles resulted in tough laminates.
A couple of conclusions can be drawn from the investigations:
1. The bulk resin modified with the reactive liquid rubber shows the expected
increase in toughness alongside the loss in modulus and Tg.
2. The hybrid bulk resins modified additionally with silica nanoparticles show
a reduced loss of the modulus.
3. The hybrid bulk resin with the highest addition level of nanosilica performs
best; exhibiting the highest toughness and modulus.
4. The laminate made from the rubber-only resin shows an increase in GIc, a
slight reduction in GIIc and ILSS as well as a reduction of the delaminated
area in impact testing alongside with an increase in CAI.
5. The hybrid laminates showed no further increase in toughness compared
to the rubber-toughened system, but rather a slight reduction of both GIc
and GIIc. ILSS was on a similar level; CAI was reduced further.
6. The reason for this unexpected lack of further increase in toughness could
be the formation of agglomerates of nanosilica during the cure.
7. The agglomeration found is caused by the presence of the reactive liquid
rubber. Very probably the fast phase separation of the rubber, pushed by
the fast cure, is the driving force of the agglomeration.
8. The conversion factor for the relative improvement in GIc versus the
relative improvement of GIc of the laminate was found to be 0.2 - in
excellent agreement with earlier findings.
The systematic investigation of other fast curing hardeners regarding

agglomerate formation in presence of reactive liquid rubber will be the subject of
future work. Furthermore, of course, hybrid systems based on core-shell rubbers
which do not phase separate upon cure will be investigated. Eventually they are
the material of choice to formulate fast curing hybrid epoxy systems.
153
3.5.6. References
[3.5.1] Regulation (EC) No 443/2009 OF THE EUROPEAN PARLIAMENT AND

OF THE COUNCIL of 23 April (2009) setting emission performance
standards for new passenger cars as part of the Community´s
integrated approach to reduce CO2 emissions from light-duty
vehicles, Official Journal of the European Union, 05.06.2009, L
140/1
[3.5.2] Estin & Co. Main dynamics of the composite industry for automotive
applications 2010 – 2015. Publ. by JEC Composites (2011); Paris;
France
[3.5.3] Hillermeier R., Hasson T., Friedrich L., Ball C. Advanced thermo-
setting resin matrix technology for next generation high volume
manufacture of automotive composite structures. SAE Technical
Paper (2013)-01-1176.; 2013; doi:10.4271/2013-01-1176
[3.5.4] Ellis, B. (Ed.) Chemistry and Technology of Epoxy Resins. Springer

Science+Business Media Dordrecht (1993), 1 – 142.
[3.5.5] Hare, C. Amine curing agents for epoxies. J. Prot. Coat. Lin. (1994),
9, 77 – 103.
[3.5.6] Hare, C. Epoxy curing agents II. J. Prot. Coat. Lin.(1994), 10, 197 –
213.
[3.5.7] Kinloch, A.J., Shaw, S.J., Tod, D.A. Deformation and fracture
behavior of a rubber-toughened epoxy: 1. Microstructure and
fracture studies. Polymer (1983), 24, 1341 – 1354.
[3.5.8] Bagheri, R., Marouf, B.T., Pearson, R. J. Rubber-toughened epoxies:

a Critical Review. Macromol. Sci. Part C: Polymer Reviews (2009),
49, 201 – 225.
[3.5.9] Sprenger, S., Eger C., Kinloch A.J., Taylor, A.C. Nanotoughening of
Epoxies. Proceedings of Stick! Conference April 9th 2003, Nürnberg,
Germany, Vincentz Verlag (2003)
[3.5.10] Picture courtesy of IVW and Polymerservice Merseburg
[3.5.11] Sprenger, S. Epoxy resin composites with surface-modified silicon

dioxide nanoparticles: a review. J. Appl. Polym. Sci. (2013), 130,
1421 – 1428.

Sprenger, S. Toughening mechanisms of nanoparticle-modified
epoxy polymers. Polymer (2007), 48, 530 - 541.
[3.5.13] Sprenger, S. Improving mechanical properties of fiber-reinforced

composites based on epoxy resins containing industrial surface-
modified silica nanoparticles: review and outlook. J. Comp. Mat.
(2014), 49, 53 – 63.
154
[3.5.14] Sprenger, S. Epoxy resins modified with elastomers and surface-
modified silica nanoparticles. Polymer (2013), 54, 4790 – 4797.
[3.5.15] EP 1457509 and WO 2004081076 (2003)
[3.5.16] Picture courtesy of Army Research Laboratory
[3.5.17] Sprenger, S. Fiber-reinforced composites based on epoxy resins

modified with elastomers and surface-modified silica nanoparticles.
J. Mater. Sci. (2014), 49, 2391 – 2402.
[3.5.18] Saxena, A., Hudak, S.J. Review and extension of compliance

information for common crack growth specimen. Int. J. of Fracture
(1978), 14, 453 – 468.
[3.5.19] Xu, S.A., Wang, G.T., Mai, Y.W. Effect of hybridization of liquid
rubber and nanosilica particles on the morphology, mechanical
properties and fracture toughness of epoxy composites. J. Mater.
Sci. (2013), 48, 3546 – 3556.
[3.5.20] Tang, Y., Ye, L., Zhang, Z., Friedrich, K. Interlaminar fracture
toughness and CAI strength of fibre-reinforced composites with
nanoparticles – a review. Comp. Sci. Tech. (2013), 86, 26 – 37.
[3.5.21] Sprenger, S., Kinloch, A.J., Taylor, A.C., Mohammed, R.D., Eger, C.
Rubber-toughened FRCs optimised by nano-Particles. JEC
Composites Magazine (2005), No 19, 73 – 76.
[3.5.22] Kinloch, A.J., Mohammed, R.D., Taylor, A.C., Sprenger, S., Egan, D.
The interlaminar toughness of carbon-fibre reinforced plastic
Composites using “hybrid-toughened” matrices. J. Mater. Sci.
(2006), 41, 5043 – 5046.
[3.5.23] Zhang, J., Deng, S., Ye, L., Zhang, Z. Interlaminar Fracture
Toughness and Fatigue Delamination Growth of CF/EP Composites
with Matrices modified by Nanosilica and CTBN rubber. Proc. 13th
Int. Conf. on Fract. (2013), June 16-21, Beijing, China
[3.5.24] Sprenger, S., Kinloch, A.J., Taylor, A.C., Mohammed, R.D. Rubber-
toughened CFRCs optimized by nanoparticles – Part III JEC
Composites Magazine (2007), No 30, 54 – 57.
[3.5.25] Caccavale, V., Wichmann, M.H.G., Quaresimin, M., Schulte, K.

Nanoparticle/rubber modified epoxy matrix systems: mechanical
performance in CFRPs Proceedings of AIAS XXXVI Convegno
Nazionale 2007, 4-8 September (2007), Ischia, Naples, Italy.
[3.5.26] Kuehn, A., Mahrholz, T., Mosch, Matrix optimization of CFRP parts
concerning fire and impact properties with process acceleration. J.
CEAS Aeronaut J. (2013), 4, 191 – 201.
155
[3.5.27] Tate, J., Trevino, E., Gaikward, S., Sprenger, S., Rosas, I., Andrews,
M. J. Low velocity impact on epoxy glass composites modified with
Rubber micro-particles and silica nano-particles. Journal of
Nanoscience, Nanoengineering, and Applications (2013), vol. 3,
issue 1, ISSN: 2231 – 1777.
[3.5.28] Manjunatha, C.M., Taylor, A.C., Kinloch, A.J., Sprenger, S. The

Tensile fatigue behavior of a glass-fiber reinforced plastic composite
Using a hybrid-toughened epoxy matrix. J. Compos. Mat. (2010),
44, No 17, 2095 – 2109.
[3.5.29] Hsieh, T.H., Kinloch, A.J., Masania, K., Sohn Lee, J., Taylor, A.C.,
Sprenger, S. The toughness of epoxy polymers and fibre composites
modified with rubber microparticles and silica nanoparticles. J.
Mater. Sci. (2010), 45, 1193 – 1210.
[3.5.30] Liang, Y.L., Pearson, R.A. The toughening mechanism in hybrid

epoxy-silica-rubber nanocomposites (HESRNs). Polymer (2010), 51,
4880 – 4890.
156
4. Summary and outlook
4.1. Summary and outlook
The properties of epoxy resins can be improved considerably by the addition of

surface-modified SiO2 nanoparticles. At higher concentrations the particles are
still monodisperse and no agglomerates are found. The rheological properties of
the modified epoxy resins remain unchanged; viscosity is increased only slightly.
The glass transition temperature of cured epoxy does not change in most
systems. A new insight is the approximative linear relationship between the
increase in modulus and the nanosilica addition. For an addition of 10 wt% SiO2
nanoparticles an increase in modulus of approx. 50 % can be expected for the
cured bulk resin. Compressive strength and modulus are increased as well.
Improvements of 10-30 % can be achieved with the addition of 10 wt%
nanosilica.
A linear relationship between the increase in fracture toughness of the cured bulk
epoxy resin and the silica nanoparticle addition level exists as well. For an
addition of 10 wt% of SiO2 nanoparticles an increase in fracture toughness of
approximatively 50 % can be expected. The mechanisms of toughening are a
void formation at the interface particle/resin matrix, followed by a plastic void
growth. The formation of local plastic shear bands contributes as well. Crack
deflection or dissipation at the particles seems to play no or only a minor role for
toughness improvements.
Fatigue performance at cyclic loading is another property of the cured bulk epoxy
resin improved by the addition of nanosilica. A linear relationship between the
nanoparticle concentration and the fatigue performance seems not to exist;
however there seems to be a maximum which depends on the resin/hardener
system. An addition of 10 wt% SiO2 improves the fatigue performance by
approximately 50-60 %.
Laminate properties are improved by a matrix modification with nanosilica as

well. The fracture toughness of a laminate is increased by 5-25 % for typical
addition levels of nanosilica of 10-20 wt%. The compressive strength and the
fatigue performance of the laminate are improved significantly. This is valid for
glass fibers as well as for carbon fibers as reinforcement.
Very interesting is the finding that the addition of SiO2 nanoparticles improves
both laminate KIc and GIc values, but no laminate improvements are found for
fast impact tests like a falling dart test.
The classic toughening of epoxy resins employing reactive liquid rubbers or core-
shell rubbers reduces the modulus of cured bulk epoxy resin systems. By the
addition of silica nanoparticles to the toughened epoxy resin matrix this can
nearly be compensated. These so-called hybrid resin systems are characterized
by a very high toughness and stiffness. Depending on the hardener of the epoxy
resin system toughness increases of 100-1200 % can be achieved. Fatigue
performance of the cured bulk resin is improved further compared to a
nanosilica-only modification of the epoxy resin.
157
New is the finding that the toughness of cured bulk hybrids based on core-shell
rubbers and nanosilica behaves additively, whereas for the combination of
reactive liquid rubber and nanosilica a synergy seems to exist. The reason is very
probably an increase in toughening efficiency due to a more ductile matrix. It is
the result of a minor distortion in rubber phase separation during the cure
caused by the presence of the silica nanoparticles and increases with increasing
addition level of nanosilica.
Bulk resin systems with high toughness and stiffness as well as excellent fatigue
performance can be formulated by a modification of 5-10 wt% core-shell rubber
or 5-15 wt% reactive liquid rubber combined with 5-10 wt% SiO2 nanoparticles.
When investigating the properties of glass fiber-reinforced or carbon fiber-

reinforced laminates, which contain in the resin matrix rubber particles as well
SiO2 nanoparticles, the improvements found are inferior in scale compared to
cured bulk epoxy resins The fatigue performance of laminates based on such
hybrid systems is outstanding, however. Up to ten times the cyclic loadings can
be achieved before the laminate fails.
New is the finding that a nearly linear relationship exists between the relative
improvement of the GIc of the cured bulk epoxy resin and the relative
improvement of the GIc of the laminate made thereof. The reinforcing fibers,
regardless if glass or carbon fibers are used, as well as their three-dimensional
orientation in the composite, seem to have little or no influence on this
relationship. A "conversion factor" can be defined which is 0.18 (± 0.02) for the
systems investigated. Means, if the toughness of a composite part shall be
increased by 100 %, the epoxy resin matrix needs to be replaced by a hybrid
resin exhibiting a toughness 555 % higher than the original unmodified bulk
resin.
Of further interest is the fact, that the linear relationship found does not have a
line through origin. A threshold value must exist, below which the toughness of
the bulk resin is increased by the hybrid modification, but no improvements will
be found in the laminate using this hybrid resin system as matrix resin.
In own experiments laminates reinforced with carbon fibers, made by RTM and
based on a fast curing epoxy resin system (DGEBA/IPD/TMD) have been
investigated. As expected the cured bulk resin system with the highest addition
level of nanosilica exhibited the best properties with the highest values for
fracture toughness and modulus.
The laminates made from the rubber-toughened epoxy resin had a somewhat
higher GIc compared to the laminates made from the unmodified control. The
values for GIIc and the interlaminar shear strength (ILSS) were somewhat lower.
The delaminated area after a 30 J impact was smaller; the value for the
compressive strength after impact (CAI) therefore higher.
158
The laminates based on hybrid epoxy resin systems, containing both reactive
liquid rubber and SiO2 nanoparticles, showed no further increase in fracture
toughness compared to the laminates made from rubber-toughened epoxy resin,
but a slight reduction in values for GIc and GIIc. The values for the ILSS were
almost identical, compressive strength after impact was lower for the hybrid
system.
The reason for this unexpected behaviour is very probably the formation of
agglomerates of nanoparticles during the cure. It seems like the fast phase
separation of the reactive liquid rubber, forced by the very fast cure, is
responsible for the agglomerate formation.
The conversion factor between the relative improvement of the value of GIc of the
cured bulk resin and the relative improvement of the laminate GIc was
determined to be 0.2 – in excellent accordance with earlier results.
The question, if the agglomerates formed have a negative influence on the

laminate performance, could not be answered definitively. Further systematic
investigations using hybrid epoxy resin systems based on reactive liquid rubber
and different hardeners with different curing speeds will be necessary to answer
this questions.
Furthermore hybrid resin systems based on core-shell rubbers in combination

with fast curing hardeners will be the subject of future investigations. As with
these systems no phase separation occurs during cure, they might be the best
choice of toughener for fast-curing hybrid resin systems.
It can be concluded that composites based on hybrid epoxy resin systems are a
new class in laminate performance – tough and stiff with a excellent fatigue
performance. No wonder they are used in an increasing number of industrial
applications.
4.2. Zusammenfassung und Ausblick
Die Eigenschaften von Epoxidharzen können durch den Zusatz von oberflächen-
modifizierten SiO2-Nanopartikeln deutlich verbessert werden. Auch bei höheren
Konzentrationen bleiben die Partikel in der Regel monodispers und es werden
keine Agglomerate gefunden. Die rheologischen Eigenschaften der Harze werden
durch den Nanopartikelzusatz nicht verändert; die Viskosität nur geringfügig
erhöht. Auch die Glasumwandlungstemperatur ändert sich in den meisten Fällen
nicht. Eine neue Erkenntnis ist der annähernd lineare Zusammenhang zwischen
der Erhöhung des Moduls und dem Zusatz an Nanopartikeln. Bei einem Zusatz
von 10 Gewichtsprozent SiO2-Nanopartikeln kann eine Steigerung des Moduls
von etwa 50 % erwartet werden. Druckfestigkeit und Druckmodul werden
ebenfalls erhöht. Verbesserungen von 10-30 % können bei einem Zusatz von 10
Gewichtsprozent SiO2-Nanopartikeln erzielt werden.
Auch zwischen der Verbesserung der Bruchzähigkeit und der Zusatzmenge an

Nanopartikeln besteht für verschiedene Härter ein linearer Zusammenhang. Bei
einem Zusatz von 10 Gewichtsprozent SiO2-Nanopartikeln kann eine
Verbesserung der Bruchzähigkeit des Reinharzes um etwa 50 % erwartet werden.
159
Die Mechanismen der Verbesserung der Schlagzähigkeit sind hauptsächlich ein
Ablösen des Epoxidharzes von den Nanopartikeln, gefolgt von einer plastischen
Hohlraumvergrößerung. Die Ausbildung lokaler plastischer Scherbänder trägt
ebenfalls zur höheren Schlagzähigkeit bei. Rissumleitung oder –streuung an den
Partikeln scheint hingegen keine oder nur eine untergeordnete Rolle zu spielen.
Eine weitere Eigenschaft, welche deutlich verbessert wird, ist das

Ermüdungsverhalten bei zyklischer Belastung. Ein linearer Zusammenhang
zwischen der Konzentration an Nanopartikeln und dem Ermüdungsverhalten
besteht jedoch nicht; vielmehr scheint ein Maximum zu existieren – in
Abhängigkeit von den Harz/Härtersystemen. Für einen Zusatz von 10
Gewichtsprozent SiO2-Nanopartikeln verbessert sich das Ermüdungsverhalten um
etwa 50-60 %.
Die Eigenschaften von Laminaten werden durch eine Modifikation der Harzmatrix
mit SiO2-Nanopartikeln ebenfalls verbessert; so etwa die Bruchzähigkeit um 5-
25 % bei typischen Zusatzmengen von 10-20 Gew% Nanosilica. Die
Kompressionsbeständigkeit und das Ermüdungsverhalten werden signifikant
besser. Dies gilt sowohl für Glasfasern als auch für Carbonfasern als Verstärkung.
Interessanterweise verbessert der Zusatz von SiO2-Nanopartikeln in aller Regel

zwar KIc und GIc der Laminate; bei einem schnellen Impact wie etwa beim
Fallbolzentest werden jedoch keine Verbesserungen gefunden.
Bei der klassischen Schlagzähmodifizierung von Epoxidharzen mit reaktiven

Flüssigkautschuken, zum Teil auch bei der Verwendung von Core-Shell-
Elastomeren reduziert sich der Modul von gehärteten Reinharzsystemen. Dies
kann durch die Kombination mit Nanopartikeln weitestgehend kompensiert
werden. Diese sogenannten Hybridsysteme zeichnen sich durch eine besonders
hohe Zähigkeit bei gleichzeitiger Steifigkeit aus. Je nach Härter können
Steigerungen der Bruchzähigkeit von 100-1200 % erreicht werden. Auch das
Ermüdungsverhalten wird im Vergleich zur reinen Nanomodifikation weiter
verbessert.
Neu ist die Erkenntnis, dass sich Hybride auf Basis Core-Shell-Elastomer und
Nanopartikeln hinsichtlich der Schlagzähigkeit additiv verhalten, während man
bei der Kombination von reaktiven Flüssigkautschuken und Nanopartikeln von
einem synergistischen Verhalten ausgehen kann. Die Ursache ist vermutlich in
einer Steigerung der Effizienz der Schlagzähmodifizierung aufgrund einer
duktileren Matrix zu suchen. Diese entsteht aufgrund einer geringfügigen
Störung der Phasenseparation während der Aushärtung durch die Präsenz der
Nanopartikel und wächst mit steigender Zusatzmenge.
Reinharzsysteme mit hoher Zähigkeit und Steifigkeit sowie sehr guter

Ermüdungsbeständigkeit können durch die Modifikation mit 5-10 Gew% Core-
Shell-Elastomer oder 5-15 Gew% reaktivem Flüssigkautschuk kombiniert mit 5-
10 Gew% SiO2-Nanopartikeln formuliert werden.
Untersucht man die Eigenschaften von glasfaserverstärkten bzw. kohlefaser-

verstärkten Laminaten, welche sowohl Elastomerteilchen als auch SiO2-
Nanopartikel im Harz enthalten, so findet man im Vergleich zu Reinharzsystemen
deutlich geringere Steigerungen.
160
Das Ermüdungsverhalten von Laminaten auf Basis von Hybridharzsystemen ist
mit Abstand am besten – es sind bis zu zehnfache zyklische Belastungen möglich,
bevor die Laminate versagen.
Neu ist die Erkenntnis, dass eine nahezu lineare Beziehung zwischen den
prozentualen Steigerungen des GIc der gehärteten Reinharzhybridsysteme und
den daraus hergestellten Laminaten existiert. Die Verstärkungsfasern,
unabhängig davon ob Glas oder Carbon, als auch ihre dreidimensionale
Orientierung im Verbund, scheinen wenig oder gar keinen Einfluss auf diese
Beziehung zu haben. Ein “Übertragungsfaktor“ kann definiert werden und beträgt
für die untersuchten Systeme 0,18 (± 0,02). D.h., wenn ein Faserverbundbauteil
um 100 % in der Zähigkeit gesteigert werden soll, muss statt des bisher
verwendeten Harzsystems ein Hybridsystem mit einem 555 % höheren GIc
eingesetzt werden.
Ebenfalls von Interesse ist die Tatsache, dass die gefundene lineare Beziehung
keine Ursprungsgerade ist. Es muss also ein Schwellenwert existieren, bei dem
zwar die Zähigkeit des ausgehärteten Reinharzes durch die Modifikation mit
Elastomeren und Nanopartikeln gesteigert wird, im daraus hergestellten Laminat
aber keine Verbesserung im Vergleich zur unmodifizierten Kontrolle zu finden ist.
In den eigenen Versuchen wurden unter Verwendung eines schnellhärtenden

Epoxidharzsystems basierend auf DGEBA/IPD/TMD kohlefaserverstärkte
Laminate mittels RTM hergestellt und untersucht. Wie erwartet zeigte das
ausgehärtete Hybridreinharzsystem mit dem höchsten Gehalt an Nanosilica die
besten Eigenschaften mit der höchsten Bruchzähigkeit und dem höchsten Modul.
Die mit kautschukmodifiziertem Epoxidharz hergestellten Laminate wiesen

gegenüber den Laminaten aus unmodifiziertem Epoxidharz einen höheren GIc auf,
der GIIc fiel, wie auch die interlaminare Scherfestigkeit (ILSS), etwas niedriger
aus. Die delaminierte Fläche nach einem 30 J Impact war im Vergleich kleiner;
die Kompressionsbeständigkeit nach dem Fallbolzentest (CAI) entsprechend
höher.
Die Laminate auf Basis Hybridharz, sowohl mit reaktivem Flüssigkautschuk als
auch mit SiO2-Nanopartikeln modifiziert zeigten jedoch keine weitere
Verbesserung der Bruchzähigkeit gegenüber der reinen Kautschukmodifikation,
sondern eher geringfügig niedrigere Werte für GIc und GIIc. Die ILSS war
vergleichbar; die Kompressionsbeständigkeit nach Impact jedoch verringert.
Der Grund für dieses unerwartete Verhalten ist möglicherweise in der Bildung
von Nanopartikel-Agglomeraten während der Härtung zu suchen. Wahrscheinlich
ist die rasche Phasentrennung des reaktiven Flüssigkautschukes, erzwungen
durch die rasche Vernetzung, verantwortlich für die Entstehung der Agglomerate.
Der Übertragungsfaktor zwischen der relativen Verbesserung des GIc des

Reinharzes und der relativen Verbesserung des GIc des Laminates wurde mit 0.2
in sehr guter Übereinstimmung mit früheren Ergebnissen ermittelt.
161
Die Frage, ob die Agglomerate einen negativen Einfluss auf die Performance des
Laminates haben, konnte nicht endgültig beantwortet werden. Weitere
systematische Untersuchungen mit Hybridsystemen auf Basis reaktiver
Flüssigkautschuke unter Verwendung von Härtersystemen mit abgestuften
Härtungsgeschwindigkeiten werden diese Frage beantworten.
Ebenfalls Gegenstand zukünftiger Untersuchungen werden Hybridsysteme auf

Basis von Core-Shell-Partikeln in Kombination mit schnellen Härtern sein. Da hier
keine Phasentrennung während der Härtung auftritt, stellen sie möglicherweise
die beste Modifikation für schnellhärtende Hybridsysteme dar.
Abschließend ist festzustellen, dass Faserverbundwerkstoffe auf Basis von

Hybridepoxysystemen eine neue Leistungsklasse darstellen - zäh und steif mit
ausgezeichnetem Ermüdungsverhalten. Es ist daher wenig verwunderlich, dass
sie in immer größerem Umfang in industriellen Anwendungen eingesetzt werden.
162
5. Bibliography
The literature cited in this work is numbered by chapter and then continuously
within the chapter. This was necessary in order to keep the integrity of the
published papers. If literature is cited in several chapters, all numbers are given
as cross-reference.
[1.1] Regulation (EC) No 443/2009 OF THE EUROPEAN

[3.5.1] PARLIAMENT AND OF THE COUNCIL of 23 April 2009
setting emission performance standards for new passenger
cars as part of the Community´s integrated approach to
reduce CO2 emissions from light-duty vehicles, Official Journal
of the European Union, 05.06.2009, L 140/1
[2.1.1] Ehrenstein, G.W. (2006) Faserverbundkunststoffe. Hanser

[3.4.1] Verlag München, Germany, 148-209
[2.1.2] Campbell, F. (2004) Manufacturing processes for advanced

composites. Elsevier Advanced Technology, Oxford, UK,
pages 104-179 and 306-357
[2.1.3] Constantino, S., Waldvogel, U. (2010) Composite processing:

[3.4.5] State of the art and future trends. In: Pascault J.P., Williams
R.J.J. (eds) Epoxy Polymers. Wiley, Weinheim, Germany,
271–287
[2.1.4] Hillermeier R., Hasson T., Friedrich L., Ball C. (2013)

[3.4.6] Advanced thermosetting resin matrix technology for next
[3.5.3] generation high volume manufacture of automotive composite
structures SAE Technical Paper 2013-01-1176.
doi:10.4271/2013-01-1176
[2.1.5] Ohlendorf, J.H., Rolbiecki, M., Schmohl, T.; Franke, J.

Thoben, K.D., Ischtschuk, L. (2013) Innovationen in der
Handhabungs- und Textiltechnik zur Rotorblattfertigung,
Lightweight Design 6 (5), 50-57
[2.1.6] Trender, L., Schromm, M. (2013) 918 Spyder. Proceedings of

Aachener Karosseriebautage, 24./25. September, RWTH,
Aachen, Germany
[2.1.7] Prepreg Technology. (2005) Company information No. FGU

017b, Hexcel Ltd., Duxford, UK
[2.1.8] De Oto, L. (2012) Lightweight inspiration from the

supersports car segment – the Lamborghini Aventador.
Proceedings of the Lightweight Car Body Conference, 18./19.
September, Neckarsulm, Germany
[2.1.9] Company brochure. (2012) Magnus Venus Plastech,

Clearwater, FL, USA
163
[2.1.10] Pulsstar brochure. (2013) Strongwell, Bristol, VA, USA
[2.1.11] Mitzler, J., Renkl, J., Würtele, M. (2011) Hoch beanspruchte

Strukturbauteile in Serie. Kunststoffe 3, 36-40
[2.1.12] Product presentation. (2013) Kegelmann Technik, Rodgau-

Jügesheim, Germany
[2.1.13] Schmid, L., Schnaufer, T. (2013) The lightweight structure of

the BMW I3. Proceedings of Aachener Karosseriebautage,
24./25. September, RWTH, Aachen, Germany
[2.1.14] Company information. (2013) Norco GRP Ltd., Holton Heath,

UK
[2.2.1] Flemming, M., Ziegmann, G., Roth, S. (1995) Faserverbund-

[3.1.44] Bauweisen. Springer Verlag Berlin, Germany, 6-179
[2.2.2] Ehrenstein, G.W. (2006) Faserverbundkunststoffe. Hanser

Verlag München, Germany, 19-49
[2.2.3] Market Research Report (2012) Global Glass Fiber Market

2012-2017: Trend, Forecast and Opportunity Analysis,
Lucintel, Las Colinas, TX, USA
[2.2.4] Jahn, B., Witten, E. (2013) Composites Market Report 2013.

Carbon Composites e.V., September 2013, Augsburg,
Germany
[2.2.5] Product presentation glas fiber fabric. (2013) Jiyuan

Wuayang Composite Materials Co. Ltd., He Nan, China
[2.2.6] Product presentation DU carbon textile (2014) Spinteks

Tekstil Insaat San. ve Tlc. A.S., Honaz, Turkey
[2.2.7] Baitinger, S. (2013) Möglichkeiten im Leichtbau durch

multiaxiale Gelege bei Auslegung und Fertigung. Proceedings
of Neue Technologien im textilbasierten Faserverbund-
Leichtbau, 18. April, DLR Stuttgart, Germany
[2.2.8] Product presentation. (2013) SGL Kümpers GmbH, Rheine/

Gellendorf, Germany
[2.2.9] BMW I3 Monocoque (2012), Composites Europe, 9.-11.

October 2012, press information Reed Exhibitions
[2.2.10] Thomason, J.L. (2013) Glass Fiber Sizings: A Review of the

Scientific Literature. ISBN 978-0-9573814-0-7, Create Space,
UK
164
[2.3.1] Hasson, T. (2013) Neue Matrixtechnologie für die CFK-
Fertigung in der Automobilindustrie. Proceedings of Neue
Technologien im textilbasierten Faserverbund-Leichtbau, 18.
April, DLR Stuttgart, Germany
[2.3.2] CEH Marketing Research Report (2010) Epoxy Resins. SRI

Consulting, Mento Park, CA, U.S.A.
[2.3.3] Ellis, B., Hrsg. (1993) Chemistry and Technology of Epoxy

Resins. Springer Science+Business Media, Dordrecht,
Netherlands
[2.3.4] Hare, C.J. (1994) Amine curing agents for epoxies. J. Prot.
[3.1.1] Coat. Lin., 9, 77-103
[3.4.3]
[3.5.5]
[2.3.5] Hare, C.J. (1994) Epoxy curing agents II. J. Prot. Coat. Lin.,
[3.1.2] 10, 197-213
[3.4.4]
[3.5.6]
[2.4.1] Okamoto, Y. (1983) Thermal Aging Study of Carboxyl-

Terminated Polybutadiene and Poly(Butadiene-Acrylonitrile)-
Reactive Liquid Polymers. Polymer Engineering and Science,
Vol.23, No 4, 222-225
[2.4.2] Kinloch, A.J., Shaw, S.J., Tod, D.A., Hunston D.L. (1983)
[3.2.1] Deformation and fracture behaviour of a rubber-toughened
[3.4.7] epoxy: 1. Microstructure and fracture studies. Polymer, 24,
[3.5.7] 1341–1354 2. Failure criteria. Polymer, 24, 1355-1363
[2.4.3] Siebert, A.R. (1984) Morphology and Dynamic Mechanical

Behaviour of Rubber-Toughened Epoxy Resins. ACS Adv. in
Chem. Series, 208, 12, 179-191
[2.4.4] Kinloch, A.J. (2003) Toughening Epoxy Adhesives to Meet

[3.2.2] Today’s Challenges. MRS Bull. June 2003, 445-448
[3.4.8]
[2.4.5] Bagheri, R., Marouf, B.T., Pearson,R.A. (2009) Rubber-

[3.2.3] Toughened Epoxies: A Critical Review. J. Macromol. Sci. Part
[3.4.9] C Polym. Rev., 49, 201-225
[3.5.8]
[2.4.6] Huang, Y., Kinloch A.J. (1992) Modelling of the toughening

mechanisms in rubber-modified epoxy polymers. Part II A
quantitative description of the microstructure-fracture
property relationships. J. Mater. Sci. 27, 2763–2769
165
[2.4.7.] Sprenger, S., Weber, C., Pulliam, L. (1997) Elastomer-
modified epoxy prepolymers - the new generations. European
Adhesives & Sealants, September 1997, 9-12
[2.4.8] Giannakopoulos, G., Masania, K., Taylor, A.C. (2011)

[3.2.4] Toughening of epoxy using core-shell particles. J. Mater. Sci.
[3.4.11] 46, 327–338
[2.4.9] Tsai, J.L., Chang, N.R. (2011) Investigating damping

[3.2.6] properties of nanocomposites and sandwich structures with
[3.4.12] nanocomposites as core materials. J. Compos. Mater. 45,
2157-2164
[2.4.10] Block, H., Pyrlik, M. (1988) Silicones are the Key: Modifying
[3.2.8] Thermosetting Resins with Silicone Elastomers. Kunstst. Ger.
Plast. 78, 1192-1196
[2.4.11] Lai, M., Friedrich, K., Botsis, J., Burkhart, T. (2010)

[3.1.17] Evaluation of residual strains in epoxy with different nano/
[3.2.5] micro-fillers using embedded fiber Bragg grating sensor.
[3.4.13] Compos. Sci. Technol. 70, 2168-2175
[2.4.12] Chen, J., Kinloch, A.J., Sprenger, S., Taylor, A.C. (2013) The
[3.2.9] mechanical properties and toughening mechanisms of an
epoxy polymer modified with polysiloxane-based core-shell
particles. Polymer 54, 4276-4289
[2.4.13] Unpublished research report of nanoresins AG,

Geesthacht, Germany
[2.4.14] Yoon, T.H., Priddy, D.B., Lyle, G.D., McGrath, J.E. (1995)
Mechanical and morphological investigations of reactive
polysulfone toughened epoxy networks. Macromol. Symp. 98,
673-686
[2.4.15] Pearson, R.A., Yee, A.F. (1993) Toughening mechanisms in

thermoplastic-modified epoxies: 1. Modification using poly-
(phenylenoxid). Polymer 34, 3658-3670
[2.4.16] WO 02/16481 A1 of 28.02.2002
[2.4.17] Product presentation (2007) Novel toughened epoxy resins.

The Dow Chemical Company, Freeport, TX, USA
[2.5.1] Odegard, G.M., Clancy, T.C., Gates, T.S. (2005) Modeling of

[3.1.3] the mechanical properties of nanoparticle/polymer
composites. Polymer 46, 553-563
[2.5.2] Sprenger, S., Eger C., Kinloch A.J., Taylor, A.C. (2003)
[3.3.1] Nanotoughening of Epoxies. Proceedings of Stick! Conference
[3.4.15] April 9th, Nürnberg, Germany, Vincentz Verlag 2003
[3.5.9]
166
[2.5.3] Picture courtesy of IVW Kaiserslautern and
[3.2.11] Polymerservice Merseburg, Germany
[3.4.19]
[3.5.10]
[2.5.4] Product information (2013) Evonik Hanse GmbH, Geesthacht,

Germany
[2.5.5] Sprenger S., Kinloch A.J., Taylor, A.C., Mohammed, R.D.
[3.1.25] (2005) Rubber-toughened GFRCs optimized by nanoparticles.
[3.2.14] JEC Compos. Magazine 21, 66-69
[3.3.3]
[3.4.21]
[2.5.6] Eger C., Schultz, P. (2005) Reinforcing epoxy resins with silica
[3.1.28] nanoparticles. In Proceedings of "High Performance Fillers
2005", March 8-9, Köln, Germany
[2.5.7] Rosso, P., Ye, L., Friedrich, K., Sprenger, S. (2006) A

[3.1.5] toughened epoxy resin by silica nanoparticle reinforcement. J.
Appl. Polym. Sci. 100, 1849-1855

[3.1.29] Sprenger, S. (2007) Toughening mechanisms of nanoparticle-
[3.5.12] modified epoxy polymers. Polymer 48, 530-541
[2.5.9] Liu, S., Zhang, H., Zhang, Z., Zhang, T., Sprenger, S. (2008)
[3.1.33] Tailoring the mechanical performance of epoxy resin by
various nanoparticles. Polymers & Polymer Composites Vol.
16, No. 8, 471-477
[2.5.10] Prasad, S., Grover T., Basu, S. (2010) Coarse-grained

molecular dynamics simulation of cross-linking of DGEBA
epoxy resin and estimation of adhesive strength. Int. J. Eng.,
Sci. Technol. 2, 17-30
[2.5.11] Gou, J., Fan, B., Song, G., Khan, A. (2006) Study of affinities
between single-walled nanotube and epoxy resin using
molecular dynamic simulation. Int. J. Nanosci. 5, 131-144
[2.5.12] Sprenger, S., Eger C. (2003) WO 2004081076 resp. EP 1 457

[3.2.12] 509
[3.4.18]
[3.5.15]
[2.5.13] Sprenger, S., Eger, C. Kinloch, A.J., Lee, J.H., Taylor, A.C.,
[3.2.13] Egan, D. (2003) Nanoadhesives: Toughness and high
strength. Adhäsion, Kleben & Dichten 03/2003, 24-30
[2.5.14] Army Research Laboratory Technical Report 4084 (2007)

[3.2.19] Aberdeen Proving Ground, MD, U.S.A. April 2007
167
[2.5.15] Picture courtesy of J. Robinette, Army Research
[3.2.20] Laboratory, MD, U.S.A.
[3.5.16]
[2.5.16] Tsai, J.-L., Huang, B.-H. Cheng, Y.-L. (2009) Enhancing

[3.2.16] Fracture Toughness of Glass/Epoxy Composites by Using
[3.3.4] Rubber Particles Together with Silica Nanoparticles. J. Comp.
[3.4.24] Mater. 43, 3107 – 3123
[2.5.17] Sprenger, S., Kinloch, A.J., Taylor, A.C., Mohammed, R.D.,

[3.2.26] Eger, C. (2005) Rubber-toughened FRCs optimized by nano-
[3.3.30] particles. JEC Comp. Mag. No 19, 73–76
[3.4.41]
[3.5.21]
[2.5.18] Final Report Stiftung Industrieforschung Projekt S 657

[3.2.38] (2005), Institut f. Verbundwerkstoffe, Kaiserslautern,
[3.4.35] Germany
[2.5.19] Liu, H.Y., Wang, G.T., Mai, Y.W. Zeng, Y. (2011) On fracture
[3.1.19] toughness of nanoparticle modified epoxy. Composites: Part
[3.2.35] B 42, 2170-2175
[2.5.20] Mahrholz, T., Herbeck, L., Riedel U. (2004) New high-

[3.3.8] performance fibre-reinforced nanocomposites. JEC Comp.
Mag. No 9, 71-75
[2.5.21] Sprenger, S., Kinloch, A.J., Taylor, A.C., Mohammed, R.D.

[3.2.24] (2007) Rubber-toughened CFRCs optimized by nanoparticles
[3.3.22] Part III. JEC Compos. Mag. No 30, 54-57
[3.4.38]
[3.5.24]
[2.5.22] Caccavale V., Wichmann, M., Quaresimin, M., Schulte, K.

[3.3.17] (2007) Nanoparticle/Rubber Modified Epoxy Matrix Systems:
[3.4.36] Mechanical Performance in CFRPs. Proceedings of AIAS XXXVI
[3.5.25] Convegno Nazionale, 4-8 September, Ischia, Neapel, Italy
[2.5.23] Kinloch, A.J., Masania, K., Taylor, A.C., Agarwal, R.,

[3.3.10] Sprenger, S., Egan, D. (2008) The fracture of glass-fibre-
[3.4.28] reinforced epoxy composites using nanoparticle-modified
matrices. J. Mater. Sci. Letters 43, 1151-1154
[2.5.24] Manjunatha, C.M., Taylor, A.C., Kinloch, A.J., Sprenger,

[3.3.14] S. (2009) The effect of rubber micro-particles and silica
[3.4.30] nanoparticles on the tensile fatigue behaviour of a glass-
fibre epoxy composite. J. Mater. Sci. Letters 44, 342-345
[2.5.25] Tsai, J.L., Huang, B-H., Cheng, Y.L. (2011) Enhancing

[3.3.5] Fracture Toughness of Glass/Epoxy Composites for Wind
[3.4.25] Blades Using Silica Nanoparticles and Rubber Particles.
Procedia Engineering 14, 1982-1987
168
[2.5.26] Tang, Y., Ye, L., Zhang, D., Deng, S. (2011) Characterization
[3.1.18] of transverse tensile, interlaminar shear and interlaminate
[3.3.19] fracture in CF/EP laminates with 10 wt% and 20 wt% silica
nanoparticles in matrix resins. Composites: Part A, 42, 1943-
1950
[2.5.27] Sprenger, S., Eger, C., Kinloch, A., Mohammed, R., Taylor, A.
[3.3.12] (2005) Rubber-toughening and Nanoparticles in Epoxies:
Synergies in FRC. Proceedings SAMPE, Paris, France, April 5 –
7, 2005
[2.5.28] Brandt, J., Drechsler, K., Schmidtke, K. (2004) Composites

[3.4.43] für den Flugzeugbau. Kunststoffe 10, 290-294
[2.5.29] Caccavale, V. (2007) Nanoparticle/rubber modified epoxy

[3.2.15] matrix systems: mechanical performance in CFRPs. Master of
[3.3.18] Science Thesis 2007, University of Padua, Italy
[3.4.37]
[2.5.30] Altstädt, V. (1991) Effect of the polymer matrix on the

properties of advanced composites. Makromolekulare Chemie,
Macromolecular Symposia 50, 137-145
[3.1.4] Sanctuary, R., Baller, J., Krüger, J.-K., Schäfer, D; Wetzel, B.,
Possart, W., Alnot, P. (2006) Complex specific heat capacity
of two nanocomposite systems. Thermochimica Acta 445, No.
2, 111-115
[3.1.6] Wetzel, B., Rosso, P., Haupert, F., Friedrich, K. (2006)

Epoxy nanocomposites – fracture and toughening
mechanisms. Engineering Fracture Mechanics 73, 2375-2398
[3.1.7] Wetzel, B. (2006) Mechanische Eigenschaften von Nano-

verbundwerkstoffen aus Epoxydharz und keramischen
Nanopartikeln. Dissertation Universität Kaiserslautern,
Germany
[3.1.8] Deng, S., Ye, L., Friedrich, K. J. (2007) Fracture behaviours of

epoxy nanocomposites with nano-silica at low and elevated
temperatures. J. Mater. Sci. 42, No. 8, 2766-2774
[3.1.9] Rosso, P., Ye, L. (2007) Epoxy/Silica Nanocomposites:

Nanoparticle-Induced Cure Kinetics and Microstructure
Macromol. Rapid Commun. 28, 121–126
[3.1.10] Walter, R., Haupert, F., Schlarb, A. (2008) Optimierung der

[3.2.39] Abrasivverschleißeigenschaften von nanopartikel- und
kautschukpartikelmodifiziertem Epoxidharz für Harzinjektions
applikationenTribologie + Schmierungstechnik 2/55, 26-30
169
[3.1.11] Dittanet, P. (2008) The use of nanosilica in epoxy resins.
Master Thesis, Lehigh University, Bethlehem, Pennsylvania,
U.S.A.
[3.1.12] Wang, G.T., Liu, H.Y., Saintier, N., Mai, Y.W. (2009) Cyclic
[3.2.34] fatigue of polymer nanocomposites. Eng. Failure Analysis
16, 2635-2645
[3.1.13] Liang, Y.L., Pearson, R.A. (2009) Toughening mechanisms in

[3.2.41] epoxy–silica nanocomposites (ESNs). Polymer 50, 4895–
4905
[3.1.14] Baller, J., Becker, N., Ziehmer, M., Thomassey, M., Zielinski,
B., Müller, U., Sanctuary, R. (2009) Interactions between
silica nanoparticles and an epoxy resin before and during
network formation. Polymer 50, 3211-3219
[3.1.15] Philipp, M., Müller, U., Jiménez Riobóo, R.J., Baller, J.,
Sanctuary, R., Possart, W., Krüger, J.K. (2009) Interphases,
gelation, vitrification, porous glasses and the generalized
Cauchy relation: epoxy/silica nanocomposites. New Journal of
Physics 11, 023015
[3.1.16] Tsai, J.L., Huang, B.H., Cheng, Y.L. (2010) Investigating

[3.3.6] Mechanical Behaviors of Silica Nanoparticle Reinforced
Composites. J. Comp. Materials 4/44, 505-524
[3.1.20] Liu, H.Y., Wang, G.T., Mai, Y.W. (2012) Cyclic fatigue crack
[3.2.36] propagation of nanoparticle modified epoxy. Comp. Sci. Tech.
72, 1530-1538
[3.1.21] Dittanet, P., Pearson, R. (2012) Effect of silica nanoparticle

size on toughening mechanisms of filled epoxy. Polymer 53,
1890-1905
[3.1.22] Hsieh, T.H., Kinloch, A.J., Masania, K., Taylor, A.C., Sprenger,
S. (2010) The mechanisms and mechanics of the toughening
of epoxy polymers modified with silica nanoparticles. Polymer
51, 6284-6294
[3.1.23] Ma, J., Mo, M.S., Du, X.S., Rosso, P., Friedrich, K. Kuan, H.C.
(2008) Effect of inorganic nanoparticles on mechanical
property, fracture toughness and toughening mechanism of
two epoxy systems. Polymer 49, 3510-3523
[3.1.24] Jajam, K.C., Tippur, H.V. (2012) Quasi-static and dynamic

fracture behavior of particulate polymer composites: A study
of nano- vs. micro-size filler and loading-rate effects.
Composites Part B 43(8), 3467–3481
170
[3.1.26] Pethrick, R., Miller, C., Rhoney, I. (2010) Influence of
nanosilica particles on the cure and physical properties of a
thermoset epoxy resin. Polym. Int. 59, 236–241
[3.1.27] Gurung, R. (2011) Effects of nanosilica filler on the thermal

and mechanical properties of an epoxy/amine resin system";
Master Thesis, Wichita State University, Wichita, Kansas,
U.S.A.
[3.1.30] Blackman, B.R.K., Kinloch, A.J., Sohn Lee, J., Taylor, A.C.,
Agarwal, R., Schueneman, G., Sprenger, S. (2007) The
fracture and fatigue behaviour of nano-modified epoxy
polymers. J. Mater. Sci. Letters 42, 7049-7051
[3.1.31] Liu, S., Zhang, H., Zhang, Z., Sprenger, S. (2008) Epoxy
Resin Filled with High Volume Content Nano-SiO2 Particles.
J. Nanosci. Nanotechnol. Vol. 8, 1-6
[3.1.32] Zjang, H., Zhang, Z., Friedrich, K., Eger, C. (2006) Property
improvements of in situ epoxy nanocomposites with reduced
interparticle distance at high nanosilica content. Acta Mater.
54, 1833-1842
[3.1.34] Hsieh, T.H., Kinloch, A.J., Masania, K., Sohn Lee, J., Taylor,
[3.2.28] A.C., Sprenger, S. (2010) The toughness of epoxy polymers
[3.3.16] and fibre composites modified with rubber microparticles and
[3.4.31] silica nanoparticles. J. Mater. Sci. 45, 1193-1210
[3.5.29]
[3.1.35] Hsieh, T.H., Kinloch, A.J., Taylor, A.C., Sprenger, S. (2011)

The Effect of Silica Nanoparticles and Carbon Nanotubes on
the Toughness of a Thermosetting Epoxy Polymer. J. Applied
Polym. Sci. 119, 2135-2142
[3.1.36] Tang, L.C., Zhang, H., Sprenger, S., Ye, L., Zhang, Z. (2012)
[3.2.30] Fracture mechanisms of epoxy-based ternary composites
filled with rigid-soft particles. Compos. Sci. Technol. 72, issue
5, 558–565
[3.1.37] Zhang, H., Tang, L.C., Zhang, Z., Friedrich, K. Sprenger, S.

(2008) Fracture behaviours of in situ silica nanoparticle-filled
epoxy at different temperatures. Polymer 49, 3816-3825
[3.1.38] Wang, Z.Z., Gu, P., Zhang, Z., Gu, L., Xu, Y.Z. (2011)
Mechanical and Tribological Behavior of Epoxy/Silica Nano-
composites at the Micro/Nano Scale. Tribol. Lett. 42, 185–
191
[3.1.39] Zhang, Y.F., Bai, S.L., Li, X.K., Zhang, Z. (2009) Viscoelastic
Properties of Nanosilica-Filled Epoxy Composites Investigated
by Dynamic Nanoindentation. J. Polym. Sci. Part B: Polym.
Phys. 47, 1030–1038
171
[3.1.40] Mahrholz, T., Stängle, J., Sinapius, M. (2009) Quantitation of
the reinforcement effect of silica nanoparticles in epoxy resins
used in liquid composites moulding processes. Composites:
Part A 40, 235-243
[3.1.41] Duwe, S., Arlt, C., Aranda, S., Riedel, U., Ziegmann, G.
(2012) A detailed thermal analysis of nanocomposites filled
with SiO2, AlN or boehmite at varied contents and a review of
selected rules of mixture. Comp. Sci. Tech. 72, 1324-1330
[3.1.42] Jumahat, A., Soutis, C., Jones, F., Hodzic, A. (2010) The
effects of nanosilica contents on thermal and mechanical
properties of epoxy polymers"; Proceedings of 10th SAMPE
Europe conference, 12-14 April, Paris, France
[3.1.43] Jumahat, A., Soutis, C., Jones, F., Hodzic, A. (2010) Effect of
silica nanoparticles on compressive properties of an epoxy
polymer. J Mater. Sci. 45, No. 21, 5973-5983
[3.1.45] Sprenger, S., Kinloch, A.J., Taylor, A.C., Hsieh, T.-H. (2009)
[3.2.33] Ultra-tough and fatigue resistant. Adhesion, Adhesives &
Sealants 10, 8–11
[3.2.7] EP 0 407 834 dated 05.07.1989

[3.4.14]
[3.2.10] Sprenger, S. (2013) Epoxy Resin Composites with Surface-

[3.3.2] Modified Silicon Dioxide Nanoparticles: A Review. J. Appl.
[3.4.16] Polym. Sci. 130, 1421-1428
[3.5.11]
[3.2.17] Pearson, R., Liang, Y. (2010) Fracture behaviour of hybrid

epoxy-silica-rubber nanocomposites. Polymer
Nanocomposites, Woodhead Publishing Ltd, Chapter 23, 773–
786
[3.2.18] Liang, Y.L., Pearson, R.A. (2010) The toughening mechanism

[3.5.30] in hybrid epoxy-silica-rubber nanocomposites (HESRNs).
Polymer 51, 4880–4890
[3.2.21] Uddin, M.F., Sun, C.T. (2010) Improved dispersion and

mechanical properties of hybrid nanocomposites. Composites
Sci. and Tech. 70, 223-230
[3.2.22] Veena, M.G., Renukappa, N.M., Suresha, B., Shivakumar,

K.N. (2011) Tribological and Electrical Properties of Silica-
Filled Epoxy Nanocomposites. Polymer Composites 32, 2038–
2050
172
[3.2.23] Veena, M.G., Renukappa, N.M., Raj, J.M., Ranganathaiah, C.,
Shivakumar, K.N. (2011) Characterization of Nanosilica-Filled
Epoxy Composites for Electrical and Insulation Applications. J.
Appl. Polym. Sci. 121, 2752–2760
[3.2.25] Singh, V.K., Gope, P.C. (2010) Silica_Styrene_Butadiene

Rubber Filled Hybrid Composites: Experimental
Characterization and Modeling. J. of Reinforced Plastics and
Composites 29, 2450–2468
[3.2.27] Manjunatha, C.M., Taylor, A.C., Kinloch, A.J., Sprenger, S.

(2009) The cyclic-fatigue behaviour of an epoxy polymer
modified with micron-rubber and nano-silica particles J.
Mater. Sci. Lett. 44, 4487–4490
[3.2.29] Tang, L.C., Zhang, H., Pei, Y.B., Chen, L.M., Song, P.,
Zhang, Z. (2013) Improved impact resistance of epoxy resin
filled with rigid and soft nanoparticles. Composites: Part A
accepted for publication
[3.2.31] Sprenger, S., Eger, C., Kinloch, A.J., Lee, J.H., Taylor, A.C.,
Egan, D. (2003) Toughening structural adhesives via nano-
and micro-phase inclusions. Journal of Adhesion 79, 867–
873
[3.2.32] Sprenger, S., Eger, C., Kinloch, A.J., Lee, J.H., Taylor, A.C.,
Egan, D. (2004) Nano-modified ambient temperature curing
epoxy adhesives. Adhäsion, Kleben & Dichten 03/2004, 17–21
[3.2.37] Palinsky, A. (2006) Epoxy Resins For High Temperature

Applications. Composites Processing 27 April 2006,
Merseyside, UK
[3.2.40] Taylor, A. (2012) personal communication, Imperial College,

London, UK
[3.2.42] Pearson, R.A., Yee, A.F. (1989) Toughening mechanisms in

elastomer-modified epoxies Part 3 The effect of cross-link
density. J. Mat. Sci. 24, 2571–2580
[3.3.7] Tate, J.S., Akinola, A.T., Sprenger, S. (2010) Mechanical

Performance of Nanomodified Epoxy/Glass Composites for
Wind Turbine Applications. Proceedings of the SAMPE
Conference May 17 - 21, Seattle, WA, U.S.A.
[3.3.9] Herbeck, L., Mahrholz, T., Mosch, J., Riedel, U.,

Röstermundt, D. (2005) Faserverstärkte Nanocomposites –
Stand der Technik und Perspektiven. GAK 58, 161-166
173
[3.3.11] Kinloch, A.J., Masania, K., Taylor, A.C., Sprenger, S. (2009)
[3.4.29] The Fracture of Nanosilica and Rubber Toughened Epoxy Fibre
Composites. Proceedings of the Composites & Polycon,
American Composites Manufacturers Association, January 15
17, Tampa, Florida, U.S.A.
[3.3.13] Kuehn, A., Mahrholz, T., Mosch, J. (2013) Matrix optimization

[3.5.26] of CFRP parts concerning fire and impact properties with
process acceleration. CEAS Aeronaut J. 4, 191–201

(2010) The tensile fatigue behaviour of a silica nanoparticle-
modified glass fibre reinforced epoxy composite. Composites
Science and Technology 70, 193-199
[3.3.20] Zeng, Y., Liu, H.Y., Mai, Y.W., Du, X.S. (2012) Improving
interlaminar fracture toughness of carbon fibre/epoxy
laminates by incorporation of nanoparticles. Composites:
Part B 43, 90-94
[3.3.21] Zhang, G., Sebastian, R., Burkhart, T., Friedrich, K. (2012)

Role of monodispersed nanoparticles on the tribological
behavior of conventional epoxy composites filled with carbon
fibers and graphite lubricants. Wear 292-293, 176-187
[3.3.23] Hackett, S.H., Nelson, J.M., Hine, A.M., Sedgwick, P., Lowe,
R.H., Goetz, D.P., Schultz, W.J. (2010) The Effect of
Nanosilica Concentration on the Enhancement of Epoxy Matrix
Resins for Prepreg Composites. Proceedings SAMPE 2010,
Oct. 11-14, Salt Lake City, UT, U.S.A.
[3.3.24] Nelson, J.M., Hackett, S.C., Goetz, D.P., Hine, A.M., Schultz,
W.J. (2012) Development of Nanosilica-Thermoset Matrix
Resins for Prepreg Composites. NSTINanotech 2012, ISBN
978-1-4665-6274-5 Vol. 1
[3.3.25] Nelson, J.M., Hine, A.M., Goetz, D.P., Sedgwick, P., Lowe,
R.H., Rexeisen, E., King, R.E., Patz, N. (2012) Nanosilica-
modified Epoxy Matrix Resin for Prepreg Composite Tooling
Applications. Proceedings SAMPE 2012, May 21-24, Baltimore,
MD, U.S.A.
[3.3.26] Nelson, M., Hine, A.M., Goetz, D.P., Sedgwick, P., Lowe,
R.H., Rexeisen, E., King, R.E., Aitken, C., Pham, Q. (2013)
Properties and Applications of Nanosilica-modified Tooling
Prepregs. SAMPE Journal 49, No. 1, 7-17
[3.3.27] Chiu, K.R., Duenas, T., Dzenis, Y., Kaser, J., Bakis, C.E.,
Roberts, J.K., Carter, D. (2013) Comparative Study of Nano-
materials for Interlaminar Reinforcement of Fiber-Composite
Panels. Proc. of SPIE 2013, 8689, 86891D1
174
[3.3.28] Jumahat, A., Soutis, C., Jones, F., Hodzic, A. (2010)
Improved Compressive Properties of a Unidirectional CFRP
Laminate Using Nanosilica Particles. Advanced Composites
Letters 19 (6), 218-221
[3.3.29] Thunhorst, K., Goetz, D.P., Hine, A.M., Sedgwick, P. (2011)

The Effect of Nanosilica Matrix Modification on the
Improvement of the Pultrusion Process and Mechanical
Properties of Pultruded Epoxy Carbon Fiber Composites.
Proc. Composites 2011, Feb. 2-4, Ft. Lauderdale, FL, U.S.A.
[3.3.31] Liu, L.Q., Li, L., Gao, Y., Tang, L., Zhang, Z. (2013) Single
carbon fiber fracture embedded in an epoxy matrix modified
by nanoparticles. Composites Science and Technology 77,
101-109
[3.3.32] Jacob, G.C., Starbuck, J.M., Fellers, J.F., Simunovic, S.,

Boeman, R.G. (2004) Strain rate effects on the mechanical
properties of polymer composite materials. J. Appl. Polym.
Sci. 94, 296-301
[3.3.33] Gao, X., Jensen, R.E., McKnight, S.H., Gillespie Jr., J.W.
(2011) Effect of colloidal silica on the strength and energy
absorption of glass fiber/epoxy interphases. Composites:
Part A 42, 1738-1747
[3.3.34] Yang, Y., Lu, C.X., Su, X.L., Wu, G.P., Wang, X.K. (2007)
Effect of nano-SiO2 modified emulsion sizing on the interfacial
adhesion of carbon fibers reinforced composites. Materials
Letters 61, 3601-3604
[3.3.35] Sprenger, S. (2013) Epoxy resins modified with elastomers

[3.4.17] and surface-modified silica nanoparticles. Polymer 54,
[3.5.14] 4790-4797
[3.4.10] Sprenger, S., Kinloch, A.J., Taylor, A.C., Mohammed, R.D.

(2008) Rubber-toughened FRCs optimised by nanoparticles
IV. JEC Compos. Mag. 38, 34–37
[3.4.20] Sprenger, S. (2015) Improving mechanical properties of fiber-

[3.5.13] reinforced composites based on epoxy resins containing
industrial surface-modified silica nanoparticles: review and
outlook. J. Comp. Mat. 49, 53-63
[3.4.22] Uddin, M.F., Sun, C.T. (2008) Strength of unidirectional glass/

epoxy composite with silica nanoparticle-enhanced matrix.
Composites Science and Technology 68, 1637–1643
175
[3.4.23] Sun, C.T. (2012) Development of toughened and multi-
functional Nanocomposites for ship structures. Final Report
for Office of Naval Research, School of Aeronautics and
Astronautics, Purdue University, West Lafayette, IN 47907,
U.S.A.
[3.4.26] Schultz, R., Tate, J.S., Gaikwad, S., Trevino, E., Jacobs, C.,
Sprenger, S. (2011) Low velocity impact on epoxy glass
composites modified with rubber micro-particles and silica
nanoparticles. Proceedings of the SAMPE TECH Conference,
October 18-19, Fort Worth, TX, USA SAMPE Covina, CA,
U.S.A.
[3.4.27] Tate, J., Trevino, E., Gaikward, S., Sprenger, S., Rosas, I.,
[3.5.27] Andrews, M. J. (2013) Low velocity impact on epoxy glass
composites modified with rubber micro-particles and silica
nanoparticles. Journal of Nanoscience, Nanoengineering, and
Applications, vol. 3, issue 1, ISSN: 2231-1777

[3.5.28] (2010) The tensile fatigue behavior of a glass-fiber reinforced
plastic composite using a hybrid-toughened epoxy matrix. J.
Compos. Mat. 44, No 17, 2095–2109
[3.4.33] Manjunatha, C.M., Jagannathan, N., Padmalatha, K. Kinloch,

A.J., Taylor, A.C. (2011) Improved variable-amplitude fatigue
behavior of a glass-fiber-reinforced hybrid-toughened epoxy
composite. J. of Reinforced Plastics & Composites 30 (21),
1783–1793
[3.4.34] Manjunatha, C.M., Bojja, R., Jagannathan, N., Kinloch, A.J.,

Taylor, A.C. (2013) Enhanced fatigue behavior of a glass fiber
reinforced hybrid particles modified epoxy nanocomposite
under WISPERX spectrum load sequence. Int. J. Fatigue
54, 25–31
[3.4.39] Kowalik, T., Hesebeck, O., Flothmeier, K., Koesling, S. (2013)

Zähelastisch modifizierte CFK Harze zur Herstellung von
Strukturbauteilen und im Reparatureinsatz. MicroCar 2013,
Feb. 25-26, Leipzig, Germany
[3.4.40] Pictures courtesy of Fraunhofer IFAM Bremen, Germany
[3.4.42] Zeng, Y., Liu, H.Y., Mai, Y.W., Du, X.S. (2012) Improving
interlaminar fracture toughness of carbon fibre/epoxy
laminates by incorporation of nanoparticles. Composites: Part
B 43, 90–94
[3.4.44] Hunston, D.L., Moulton, R.J., Johnston, N.J., Bascom, W.D.

(1987) Tough Composites. STP 937, ASTM, Philadelphia,
U.S.A., 74–94
176
[3.4.45] Jordan, W.M., Bradley, W.L. (1988) The relationship between
resin mechanical properties and Mode I delamination fracture
toughness. J. Mat. Sci. Lett., 7, 1362–1364
[3.5.2] Estin & Co. (2011) Main dynamics of the composite industry
for automotive applications 2010-2015. Publ. by JEC
Composites, Paris, France
[3.5.4] Ellis, B. (Ed.) (1993) Chemistry and Technology of Epoxy

Resins. Springer Science + Business Media Dordrecht,
Netherlands, 1-142
[3.5.17] Sprenger, S. (2014) Fiber-reinforced composites based on

epoxy resins modified with elastomers and surface-modified
silica nanoparticles. J. Mater. Sci. 49, 2391-2402
[3.5.18] Saxena, A., Hudak, S.J. (1978) Review and extension of

compliance information for common crack growth specimen.
Int. J. of Fracture 14, 453-468
[3.5.19] Xu, S.A., Qang, G.T., Mai, Y.W. (2013) Effect of hybridization
of liquid rubber and nanosilica particles on the morphology,
mechanical properties and fracture toughness of epoxy
composites. J. Mater. Sci. 48, 3546-3556
[3.5.20] Tang, Y., Ye, L., Zhang, Z., Friedrich, K. (2013) Interlaminar
fracture toughness and CAI strength of fibre-reinforced
composites with nanoparticles - a review. Comp. Sci. Tech.
86, 26-37
[3.5.22] Kinloch, A.J., Mohammed, R.D., Taylor, A.C., Sprenger, S.,

Egan, D. (2006) The interlaminar toughness of carbon-fibre
reinforced plastic composites using "hybrid-toughened"
matrices. J. Mater. Sci. 41, 5043-5046
[3.5.23] Zhang, J., Deng, S., Ye, L., Zhang, Z. (2013) Interlaminar
Fracture Toughness and Fatigue Delamination Growth of
CF/EP Composites with Matrices Modified by Nano-silica and
CTBN rubber. Proc. 13th Int. Conf. on Fract., June 16-21,
Beijing, China
177

The Effects of Silica Nanoparticles in Toughened Epoxy Resins and Fiber-Reinforced Composites

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Effects of Silica Nanoparticles in Toughened Epoxy Resins and Fiber-Reinforced Composites

Uploaded by

Copyright:

Available Formats

Sprenger

The Effects of Silica Nanoparticles in Toughened

The Effects of Silica

Hanser Publishers, Munich Hanser Publications, Cincinnati

Distributed in North and South America by:

Cataloging-in-Publication Data is on file with the Library of Congress

© 2016 Carl Hanser Verlag, Munich

1.2. Objectives and structure 9

2. State-of-the-art of the science and technology 10

2.1. Manufacturing technologies for fiber-reinforced composites 10

2.1.1. Prepreg methods 11

2.1.2. Filament winding 12

2.1.4. Injection processes via a mold 13

2.1.5. Infusion processes via a vacuum bag 15

2.2. Fiber reinforcements for epoxy resin systems 17

2.2.1. Woven fabrics 17

2.2.2. Unidirectional fabrics 18

2.2.3. Multiaxial fabrics 19

2.2.5. Fiber sizing 20

2.3.1. Epoxy resins 22

2.3.2. Curing agents for epoxy resins 24

2.3.2.1. Amine curing agents 24

2.3.2.2. Anhydride curing agents 24

2.3.3. Epoxy resin systems 26

2.4. Toughening of epoxy resins 28

2.4.1. Differentiation between toughening and increasing

2.4.2. Reactive liquid rubbers 28

2.4.3. Core-shell rubber particles 30

2.4.4. Thermoplastic particles 31

2.4.5. Self-organizing block copolymers 32

2.5. Modification of epoxy resins with SiO2 nanoparticles 35

2.5.1. SiO2 nanoparticles 35

2.5.2. Property improvements of cured bulk resin

2.5.3. Property improvements of cured bulk

2.5.4. Property improvements of fiber-reinforced

3. Results and Discussion 48

3.1. Property improvements of epoxy resins modified with SiO2

3.1.3.1. Amine-cured epoxy resins 51

3.1.3.2. Anhydride-cured epoxy resins 56

3.1.3.3. Comparison between different types of hardeners 60

3.2. Property improvements of epoxy resins modified with SiO2

3.2.3.1. Epoxy resins modified with reactive liquid rubbers

3.2.3.2. Epoxy resins modified with styrene-butadiene rubber

3.2.3.3. Epoxy resins modified with reactive liquid rubbers

3.2.3.4. Epoxy resins modified with silica nanoparticles, cured

3.2.3.5. Epoxy resins modified with core-shell elastomers

3.2.3.6. Epoxy resins modified with core-shell elastomers

3.2.3.7. Short overview of improvements achieved 79

3.2.3.8. Synergy or no synergy? 80

3.3. Property improvements of fiber-reinforced composites based on

3.3.3.1. Glass fiber-reinforced epoxy resin composites 90

3.3.3.2. Carbon fiber-reinforced epoxy resin composites 94

3.3.3.3. Comparision between bulk and laminate property

3.3.3.4. Mechanisms of toughening 99

3.3.3.5. How much silica nanoparticles do you really

3.3.4. Conclusions 101

3.3.5. References 102

3.4. Property improvements of fiber-reinforced composites based on

3.4.1. Abstract 106

3.4.2. Introduction 106

3.4.3. Discussion 110

3.4.3.1. Glass fiber-reinforced epoxy resin composites 111

3.4.3.2. Carbon fiber-reinforced epoxy resin composites 115