photopymerized micro- and nano-composites - les thèses de...

N° d’ordre 2004ISAL0011 Année 2004

Thèse

PHOTOPOLYMERIZED MICRO- AND NANO-

COMPOSITES:

INTERFACE CHEMISTRY AND ITS ROLE ON

INTERFACIAL ADHESION

présentée devant L’Institut National des Sciences Appliquées de Lyon

pour obtenir

le grade de docteur

Ecole doctorale : MATERIAUX DE LYON

Spécialité : MATERIAUX POLYMERES ET COMPOSITES

par Francesca PEDITTO

Soutenue le 20 FEVRIER 2004 devant la Commission d’examen

Jury

PRIOLA Aldo Professeur Directeur GERARD Jean François Professeur Directeur PILATI Francesco Professeur Rapporteur COSTA Giovanna Directeur de Recherche CNR Rapporteur Laboratoire de recherche : LMM-IMP UMR CNRS 5627-INSA Lyon

i

RESUME

Introduction

La recherche et la production industrielle de matériaux composites à base de matrices

polymère ont augmenté rapidement dans les dernières décennies compte tenu des

caractéristiques pouvant être atteintes par ce type de matériaux en comparaison avec des

matériaux traditionnels.

Par ailleurs, parmi les procédés d’élaboration utilisés pour les polymères, la

technologie de polymérisation UV ou ‘UV-curing’ a connu un essor très rapide et

remplace des techniques traditionnelles de cuisson de systèmes réactifs grâce à la

vitesse importante associée au processus, le coût faible et le respect de l'environnement:

puisque celui permet de s’affranchir de la présence de solvants. Aussi, le développement

du procédé de photopolymérisation pour la production de matériaux composites paraît

prometteur.

Ce travail décrira alors la préparation de matériaux composites à matrice époxy

cycloaliphatique et nano-silice (notés nanocomposites) ou fibres de verre (notés

microcomposites) comme agents de renforcement. Des matériaux composites seront

alors élaborés en utilisant la photopolymérisation cationique. Dans une première partie,

une réaction de modification de surface des agents de renforcements inorganiques,

nanoparticules de silice ou fibres de verre, sera mise au point et l’analyse des

mécanismes de greffage sera alors développée. L’emploi des nanoparticules de silice et

de fibres de verre greffées sera l’objet de la préparation de matériaux composites à

matrice époxy photopolymérisable. L’influence sur la réaction de polymérisation aussi

bien que les propriétés des composites obtenues feront l’objet d’une attention

particulière dans une seconde partie du travail.

Dans l'Appendice, les descriptions des techniques expérimentales utilisées dans ce

travail sont rassemblées.

ii

Partie 1 : Photopolymérisation & Matériaux Composites.

Dans la première partie de ce manuscrit sont rassemblées les informations générales

sur les matériaux composites, la technologie de polymérisation UV et son emploi pour

la préparation de composites à matrices polymère.

CHAP. 1 : MATERIAUX COMPOSITES/INTERFACE

1.1 Caractéristiques et propriétés de matériaux composites1-3

Dans la quête continue pour les performances améliorées, les matériaux traditionnels

sont remplacées de plus en plus par les matériaux composites synthétiques faisant appel

à l’association d’une matrice polymère et d’agents de renforcement comme des charges

particulaires (nanométriques ou micrométriques) et des fibres de renfort comme les

fibres de verre ou de carbone.

Les matériaux composites sont constitués de phases chimiquement différentes sur

une échelle microscopique, séparée par une interface distincte. Le composant qui

constitue la phase continue et présent en plus grande quantité est appelé la matrice. Le

deuxième composant est connu sous le nom de la phase renforçante, ou renforcement,

puisque généralement les propriétés mécaniques de cette phase sont supérieures à celles

de la matrice.

Les paramètres géométriques liés à la phase renforçante (facteur de forme, surface

spécifique, etc.) sont essentiels pour déterminer les caractéristiques des matériaux

composites qui en seront issus.

1.2 Interface/interphase: structure et propriétés

Les propriétés de composites également sont contrôlées par les caractéristiques de

l’interface1,3,4, région à deux dimensions (interface) ou plutôt de l’interphase, zone à

trois dimensions (phase intermédiaire aux propriétés spécifiques). Une forte liaison à

l’interface ou via l’interphase entre la matrice et la fibre assure alors le transfert de

charge de la matrice (renforcement) et est une des conditions essentielles pour conduire

à des propriétés de renforcement associées à l’association d’une matrice polymère et

d’un composant de fort comportement élastique comme des fibres ou des charges

particulaires. La résistance à la fracture (choc ou impact, propagation de fissures),

iii

comme d’autres propriétés des matériaux composites (résistance à fatigue et durabilité

hygrothermique), sont également conditionnées par la nature de l’interface et les

interactions/liaisons qui y sont développées. Dans ce contexte, il est très important de

prendre en compte l’élaboration de l’interface ou la génération des interphases lors de la

mise en œuvre des matériaux composites (processing). Intervient alors le concept de

mouillabilité de la surface inorganique de la fibre ou plus généralement du renfort.

Celui-ci définit l’aptitude avec laquelle un liquide s'étendra sur cette surface solide : une

"mouillabilité parfaite" signifiera alors que le liquide (dans ce cas la matrice à l’état

fondu –thermoplastique- ou sous forme d’une système réactif, mélange de monomères)

interagira fortement avec la surface. Si ce critère est respecté lors de l’élaboration est

observé, des liaisons de type physique (van der Waals) ou covalentes pourront alors

s’établir.

Dans le cas des renforts de verre ou de silice, des organosilanes sont utilisés comme

intermédiaires couplants entre les groupements de surface de la surface inorganique et la

matrice polymère en formant des liaisons fortes (liaisons covalentes)5,6, ou ponts

siloxane. Cette réactivité avec la surface inorganique et la matrice polymère par liaisons

covalentes conduit à un continuum moléculaire à l'interface renfort/matrice polymère

mais aussi à la génération d’une interphase à morphologie complexe (zone de nature

organique/inorganique).

CHAP. 2 : PHOTOPOLYMERISATION

2.1 Généralités relatives à la photopolymérisation1-5

La polymérisation UV est défini comme:

TRANSFORMATION RAPIDE DE 100% LIQUIDES REACTIFS SPECIALEMENT

FORMULES, EN SOLIDES PAR L’ACTION DES PHOTONS UV.

Les photons produits par la radiation UV sont absorbés par le site chromophore d'une

molécule; cette molécule produit alors l’ « espèce active » (radicaux ou protons),

conduisant en une transformation rapide (gamme de temps 10-2-1 s) du liquide en

solide.

Une formulation pour polymérisation UV comprend trois composants de base :

1. le photoamorceur, qui absorbe la radiation incidente et produite l'espèce réactive ;

iv

2. l’oligomère fonctionnalisé, structure de base du futur réseau du polymère ;

3. un monomère mono- ou multifonctionnel qui est alors un diluant réactif et sera

incorporé dans le réseau.

Les principaux domaines d’applications industriels dans lesquels la technologie UV

est employée sont: les arts graphique et coatings, les adhésifs, l’électronique, la stéréo-

lithographie, les matériaux composites ou ciments utilisés des applications dentaires.

2.2 Photopolymérisation radicalaire1,2,4

Le mécanisme de la polymérisation radicalaire est représenté dans la Fig. 2.1:

Fig. 2.1: Mécanisme de polymérisation radicalaire.

R est l’espèce active produite par photodécomposition de l'amorceur.

Les classes principales de systèmes réactifs qui peuvent être polymérisés en

polymérisation radicalaire sont: les monomères acrylate et méthacrylate, les systèmes

thiol-ène et les résines polyester insaturées.

2.3 Photopolymérisation cationique6-9

Le mécanisme de la polymérisation cationique est représenté dans la Fig. 2.2:

Fig. 2.2: Mécanisme de polymérisation cationique.

H+ est l’espèce active produite par photodécomposition de l'amorceur.

Les classes les plus intéressantes pour la photopolymérisation cationique sont les

vinyles éthers et les époxydes multifonctionnels puisque très réactifs et communément

disponibles.

R CH2 CHR' R CH2 CHR'

H+ CH2 CHR CH3 CHR+

v

CHAP. 3: PHOTOPOLYMERIZATION ET MATERIAUX COMPOSITES

L'usage de la technologie UV pour mettre en œuvre des matériaux composites n'a pas

été étudié largement puisque la technologie communément employée est la

polymérisation thermique.

Les limites principales4 de la polymérisation UV dans le cas des composites sont :

l'épaisseur des pièces contrairement aux applications pour revêtements.

la transparence du renfort à la radiation UV.

l'influence des traitements de surface comme l’ensimage du renfort qui peuvent

de part les espèces présentes intervenir sur les mécanismes de polymérisation

les propriétés mécaniques limitées obtenues4.

En prenant en compte ces limitations de la technique UV dans la l’élaboration de

matériaux composites, les solutions suivantes sont proposées :

La technologie UV peut être utilisée dans la préparation de composites pour les

applications pour lesquelles les performances notamment mécaniques recherchées sont

peu importantes.

La technologie UV peut être utilisée pour la réparation de structures composites.

Une cuisson thermique peut être associée après polymérisation UV pour

compléter la polymérisation dans les échantillons de forte épaisseur.

vi

Partie 2 : Photopolymérisation & Micro-/Nano- Composites

La deuxième partie de ce manuscrit est consacrée à la présentation et discussion des

données expérimentales obtenues pour des matériaux à renfort fibre de verre unitaire

(microcomposites) et nanoparticules de silice (nanocomposites).

CHAP. 1: MATERIAUX ETUDIES

1.1 Nano-silice pyrogénée 1-3

La nano-silice est constituée par la silice amorphe SiO2 pur dans forme de particules

ayant haute surface spécifique (250 m2/g). L’Aerosil® 200 (fournie par Degussa) est

une silice de pyrogénation obtenue dans un processus hydrothermique à partir du

tétrachlorure du silicium SiCl4 dans une flamme d’oxygène et hydrogène à 1200-

1600°C.

1.2 Fibres de verre4-6

La fibre de verre retenue pour ce travail est une fibre de type "E" (électrique), type

communément utilisé pour des applications de matériaux composites structuraux. Ils

sont basés sur le système CaO-Al2O3-SiO2. Le diamètre de ces fibres est de 18 µm.

1.3 Organosilanes7,8

Les silanes (Fig. 1.1) sont molécules hybrides organiques-inorganiques qui à

l'interface entre la surface minérale de verre et un polymère permettent d’établir des

liaisons covalentes.

X

XX Si R Y

X = hydrolysable functional group (CH3O )

R Y = functional group which can react with the matrix

Fig. 1.1: Structure générale des organosilanes.

Les groupes hydrolysables X conduiront à la formation de silanols pour réagir à la

surface minérale avec les groupements silanol du verre. Les groupements organo-

fonctionnels R-Y sont choisis pour leur réactivité ou compatibilité avec le polymère.

vii

Les organosilanes sont appliqués aux surfaces inorganiques à partir de solutions

aqueuses ou hydro-alcooliques. Leur hydrolyse dans l'eau dépend de la nature du

groupement R-Y9. Néanmoins, cette réaction d’hydrolyse est rapide et peut être

considéré complète en 1-30 minutes (à pH acide de 3-4) Les silanols de l’organosilane

se condensent pour former des oligomères par réaction beaucoup plus lente (heures) et

dépendant de la température (100-110°C). Dans le cas idéal, une monocouche peut être

obtenue sur la surface du minéral par condensation silanol de surface-silanol des

espèces hydrolysées. Expérimentalement, une structure de plusieurs couches non

complètement condensées est obtenue en surface. La réaction de greffage est

schématisée dans la Fig. 1.2.

X Si

X

X

R YH2O

Si

OH

OH

HO R Y

Hydrolysis

Condensation

SiOH

SiOH

SiOH Si

OH

OH

HO R Y

Si OSi O Si R Y

Si O

3 H2Oinorganicsurface

inorganicsurface

Fig. 1.2: Hydrolyse et condensation d’une molécule d’organo-silane sur une surface

inorganique.

Plusieurs types d’organosilanes ont été utilisés dans ce travail:

Epoxycyclohexyl-éthyle triméthoxy silane (noté CETS, fourni par WITCO) pour

modifier les surfaces inorganiques de silice ou de verre E pour les composites à matrice

à base de monomère CE, diépoxycycloaliphatique.

Glycidoxypropyl triméthoxy silane (noté GPTS, fourni par Aldrich) pour


à base de monomère DGE.

viii

Triméthoxysilyl Propyl-méthacrylate (noté MÉMO, fourni par Aldrich) pour


à base d’oligomère SOA.

n-propyl triméthoxysilane (noté C3, fourni par Petrarch System Inc.) a été utilisé

pour modifier des surfaces inorganiques de verre E ou de silice, par greffage de ligands

hydrophobes non réactifs.

1.4 Systèmes thermodurcissables (photopolymérisables)

Divers de types de monomères ou d’oligomères fonctionnels ont été retenus pour ce

travail afin de les combiner avec les surfaces de silice ou de verre E fonctionnalisées par

les organosilanes fonctionnels précédemment présentés.

3,4-époxycyclohexyl méthyl-3',4'-époxycyclohexane carboxylate (noté CE,

fourni par DOW Corp.).

1,4-cyclohexane diméthanol diglycidyl éther (noté DGE, fourni par Aldrich).

Huile soja époxydée et acrylate (notée SOA, fourni par Aldrich).

1.5 Photoamorceurs

Compte tenu des différents monomères retenus pour cette étude, deux types de

photoamorceurs ont été utilisés:

Triphénylsulphonium hexafluoroantimonate (fourni par DOW Corp.) a été

utilisé comme photoamorceur cationique pour polymériser les monomères CE et DGE.

SOA a été polymérise par mécanisme radicalaire, en utilisant le 2-hydroxy-2-

méthyl-1-phényl-propane-1-one (fourni par Ciba Specialty Chem.) comme

photoamorceur radicalaire.

CHAP. 2: MODIFICATION DE SURFACES INORGANIQUES PAR

ORGANOSILANES

2.1 Introduction

La modification de surfaces inorganiques de silice ou de verre E conduira à une

meilleure adhésion interfaciale et ainsi des propriétés améliorées aux matériaux

ix

composites. Il autorise former des liaisons entre les deux phases qui ont une structure

différente et augmenter la compatibilité des systèmes.

2.2 Protocole expérimental adopté pour les surfaces de nanosilice et des fibres

de verre E

La modification de surface inorganique a été effectuée sur silice sous forme de

poudre et sur les fibres de verre E non ensimées.

La procédure de greffage adoptée pour nos systèmes peut être résumée comme suit:

Matériaux:

Nano-silice = 2g

Silane, CETS ou GPTS (pour 2g de silice) = 1mL

Solvant de réaction, eau distillée = 100 mL

Détails expérimentaux:

pH = 4 (CH3COOH)

T = température ambiante, 25°C

Temps = 2 heures

Agitation par ultrasons pendant 10 min

Addition de la silice et agitation US pendant 2 heures

Filtration

Réaction de condensation dans un four à 120°C pour 4 hrs

Lavage avec eau distillée

Séchage à 120°C pendant 2 heures.

Des analyses par analyse thermogravimétrique (ATG) ont été faites pour caractériser

la surface de la silice modifiée. Des expériences de répétabilité ont été faites pour la

silice greffée avec l’organosilane CETS mettent en évidence une très bonne

reproductibilité de l'expérience de greffage (Fig. 2.1).

x

Fig. 2.1: Courbes ATG de silices greffées avec le CETS (1% v/v).

La modification de la surface a été détectée en mesurant les angles de contact (i/ sur

des surfaces de wafers de silicium oxydé et traitée/non traitée, utilisées comme modèles

pour la silice nanométrique, ii/ fibres de verre et iii/ plaques en verre ‘float’) avant et

après traitement de la surface. Les résultats (Tab. 2.1) mettent clairement en évidence la

modification de surface par greffage.

Tab. 2.1: Angles de contact (avancée et retrait –en degrés-) avec eau obtenus sur les

fibres de verre E et plaques de verre float.

ϑADV ϑREC

fibre de verre non traitée 35 ± 10 28.8± 10

fibre de verre traitée CETS 85.5 ± 12 57.7 ± 12

fibre de verre traitée GPTS 83.1± 5 56.9 ± 5

Plaque de verre float non traité 29.5 ± 2 -

Plaque de verre float traité CETS 95.2 ± 1 67.1± 1

Plaque de verre float traité GPTS 79.9 ± 3 52.9 ± 3

Plaque de verre float traité C3 95.4 ± 1 65.2 ± 1

96

97

98

99

100

0 200 400 600 800 1000

Temperature (°C)

%

xi

TOPOGRAPHIE

Des observations par Microscopie à Force Atomique AFM ont été exécutées sur les

surfaces de wafer de silicium oxydé et traitée/non traité et sur la surface des fibres de

verre E traitées/non traitées. Sur les surfaces inorganiques sont clairement visibles (Fig.

2.3 et Fig. 2.5) les agglomérats décrits dans littérature6 comme "îlots de silane", associés

à la réaction de greffage et à une modification non homogène de la surface. D’autres

évidences expérimentales de changements dans la morphologie de la surface, après

réaction de greffage, sont mises en évidence par analyses en microscopie électronique

MEB sur les fibres de verre.

Fig. 2.3 : Image AFM de la surface d’un wafer de silicium oxydé, non traité

(référence)

Fig. 2.5 : Image AFM de la surface d’un wafer de silicium oxydé et traité avec

l’organosilane CETS 1% v/v.

xii

CHAP. 3: POLYMERISATION UV EN PRÉSENCE DE NANOSILICE OU

FIBRES.

3.1 Introduction

L'influence de nanosilice (développant une grande surface spécifique et donc une

grande quantité d’intreface) et de fibres de verre sur la polymérisation UV, en termes de

cinétique de polymérisation et de conversions finales, a été étudiée.

3.2 Partie expérimentale

Les résultats rapportés indiquent que la silice interagit en surface avec le

photoamorceur cationique (Tab. 3.1).

Tab. 3.1 : Absorbance UV sur les systèmes photoamorceur/nanosilice.

* Ph3S+SbF6-, 5.56 10-5 M dans propylène carbonate.

Le photoamorceur adsorbé pourrait donc avoir une activité inférieure sous irradiation

UV. Dans ce cadre, nous pouvons expliquer les ralentissements observés de la cinétique

du photopolymérisation du système réactif CE en présence de nanosilice (Fig. 3.1,

Cinétiques par spectroscopie FT-IR et Fig. 3.2 Enregistrements par photo-calorimétrie

DSC).

Specimen Abs310 nm

photoamorceur * 0.713

photoamorceur * ajouté de 5% w/w de silice non traité ≈ 0

photoamorceur * ajouté de 5% w/w de silice traité (CETS) ≈ 0

photoamorceur * ajouté de 5% w/w de silice traité (GPTS) ≈ 0

xiii

0102030405060708090

0 20 40 60 80 100time (sec.)

% c

onv.

CE

CE + 10% wttreated silica

CE + 10% wtuntreatedsilica

Fig. 3.1: Cinétiques de polymérisation obtenues par spectroscopie FT-IR pour le

système réactif CE en présence silice traitée/non traitée par l’organosilane CETS

(I = 51 mW/cm2).

150170190210230250270290310

0 5 10 15 20

% silica

-del

taH

(J/g

)

CE + untreated silicaCE + treated silica

Fig. 3.2: Enthalpie de réaction ∆H de la reaction de photopolymérisation en fonction

de la fraction massique de nanosilice traitée CETS et non traitée dans le système réactif

CE.

xiv

CHAP. 4: MICRO-COMPOSITES ET NANO-COMPOSITES

PHOTOPOLYMERISES

Les nano-composites ont été étudiés par analyse thermomécanique dynamique

(DMTA)1 compte tenu de la grande sensibilité de cette méthode dans la zone de

transition principale (vitreuse).

Les résultats expérimentaux DMTA (Tab. 4.1) mettent en évidence une diminution

de la température de transition principale, Tα , associée à Tg quand la silice est présente

Ces résultats sont en accord avec les données cinétiques déjà rapportées, qui montrent

une réduction de la conversion des groupements époxy et donc de la densité de

réticulation de la matrice en présence de silice.

Tab. 4.1: Tα des systèmes photopolymérisés en présence de nanosilice.

Les microcomposites, constitués à base d’une fibre unique sur laquelle a été formée

une microgoutte de matrice, ont alors été étudiés en utilisant la technique2-5 de la

microgoutte.

Il est démontré (Fig. 4.1) qu'un post-traitement thermique des microcomposites

augmente l'adhésion interfaciale quand les fibres de verre utilisées sont traitées avec une

organosilane fonctionnel. Ce résultat suggère que le cycle thermique auquel est traité

l'échantillon est responsable de réactions (pontages)9,10 à l’interface verre-polymère

matrice.

Echantillon Tα matrice Tα matrice + 10% silice non traitée

CE 214°C 182°C

DGE 53°C 37°C

xv

1,5

2

2,5

3

3,5

4

4,5

5

0 0,1 0,2 0,3 0,4 0,5% CETS

IFSS

t=0

t=4 (80°C)

t=4 (80°C+95%RH)t=4 (80°C)+(80°C+95%RH)

Fig. 4.1: Contrainte de cisaillement interfacial moyenne (IFSS) mesurée sur

microcomposites matrice CE /fibre de verre E traitée CETS après 4 jours de différents

types de vieillissement thermique et hydrothermique.

Des mesures d’adhérence (méthode « cross-cut » ASTM D3359) ont été exécutées

sur les plaques en verre float traitées par les organosilanes/non traitées, utilisées comme

modèles de systèmes des fibres de verre. Les résultats, bien qu'ils soient exécutés en

conditions très différentes, sont en accord avec ceux obtenus en utilisant la méthode de

déchaussement de la microgoutte.

CHAP. 5: CONCLUSIONS

Dans ce travail, la préparation de composites à matrice polymère par

photopolymérisation cationique a été étudiée ainsi que les propriétés des matériaux

obtenus.

La réaction de greffage pour modifier les propriétés de la surface des renforts

inorganiques (nanosilice et fibres de verre E) a été mise au point et optimisée afin

d’améliorer les interactions développées à l’interface avec la matrice polymère. Son

efficacité a été confirmée par analyses gravimétriques ATG et évaluation des propriétés

de surface (énergie de surface).

L'influence des agents de couplage et des espèces greffées en surface de renfort sur la

réaction de polymérisation UV a été analysée en évaluant la cinétique et les conversions

xvi

finales. Il a été mis en évidence des interactions entre la nanosilice et les espèces actives

photopolymérisables pendant la réaction UV de part la grande surface développée par ce

type de nanocharge.

Les caractéristiques des zones interfaciales ont été étudiées en utilisant des mesures

d’adhésion conduites avec la technique de la microgoutte (déchaussement). Les résultats

expérimentaux obtenus montrent la présence d'une corrélation entre adhésion et

épaisseur de l'interphase. Comme conséquence du traitement appliqué, l'adhésion entre

les deux phases conduit à des propriétés mécaniques améliorées et constantes dans des

conditions de vieillissement hydrolytique.

Des perspectives à cette recherche s’ouvrent alors:

Approfondir l'étude des interactions entre les espèces actives en

photopolymérisation et la surface inorganique pour les contrôler ou réduire le

ralentissement de la cinétique de réaction radicalaire.

Les résultats des mesures d’adhésion/adhérence suggèrent qu’étudier les

réactions aux interfaces (dans l'interphase) quand un traitement thermique est appliqué à

la suite de la polymérisation UV afin de mieux comprendre l’implication de l'interphase

et optimiser la procédure de polymérisation.

PhD thesis in Material Science and Technology

Francesca Peditto

PHOTOPOLYMERIZED MICRO- AND

NANO-COMPOSITES:

INTERFACE CHEMISTRY AND ITS ROLE

ON INTERFACIAL ADHESION

Prof. Aldo Priola Politecnico di Torino

Prof. Jean François Gerard INSA, Lyon

«La science est infaillible; mais les savants se trompent toujours.»

Anatole France

Ringraziamenti/Remerciements

Desidero ringraziare il Prof. Aldo Priola e tutto il Gruppo Polimeri del Politecnico di

Torino per avermi dato l’opportunità di intraprendere il cammino del dottorato di

ricerca; ringrazio inoltre tutte le persone che ho conosciuto in questi tre anni e che mi

sono state vicine nei momenti di lavoro e in quelli di svago.

Je veux remercier Monsieur le Professeur Jean François Gérard et les autres

permanents du laboratoire des Matériaux Macromoléculaires de l’INSA de Lyon pour

m’avoir accueillie dans son laboratoire, encouragée et conseillée chaque jour de ma

présence à Lyon; les thésards (chaque un de vous!), les techniciens (Nat et Hervé) et les

secrétaires (Isa et Mallou) que j’ai connu et avec lesquelles j’ai partagé boulot et

divertissement.

J’exprime ma profonde reconnaissance à le Dr. Alain Roche, qui m’a suivi avec

patience pendant tous mon travail et surtout qui m’a appris comme travailler avec

rigueur et précision : ses enseignements resteront avec moi n’importe quel chemin je

vais poursuivre.

Merci à tout le monde, l’expérience que j’ai fait chez vous restera toujours avec moi!

Questa tesi è dedicata al protochimico.

Index

Photopolymerized micro- and nano-composites:

interface chemistry and its role on interfacial adhesion

Introduction

Part 1- Photopolymerization & composite materials – Introduction

Chap. 1 COMPOSITE MATERIALS – INTERFACE…………….………….......1

1.1 “Composite material”: characteristics and properties……………..…………….1

1.2 “Interface/interphase”: structure and properties………………………………….8

Chap. 2 PHOTOPOLYMERIZATION.…………………………………………...16

2.1 Description of photopolymerization; applications………………………………16

2.2 Radical photopolymerization………………………….………………………...23

2.3 Cationic photopolymerization…………….……………………………..………29

2.4 Why using cationic photopolymerization?...........................................................38

Chap. 3 PHOTOPOLYMERIZATION OF COMPOSITE MATERIALS……........39

3.1 State-of-the-art of photocuring for composite materials…………….….……….39

3.2 UV polymerization limits in the case of composite materials…………………..40

3.3 Solutions reported in literature…………………………………………………..43

Part 2 – Photopolymerized NANO- and MICRO-composites –

Experimental

Chap. 1 MATERIALS……………………………………………………………...44

1.1 Silica nanoparticles-fumed silica……………..………………………………...44

1.2 Glass fibres……………………………………………………………………..47

1.3 Organosilanes…………………………………………………………………...50

1.4 Thermoset matrices…………………..…………………………………………54

1.5 Photoinitiators…………………………………………………………………..56

1.6 UV-lamps and reactive formulations selected..………………………………...57

Chap. 2 MODIFICATION OF INORGANIC SURFACES BY

ORGANOSILANES……………………………………………………………………59

2.1 Introduction……………………………………………………………………..59

2.2 Experimental: protocols for nanosilica and glass fibres………………………..59

2.3 Conclusions……………………………………………………………………..76

Chap. 3 UV-POLYMERIZATION IN THE PRESENCE OF

NANOFILLERS……………………..…………………………………………………77

3.1 Introduction……………………………………………………………………..77

3.2 Experimental……………………………………………………………………77

3.3 Reaction kinetics: Results and discussion………………...…..………………...79

3.4 Conclusions……………………………………………………………………..86

Chap. 4 UV-POLYMERIZED MICRO- AND NANO-COMPOSITES…………...87

4.1 Introduction……………………………………………………………………..87

4.2 Experimental……………………………………………………………………87

4.3 Results and discussion………………………………………………………….91

4.4 Conclusions……………………………………………………………………107

Chap. 5 CONCLUSIONS…………………………………………………………109

Experimental Techniques…………………………………………………...111

Spectroscopic analysis: FT-IR, UV-vis…………….……………………………...111

Surface analysis: dynamic contact angle, Chan balance…………………………...112

Calorimetric techniques: TGA, DSC, photo-DSC, DMTA………………………..115

Microscopy: AFM, SEM…………………………………………………………..119

Interface mechanical analysis (microbond technique).………….…………………122

Others: UV lamp devices…………………………………………………………..123

REFERENCES………….………………………………………………………...124

INTRODUCTION

The research on polymer-based composite materials is still a growing sector,

regarding both the improvements on existing products as well as the developments of

new ones.

The interest and the industrial production of composites based on polymeric matrices

have been rapidly increased thanks to their peculiar qualities, with respect to the

traditional materials, such as: low weight, low cost, ease of production, and very

specific application fields.

Among the various production processes used for polymers, the UV-curing

technology is growing rapidly even outside of its classical applications (ex. the coating

industry), and somewhere it is replacing the traditional curing techniques thanks to its

process speed, low costs and environmental friendly character.

From this point of view the development of UV-curing processes for the production

of polymer-based composites looks promising.

In this work the preparation of epoxy matrix-based composites with nanosilica and

glass fibers as reinforcing agents through cationic photopolymerization was studied.

First was set up a reaction to modify the surface of the inorganic fillers, which will be

used in the preparation of polymeric composites. The influence of surface treatment on

the curing process as well as the composite final properties was investigated.

INTRODUCTION

Photopolymerized micro- and nano-composites: interface chemistry and its role on interfacial adhesion

II

The manuscript structure is presented below, with evidences of the content of each

chapter.

In the first part of this work general information are collected on composite

materials, UV-curing technology, and reactions as well as its use for the processing of

polymer-based composites.

In Chap. 1, polymer-based nano- and micro-composites are presented and their

specific properties are described and related to the ones of traditional materials.

Examples of their main applications fields are reported. In this chapter is also described

the specificity of composites materials, i.e. the interphase region, and the principal

theories regarding its formation and properties.

In Chap. 2, the UV-curing technique and its application fields are described. The

reaction mechanisms of radical and cationic photopolymerization are analyzed and

commented. The UV-curing technique is then compared to the traditional thermal

curing process, explaining differences, advantages, and disadvantages of the two

processes; evidences of the advantages in using cationic photopolymerization are given.

In Chap. 3 are collected the information found in literature regarding the application

of the UV-curing technology in the preparation of composites, evidencing that it is still

a quite unexplored field.

The second part of this work is dedicated to the presentation and comment of the

obtained experimental data.

In Chap. 1 are presented and described, from a physic-chemical point of view, all the

materials used, i.e. the inorganic reinforcing agents, the silane coupling agents used to

modify their surface, the monomers and the photoinitiators.

In Chap. 2 the set up of the experimental grafting procedure is described: the

inorganic surfaces have been modified to improve their compatibility with the selected

polymeric matrices; afterwards they were characterized by various means. In particular,

the atomic force microscopy, AFM, and scanning electron microscopy, SEM,

measurements done at the INSA-Lyon Laboratory (LMM IMP UMR CNRS 5627) as a

part of the cooperation between INSA-Lyon and Politecnico di Torino are reported.

INTRODUCTION


III

In Chap. 3, the influence of the reinforcing agents on the photopolymerization

kinetics was investigated by using also “real time” techniques (photo-calorimetric

technique).

In Chap. 4, the properties of the obtained composites are analyzed. Particular

attention is given to the adhesion measurements which allow evaluating the interface

properties and their changes when the composite is exposed to a hostile environment

(hydro-thermal ageing).

In the Appendix the descriptions and main characteristics of the experimental

techniques used in this work are collected.

PART 1

PHOTOPOLYMERIZATION

AND

COMPOSITE MATERIALS

INTRODUCTION

CHAP. 1 COMPOSITE MATERIALS-INTERFACE

1.1 “Composite materials”: characteristics and properties 1-3

In the continuing quest for improved performances, traditional materials are more

and more replaced by composite materials.

Composites are known from ages, for example concrete (a mixture of stones held

together by cement) is a familiar material for building; other examples are wood and

bone, as natural composites.

During the last 40 years the production of synthetic composites has been rapidly

increased. The spur of this rapid expansion over the last few decades was the

development in the UK of carbon fibers and in the USA of boron fibers in the early

1960s. These new fibers, with high elastic constants, gave a significant increase in the

stiffness of composites and hence made possible a wide range of applications. One of

the key factors was the very high modulus-to-weight and stiffness-to-weight ratio

presented by these composites.

In Fig. 1.1 the importance of the various classes of materials used in engineering is

illustrated: composites are present besides the traditional materials.

A composite is defined as a material having two or more distinct constituents or

phases; they have to be present in reasonable proportions (greater than 5%) and they

must have different properties, hence the composite properties are noticeably different



2

from the properties of the phases; lastly a composite is usually produced by intimately

mixing and combining the constituents by various means.

Composites are made by chemically different phases on a microscopic scale,

separated by a distinct interface. The constituent that is continuous and is often, but not

always, present in the greater quantity is termed matrix. The second constituent is

referred to as the reinforced phase, or reinforcement, as it enhances the mechanical

properties of the matrix. In most cases, the reinforcement is stiffer than the matrix.

Fig. 1.1: Relative importance of the four classes of materials (ceramics,

composites, polymers and metals) in mechanical and civil engineering as a function

of time1.

Geometry of reinforcement phase, which has at least one of the dimensions less than

500 µm, is one of the major parameters in determining the effectiveness of the

reinforcement.

Usually it is possible to describe reinforcement as being fibrous or particulate and its

arrangement may be random or with a preferred orientation.

A fibrous reinforcement is characterized by its length being much greater than its

cross-sectional dimension. When fibers are used, matrix properties are chosen to be

complementary to the properties of fibers: for example great toughness in a matrix

complements the tensile strength of the fibers. The resulting combination may then

achieve high strength and stiffness (due to the fibers) and resistance to crack



3

propagation (due to the interaction between fibers and matrix). Strong fibers have also

the great advantage of restraining cracking in what are called brittle matrices.

The most common materials used as matrices in composites preparation are

polymers. Their mechanical properties are generally inadequate for many structural

purposes, so it is a benefit to have reinforced polymers. Beside that, processing of

polymer matrix composites (PMCs) does not require high temperatures or high

pressures, so problems associated with degradation of the reinforcement during

manufacture are less important compared with the fabrication of metal and ceramic

matrix composites, hence reinforcements with low temperature capabilities (organic and

glass fibers) may be used. The equipment needed is simpler than in the case of other

matrices.

PMCs were used at the beginning during World War II and started to diffuse widely

immediately after. Today PMCs are used in many fields, as it can see from Tab. 1.1.

Tab. 1.1: Main industrial applications of composite materials.

Industrial sector Examples

Aerospace wings, fuselage, radomes, antennae, tail-planes, helicopter, blades, landing gears, seats, floors, interior panels, fuel tanks, rocket motors cases, nose cones, launch tubes

Automobile body panels, cabs, spoilers, consoles, instrument panels, lamp-housings, bumpers, leaf springs, drive shafts, gears, bearings

Boats hulls, decks, masts, engine shrouds, interior panels

Chemical pipes, tanks, pressure wessels, hoppers, valves, pumps, impellers

Domestic interior and exterior panels, chairs, tables, baths, shower units, ladders

Electrical panels, housings, switchgear, insulators, connectors

Leisure motor homes, caravans, trailers, golf clubs, racquets, protective helmets, skis, archery bows, surfboards, fishing rods, canoes, pools, diving boards, playground equipment.



4

Among the PMCs, the ones with thermoset matrices developed first, due to their

quite simple production. The success of PMCs and in particular of fiber composites

with thermosetting matrices results from the much improved mechanical properties of

the composite compared with the matrix material. PMCs reinforced with glass fibers do

not have clear advantages on conventional materials in term, for example, of elongation

to fracture (lower than for metallic alloy), but their main advantage over metals is linked

to their low density, ρ, particularly when one considers the Young’s modulus per unit

mass E/ρ (specific modulus) and tensile strength per unit mass σ1/ρ (specific strength).

Higher specific modulus and specific strength of PMCs composites means that the

weight of certain components can be reduced, with a significant reduction of transport

costs. Today glass-reinforced polymers are by far the most used composite material in

terms of volume with the exception of concrete. In Tab. 1.2 are presented some

properties of typical PMCs.

The bonds of the links are covalents, as chain bonds (Fig. 1.2). These strong bonds

have the effect of pulling the chains together; the resulting three-dimensional network

gives to thermosets significant advantages over thermoplastics such as greater

dimensional stability, less flow under stress, greater resistance to solvents and a lower

coefficient of thermal expansion. Properties achieved by composites with thermoset

matrix compared with the ones achieved by thermoplastic matrix are presented in Tab.

1.3.

Tab. 1.3: Properties of thermosets and thermoplastics matrices.

Fibrous reinforcements used in thermoset composites are usually coated with a sizing

solution. The nature of the size depends from the chemistry of the matrix. Among the

thermosets, polyester resins dominate the market whereas epoxy resins offer new high-

performances required in advanced composites applications. For example for

Thermosets Thermoplastics Young’s modulus (GPa) 1.3-6.0 1.0-3.8 Tensile strength (MPa) 20-180 40-190

Fracture toughness KIC (MPa m1/2) 0.5-1.0 1.5-6.0

GIC (kJ/m2) 0.02-0.2 0.7-6.5 Max. service temperature (°C) 50-450 25-230

CH

AP. 1

C

OM

POSI

TE M

ATER

IALS

-IN

TERF

ACE

Phot

opol

ymer

ized

mic

ro- a

nd n

ano-

com

posi

tes:

inte

rfac

e ch

emis

try

and

its ro

le o

n in

terf

acia

l ad

hesi

on

5

Tab.

1.2

: Pro

pert

ies o

f pol

ymer

com

posi

tes-

Exam

ples

.

D

ensi

ty

(Mg/

m3 )

You

ng’s

m

odul

us

(GPa

)

Ten

sile

st

reng

th

(MPa

)

Elo

ngat

ion

at b

reak

(%

)

Flex

ural

st

reng

th

(MPa

)

Spec

ific

mod

ulus

[(

GPa

)/( M

g m

3 )]

Spec

ific

stre

ngth

[(

MPa

)/( M

g/m

3 )]

Poly

amid

e 66

+ 4

0%

carb

on fi

ber

1.34

22

24

6 1.

7 41

3 16

18

4

Epo

xy +

70%

gla

ss

fiber

s

Uni

dire

ctio

nal

long

itudi

nal

1.90

42

75

0

1200

22

39

5

Uni

dire

ctio

nal

trans

vers

e 1.

90

12

50

6 26

Epo

xy +

60%

ar

amid

e fib

ers

1.40

77

18

00

55

1286

Poly

ethe

r im

ide

+ 52

% a

ram

ide

fiber

s

54

253

Poly

este

r +

glas

s 1.

50

7.7

95

17

0 5

63

Poly

este

r +

50%

gl

ass f

iber

s

unid

irect

iona

l lo

ngitu

dina

l 1.

93

38

750

1.8

20

38

9

unid

irect

iona

l tra

nsve

rse

1.93

10

22

0.

2

5 11



6

aerospace applications hot wet properties, damage tolerance, higher Tg, greater

toughness and environmental resistance are required.

Fig. 1.2: Arrangements of polymer chains: (a) cross-linked; (b) linear; (c) branched.

A comparison of main properties of polyesters and epoxies is presented in Tab. 1.4.

Tab. 1.4: Properties of some polymeric matrices.

Epoxy Polyester Phenolics Polyimides Density (Mg/m3) 1.1-1.4 1.1-1.5 1.3 1.2-1.9

Young’s modulus (GPa) 2.1-6.0 1.3-4.5 4.4 3-3.1 Tensile strength (MPa) 35-90 45-85 50-60 80-190

Fracture toughness KIC (MPa m1/2) 0.6-1.0 0.5

GIC (kJ/m2) 0.02 0.3-0.39 Tg (°C) 120-190

Thermal expansion coefficient (10-6K-1)

55-110 100-200 45-100 14-90



7

Glass fibers are the most common reinforcement for PMCs, in particular glass fiber-

reinforced epoxies (GREs) are employed in technological applications.

In Tab. 1.5 the principal characteristics of GREs in comparison with glass fiber-

reinforced polyesters (GRPs) are illustrated.

Tab. 1.5: Main characteristics of GREs and GRPs (fibers content = 50-80% vol.).

Better mechanical properties of GREs are a result of the good strength and stiffness

of the epoxy matrix and the strong bonding of glass fiber to epoxy. In fact, epoxies

adhere efficiently to glass fiber surface compared to than any other thermoset resin

commonly used. This good bonding leads to high interlaminar shear strength, as shown

in Tab. 1.6.

Tab. 1.6: Properties of epoxy composites based on different types of fibers.

source: Dow Chemical Company.

Epoxy Polyesters Density (Mg/m3) 1.6-2.0 1.6-2.0

Tensile modulus (GPa) 30-55 12-40 Flexural modulus (GPa) 10-35

Tensile strength (MPa) 600-1165 140-690

Flexural strength (MPa) 1000-1500 205-690

Compressive strength (MPa) 150-825 140-410 Interlaminar shear (MPa) 30-75

Fiber Strength (MPa) Young’s modulus (GPa)

Density (Mg/m3)

Tensile Compressive

E-glass 1165 490 50 1.99

S-glass 1750 495 60 1.99 Carbon (AS4) 1480 1225 145 1.55

Carbon (HMS) 1275 1020 205 1.63

Aramid 1310 290 85 1.38



8

1.2 “Interface/interphase”: structure and properties

The properties of composites are controlled by the interfacial region1,3,4, a two-

dimensional (interface) or three-dimensional (interphase) area between the reinforcing

agent and the matrix. A good interface bonding to ensure the load transfer from the

matrix to the reinforcement is a primary requirement for the effective use of

reinforcement properties.

Since the load acting on the matrix has to be transferred to the reinforcement via

interface, it is clear that reinforcement and matrix should be strongly bonded one to the

other in order to impart to the composite the high strength and stiffness of

reinforcement.

The fracture behavior is also dependent on the strength of interface: a weak interface

results in low stiffness and strength, but high resistance to fracture, while a strong

interface produces high strength and stiffness, but brittle behavior. Other properties of

composites (resistance to creep, fatigue, and environmental degradation) are affected by

the characteristics of interface.

Interfacial bonding is due to adhesion between reinforcement and matrix that in some

stages of the processing of composite must be in intimate contact. In particular, in these

stages, the matrix is often liquid or in a condition where is capable to flow on

reinforcement surface. In this context, the concept of wettability is very important.

Wettability defines the extent to which a liquid will spread over a solid surface, so

“good wettability” means that the liquid (in this case the matrix) will flow over the

reinforcement covering the entire surface and displacing all the air; this can be obtained

only if the matrix is not too viscous and if wetting results in a decrease of the free

energy of the system.

Considering a thin film of liquid (matrix) on the solid surface (reinforcement): all

surface have an associated energy and the free energy per unit area of the solid-gas,

liquid-gas and solid-liquid interfaces are: γSG, γLG, and γSL respectively.

For an increment of area dA covered by the spreading film, extra energy is required

for the creation of new interface areas (solid-liquid and liquid-gas).

This extra energy is (γSLdA+γLGdA); for a spontaneous spreading we must have



9

dAdAdA SGLGSL γγγ ≤+

or, dividing by dA,

SGLGSL γγγ ≤+

So the spreading coefficient, SC, can be defined as

( )LGSLSGSC γγγ +−=

If SC > 0, wettability is a spontaneous process and if γSG is similar to or less than γLG

wetting will not occur.

In Fig. 1.3 an example of a drop of liquid which has been allowed to reach

equilibrium and has partially wet the solid is illustrated.

Fig. 1.3: A liquid in equilibrium with a solid with a contact angleϑ.

Free energy of an interface is measured in J/m2 and can be shown to be equal to the

surface tension, which is expressed in N/m.

At the equilibrium we have

θγγγ cosLGSLSG +=

ϑ is called contact angle, it is used as a measure of the degree of wettability.

If ϑ = 180°, the drop is spherical with only one point of contact with the solid and no

wetting takes place. If ϑ = 0 we have perfect wetting. For 0° < ϑ < 180°, the degree of



10

wetting increases as ϑ decreases. Often is considered that the liquid does not wet the

solid if ϑ > 90°.

Once the matrix has wet the reinforcement, bonding will occur; the type of bonding

varies from system to system and often more than one bonding mechanism may be

operative at the same time.

It is possible to describe at least four types of different bonding1:

Mechanical Bonding: it is usually present with other types of bond, it is more

effective if the surfaces are rough and it is promoted from contraction of the matrix onto

the reinforcement or when the force is applied parallel to the interface.

Electrostatic Bonding: in this case bond is possible between two surfaces with

different charges, even if it is a short range interaction.

Chemical Bonding: it is formed between chemicals groups on the reinforcement

surface and compatible groups in the matrix; its strength depends on the number and the

type of bonds per unit area.

Reaction or Interdiffusion Bonding: atoms or molecules of the two components may

interdiffuse at interface to give interdiffusion bonding, for example in the case of

polymers it can be due to the intertwining of molecules.

In Fig. 1.4 all the different types of bonding described are presented.

When we speak of interfacial region, normally we consider a region of finite

thickness which is different in composition from both the reinforcing agents and the

matrix: this is what nowadays is called interphase.

Interphase is so defined as a region with finite volume which may possess chemical,

physical, microstructure and mechanical properties that differ from those of the bulk

reinforcing agent and matrix, but its material properties and effective thickness are

largely unknown because of its microscopic or even nanoscopic scale, often buried with

in the composite body.



11

Fig. 1.4: Schematic diagrams of the interfacial bonding mechanism: (a) mechanical

bonding, (b) electrostatic bonding, (c) chemical bonding, (d) chemical bonding as

applied to a silane coupling agent, (e) reaction bonding involving polymers, (f)

interfacial layer formed by interdiffusion.



12

Theories have been proposed to explain the formation of this interfacial region and

the interactions between reinforcing agents and matrix. A popular theory, the Chemical

Bonding Theory5,6, states that a chemical reaction exists between the silane and the

inorganic filler and between the silane and the matrix leading to increased adhesive

bond stability.

The bifunctional silane molecules act as a link between the inorganic surface and the

resin by forming a chemical bond with the surface of the inorganic filler through a

siloxane bridge, while its organofunctional group bonds to the polymer resin. This co-

reactivity with the inorganic surface and the polymer matrix via covalent primary bonds

provides molecular continuity across the interface region of the composite.

The Chemical Bonding Theory explains successfully many phenomena observed for

composites made using silane treated glass fibres. However, a layer of silane agent

usually does not produce an optimum mechanical strength, and there must be other

important mechanisms taking place at the interfacial region. An established view is that

bonding through silane other than simple chemical reactivity is best explained by

interdiffusion and interpenetrating network, IPN, or an hybrid organic-inorganic

material formation as the interphase region7,8.

The fillers coverage by the organosilane is usually equivalent to several monolayers.

The hydrolyzed silane condenses to oligomeric siloxanols that are solubles until they

condense to a rigid cross-linked structure.

Sizing inorganic surfaces with silane leads to the formation of three layers with

different chemical structures and physical properties:

1. Physisorbed region, the outermost layer that consist mainly of bulk of the

deposited silane;

2. Chemisorbed region, the next layer, that possesses a better resistance to

hydrolysis than physisorbed region;

3. Chemically reacted region, the innermost region next to the inorganic surface

that consists of a three-dimensional network of siloxane, which is very stable and

resistant to extraction even by hot water.



13

However contact with the polymer-based film former and matrix is made while

siloxanols still have a certain degree of solubility. In case of thermoset resins, coupling

with silane-treated fillers is something in between the following two cases:

1. oligomeric siloxanol layer may be compatible in the liquid resin and form a true

copolymer during resin cure, i.e. a mixed phase (organic-inorganic).

2. siloxanols and resin cure with a limited amount of co-polymerization, this is the

case of only partial solution compatibility.

So the formation of an interfacial region comes from interdiffusion, taking place in

the coupling agent–polymer resin interface region, due to penetration of the matrix

monomer into the chemisorbed/physisorbed condensed silane layer followed by a

possible migration of the physisorbed condensed silane oligomer molecules into the

matrix. The migration and intermixing of silane with polymer create an interphase of

substantial thickness.

The combination of chemical reaction and IPN theories is of particular importance in

composites based on thermoset matrices.

To characterize the interphase means to quantify the rates of interdiffusion and

chemical reactions between silane and polymer matrix systems. This can be very

difficult because of the already told reduced dimension of this region.

A great number of techniques have been employed in the analysis of surface layers,

based on elemental chemistry, physics, and mechanical means.

As an efficient method, FTIR spectroscopy has been employed since 1970s in order

to detect and analyze complex chemical reactions taking place between the inorganic

filler and the silane agent9-12. It was possible to see the effective presence of a multilayer

on filler surface as well as the formation of interphase between glass fibers sized with

organosilane and bulk matrix via copolymerization. These results indicate that there are

complex reactions at the interface fiber/silane and at the interphase

fiber/silane/matrix7,8.

Other spectroscopic techniques13,14 like ion scattering spectroscopy, ISS, and

secondary ion mass spectroscopy, SIMS, show that the coating on the inorganic surface

consisted of three layers with distinct properties:



14

1. a stiff layer from the free surface to 140 Å, i.e. highly condensed inorganic

network;

2. from 140 Å to 240 Å, a soft oligomeric layer containing incompletely condensed

siloxane;

3. from 240 Å to the inorganic surface, a high molecular weight siloxane layer at

least partially covalently bonded to the surface and different from the two outermost

layers.

These results are in good agreement with the theory of three different chemical

structure (physisorbed, chemisorbed, and chemically-reacted) regions of the silane

treated interface.

More recently, X-ray photoelectron spectroscopy (XPS)14,15 has been used to

characterize glass fiber coatings, showing that the concentration of silicon in the surface

layer further increases due to the presence of size containing silane coupling agents.

Nanoindentation and nanoscratch techniques15 to measure the mechanical properties

of interphases are very useful because of the sub- or near-micron size of the interphase

itself.

These techniques are very powerful for the measure of the effective interphase

thickness and modulus: they both depend from the silane type and concentration. For

example by increasing silane concentration, the effective interphase thickness increases,

as well as interfacial bond strength and composite tensile strength.

In general it is possible to observe that interphase properties are really different from

the bulk matrix: the interphase has a lower Tg, a higher tensile strength and modulus,

and lower fracture toughness.

Nanoscratch tests are also useful to evaluate samples after water-aging: the hard part

of the interphase forms an extended region of reacted silane molecules after aging.

With atomic force microscopy14,15, AFM, it was possible to characterize the surface

of glass fiber before and after treatment with silane: topographic images of treated

surface exhibit a rougher surface, with the characteristic agglomerates of sizing agent,

the “silane islets”, than unsized glass fibers. Besides these qualitative characterizations,



15

AFM makes also possible to evaluate the interphase thickness (for example for the

unsized glassy-epoxy system the thickness was approximated to be 1 µm) and its

ductility when treated.

Anyway, experience has shown that proper characterization of an interphase, for

chemical, physical or mechanical properties, is very difficult because the intephase is on

the microscopic or nanoscopic scale and is buried within the composite body.

Furthermore it is not easy to find out distinct boundaries that would allow the interphase

between the bulk reinforcing agent and matrix to be defined.

These are the reasons for which it is necessary to use techniques of ultra high

magnification and resolution. Nowadays techniques like AFM and nanoindentation tests

provided interesting results from which it is possible to conclude that the formation of

both softer and harder interphase is possible, depending on the combination of the

system applied. Besides the influence of reinforcement, matrix and coupling agent, there

are also environmental factors, as well as thermal, chemical, physical and mechanical

phenomena that can be equally important in the formation of interphase.

CHAP. 2 PHOTOPOLYMERIZATION

2.1 Description of photopolymerization; applications 1-5

The transformation of a reactive liquid into a solid, by UV-radiation, leading to

polymerization and cross-linking is termed photopolymerization or UV-curing.

UV-curing is defined as:

FAST TRANSFORMATION OF 100% REACTIVE, SPECIALLY FORMULATED,

LIQUIDS INTO SOLIDS BY UV PHOTONS.

Photons generated by UV-light are absorbed by the chromophoric site of a molecule

in a single event; this molecule generates radicals or protons, the initiating species that

promote the fast transformation (time range 10-2-1 s) from the liquid into the solid. As a

result of the curing process, a solid polymer network, totally insoluble in the organic

solvents and very resistant to heat and mechanical treatments, is formed from a 100%

reactive liquid.

The entire process is schematized in Fig. 2.1.



17

Fig. 2.1: Schematic representation of photocuring process.

As shown in Fig. 2.1, a UV-curable formulation is made of three basic components:

1. photoinitiator, which absorbs the incident light and readily generates reactive

radicals or ions;

2. functionalized oligomer, which, by polymerizing, will constitute the back-bone of

the three-dimensional polymer network formed;

3. a mono- or multifunctional monomer, which acts as a reactive diluent and will be

incorporated into the network.

The photoinitiator is the key of all process, because it determines both the rate of

initiation and the penetration of the incident light into the sample, governing in this case

also the depth of cure.

Depending on the photoinitiator used, the reactive species generated can be radicals

or ions, so the process can be named radical or cationic photopolymerization. As

described in the following paragraphs, radical and cationic photopolymerizations are

very different not only for the active species that start the reaction, but also for the types

of monomers used and for the cure mechanism and experimental conditions in which

the process is performed. In Fig. 2.2, the differences in the reactive species generated

and the initiation step for the two processes are schematically illustrated.

During the initial part of the reaction, polymerization rate depends on the reactivity

and concentration of the functional group as well as on the viscosity of the matrix

medium. Other important parameters are chemical micro-structure and functionality of

monomers and/or oligomers: they will determine the final degree of polymerization,

physical, and chemical characteristics of the final polymer.

Photoinitiator

UV radiation

Reactive species(radicals or ions)

Multifunctional monomer

Crosslinked polymer



18

Fig. 2.2: Initiation step for radical (I) and cationic (II) photopolymerization.

APPLICATIONS

Nowadays, UV-curing technology is well established in many industrial fields and in

particular applications, it offers new possibilities of development. The principal

industrial use of UV-curing technology is in the coating industry for the surface

protection of all kind of materials, due to high speed process and good energy yield. A

typical industrial line UV processor is made of two parts: the coating machine, where

the UV-curable resin is applied on the substrate, and the UV oven, where the liquid

resin is dried within a fraction of a second by passing under a powerful lamp.

In Fig. 2.3 an industrial processor for the UV-curing of organic coatings is

schematically presented.

Fig. 2.3: UV-curing industrial processor for coatings.

R CH2 CHR' R CH2 CHR'

(I)H+ CH2 CHR CH3 CHR+

(II)



19

Acrylate resins, cured with radical photopolymerization, are the most widely used

UV-systems, with a total annual production of approximately 60000 tons, while

cationic-type resins, cured with cationic photopolymerization, represent a minor part,

i.e. about 2000 tons, but in continuous growth (10-12% per year).

Here are reported the main industrial fields in which UV-curing technology is

employed1,2.

Graphic arts/Coatings

Adhesives

Electronics

Stereolithography

Dental composite materials

Graphic Arts

UV-curing is used both in the pre-press part to produce the printing plate as well as

in the printing process itself, thanks to the development of fast-drying UV-curable inks.

The printing process consists of the rapid transfer through an ink of a given image

from a printing plate to the substrate (usually paper), thus allowing a fast production of

prints.

The main printing processes in which UV-curing is involved are: letterpress, gravure,

flexography, screen printing and lithography. On the other side new UV inks have been

developed. They present a number of advantages over conventional solvent-based inks:

the higher viscosity allows several colours to be applied successively;

their solvent-free formulations lead to a better print definition and high gloss

images;

the UV process is more economic because it requires less energy and

achieves a higher productivity; the entire process is performed at ambient temperature,

without any solvent emission, which makes it a environmental friendly procedure.

Finally, for some specific applications, it is necessary to further improve surface

properties of printed material (ex. gloss, smoothness, and abrasion and scratch

resistances, weathering resistance). This can be achieved by applying a thin layer of a

UV-curable varnish, which is known to give high gloss and smooth surface.



20

Coatings

Coatings are applied to a surface. They can be divided into:

Functional coatings, improving the surface by:

protecting it from abrasion, scratch, mar, chemicals;

providing different properties such as release, slip, adhesion, electrical

conductivity or insulation, antifogging, flame retardance;

acting as a barrier to various liquids or gasses.

Decorative coatings are applied to:

change appearance ( colour, gloss or mat finish, texture);

hide surface (imperfections, electrical circuitry, etc.).

Usually coatings are classified according to the substrate they are applied to:

paper and paperboard

wood

plastics

metal

glass and ceramic

miscellaneous.

This type of employ of UV-curable varnishes is increasingly used to obtain highly

resistant coatings to protect any substrate: wood, plastic, metal, glass, optical fibres,

paper, leather, fabrics, etc.; the film thickness is of the order of 20-100 µm to assure a

long-lasting protection.

Adhesives

Radiation curing has two main areas of application in the field of adhesion:

1. to bond together two parts of a laminate, acting as a quick-setting glue. In this

case the use is limited by the UV-transparency of one of the two parts of the laminate.

The whole process is divided in three steps: applying of the adhesive in the liquid state;

assemblage of the two parts; exposure of the assembly to UV-light.

UV-cured laminates show a great potential because they are produced by a process

that is faster, cheaper and easier to work out than the usual thermal cure carried out for

hours under high pressure.



21

2. To produce pressure-sensitive adhesives and release coatings. It consists in a

rapid photoinitiated crosslinking producing a viscoelastic system with predetermined

properties.

Electronics

Here UV-curable systems have found applications as photoresists in the imaging

step, fast drying adhesive and conformal coatings.

Stereolithography

This new technology is based mainly on the capability of UV-curable systems to give

three-dimensional solid objects, by scanning the surface of a resin with a laser to form a

thin solid pattern, and building up the model step-by-step by adding one layer on top of

another. Complex parts can be obtained faster, with great precision, and more flexible

processing than with conventional modelling techniques. Besides it allows the direct use

of digital design information to guide the formation of a model that closely represents

the original design.

Dental Composite Materials

Adding mineral fillers such as glass or silica particles to UV formulations is possible

to obtain extremely hard and abrasion-resistant composite materials. These types of

resins present a number of advantages over conventional systems: immediate readiness

for use, extended working time, higher polymerization rate, and short setting time,

better adhesion of the filler particles to the matrix.

The curing of these systems has to be performed at visible light and it is necessary to

take into account that inert filler can be up to 60% in volume, so the penetration of light

in these composite resins is limited therefore it has to be carried a multiple step process.

PRINCIPAL ADVANTAGES/DISADVANTAGES1,2

The main advantages of UV-curing technique are better understood if compared with

the traditional thermal-curing polymerization (Tab. 2.1).



22

Tab. 2.1: Comparison of UV and thermal curing1.

Parameter UV Thermal

Commercial

Capital cost + -

Operational cost + -

Formulation cost - +

Floor space + -

Cure speed + -

Skill level required 0 +

Environmental

No solvent release + -

Energy consumption + -

Technical

Chemical resistance + -

Formulation variety 0 +

Curing of pigmented films - +

No substrate damage + 0

Low cure temperature 0 -

Sensitivity to oxygen + +

Health & safety

Fire hazard + -

Radiation hazard 0 +

Irritant raw materials - +

+ = advantage - = disadvantage 0 = intermediate



23

Arguments in favour of the replacement of thermal curing by UV-curing are mainly

lower capital and running costs, lower floor space requirements, higher running speeds,

less substrate heating, the high quality of the cured coating or ink, no solvent release

during curing and the development of new curable formulations having less or no skin

irritant raw materials.

There are also economic and ecological factors that encourage the continuous growth

of radiation curing technology such as:

Raw materials containing a low amount of volatiles, less or no skin irritant, have

been developed and increase the range of formulation variety.

Low-viscosity monomer-free oligomers and water reducible oligomers can be

used in spray coating applications.

New applications in metal and glass coatings are possible thanks to oligomers

that adhere well to critical substrates.

Weather resistant products are available for outdoor applications.

More reactive photoinitiators allow lower concentrations in formulations or less

powerful UV sources to be used.

Photoinitiator-free UV-curable systems appear on the market.

New monochromatic UV-sources were introduced.

On the other side thermal curing still holds a strong position due to the advantage in

formulation costs and variety, the avoidance of radiation and the lower skill level

required. Moreover the thickness of the sample that can be photocured is normally very

thin if compared to a thermal cured one. Mainly for this reason UV-cure technology is

still not widespread in the composites industry.

2.2 Radical photopolymerization1,2,4

The radical polymerization mechanism can be schematically represented in Fig. 2.4:

Fig. 2.4: Radical polymerization mechanism. R CH2 CHR' R CH2 CHR'



24

R is the active specie generated by photodecomposition of the initiator.

It should be pointed out that it is only the initiation step, radical formation from the

photoinitiator, which is different from thermal polymerization.

Radical photoinitiators can be divided into two groups according to the way the

active species are generated2:

1. by photocleavage, if radicals are generated by a intramolecular scission ;

2. by hydrogen abstraction, if radicals are generated by the abstraction of an atom of

hydrogen from a donor molecule.

In Fig. 2.5 are illustrated the two ways of radicals’ generation.

A-B* → A• + B• A* + RH → AH• + R• homolitic cleavage hydrogen abstraction

Fig. 2.5: Mechanism of radicals’ generation in radical photopolymerization.

1. Photocleavage: in this class we found aromatic carbonyl compounds that

undergo to homolytic C-C bond scission upon UV exposure, with the formation of two

radical fragments; the benzoyl radical was shown to be the major initiating species.

The process is schematized in Fig. 2.6; examples of photoinitiators belong to this

class are: benzoin ethers derivatives, benzilketals, hydroxyalkylphenones, α-amino

ketones, and acylphosphine oxides.

Fig. 2.6: Radical formation reaction for aromatic carbonyl compounds.

2. Hydrogen abstraction: this is a typical reaction of some aromatic ketones, like

benzophenone, thioxanthone, or camphorquinone. Under UV irradiation, they do not

undergo fragmentation, but abstract a hydrogen atom from an H-donor molecule to

generate a ketyl radical and the donor radical.

C

O

C Xhv

O

C XC



25

The process is schematized in Fig. 2.7:

O

Cvh

C

O

*

RHC

OH

R

Fig. 2.7: Radical formation reaction for aromatic ketones.

In this case initiation of polymerization occurs through the H-donor radical. The

most frequently used H-donor molecules are tertiary amines, because of the high

reactivity of the α-amino alkyl radical towards the double bond, as shown in Fig. 2.8:

Fig. 2.8: Radical formation reaction in case of tertiary amine used as co-initiator.

This latter class of photoinitiators have also the advantage of reducing the inhibition

effect of oxygen because they promote a peroxidation mechanism that consumes the

oxygen present in the monomer.

In Fig. 2.9 are listed the principal classes of radical photoinitiators.

C

O

C

OR

R'

benzoin derivatives

benzil ketals C

O

C

OR

OR'

hydroxyalkylphenone C

O

C

R

OH

R'

acylphosphine oxides C

O

P

O

benzophenone derivatives C

O

thioxanthone derivativesC

S

O

Fig. 2.9: Radical photoinitiators commonly used.

Ar2C O N CH2hv CHNAr2C OH

CH2 CHN CH CH2



26

The main classes of resins that can be cured with radical system are: acrylate and

methacrylate monomers, thiol-ene systems, and unsatured polyester resins.

Acrylate and methacrylate monomers are by far the most used in industry because

they are very reactive and can be used to create a large variety of crosslinked polymers

with tailor-made properties. Their polymerization is very fast at the beginning, but

progressively slows down when gelification and vitrification occur; for this reason there

are always some residual unreacted insaturations trapped in the polymer network.

They can be divided into:

functionalized oligomers

mono- or poly-functional monomers

The most important types of functionalized oligomers are:

epoxy acrylic resins

urethane acrylic resins

polyalkylene glycol diacrylates

polyester diacrylates

The most important monomers are:

diethylene glycol diacrylate

hexanediol diacrylate

trimethylolpropane triacrylate.

In Fig. 2.10 is presented the typical reaction scheme for this class of monomers.

Epoxy acrylates are highly reactive and produce hard and chemically resistant

coatings, so they are used in wood finishing applications, varnishes for paper, and

cardboard as well as for hard coatings2,4.

Polyesters acrylates are often applied in wood coatings, varnishes, lithographic and

screen inks.

Methacrylates monomers have similar reactivity to acrylates monomers, but with a

lower propagation rate2,4.



27

C

O

CH2 CH C

O

O C

O

CH2 CH C

O

O

Propagation

C

O

CH2 CH C

O

Omonomer

C

O

CH2 CH

C

CH2 CH CH2 CH

CO

C O

CHCH2 CH2

O

C O

CH CH2CH CH2CH

Termination

Pn Pm PnPm

Pn vitrification

Initiation

Fig. 2.10: Polymerization reaction of acrylates.

The most important advantages of acrylate formulations are high reactivity and

adjustable viscosity. Rapid cure speed and low viscosity combined with brittleness and

poor adhesion are obtained when acrylate monomers are used; acrylate oligomers have

higher viscosity and lower reactivity than monomers, but they guarantee a broad range

of coating property requirements. Therefore radiation curable formulations usually

consist of monomers as reactive thinners and oligomers as binders.

Thiol-ene systems are used in many applications such as coatings, adhesives,

sealants, etc.

Their polymerization reaction can be represented as follows (Fig. 2.11):

Fig. 2.11: Polymerization reaction of thiol-ene systems.

Using multifunctional monomers it is possible to obtain a three-dimensional network

in which connecting chains are made of alternating copolymer. It should be noticed that

OAr2C RSHvh Ar2C OH RS

RS CH2 CH R' RS CH2 CH R'

R'RS CH2 CH RSH RS CH2 CH2 R' RS



28

thiol-ene systems are less sensitive to air inhibition that other radical systems because

peroxy radicals are also capable to extract H from the thiol (Fig. 2.12) forming the thiil

radicals which continue the polymerisation process.

Fig. 2.12: Hydrogen abstraction from thiol molecule.

Unsatured polyester resins are mainly employed in the wood finishing industry; the

radical-initiated crosslinking occurs by direct addition copolymerization of the vinyl

monomer with the unsaturations at the polyester backbone, as shown in Fig. 2.13:

Fig. 2.13: Polymerization of unsatured polyesters.

In Fig. 2.14 the principal classes of radical monomers are listed.

polyester/styrene C

O

CH CH C

O

CH CH2

thiol/ene C(R SH)4 CH2 CH R' CH CH2

acrylates

(CH2 CH C

O

O CH2)3 CH2 CH2 CH3

CH2 CH C

O

O R O C

O

CH CH2

R = polyester, polyether, polyurethane, polysiloxane

Fig. 2.14: Radical monomers commonly used.

RSCHCH2

P PO2RSH

PO2H RS

R O C

O

CH CH C

O

O CH CH2 crosslinked polymer



29

2.3 Cationic photopolymerization6-9

The cationic polymerization mechanism is schematically represented in Fig. 2.15:

Fig. 2.15: Cationic polymerization mechanism.

H+ is the active specie generated by photodecomposition of the initiator.

Photoinitiators for cationic photopolymerization can be divided into three groups:

Aryl diazonium salts

Ferrocenium salts

Diaryliodonium/triarylsulfonium salts.

The latter are named “onium salts” and are nowadays the photoinitiator class most

used in cationic polymerization. They are stable crystalline compounds, readily soluble

in a wide variety of common polar solvents and cationically polymerizable monomers

and absorb strongly in the UV region. In Fig. 2.16, their structure is represented.

Fig. 2.16: General structure of “onium salts”:

diaryliodonium (I) and tryarylsulfonium (II) salt.

Under UV light, they are subjected to photolysis through a quite complex

mechanism. In the case of diaryliodonium salts, one can have photoexcitation of the salt

and after the decay of the resulting excited singlet with heterolytic and homolytic

H+ CH2 CHR CH3 CHR+

I

MtXn

S

MtXn

(I) (II)



30

cleavages of carbon-iodine bond. Free-radicals, cationic and cation-radical fragments

are produced according to the scheme reported in Fig. 2.17.

*

HMtXnArArI MtXnAr2I MtXnhv MtXnAr2I ZH ArI Z

Fig. 2.17: Photolysis of diaryliodonium salt under UV light.

Protonic acids, denoted as HMtXn, derive from the reaction between the aryl cations

and aryliodine cation radicals with solvents, monomers, or impurities. HMtXn is the real

initiator of cationic polymerization, as shown in Fig. 2.18.

nM

MHMtXn H M+ MtXn

H M+ MtXn H (M)nM+ MtXn Fig. 2.18: Initiation mechanism for cationic polymerization.

For triarylsulfonium salts the photolysis is similar, but the heterolytic cleavage is

dominant on homolytic cleavage.

The anion generally indicated as MtX-n must have non-nucleophilic characteristics

because any cationic species generated during photolysis or by addition to a monomer

would give combination with a nucleophilic anion and, as result, retardation or

complete suppression of polymerization reaction. According to their non-

nucleophilicity, the most useful anions are: PF-6, AsF-

6, and SbF-6.

The type of anion determines also the strength of the Brønsted acid generated via

photolysis: bigger anions generate stronger acids, so the reactivity order is:

SbF-6 > AsF-

6 > PF-6 > BF-

4.

In Fig. 2.19 are shown the differences observed changing the anion on the kinetics of

photopolymerization of cyclohexene oxide.



31

Fig. 2.19: Photopolymerization of cyclohexene oxide using 0.02% mol of (C6H6)3S+X- salts6.

The onium salts show a very high degree of thermal stabilty due to their cation part

which is stabilized by the resonance of benzenic rings and by the d-orbital of central

atom. As a result of this stability, they undergo thermal decomposition at very high

temperatures, as shown in Fig. 2.20.

Fig. 2.20: TGA analysis of (C6H6)3S+ AsF-

6 in nitrogen and air during an heating

ramp of a rate of 10 C/min6.

In Fig. 2.21 are summarized the various critical functions that can be assigned to the

cation and anion portion of an onium salt.



32

Fig. 2.21: “Anatomy” of an onium salt photoinitiator.

Studies made on reactive systems using photo-calorimetric technique6, i.e. photo-

DSC, have revealed that there are other parameters controlling the reaction:

Concentration of photoinitiator, for each of them is possible to observe that there is a

specific concentration for which is obtained an optimum cure rate. Further increase in

photoinitiator level does not produce a corresponding increase in the cure rate, possibly

due to the light screening effects by the triarylsulfonium salt itself or its photolysis

products.

UV-light intensity, because the system is limited by the absorption of the

photoinitiator, so it is useless to have very high light intensities. At very low intensities

there appears to be some type of inhibition effect.

Temperature effect, it has been observed that in all cationic systems cure at the

highest temperature the substrate give the highest cure rate, of course this is not always

possible.

CATION

DETERMINES PHOTOCHEMISTRY

λmax

molar absorption coefficient

quantum yield

photosensitization

thermal stability

ANION

DETERMINES POLYMER CHEMISTRY

acid strength

nucleophilicity

anion stability

initiation efficiency

propagation rate constants

MtXnI



33

Water effect, because the presence of water (or other hydroxyl containing impurities)

can change both the rate and the extent of polymerization of epoxy monomers.

Two other classes of cationic photoinitiators have been mentioned above:

Aryldiazonium salts

Ferrocenium salts.

Aryldiazonium salts were the first class of cationic photoinitiators developed in the

1970s. They can be used in the ring opening polymerization of epoxides through the

reaction scheme represented in Fig. 2.22.

Fig. 2.22: Photolysis mechanism of diaryldiazonium salt and cationic polymerization

of an epoxy monomer.

This class of cationic photoinitiators had no success essentially for two reasons:

1. the thermal instability of aryldiazonium salt leads to poor latency so that the

systems spontaneously gelled in few hours even in absence of light.

2. The generation of nitrogen gas as photolysis product leads to film defects.

Ferrocenium salts are a very different class of cationic photoinitiators. They undergo

photolysis to generate an iron-based Lewis acid with the loss of the arene ligand. This

species coordinates to an epoxy monomer to give ring-opening polymerization as shown

in Fig. 2.23.

Ar N2 BF4hv Ar F BF3 N2

O BF3

H2O

O

n



34

R1

Fe X hv XFe

R1

O

R1Fe

O

R1

( )3

X- R1

nO

R1

polymer Fig. 2.23: Photolysis mechanism of ferrocenium salt and cationic polymerization of

an epoxy monomer.

The use of this class of photoinitiator is limited to the monomers that can bond

effectively with the photogenerated coordinatively unsatured ion center.

Cationic photopolymerization is used to cure monomers that are reactive towards

cationic species. In Fig. 2.24, the most important monomers that can be UV-cured in the

cationic way are scheduled. Among all the monomers presented, the most interesting

classes for cationic photopolymerization are multifunctional vinyl ethers and epoxides

because they are very reactive and commonly available.

Fig. 2.24: Polymerizable monomers with cationic photoinitiators.

Cationic Photoinitiators

hv

nCH

R

CH2 O

O

R

nCH2 CH2 S

S

nCH

OR

CH2OR

nN

C O

R

CH2 CH2

N

OR

n(CH2)4 O

O

n(CH2)5 O C

O

O

O

nCH

R

CH2

R

nCH2O CH2O CH2O

O

O O



35

The epoxy monomers can be UV-cured through the opening of the epoxy ring,

catalyzed by the acid species generated by photolysis of the initiator. The reaction

mechanism is presented in Fig. 2.25.

X OCH

CH

R

R'X O

CH

CH

R

R'oxonium ion

monomer

X (O CH

R

CH

R'

)n OCH

CH

R

R'

Fig. 2.25: Polymerization scheme for an epoxy monomer.

In presence of difunctional epoxides UV-curing leads to a crosslinked polymer.

The reactivity of this class of monomers is quite broad, for example monomers

containing the epoxycyclohexane group are much more reactive than glycidyl ethers or

glycidyl esters, due to steric and electronic factors.

Two examples of epoxy monomers commonly used are cyclohexane dimethanol

diglycidylether, denoted DGE and 3,4-epoxycyclohexyl-3’,4’-

epoxycyclohexanecarboxilate, denoted CE. Their structures are given in Fig. 2.26.

Fig. 2.26: Chemical structures of DGE and CE.

O

O

OO

CE

CH2

CH2

O CH2

O CH2

O

O

DGE



36

Vinyl-ethers monomers are the most reactive towards cationic photopolymerization,

giving a three-dimensional polymer network with a low number of residual

insaturations. The high reactivity of these monomers is due to the presence of the

double bond C=C that, with the oxygen atom, stabilizes the cation through the chain

growth (Fig. 2.27).

R CH2 CH OR CH2 CH OCH2 CH OR+

Fig. 2.27: Growth of the polymer chain and its stabilization by resonance.

Even if vinyl ethers are ideally suited for cationic photopolymerization, their use in

industry is limited by their high cost and the hazards of using acetylene under high

pressure during their synthesis.

PRINCIPAL ADVANTAGES/DISADVANTAGES

Effect of oxygen

One of the main advantages of cationic-initiated polymerization, if compared to the

radical induced process, is that the former is not sensitive to oxygen, thus allowing

coatings to be cured rapidly even in the presence of air.

Influence of film thickness

In thin films the photopolymerization develops at the same rate, but as the film

thickness is increased, the propagation rate value, Rp, drops, due to the UV filter effect

of the top layer (Fig. 2.28). Film thickness has a pronounced effect also on the

maximum conversion level. Moreover atmospheric oxygen will diffuse less rapidly in

thick coatings.



37

Fig. 2.28: Influence of the film thickness on the photopolymerization of a

cycloaliphatic diepoxy7.

Post-polymerization

One of the distinct features of cationic photopolymerization, compared with radical-

induced process, is the post-cure phenomena: it consists in a further and not negligible

polymerization taking place once the light has been switched off. Such an important

post polymerization is due to the fact that two cations cannot interact to undergo

coupling or disproportionation, so that the living polymer chain continues to grow in the

dark, until termination occurs by transfer reaction or bimolecular interaction with

another species present in the polymerization mixture (as water, bases, or another

portion of polymer chain).

Fig. 2.29 shows some typical conversion vs. time curves recorded after exposure,

compared to continuous irradiation: post-polymerization is relatively more important in

the early stages of the reaction, but a significant increase of the degree of conversion

could be noticed even after 20 minutes of storage in the dark.



38

Fig. 2.29: Polymerization profiles recorded after UV exposure of various durations for a cycloaliphatic diepoxy7.

2.4 Why using cationic photopolymerization?

The main differences between radical and cationic photopolymerization has been

described and it becomes evident that the cationic UV-curing process offers many

important advantages that are summarized here:

the initiating species is a stable compound only consumed by anions or

nucleophiles;

after UV exposure, cationic polymerization continues for a long time;

since no radicals are involved, cationic photopolymerization is not sensitive to

oxygen;

films made from cationic formulations show low shrinkage and good adhesion.

CHAP. 3 PHOTOPOLYMERIZATION OF

COMPOSITE MATERIALS

3.1 State-of-the-art of photocuring for composite materials 1-5

In the field of composites the main technology actually employed is thermal curing,

but the energy required for curing can be supplied also via high energy electrons (EB)

or ultra-violet radiation (UV).

The use of UV-curing technology to prepare composite materials has not been

extensively studied, so that only scant information is found in the literature. UV-light

induced free radical crosslinking polymerization has been employed in the preparation

of fiberglass-reinforced unsatured polyesters composites, leading to a product having

mechanical properties comparable to the thermal cured ones but obtained in a rapid

process. The same type of technique is employed in the fabrication of reinforced dental

composites.

UV radiation technology can successfully be employed in the preparation of

nanocomposites because it allows an intercalative polymerization in situ that is the most

appropriate technique to prepare polymer layered silicate nanocomposites. The process

is simple, UV irradiation of the layered silicate swollen in the liquid monomer

containing a photoinitiator. It assures not only all the advantages typical of UV-cure

(Par. 2.1), but also specific benefits to the obtained nanocomposites, such as a fine

CHAP. 3 PHOTOPOLYMERIZATION OF COMPOSITE MATERIALS


40

control of the swelling time to ensure a perfect interpenetration of the resin into the

interlayer galleries of the mineral. UV-cure has been employed to produce

nanocomposite adhesives for the applications in integrated optics, realized irradiating an

epoxy-based resin containing nanosized silica particles. Other works on UV-cured

coatings containing functionalized colloidal silica evidence the great improvement

obtained in abrasion and scratching resistance properties.

3.2 UV polymerization limits in the case of composite materials4

As seen above, the importance of UV radiation curing in composite production is

increasing nearby the classical thermal curing. Therefore it is important to analyze and

compare the cure processes in order to understand their advantages and disadvantages.

In a thermally initiated polymerization, decomposition of initiator is obtained by

heating. Reaction exothermicity adds to the thermal energy supplied by the oven,

leading to high temperatures in the core of the structure. Therefore complex temperature

programs are necessary to dissipate the heat of reaction, because it can lead thermal

degradation, internal stresses, etc., in the composite. The procedure is quite complex,

requiring long processing times at high temperatures, as well as complex curing

equipments.

In UV-cure process the initiator decomposes when UV-irradiated, producing reactive

species that propagate the crosslinking reaction. Therefore there is a cure front that

propagates throughout the sample thickness as the reaction proceeds. There is a gradient

of cure in the sample during the reaction and a homogenous degree of cure in the cured

sample. This technique allows curing of thick layers of resins if the reinforcing agent

and the resin do not absorb at the selected wavelengths.

Photopolymerization technique can be successfully applied in the production of

composites for low-performance applications from inexpensive starting materials by

relatively unskilled workers.

The main limits of UV polymerization in the case of composites are4:

Sample thickness: in fact the depth of penetration of the UV radiation represents

in most cases also the limit of sample thickness. Otherwise it is necessary to proceed to



41

a step process where the sample is irradiated from both sides. Fig. 3.1 illustrates that the

radiation received by the bottom layer decreases exponentially with the thickness of the

sample.

Filler transparency to UV-radiation: in order to assure a complete

photopolymerization through all the depth of sample, the reinforcing material should be

reasonably transparent to UV light.

Sizing agents’ influence: it is well known that the majority of inorganic

reinforcements (ex. glass fibers) are treated with sizing materials in order to protect

them during their use and to promote specific characteristics; these sizing agents can

include also starch, oils, and gelatin and fatty amines, materials that can act as inhibiter

in particular towards cationic photopolymerization.

Fig.3.1: Depth of cure versus incident radiation intensity in a UV-cured glass

fiber/epoxidized linseed oil composite4.

Obtained mechanical properties: literature4 reports examples in which the tensile

properties of UV-cured composites are lower that the ones of composites prepared from

the same monomers, but using conventional thermal curing methods.



42

However, the main limit to the use of UV-curing in composites preparation remains

the maximum thickness of a composite that can be irradiated, since composites gain

their mechanical properties from multiple layers of resin and reinforcing agents.

The depth of cure is dependent from a complex set of variables, such as light

intensity and wavelength, length of irradiation duration as well as photosensitivity of

photoinitiator, the UV absorption characteristics and reactivity, configuration of the

sample.

In the case of nanocomposites preparation3, it is also important to evaluate if the

presence of mineral filler affects the polymerization kinetics, with respect to the

reaction rate and cure extent.

In Tab. 3.1 results obtained for glass fibers-acrylate/methacrylate composites UV- or

thermal- cured5 are collected and compared.

Tab. 3.1: Comparison between UV- and thermal-curing in the preparation of glass

fiber-reinforced composites.

UV-cured composites Thermal-cured

composites

propagation rate limited by photo-bleaching

of the photoinitiator

limited by the induction

time, but after very fast due

to auto-acceleration

T°max reached in the

sample low high

Tg not very high high

% of residual

insaturations quite high low

yellowing present present

thermal shock - present (samples cracked)



43

3.3 Solutions reported in literature

Considering all the disadvantages related to the use of UV-curing technique in the

preparation of composites, the following solutions are proposed:

UV-curing can be successfully used in the preparation of composites for low-

performance applications, combining inexpensive starting materials to a simple

production technique.

UV-curing can be used for repairing large composites structure, taking into

account the simplicity of the equipment and protection needed as well as the short

curing time.

A thermal bake can be coupled after the radiation curing step to complete the

cure and accelerate the post-cure in thick samples.

PART 2

PHOTOPOLYMERIZED

MICRO- AND NANO-COMPOSITES

EXPERIMENTAL

CHAP. 1 MATERIALS

1.1 Silica nanoparticles-fumed silica 1-3

Silica is constituted by pure SiO2 in form of high surface area particles. It is colorless

and maintains its isolating properties constant at high temperature.

Applications of silica powders are based on porosity, active surface, hardness,

thixotropic, and viscosity management. If the chemical structure of silica surface is

altered, these properties may be combined with specific chemical or physical interaction

capacities. Modified silica is used as reinforcing material and its presence can improve

mainly tear, tensile, and abrasive resistance.

Silica can be divided in pyrogenic silica, i.e. fumed silica, and silica made by wet

methods, i.e. precipitated silica. In Tab. 1.1 are collected the main properties of different

types of silica. Precipitated silica includes a wide range of silica with a variety of

structural characteristics. In general, the formation involves an acid-precipitation of

aqueous solutions of alkaline silicates. As an overall definition it can be assumed that

precipitated silica is dry silica with no long or short distance characteristic structure.

Fumed silica is widely used in industry as an active filler for reinforcement of

elastomers, as a rheological additive in fluids and as a free flow agent in powders.

It is a synthetic amorphous form of silicon dioxide produced in a hydrothermal

process by burning silicon tetrachloride in an oxygen-hydrogen flame at 1,200-1,600°C.

At these high temperatures viscous droplets of amorphous silicon dioxide are formed,

CH

AP. 1

M

ATER

IALS

Phot

opol

ymer

ized

mic

ro- a

nd n

ano-

com

posi

tes:

inte

rfac

e ch

emis

try

and

its ro

le o

n in

terf

acia

l adh

esio

n

44

Tab.

1.1

: Phy

sica

l pro

pert

ies o

f var

ious

silic

as.

Cha

ract

eris

tics

Py

roge

nic

silic

a Si

lica

mad

e by

wet

met

hods

Fum

ed si

lica

Arc

silic

a Pr

ecip

itate

d

silic

a xe

roge

ls

aero

gels

Spec

ific

BE

T a

rea

m2 /g

50

to 6

00

25 to

300

30

to 8

00

250

to 1

000

250

to 4

00

Size

pri

mar

y pa

rtic

les

nm

5 to

50

5 to

500

5

to 1

00

3 to

20

3 to

20

Size

aggr

egat

ions

/agg

lom

erat

ions

µm

*

2 to

15

1 to

40

1 to

20

1 to

15

Den

sity

g/

cm3

2.2

2.2

1.9

to 2

.1

2.0

2.0

Vol

ume

ml/1

00g

1,0

00 to

2,0

00

500

to 1

000

200

to 2

,000

10

0 to

200

8

00 to

2,0

00

Mea

n po

re d

iam

eter

nm

no

n po

rous

till

300

m2 /g

no

n po

rous

>

30

2 To

20

> 25

Pore

dia

met

er d

istr

ibut

ion

*

* ve

ry b

road

na

rrow

na

rrow

Shap

e of

inte

rior

surf

ace

0

0 po

or

very

muc

h m

uch

Agg

rega

tion

and

aggl

omer

atio

n

stru

ctur

e

chai

n-lik

e

aggl

omer

atio

n

(ope

n su

rfac

e)

dens

e sp

heric

al

aggr

egat

es/p

artic

les

non-

aggl

omer

ated

slig

htly

aggr

egat

ed

near

ly

sphe

rical

parti

cles

high

ly p

orou

s

aggl

omer

ated

parti

cles

mac

ropo

rous

aggl

omer

ated

parti

cles

CHAP. 1 MATERIALS


46

the so-called primary particles, which collide and are fused together to build up stable

aggregates.

A widely used fumed silica is Aerosil® that was employed in our study. Its properties

are presented in Tab. 1.2.

Tab. 1.2: Physical and chemical data of Aerosil® 200.

*ex plant

Aerosil® 200 is mainly employed in paints and coatings, unsatured polyester resins,

laminated resins and gel coats, silicon rubber, adhesives and sealants, printing inks,

cable compounds and cable gel, plant protection, food, and cosmetics. It is used as an

anti-settling, thickening, anti-sagging agent and it improves free flow and anti-caking

characteristics of powders.

Properties Typical value

specific surface area (BET) 200±25 m2/g

average primary particle size 12 nm

tapped density*

acc. to DIN ISO 787/XI, Aug. 1983 50 g/l

bulk density*

ACM 104 30 g/l

moisture*

2 hours at 105°C ≤ 1.5%

ignition loss

2 hours at 1000°C based on material dried

for 2 hours at 105°C

≤ 1.0%

pH

in 4% dispersion 3.7-4.7

SiO2-content

based on ignited material > 99.8%

CHAP. 1 MATERIALS


47

1.2 Glass fibers4-6

Glass is a non-crystalline material with a short range network structure, it has no

distinctive microstructure and its mechanical properties (determined by composition and

surface finish) are isotropic. Glass fibers are designated normally by alphabetical codes.

The main fibers used are “E” (electrical) glass fibers, which amount to 90% of the

market. They are based on the CaO-Al2O3-SiO2 system, but there is no “standard”

composition. It is only possible to give some indications on their constituents

proportions (Tab. 1.3) and their thermal, mechanical and physics properties (Tab. 1.4).

S-glass (in Europe R-glass) is based on the SiO2-Al2O3-MgO system; these fibers

have higher stiffness and strength (S) than E-glass. Its properties are stable even to high

temperatures, but it is more difficult to produce the fibers because of its limited working

range, so they are more expensive and used only for some specific applications,

nowadays replaced by aramidic and carbon fibers.

Other more resistant types of glasses have been developed: C-glass (chemical), E-

CR-glass (electrical-corrosion resistant glass) and AR-glass (alkali resistant).

Tab. 1.3: Composition of glass fibers E-type.

constituents % weight

SiO2 53-54

Al2O3 14-15.5

CaO 20

MgO

B2O3 6.5-9

F 0-0.7

Fe2O3 <1

TiO2

Na2O <1

K2O

CHAP. 1 MATERIALS


48

Tab. 1.4: Thermal, mechanical, physical properties of glass fibers E-type.

properties

thermal capacity 0.8 J g-1 K-1

thermal conductivity 1.0 W m-1 K-1

linear dilatation coefficient 5.10-6 K-1

traction resistance 3.4 GPa

elasticity modulus 73 GPa

Poisson coefficient 0.22

extensibility 4.4%

volume 2.6 g cm3

humidity resumption <0.1%

The glass-fiber reinforced plastics marketed are predominately based on one type of

glass fiber, the E-type, but a wide variety of fiber formats (mats, fabric, unidirectional

roving), resin types filler/additives and process techniques are available. Fiber lengths

can vary from different length discontinuous fibers (milled, short, and long) to

continuous fibers in swirled mats, fabrics, non-crimped fabrics and unidirectional plies.

The major use of glass fibers is still as chopped strand mats of 25-50 mm length. The

different formats are often used together. In Tab. 1.5 are listed the different processes

routes and uses associated with formats.

Molten glass is extruded under gravity from a melting tank through an orifice and

rapidly pulled to draw it down to a 10 µm diameter fiber. Coatings are used to promote

adhesion and protection as well as to enhance wetting and bonding between fibers and

matrix. They are made mainly by water (85-95%), the other components are:

3-15% of sealants, such as vinyl polyacetates, polyesters used to protect;

0.5-2% of lubricants, such as cationic surfactants, to protect and lubricate fibers

surface during their use;

antistatic agent;

0.5-1.5% of coupling agents, typically organosilanes, to lay organic matrix to

glass surface.

CH

AP. 1

M

ATER

IALS

Phot

opol

ymer

ized

mic

ro- a

nd n

ano-

com

posi

tes:

inte

rfac

e ch

emis

try

and

its ro

le o

n in

terf

acia

l adh

esio

n

46

Tab.

1.5

: Gla

ss fi

ber f

orm

ats w

ith c

orre

spon

ding

com

posi

te m

ater

ials

, man

ufac

turi

ng p

roce

sses

and

end

use

s.

Fibe

r fo

rmat

Fi

ber

leng

th (m

m)

Com

posi

te m

ater

ials

type

s N

orm

al p

roce

ss r

oute

s T

ypic

al a

pplic

atio

ns

Mill

ed

< 0.

1 M

oldi

ng c

ompo

unds

In

ject

ion

mol

ded

Elec

trica

l, au

tom

obile

D

isco

ntin

uous

-sho

rt <

1 M

oldi

ng c

ompo

unds

In

ject

ion

mol

ded

Elec

trica

l, au

tom

obile

D

isco

ntin

uous

-long

<

7.5

Mol

ding

com

poun

ds

Inje

ctio

n m

olde

d El

ectri

cal,

auto

mob

ile

Cho

pped

stra

nd m

at

7.5-

50

CSM

, Dou

gh m

oldi

ng

com

poun

ds (D

MC

), sh

eet

mol

ding

com

poun

ds (S

MC

)

Han

d la

y-up

, spr

ay la

y-up

, co

mpr

essi

on m

olde

d M

arin

e, c

hem

ical

tank

s,

gene

ral t

rade

mol

ding

Swirl

ed m

at

Con

tinuo

us

Gla

ss m

at th

erm

opla

stic

s (G

MT)

, pul

trude

d pr

ofile

s Th

erm

ofor

med

, pul

trusi

on

Aut

omob

ile c

ompo

nent

s,

acce

ss e

ngin

eerin

g, c

able

tra

ys

Stitc

hed,

pin

ned,

ne

edle

d pr

oduc

ts

All

Any

M

ost

All

Wov

en fa

bric

s C

ontin

uous

Li

ghte

r wei

ght c

loth

s. D

iffer

ent s

tyle

s R

esin

inje

ctio

n, h

and

lay-

up,

pres

s mol

ded

Gen

eral

eng

inee

ring,

pr

essu

re v

esse

ls, m

arin

e

Wov

en ro

ving

s C

ontin

uous

H

eavi

er w

eigh

t clo

ths

Han

d an

d m

achi

ne la

y-up

, pu

ltrus

ion

Hea

vy m

arin

e

Kni

tted

Con

tinuo

us

2-D

and

3-D

fabr

ics

Res

in in

ject

ion

Con

stru

ctio

n, ra

ndom

es,

prop

elle

rs

Non

crim

p fa

bric

s N

CF

Con

tinuo

us

Bi-,

tri-

and

quad

ric-a

xial

R

esin

inje

ctio

n, p

ress

m

olde

d, H

and

lay-

up

Mar

ine,

con

stru

ctio

n,

auto

mob

ile

Mul

tidire

ctio

nal

Con

tinuo

us

Prei

mpr

egna

tes,

rovi

ngs

Pres

s mol

ded,

fila

men

t w

indi

ng

Hig

h pe

rfor

man

ce

aero

spac

e,

F1

raci

ng, p

ipes

, tor

que

tube

s, ro

cket

mot

or c

ases

U

nidi

rect

iona

l C

ontin

uous

Pr

eim

preg

nate

s Pr

ess m

olde

d, p

ultru

ded

strip

bar

W

ind

turb

ine

blad

es

CHAP. 1 MATERIALS


50

Glass fibers used in this work have been supplied by Vetrotex International, treated

with a coating made by water and antistatic agents. Their characteristics are resumed in

Tab. 1.6.

Tab. 1.6: Properties of glass fibers E-type used in this work.

name coating type weight

(g/km of fiber)

fiber diameter

announced

(µm)

fiber diameter

measured

(µm)

untreated

fibers

Water +

antistatic

agents

480 17 18.1 ± 1.4

1.3 Organosilanes7,8

A coupling agent can be defined as a material that improves the retention of

properties of the chemical bond across the interface between a mineral surface and an

organic resin, in presence of moisture. Since silane organofunctional silicones are

hybrids of silica and of organic materials related to resins, they could be called silane

coupling agents.

The effectiveness of a silane as coupling agent is related to the reactivity of its

organofunctional group towards the resin.

The chemical structure of a silane coupling agent can be represented as follows (Fig.

1.1):

X

XX Si R Y

X = hydrolysable functional group (CH3O )

R Y = functional group which can react with the matrix monomers

Fig. 1.1: General structure of silane coupling agent.

The hydrolysable groups X are intermediates in the formation of silanol groups for

bonding to the mineral surface; the organofunctional group R-Y is chosen for its

CHAP. 1 MATERIALS


51

reactivity or compatibility with the polymer. The coupling action depends on a stable

link between the X and Y groups.

Usually silanes are applied to inorganic surfaces from a water solution. Their

hydrolysis in water is dependent of the nature of the R-Y group9, but it is relatively fast

and can be considered complete in 1-30 minutes (at pH 3-4); afterwards the silane triols

condense to oligomers. The grafting reaction is schematized in Fig. 1.2.

In the ideal case a monolayer should be obtained on the surface, but what is realized

experimentally is a structure made of non-fully condensed grafted polysiloxane layers.

When silane is deposed on the surface of the inorganic filler, it forms a covalent

bond through a condensation reaction which is much slower (hours) and dependent on

the temperature (usually around 100-110°C).

X Si

X

X

R YH2O

Si

OH

OH

HO R Y

Hydrolysis

Condensation

SiOH

SiOH

SiOH Si

OH

OH

HO R YSi OSi O Si R Y

Si O3 H2O

inorganicsurface

inorganicsurface

Fig. 1.2: Scheme of hydrolysis and condensation reaction of organo-silane molecule on

inorganic surface.

The thickness of the silane layer can be estimated by various means, i.e. total carbon

analysis, ignition weight loss, scanning electron microscopy, contact angle

measurements, FT-IR spectroscopy, etc.

In particular FT-IR studies made by Ishida and Koenig on high surface fumed silica

and E-glass fibres10,11 had demonstrated the mechanisms of organization of silane

molecules on mineral surface: silanes adsorbed from water solution tend to give a

monolayer coverage on silica surface, while they form a film on E-glass fibres. It should

be noticed that in the case of multilayers, we have a high degree of condensation of the

silanols to siloxanes after the drying treatment. These siloxanes are initially soluble and

CHAP. 1 MATERIALS


52

fusible, so the interaction with the polymer should be made at that moment because

after a rigid cross-linked structure is obtained.

Thermosetting resins react with a combination of the two mechanisms, while

thermoplastics go often through the formation of an IPN.

The inorganic surfaces used in this work have been modified using different types of

alkoxy-silanes; they have been chosen taking in account their specific functionality to

couple with the polymeric matrices.

According to this assumption, two alkoxy-silanes with epoxy-functionality (to match

with epoxy resines) and an acrylate-terminated one (to match with acrylated resin) were

used. The hydrolysable ligand was selected to be the same for all the organofunctional

silanes employed in this work.

Their structures are:

Epoxycyclohexil-ethyl trimethoxysilane (CETS), supplied by WITCO, has been

used to modify inorganic surfaces for the preparation of cycloaliphatic-matrix

composites, as it can be seen from its structure, Fig. 1.3.

H3CO

H3COH3CO Si CH2 CH2

O

CETS Fig. 1.3: CETS structure.

Glycidoxypropyl trimethoxysilane (GPTS), supplied by Aldrich, has been used

to modify inorganic surfaces for the preparation of diglycidyl ether-based composites,

as it can be seen from its structure, Fig. 1.4.

CHAP. 1 MATERIALS


53

H3CO Si CH2 CH2 CH2 O CH2

O

H3CO

H3CO

GPTS Fig. 1.4: GPTS structure.

Trimethoxysilyl propyl-methacrylate (MEMO), supplied by Aldrich, has been

used to modify inorganic surfaces for the preparation of epoxidized acrylate soybean oil

-based composites, as it can be seen from its structure, Fig. 1.5.

H3CO Si CH2 CH2 CH2 O C

O

C

CH3

CH2

H3CO

H3CO

MEMO Fig. 1.5: MEMO structure.

An alkyl alkoxy-silane has been used to modify inorganic surfaces inducing

hydrophobic characteristics:

n-propyl trimethoxysilane (C3), supplied by Petrarch System Inc.; its structure is

represented in Fig. 1.6.

H3CO

H3COH3CO Si CH2 CH2 CH3

C3 Fig. 1.6: C3 structure.

CHAP. 1 MATERIALS


54

1.4 Thermoset matrices

In this work three thermoset matrices have been used.

3,4-epoxycyclohexylmethyl-3’,4’-epoxycyclohexanecarboxylate (CE), Fig. 1.7,

Cyracure® UVR 6110, supplied by DOW Corporation.

O

O

OO

CE Fig. 1.7: CE structure.

1,4-cyclohexane dimethanol diglycidyl ether (DGE) supplied by Aldrich.

Its structure is represented in Fig. 1.8.

CH2

CH2

O CH2

O CH2

O

O

DGE

Fig. 1.8: DGE structure.

Soybean oil epoxidized acrilate (SOA) supplied by Aldrich.

SOA was obtained by acrilation of the corresponding epoxidized product13 (Fig. 1.9);

CHAP. 1 MATERIALS


55

CH2 O C

O

(CH2)7CH CH CH2 CH CH (CH2)4CH3

CH O C

O

(CH2)4CH CH CH2 CH CH CH2 CH CH (CH2)4CH3

CH2 O

C

O

(CH2)7CH CH (CH2)7CH3

O

O

O

epoxidized soybean oil Fig. 1.9: Epoxidized soybean oil structure.

For each molecule of epoxidized product, two molecules of acrylic acid, on average,

have been introduced.

According to C13-NMR, the proposed structure of SOA is represented in Fig. 1.10.

CH2 O C

O

(CH2)7CH CH CH2 CH

OH

CH

O C

O

CH CH2

(CH2)4CH3

CH O C

O

(CH2)4CH CH CH2 CH CH CH2 CH

OH

CH

O C

O

CH CH2

(CH2)4CH3

CH2 O

C

O

(CH2)7CH CH (CH2)7CH3

O

SOA Fig. 1.10: SOA proposed structure.

CHAP. 1 MATERIALS


56

1.5 Photoinitiators

Triphenylsulphoniumhexafluoroantimonate, (Cyracure® UVI 6974) supplied by

DOW Corporation was used as cationic photoinitiator to polymerize CE and DGE

monomers. Its structure is represented in Fig. 1.11 (I).

SOA has been cured via radical mechanism, using 2-hydroxy-2-methyl-1-

phenyl-propan-1-one, (Darocur® 1173) supplied by Ciba Specialty Chem. as radical

photoinitiator. Its structure is also represented in Fig. 1.11 (II).

S

SbF6-

(I)

C

O

C

CH3

CH3

OH

(II)

Fig. 1.11: Cyracure® UVI 6974 structure (I), Darocur® 1173 structure (II).

In Tab. 1.7 are reported the UV absorption peaks for the two photoinitiators.

Tab. 1.7: UV absorption peaks for the photoinitiors used.

photointiator UV absorption peak (nm)

Cyracure® UVI 6974 (5.56 10-5 M in propylene carbonate)

300-310

Darocur® 1173 (4 10-5 g/ml in methanol)14

265-28014

Both the photoinitiators are activated by standard mercury-filled UV bulbs, such as

Fusion® lamp, because they efficiently absorb some of the major emission bands of the

standard bulbs, as shown in Fig. 1.12 for the cationic photoinitiator.

CHAP. 1 MATERIALS


57

Fig. 1.12: Overlap of cationic photoinitiator absorption and standard UV bulb

emission spectra12.

1.6 UV-lamps and reactive formulations selected

Different UV-lamps have been used:

Helios Italquartz lamp: laboratory UV-lamp, equipped with a support which

allows changing the distance of the sample from the lamp; there is possibility of

operating in nitrogen atmosphere. Lamp intensity (I) = 10 or 50 mW/cm2.

Fusion lamp: industrial UV-lamp, equipped with a belt conveyor giving to the

sample an adjustable speed. Lamp intensity (I) = 371.8 mW/cm2.

UV Perkin-Elmer DPA 7 XBO 450 W equipped with a monocromator to select

the wavelength; it was used for the photo-DSC measurements. The intensity is

dependent from the selected wavelength: at 300 nm, Iref = 0.581 µW/cm2 and Isample =

0.579 µW/cm2.

Reactive systems:

The different components were mixed in the proper concentration12,14,15 and UV-

cured.

CHAP. 1 MATERIALS


58

Epoxy monomers, CE and DGE, were UV-cured through a cationic mechanism using

2% w/w of cationic photoinitiator. Curing was performed in air when Fusion or Helios

lamp has been used; in nitrogen during photo-DSC measurements, to assure a uniform

heat transfer during the experiment.

SOA monomer was UV-cured through a radical mechanism using 4% w/w of radical

photoinitiator. Curing was performed in air using Fusion lamp: the high intensity of the

lamp made possible to cure in air, overcoming oxygen inhibition.

CHAP. 2 MODIFICATION OF INORGANIC

SURFACES BY ORGANOSILANES

2.1 Introduction

The modification of inorganic surfaces is very important and usually performed in

order to assure a better adhesion and improved properties to the composite materials.

The surface modification allows to form bonds (physical or chemical) between the two

phases, which have different structure, and to increase compatibility of the systems.

2.2 Experimental: protocols for nanosilica and glass fibers

The modification of inorganic surface was carried out both on silica powder and on

glass fibers. In this paragraph the experimental procedure is reported; the starting

information was that found in the literature1-5, but it has been modified in order to

increase the adhesion properties.

All the experiments follow some common features: silane coupling agents are

dissolved in water solution at pH = 4 in order to have a fast hydrolysis of the

organosilane molecules3 (and, as a consequence, a lower rate of condensation reactions

between hydrolyzed species in the solution).

CHAP. 2 MODIFICATION OF INORGANIC SURFACES BY ORGANOSILANES


60

The hydrolysis reaction in these conditions is known to be very fast (from 1 to 10

min. depending on the reactivity of the silane); then the mixture is ready to perform the

grafting of inorganic surfaces.

The first step is the absorption of the silane on the inorganic surface; then a

condensation reaction is performed in order to link the silane molecule on the surface by

forming strong chemical bonds3,5.

In this step, done at high temperatures, silane molecules form siloxane bonds with

inorganic surface.

In Fig. 2.1 the scheme of the reactions involved in aqueous systems is reported.

X Si

X

X

R YH2O

Si

OH

OH

HO R Y

Hydrolysis

Condensation

SiOH

SiOH

SiOH Si

OH

OH

HO R YSi OSi O Si R Y

Si O3 H2O

inorganicsurface

inorganicsurface

Fig. 2.1: Scheme of hydrolysis and condensation reaction of organo-silane molecule on

inorganic surface.

The grafting reaction was carried out in different experimental conditions to evaluate

the best ones in order to obtain high adhesion values.

Before any grafting treatment, the inorganic products were dried by putting in an

oven at 150°C for 18 hours for silica powder, and at 250°C for 48 hours for glass fibers.



61

GRAFTING OF SILICA NANOPARTICLES

Grafting of silica was performed in different solvents, namely water, CH3COCH3 and

H2O/EtOH (50/50 v/v) at pH = 4 (CH3COOH) by using a silane concentration of 1 mL

for 100 mL of solvent.

Silica powder was introduced in the silane solution and left for two hours, at room

temperature, under ultrasonic stirring. Then it was filtered and treated in an oven at

120°C for 4 hours in order to promote the condensation reaction. The product was then

washed, to remove all the unreacted species, and again dried in an oven at 120°C for 2

hours.

Modified silica powder was characterized using thermogravimetric (TGA) analysis.

TGA were performed in the interval 50°-750°C, at heating rate of 10 C/min, in

nitrogen atmosphere, using 10-20 mg of sample and a Mettler-Toledo instrument.

GRAFTING PROCEDURE OPTIMIZATION

In Tab. 2.1 the obtained TGA results in the different solvents are reported. They

indicate in all cases a similar weight loss values. These results are in agreement with

those reported in literature1.

Tab. 2.1: TGA results on silica grafted using different solvents*.

solvent % weight loss (TGA)

H2O 5.8

H2O/EtOH 5.4

CH3COCH3 5.9 *silica grafted using CETS 1% v/v.

Water was selected as solvent for the grafting reaction in all the experiments.

Then the duration of the condensation step was investigated. The experiments were

executed at 120°C, changing the time from 4 to 18 hours. TGA analysis (Tab. 2.2)



62

showed that weight loss is practically, as expected, the same. Therefore a reaction time

of 4 hours was adopted for the grafting procedures.

Tab. 2.2: TGA results on silica powder*after different condensation time.

time (hrs) % weight loss (TGA)

4 8.8

18 9.3 *silica grafted using CETS 1% v/v in CH3COCH3.

In order to investigate the influence of washing protocol of the surface, experiments

were performed washing the silica powder immediately after grafting or after a 4 hrs

condensation in an oven at 120°C (Tab. 2.3).

Tab. 2.3: TGA results on silica powder*after different washing treatments.

sample % weight loss (TGA)

washed without condensation 0.43

washed after condensation (4 hrs) 7.03 *silica grafted using CETS 1% v/v in CH3COCH3.

The results indicate the importance of performing the thermal condensation reaction

in order to obtain grafting, i.e. covalent bonding of silane molecules to the surface.

The results of Tab. 2.2 and Tab. 2.3 are slightly different from those reported in Tab.

2.1, because they were performed using a different TGA instrument.

In conclusion the grafting procedure adopted for our systems can be summarized as

follows:

Materials:

silica powder = 2g

silane coupling agent, CETS or GPTS (for 2g of silica powder) = 1ml

reaction solvent, distilled water = 100 ml



63

Experimental:

pH = 4 (CH3COOH)

T = room temperature, i.e. 25°C

Time = 2 hours

Ultrasonic stirring, 10 min

Silica powder addition and ultrasonic stirring, 2 hrs

Filtration

Condensation reaction in an oven for 120°C for 4 hrs

Washing with distilled water

Drying at 120°C for 2 hrs.

GRAFTING OF SILICA WITH CETS AND GPTS SILANES

In order to characterize the modified silica surface, TGA analyses were carried out.

The analyses were carried out in different experimental conditions from the previous

ones (LECO TGA-601 instrument):

Temperature range = 50-900°C

Heating rate = 1°C/min

Atmosphere = air

Sample weight = 500 mg.

The results obtained are listed in Tab. 2.4 and Tab. 2.5 referred to the use of CETS

and GPTS silanes in the standard grafting conditions.

In Fig. 2.2 are presented typical TGA curves obtained for silica grafted with CETS:

they give evidence of the reproducibility of the experiment.

Tab. 2.4: Weight loss of silica grafted with CETS (1% v/v, 2g silica powder).

%weight loss

2.79; 2.84

2.87; 2.85

2.58; 2.57



64

Tab. 2.5: Weight loss of silica grafted with GPTS (1% v/v, 1g silica powder).

%weight loss

5.22; 5.26

5.57; 5.58

5.35; 5.44

The scattering of the reported results obtained in Tab. 2.4 and Tab. 2.5 can be

attributed to the different quantities of silica analyzed in these experiments.

Fig. 2.2: TGA reproducibility curves of silica powder grafted with CETS (1% v/v).

GRAFTING OF E-GLASS FIBERS

The grafting conditions and the experimental procedure used were the same as for

silica nanoparticles. Glass fibers were modified using different concentrations of silane

coupling agent in solvent (CH3COOH; pH = 4), in the interval 0.1-1% v/v, in order to

obtain silane layers of different thickness. In fact, in the literature6 it is reported that

adhesion between glass fibers and polymer matrix increases by decreasing the thickness

of silane layer.

TGA analyses were carried out on glass fibers, treated with the standard procedure

described previously.

96

97

98

99

100

0 200 400 600 800 1000

Temperature (°C)

%



65

In Tab. 2.6 the results obtained for glass fibers grafted using 0.5% v/v of silane agent

are reported.

Tab.2.6: TGA analyses of grafted E-glass fibers.

silane agent silane concentration %weight loss

0.1 % v/v in H2O 0.62 CETS

0.5 % v/v in H2O 1.25

0.1 % v/v in H2O 0.65 GPTS

0.5 % v/v in H2O 1.20

It can be noticed that practically the same weight loss is obtained with the two silane

agents; moreover the weight loss values are clearly lower compared with the ones

resulting from silica nanoparticles; these results can be explained mainly considering

the lower specific surface of E-glass fibres compared to silica.

SURFACE WETTING MEASUREMENTS

Surface modification can be detected by measuring the contact angles of the surfaces

before and after surface treatment with probe liquids. As far as silica nanoparticles are

concerned, the surface modification was simulated by using an oxidized,

treated/untreated silicon wafer surface.

Silicon wafer surfaces were cleaned at 550°C for 2 hrs and then oxidized by treating

with a mixture of H2SO4 (98%): H2O2 (30 vol.): H2O = 1: 1: 3 for 15 min at 100°C.

The wetting measurements were done using bi-distilled deionized water. A dynamic

contact angle instrument or a Cahn balance was used.

In Tab. 2.7 the results obtained from dynamic contact angle measurements performed

on silicon wafer surface are collected.



66

Tab. 2.7: Contact angle with water on oxidized grafted silicon wafer.

sample grafting solutions ϑADV ϑREC

oxidized, untreated

silicon wafer - 0 -

CETS 1% v/v in H2O 94.6 72.3

CETS 1% v/v in CH3COCH3 93 60.3

CETS 1% v/v in H2O/EtOH 90.7 52.8

GPTS 1% v/v in H2O 92.2 69.6

GPTS 1% v/v in CH3COCH3 88.6 68.5

GPTS 1% v/v in H2O/EtOH 85.6 67.4

The results of Tab. 2.7 indicate clearly a strong modification of the surface which

becomes more hydrophobic after the grafting reaction. The contact angles slightly

changes by changing the grafting solutions; in any case they are higher when water is

used as solvent.

As far as glass fibers are concerned, the Cahn balance was used both on fibers, and

glass slides (models of the glass fibers surface). Results are presented in Tab. 2.8.

Tab. 2.8: Contact angles with water obtained with Cahn balance on grafted* glass

fibers and glass slides.

ϑADV ϑREC

untreated glass fiber 35 ± 10 28.8± 10

CETS treated glass fiber 85.5 ± 12 57.7 ± 12

GPTS treated glass fiber 83.1± 5 56.9 ± 5

untreated glass slide 29.5 ± 2 -

CETS treated glass slide 95.2 ± 1 67.1± 1

GPTS treated glass slide 79.9 ± 3 52.9 ± 3

C3 treated glass slide 95.4 ± 1 65.2 ± 1 * 1% v/v silane in water.



67

It should be noticed that results obtained on glass fibers, (Tab. 2.8), are done at the

sensibility limits of the instrument; therefore the experimental error is quite high.

It is evident that the surfaces of both glass fibers and glass slides have been modified

making them more hydrophobic.

TOPOGRAPHY

AFM measurements, using the tapping mode, were performed on treated/untreated

silicon wafer surface (used as model for silica powder) and on treated/untreated glass

fibers surface.

Fig. 2.3 shows the AFM image and profile of oxidized untreated silicon wafer

surface; Fig. 2.4 and Fig. 2.5 show AFM images and profiles of silicon wafer surface

after oxidation and treatment with GPTS 1% v/v and CETS 1% v/v, respectively. With

any grafting treatment performed, on wafer surface are clearly visible the agglomerates

described in literature6 as “silane islands”, due to the grafting reaction that leads to a

non-homogeneous surface modification.



68

Fig. 2.3: AFM image and profile of untreated oxidized silicon wafer surface.

Fig. 2.4: AFM image and profile of silicon wafer surface oxidized and grafted with

GPTS 1% v/v.



69

Fig. 2.5: AFM image and profile of silicon wafer surface oxidized and grafted with

CETS 1% v/v.

AFM images on E-glass fibers surface were recorded taking into account the

cylindrical shape of the sample. A portion of the fiber was selected in order to assure the

best adhesion of the AFM cantilever tip to the sample during all the measurement (Fig.

2.6).



70

Fig. 2.6: Schematic representation of AFM analysis on E-glass fiber surface.

Fig. 2.7 shows the AFM image and profile of untreated E-glass fiber, while Fig. 2.8

and Fig. 2.9 show the images and profile of the surface of E-glass fiber grafted with

CETS in different percentages; Fig. 2.10 and Fig. 2.11 show the images and profiles of

the surface of E-glass fiber grafted with GPTS in different percentages.

The glass fiber surface after the grafting presents a different morphology from the

untreated one, more homogeneous and with the characteristic “silane islands” already

observed onto the grafted silicon wafer surface.

CANTILEVER TIP

GLASS FIBER



71

Fig. 2.7: AFM image and profile of untreated E-glass fiber surface.

Fig. 2.8: AFM image and profile of E-glass fiber surface treated with CETS 0.1%

v/v.



72

Fig. 2.9: AFM image and profile of E-glass fiber surface treated with CETS 0.5%

v/v.



73

Fig. 2.10: AFM image and profile of E-glass fiber surface treated with GPTS 0.1%

v/v.



74

Fig. 2.11: AFM image and profile of E-glass fiber surface treated with GPTS 0.5%

v/v.

Other experimental evidences of changes in surface morphology, after grafting

reaction, are visible from the results of analyses performed with SEM on glass fibers.

In Fig. 2.12 the untreated glass fiber surface is presented, while in Fig. 2.13 and Fig.

2.14 glass fiber surface after treatment with silane (GPTS 0.5% v/v).

Also from these images is evident the presence of agglomerates (“silane islands”) on

inorganic surface, moreover it is possible to observe that after treatment, surface is

much more homogeneous7, the defects present are somehow “repaired”.



75

Fig. 2.12: SEM image of untreated E-glass fiber.

Fig. 2.13: SEM image of treated E-glass fiber (GPTS 0.5% v/v).



76

Fig. 2.14: SEM image of treatedE- glass fiber (GPTS 0.5% v/v).

2.3 Conclusions

Grafting of silica powder was performed in different solvents (H2O, CH3COCH3,

H2O/EtOH 50/50 v/v); the weight loss values determined by TGA were practically the

same. Contact angle measurements performed on oxidized and grafted silicon wafers,

using the same solvents indicate a clear increase of the hydrophobicity of the surface;

the values obtained were the same using the different solvents, indicating that the

grafting reaction is not influenced by the solvent in the adopted conditions.

The condensation reaction gives a constant weight loss after 4 hrs treatment at

120°C. If the thermal treatment is not performed the weight loss was negligible, thus

indicating the absence of condensation reaction.

Grafting of glass fibers was performed in the same conditions adopted for silica

powder. The weight loss values determined by TGA were about one order of magnitude

lower than those obtained with silica powder; nevertheless the contact angle values with

water were increased sharply, indicating the formation of hydrophobic surfaces.

CHAP. 3 UV-POLYMERIZATION IN THE

PRESENCE OF NANOFILLERS

3.1 Introduction

The first problem in preparing UV-cured composites is the filler transparency

towards UV-light. Otherwise it will be a competition of the filler in absorbing the UV-

radiation. In this work the influence of nanosilica on the curing reaction was

investigated in order to verify if it can modify kinetics as well as total conversion of

monomer during the UV-curing process and if the surface modification has any

influence on the reaction1,2. Only nanosilica-filled systems were investigated, because

its higher surface area should lead to more evident interaction effects in polymerization

kinetics. In the following sections the results obtained on the filler/polymer systems are

presented.

3.2 Experimental

UV-CURING

Different types of UV-lamps were used to cure the composites:

Fusion lamp: thick (2 mm) samples, used for mechanical properties

measurements, were irradiated for 21 s each side at I = 371 mW/cm2.

CHAP. 3 UV-POLYMERIZATION IN THE PRESENCE OF NANOFILLERS


78

Helios Italquartz lamp: it was used for FT-IR kinetic measurements. The

samples were put between two disks of KBr and irradiated for 5 s intervals at I = 51

mW/cm2.

Photo-DSC lamp: samples were irradiated for 33 min. at I = 0.58 µW/cm2 (total

time run = 36 min). The optimal operative conditions for these systems were established

with tests at different wavelengths, with different sample weights and different amounts

of treated/untreated nanofillers; the obtained conditions, reported Tab. 3.1, were used

for all the experiments.

Tab. 3.1: Operative conditions for photo-calorimetric experiments.

CE CE + 10% w/w silica

optimal weight (mg) 0.390 0.370*

optimal wavelength (nm) 280-290 280

DGE DGE + 10% w/w silica

optimal weight (mg) 0.400 0.400*

optimal wavelength (nm) 300 300

The weight values indicated with * represent the total weight of the sample present in

the DSC aluminum pan; all the ∆H values obtained during experiments with nanofillers

have been corrected to taking into account the presence of silica.

Several experiments were done for each concentration to control their

reproducibility.

FT-IR and photo-DSC techniques were used to follow the reaction kinetics of the

reactive systems filled with different amounts of nanosilica concentrations.

FT-IR measurements

FT-IR instrument was used to follow the reaction kinetics by measuring the decrease

of the band at 750 cm-1 due to the epoxy group polymerization. The measurement was

discontinue; irradiation was performed for 5 s at the beginning and for 10 s until the end

of the reaction. The absorbance data were plotted against the reaction time.



79

Photo-DSC measurements

Photo-DSC was used to follow the reaction through a real-time measurement. Each

experiment was done at 30°C with an isothermal analysis, so that the ∆H curve was

obtained as a function of time. ∆H is related to the heat developed during the

polymerization reaction.

3.3 Reaction kinetics: Results and discussion

FT-IR RESULTS

The reaction kinetics were followed by monitoring the decrease of the epoxy band at

750 cm-1 (Fig. 3.1); normalization was made using the C=O band at 1730 cm-1.

40050060070080090010001100

Wavenumbers

Absorbance

Fig. 3.1: Typical FT-IR spectrum of neat CE.

In Fig. 3.2 examples of the data obtained for kinetic curves by using CE system are

reported. They concern the pure monomer and its mixtures with 10% w/w of treated

silica and with 10% w/w of untreated silica.

0 s

60 s



80

0

10

20

30

40

50

60

70

80

90

0 20 40 60 80 100time (sec.)

% c

onv.

CE

CE + 10% w/wtreated silica

CE + 10% w/wuntreated silica

Fig. 3.2: FT-IR kinetic curves of CE system filled with untreated or treated silica

(I = 51 mW/cm2).

A decrease of both kinetic and total conversion values when treated/untreated silica

is added to the photopolymerizable system is evident. The grafting of silica seems not to

have an important effect on the UV-curing reaction kinetics.

As far as the DGE system is concerned (Fig. 3.3), similar results were obtained when

DGE is photopolymerized with 5% w/w of untreated silica3.

Fig. 3.3: FT-IR kinetic curves of unfilled and silica filled DGE system3

(I = 8 mW/cm2).

DGE

DGE + 5% untreated silica



81

The first explanation proposed for the effect of silica on the photopolymerization

kinetics could be the scattering of the UV radiation by silica particles.

In typical experiments, in the presence of 5% w/w of silica, the light transmittance

curve of the mixture decreases of 15%. When the photopolymerization was performed

reducing the intensity of the emitted light of this value, a very small variation of the

kinetic was observed. Therefore the scattering of UV-light by the dispersed silica

particles does not justify the modification of the curing process.

Taking into account the high specific surface of silica (200 m2g-1), we suggest that

the particles could adsorb the cationic species of the photoinitiator molecules, thus

decreasing their activity in the photopolymerization. In fact the photoinitiator species

are polar molecules which can strongly interact with the silica surface.

In order to check this possibility, UV measurements were performed on solutions of

the cationic photoinitiator in the presence of silica. The photoinitiator was dissolved in

propylene carbonate and its absorbance at 300 nm was evaluated. 5% w/w of treated

and untreated nanosilica was added to the solution. The mixture was centrifuged in

order to separate silica particles and the solution was examined by UV spectroscopy.

Results are collected in Tab. 3.2.

Tab. 3.2: Results of UV absorptions on photoinitiator-silica systems at 300 nm.

*Ph3S+SbF6-, 5.56 10-5 M in propylene carbonate.

The results above reported indicate that silica interacts deeply with the cationic

photoinitiator by adsorbing it on its surface. The adsorbed photoinitiator could have

lower activity under UV-irradiation. In this way, we can explain the decrease of the

photopolymerization kinetics in the presence of silica. The photoinitiator-silica

interaction will be further study in depth.

sample Abs310 nm

photoinitiator* 0.713

photoinitiator* added of 5% w/w untreated silica ≈ 0

photoinitiator* added of 5% w/w treated silica (CETS) ≈ 0

photoinitiator* added of 5% w/w treated silica (GPTS) ≈ 0



82

It should be noticed that in radical systems, no decrease of UV-curing kinetics in the

presence of silica was observed4.

PHOTO-DSC RESULTS

Examples of typical photo-DSC curves for the systems based on CE and DGE are

presented in Fig. 3.4 and Fig. 3.5.

In each figure are reported two different thermograms: one is related to the

polymerization of the pure monomer and the other to the polymerization of the

monomer filled with 20% treated silica.

Even if the most part of the photopolymerization occurs in few minutes; the samples

were irradiated for 33 minutes to assure the completion of the reaction.

Fig. 3.4: Photo-DSC traces of neat CE and CE monomer filled with silica

(treated CETS 1% v/v).

19,319,419,519,619,719,819,9

2020,1

0 10 20 30 40

time (min)

delta

H (J

/g)

CE

CE + 20% w/wtreated silica



83

Fig. 3.5: Photo-DSC traces of neat DGE and DGE monomer filled with silica

(treated GPTS 1% v/v).

To obtain the ∆H (J/g) values of each UV-DSC trace, integration of the area below

the base line (Fig. 3.6) was made. The integration limits were chosen in correspondence

of the switch on/off of UV lamp. The obtained value was corrected in the case of

composites, taking into account the different percentages of silica added to the sample.

19,319,419,519,619,719,819,9

2020,1

0 10 20 30 40

time (min)

delta

H (J

/g)

Fig. 3.6: Method for integration on UV-DSC trace.

19,319,419,519,619,719,819,9

2020,1

0 10 20 30 40time (min)

delta

H (J

/g)

DGE

DGE + 20%w/w treatedsilica



84

In Fig. 3.7 and Fig. 3.8 the results obtained in the kinetics experiments on CE and

DGE in the presence of treated or untreated silica are reported (for each concentration

three ∆H values were considered).

The figures indicate for both CE and DGE system a decrease of ∆H of

polymerization in the presence of different amounts of silica. Moreover the results

indicate no clear influence of the surface treatment of grafting on the

photopolymerization reaction.

Fig. 3.9 reports the kinetic curves calculated from the thermograms obtained using

CE added of different percentages of treated silica: they show the decrease of the ∆H

values when the amount of silica present in the system is increased.

The results of photocalorimetric experiments fully confirm those obtained by FT-IR

measurements indicating a decrease of the rate of reaction and of the final conversion in

the presence of silica nanoparticles.

150170190210230250270290310

0 5 10 15 20

% w/w silica

-del

taH

(J/g

)

CE + untreated silicaCE + treated silica

Fig. 3.7: ∆H of reaction as a function of percentage oft silica introduced in the CE

monomer.



85

200

250

300

350

400

450

500

550

0 5 10 15 20

% w/w silica

-del

atH

(J/g

)DGE + untreated silicaDGE + treated silica

Fig. 3.8: ∆H of reaction as a function of percentage oft silica introduced in the DGE

monomer.

0

10

20

30

40

50

60

70

80

90

0 5 10 15 20 25

t (min)

-del

taH

(J/g

)

CE+20% w/w treated silicaCE+15% w/w treated silicaCE+10% w/w treated silicaCE+5% w/w treated silicaCE

Fig. 3.9: Dependence of polymerization of CE monomer filled with various percentages

of grafted silica.



86

3.4 Conclusions

FT-IR and photo-DSC measurements indicate that the addition of nanosilica leads to

a modification of the photopolymerization reaction and to a decrease in both kinetics

and total conversions of reactive groups.

As far as the influence of the surface treatment is concerned, the results obtained do

not indicate a clear influence on the photopolymerization reaction.

We explain the observed decrease of the photopolymerization kinetics in the

presence of silica on the basis of the interaction between photoinitiator and silica, as

previously reported in this chapter. In fact, such nanofillers display a large amount of

interface as the specific surface of silica is very important. One can aspects a minor

effect on the polymerization kinetics for the fiber-based composites, as the surface

displayed by the filler is very low compared to nanosilica.

CHAP. 4 UV-POLYMERIZED MICRO- AND

NANO-COMPOSITES

4.1 Introduction

In this chapter the preparation of micro- and nano-composites using modified or

unmodified fillers is reported. The cured products were subjected to thermal and

mechanical tests.

Nano-composites were investigated by using dynamic-mechanical analysis

(DMTA)1.

Micro-composites were investigated by using the microbond technique2-5 in order to

evaluate the interfacial adhesion as a function of the grafted species.

4.2 Experimental

DMTA analysis

The samples were prepared using 10% w/w of silica. Photopolymerization was

performed by irradiating with a Fusion lamp for 21 s at I = 371 mW/cm2 on each side of

the sample. Testing of the samples was performed in the bending mode using a

CHAP. 4 UV-POLYMERIZED MICRO- AND NANO-COMPOSITES


88

Rheometric Scientific MKIII apparatus. Samples were tested in bending

configuration, single cantilever; the temperature range was from 0° to +250°C for the

CE/silica systems and from -50° to +80°C for the DGE/silica systems. Measurements

were carried out at 1Hz frequency.

Microbond technique

Microcomposites were prepared by putting microdroplets of photopolymerizable

monomer (+ photoinitiator) on E-glass fibers and by curing with a Fusion lamp for 21 s

at I = 371 mW/cm2.

The interfacial adhesion of samples was tested by using a dynamometer (cell load =

10 N, v = 0.1 mm/min); the experiments were followed using a micro camera.

Preparation of samples: single treated or untreated glass fibers were fixed on a frame

under small tension. By using a copper filament, microdroplets of photocurable mixture

were deposed on fibers and UV-cured.

In Fig. 4.1 the preparation process is schematically reported.

Fig. 4.1: Preparation of microcomposites for the microbond test.

After UV-curing, fibers were cut in 1 cm length pieces taking care of having in each

piece at least one cured droplet; each segment was fixed with glue on a triangle made by

PET (Fig. 4.2) and tested with the dynamometer.

FRAME

GLASS FIBERS

MICRODROPLETS

COPPER FILAMENT



89

Fig. 4.2: Sample for microbond test.

Each mechanical test gives a diagram similar to the one presented in Fig. 4.3.

The peak in the curve indicates the maximum force registered opposed by the sample

immediately before the detachment of the droplet.

Fig. 4.3: Typical curve obtained from a microbond test.

As each force value is related to the droplet dimensions (length and diameter), it is

necessary to know them exactly before any measure. They were evaluated using an

optical microscope equipped with a device that allows to measure and to express them

in µm.

About 30 measurements for each type of sample should be performed in order to

overcome the high degree of dispersion of data which is connected with the type of test

and the type of fibers used.

Fig. 4.4 represents a typical graph obtained; it is evident the difficulty in having a

good reproducibility, therefore a comparison of the data obtained in this form.

Force

FrictionForce

Maximum Force

Extension

Force

FrictionForce

Maximum Force

Extension



90

0

0,1

0,2

0,3

0,4

50 100 150 200Le (micron)

F (N

)

Fig. 4.4: Debonding force (F) vs. embedded length (Le) curve obtained from

microbond measurements (CE + untreated glass fibers).

For these reasons, the experimental values were treated using the Kelly-Tyson

formula for the average interfacial shear strength at the interface (IFSS) 6, dτ :

ef

dd

lrFπ

τ2

=

Where: Fd = force at debonding

le = embedded length

2rf = fiber diameter.

The dτ value is a measure of the interfacial adhesion assuming a constant shear stress

along the embedded length, i.e. a plastic behavior of the interface7. It allows evaluating

the change of the adhesion between the matrix and the glass fibers as a function of the

surface treatment.

Higher values reflect a more effective interface, thus a better silane performance.



91

4.3 Results and discussion

DMTA ANALYSES

The measured Tα, i.e. the temperature position for the maximum of the main

elongation peak related to Tg, values for CE and DGE systems in presence of silica are

reported in Tab. 4.1 and compared with those obtained from the pure monomers.

Tab. 4.1: Tα, values for UV-cured systems in presence of silica at 1 Hz.

The experimental results presented in Tab. 4.1 evidence a decrease of Tα for both the

systems (CE, DGE) when silica is added. These results are in agreement with the

polymerization kinetic data reported in the previous chapter indicating a reduction of the

epoxy group conversion and thus of crosslinking density of the matrix in the presence of

silica.

In Fig. 4.5 and Fig. 4.6, the DMTA curves related to CE system are presented. This

one cannot be easily obtained from the value of the storage modulus in the rubbery

plateau as this network is intrinsically heterogeneous, i.e. proceeding from the

percolation of microgels formed from the early stages of the polymerization.

sample Tα matrix Tα matrix + 10% w/w untreated silica

CE 214°C 182°C

DGE 53°C 37°C



92

Fig. 4.5: DMTA curves of CE neat matrix.

Fig. 4.6: DMTA curves of CE (matrix + 10% w/w untreated silica).

DMTA tests were performed also on CE added with untreated glass fibers finely

pulverized and cured with the same technique used for silica composites.

The results are presented in Tab. 4.2.



93

Tab. 4.2: DMTA analyses performed on photocured neat and pulverized glass fibers

filled CE composites.

CE neat matrix CE + 50% untreated glass fibers

Tα* 214 °C 204 °C

E’25°C 287 MPa 973 Mpa

E’250°C 7.71 Mpa 56.4 Mpa

* Tα is given at 1 Hz.

It can be seen that Tg is almost practically not affected by the addition of fillers, even

in great quantities. This behavior is very different from that obtained in the presence of

silica powder.

In the previous chapter it was proved that silica interacts with the photoinitiator

causing a decrease of the curing kinetics. In the case of pulverized glass fibers this

interaction does not occur, as the amount of inorganic surface which can interact with

photoinitiator molecules is lower than for nanosilica.

As expected, the storage tensile module increases with respect to the pure monomer

due to the high stiffness of the inorganic fillers.

MICROBOND MEASUREMENTS

As far as the microbond test is concerned, microcomposites were prepared using

glass fibers treated with different concentrations of the silane agent. The experimental

conditions are listed in Tab. 4.3.

In Fig. 4.7 and Fig. 4.8 are reported the variations of calculated IFSS, dτ , as a

function of the silane agent percentage for DGE and CE matrix-based microcomposites.



94

Tab. 4.3: Concentration of silane agent used for grafting E-glass fibers used in

microbond tests.

CE matrix microcomposites DGE matrix microcomposites

silane

concentration of

silane solution

(% v/v)

silane

concentration of

silane solution

(% v/v)

0 0

0.1 0.1

0.25 0.25

0.5 0.5

CETS

1

GPTS

1

00,20,40,60,8

11,2

0 0,5 1

% GPTS

IFSS

(MPa

)

Fig. 4.7: IFSS of DGE matrix-based microcomposites as a function of the silane

concentration.



95

The adhesion results obtained using DGE as polymer matrix show that increase of

IFSS is reached in the presence of a much diluted silane agent concentration (0.1%),

then IFSS decreases.

1

1,5

2

2,5

3

3,5

0 0,5 1

% CETS

IFSS

(MPa

)

Fig. 4.8: IFSS of CE matrix microcomposites as a function of the silane

concentration.

The results obtained using CE as polymer matrix show good adhesion properties

even on untreated glass surfaces, in agreement with the literature8,9. These values can be

explained on the basis of the polar interactions between the two phases. Nevertheless,

interfacial shear strength of the interface remains very low compared to that obtained

for thermoset-glass fibers interfaces7.

By using glass fibers treated with CETS, the IFSS values decrease. This result could

be attributed to a decrease of the polarity of the surface in the presence of CETS even if

strong interactions, i.e. covalent bonding, are expected from the reactions of

cycloaliphatic epoxy groups from the grafted silane and CE matrix.

In fact when glass fiber surface is treated with a less polar silane coupling agent, as

propyltrimethoxy silane (C3) the IFSS values decrease sharply. The experimental

results are presented in Tab. 4.4.



96

Tab. 4.4: Results of interfacial shear strength of the interface from microbond test on

CE matrix-based microcomposites

(E-glass fibers treated with the different organosilane agents).

sample characteristics IFSS (MPa)

silane concentration (% v/v)

CETS 0 3.33

CETS 1 2.74

C3 1 1.32

As literature8 gives evidences of the benefits of treating fillers surface in order to

improve interface resistance during hydrothermal exposure, mechanical properties tests

were performed on samples before and after thermal or hydro-thermal (RH-controlled

chamber) ageing. In fact, hydrothermal ageing could be used to proof the existence of

covalent bonds based interfaces vs. secondary interactions based interfaces. Covalent

bonds remain after aging as physical ones are destroyed. This behavior could be

explained taking into account the interactions developed by epoxy-based polymer

networks and water molecules during the hydrothermal ageing9-11.

These interactions are mainly due to the diffusion of water molecules in the

interfacial region. Once there, water molecules break the weak polar-polar interaction

created between polymer matrix and inorganic surface, thus decreasing adhesion

between epoxy matrices (obtained mainly by polycondensation reactions) and untreated

glass surfaces decreases after hydrothermal ageing12.

The ageing conditions used are listed in Tab. 4.5.

In Fig. 4.9 and Fig. 4.10 the values of IFSS vs. the different percentages of silane

agent for the CE and DGE matrix microcomposites, after 7 and 14 days of ageing at

40°C and 95% RH are shown.



97

Tab. 4.5: Hydrothermal ageing conditions used for microcomposites.

time (days) temperature (°C) humidity (%RH)

7 40 50

7 40 95

14 40 95

4 80 50

4 80 95

4+4 80+80 50+95

The data presented in Fig. 4.9 and Fig. 4.10 evidence an increase of the IFSS

proportional to the ageing time.

The glass transition temperature of the DGE neat matrix: Tg DSC = 38°C, Tα DMTA

= 53°C, probably also affects the values of IFSS presented in Fig. 4.10. In this case, the

water adsorbed by the matrix acts as a plasticizer inducing a layer molecular mobility13,

i.e. decreasing Tg. The ageing at 40°C for long time period could modify the physical

behavior of the matrix itself, i.e. its ability to transfer the stress to the fiber through the

interface.

Confirms of the changes in the matrix mechanical behavior were clearly visible

during the test: the droplet had lost its elasticity (consequence: brittleness) to give

plastic deformation. In this case, an increase in the numerical value of the maximum

force before the detachment should be attributed to the absorption of energy from

plastic deformation from shear yielding of matrix network. In these conditions it

becomes difficult to quantify how much energy is spent for plastic deformation and how

much for the debonding. This effect of matrix plasticization is even more important in

our type of cycloaliphatic epoxy-based matrices compared to other types of matrices

such as the epoxy-amine or epoxy-anhydride matrices.



98

11,5

22,5

33,5

4

0 0,5 1% CETS

IFSS

(MPa

)

t=0t=7t=14

Fig. 4.9: IFSS of CE matrix-based microcomposites after 7 and 14 days of ageing at

40°C + 95% RH.

0

0,5

1

1,5

0 0,5 1

% GPTS

IFSS

(MPa

)

t=0t=7t=14

Fig. 4.10: IFSS of DGE matrix-based microcomposites after 7 and 14 days of ageing

at 40°C + 95% RH.

Taking into account these results, experimental parameters were made more severe.

Samples were aged at 80°C + 95% RH for 4 days. These conditions could not be

applied to DGE matrix microcomposites. In Fig. 4.11 are reported the values of IFSS vs.

the different percentages of silane used to graft glass fibers surface for CE matrix

microcomposites, after 4 days of ageing at 80°C and 95% RH.



99

1,5

2

2,5

3

3,5

0 0,1 0,2 0,3 0,4 0,5% CETS

IFSS

(MPa

)

t=0

t=4 (80°C+95% RH)

Fig. 4.11: IFSS of CE matrix-based microcomposites after 4 days of ageing at 80°C

+ 95% RH.

The same type of behavior already observed at 40°C + 95% RH (Fig. 4.9) is found.

In order to better understand these results, the number of the experimental variables

was cut down: only one parameter, temperature or humidity, was changed. CE matrix

microcomposites were thermally treated at 80°C for 4 days; in Fig. 4.12 the obtained

data are shown and related to those reported in Fig. 4.11.

1,52

2,53

3,54

4,55

0 0,1 0,2 0,3 0,4 0,5% CETS

IFSS

(MPa

)

t=0t=4 (80°C)t=4 (80°C+95% RH)

Fig. 4.12: IFSS of CE matrix-based microcmposites after 4 days of different types of

ageing.



100

Experimental evidences show that the thermal treatment increases the adhesion

compared with both non-aged and hydro-thermal aged samples. This result, correlated

with what reported in literature14,15, suggests that the thermal cycle at which is treated

the sample governs interfacial reactions. In fact thermal treatment has no effect on

microcomposites made with untreated glass fibers. These reactions are comparable to

the ones we have during thermal curing in creating an interface with enhanced

properties according to the scheme proposed in Fig. 4.13.

R Si

OH

O

O Si

OH

OH

R

Si

R Si

OH

OH

O Si

OH

OH

R R Si

OH

O

O Si

O

OH

R

R Si

OH

O Si

OH

OH

R

Si

silica or glass silica or glass

Fig. 4.13: Schematic representation of the reactions occurring between siloxane chains

during thermal curing.

To verify this assumption we chose to couple the thermal treatment to the

hydrothermal aging, following this schedule:

microcomposite preparation

↓

1st: thermal treatment: 80°C, 4 days

↓

2nd: hydrothermal aging: 80°C+95% RH, 4 days.



101

This multiple treatment has been applied to the CE matrix/ treated glass fibers (CETS

0.1% v/v) system, since it displays the highest interfacial adhesion values. Results are

presented in Fig. 4.14 and compared to the previous ones.

1,52

2,53

3,54

4,55

0 0,1 0,2 0,3 0,4 0,5% CETS

IFSS

(MPa

)

t=0

t=4 (80°C)

t=4 (80°C+95%RH)

t=4 (80°C)+(80°C+95%RH)

Fig. 4.14: IFSS of CE matrix-based microcomposites after 4 days of different types of

ageing.

From Fig. 4.14 it can be seen that the IFSS value for the latter system is lower than

the IFSS of the thermal treated one. This means that during the 2nd step of treatment, a

degradation process occurred at interface, due to the diffusion of water molecules.

Moreover this IFSS value is higher than the one measured after the hydro-thermal aging

at 80°C + 95% HR. This means that during the 1st step of the treatment reactions occur

in the interfacial region leading to enhanced mechanical properties of the interface. The

same test has been performed on the CE systems at 40°C for 7 days, obtaining similar

results (Fig. 4.15).

Considering all the experimental values, we can conclude that, in order to have a real

improvement of adhesion between polymer matrix and inorganic filler, a thermal

treatment after the grafting is necessary.



102

1,5

2

2,5

3

3,5

4

0 0,1 0,2 0,3 0,4 0,5% CETS

IFSS

(MPa

)

t=0t=7 (40°C+95% RH)t=7 (40°C)

Fig. 4.15: IFSS of CE matrix-based microcmposites after 7 days of different types

of ageing.

All the UV-cured epoxy systems described (CE, DGE) present polar characteristics,

therefore they have good adhesion on inorganic surfaces (glass) even in absence of

chemical surface treatments. As already discussed, the adhesion in this case can be

related to the polar-polar interactions created at interface between glass surface and

epoxy matrix.

In order to complete our informations on different composites, we have investigated

a different system, constitued by a non-polar matrix and glass fibers.

The matrix chosen was epoxidized acrylate soybean oil (SOA) combined with glass

fibres treated with a silane molecule having an acrylic functionality: 3-(trimetoxysilil)

propyl methacrylate (MEMO). Glass fibers modification with MEMO (1% v/v) was

carried out using the same grafting protocol already described for epoxysilanes grafted

fibers.

The same procedure and the same measurement technique adopted with the epoxy

systems were followed to prepare samples for microbond tests.

Microcomposites made with untreated and treated glass fibers were tested before and

after hydro-thermal ageing.

The ageing conditions used are described:

time (days) = 4

temperature (°C) = 60



103

humidity (%RH) = 95.

The data obtained were treated with the Kelly-Tyson relationship. The IFSS values

related to each system are reported in Tab. 4.6.

Tab. 4.6: IFSS values of SOA matrix microcomposites before and after hydro-

thermal ageing.

sample IFSS (MPa)

silane concentration

(% v/v)

0 0.54

1 0.7

0 0.63* MEMO

1 0.9* *after hydrothermal ageing (4 days at 60°C and 95% HR).

The surface treatment improves clearly the adhesion on fibers surface due to the

formation of stronger bonds between the non-polar matrix SOA and the modified

inorganic surface.

This improvement is still present after aging, thus confirming the formation of

covalent bonds.

It should be noticed that also in this case IFSS values increase after treatment, as

already seen for the UV-cured epoxy systems. It can be suggested that also in this case

the IFSS value increase is due to a thermal post-curing process which allows enough

mobility to lead to covalent bonding from reaction of grafted species.

MORPHOLOGY OF MICROBOND SAMPLES

In this section, the results of analyses performed by scanning electron microscopy,

SEM, on micro-composites after the microbond measurement are presented and

discussed.



104

Fig. 4.16 gives an “overall” looking at the samples used in this type of test: in the

picture it can be seen a typical section of an E-glass fiber with the cured microdroplets.

The red frame evidences a droplet after the detachment.

Fig. 4.16: SEM micrograph of a typical E-glass fiber microcomposite used in the

microbond test.

Fig. 4.17 shows a droplet profile after the detachment: the microcomposite is CE-

based matrix, on untreated E-glass fiber and the test was performed after 7 days of

ageing at 40°C. This picture has to be compared to Fig. 4.18, showing a CE-based

matrix microcomposite with treated (CETS 0.1% v/v) E-glass fiber, submitted to the

same ageing treatment. The grafting treatment performed assures the retention of

interface properties even after exposure in hostile environment: in fact, in these

conditions, the droplet debonding was caused by the rupture of the polymer matrix

while the interface displays good adhesion properties, as shown in Fig. 4.19.

These results can be attributed to the formation a strong interface, thanks to the

grafting procedure as well as to the thermal post-curing treatment.



105

Fig. 4.17: SEM micrograph of microcomposite CE-based matrix with untreated glass

fiber, aged for 7 days at 40°C.

Fig. 4.18: SEM micrograph of microcomposite CE-based matrix with treated glass

fiber (CETS 0.1% v/v), aged for 7 days at 40°C.



106

Fig. 4.19: Particular of SEM micrograph of microcomposite CE-based matrix with

treated glass fiber (CETS 0.1% v/v), aged for 7 days at 40°C.

ADHESION ON GLASS SHEETS EXPERIMENTS

Adhesion measurements were performed also on treated or untreated glass sheets

used as models of glass fibers systems.

Adhesion was measured by using the standard cross-cut method ASTM D3359.

The results obtained are reported in Tab. 4.7. They confirm that UV-epoxy systems

display good adhesion even on untreated glass surface. Only after the C3 treatment the

adhesion is absent due to the weak interactions between the apolar grafted alkyl chains

from the silane and the polar epoxy matrix.



107

The hydrothermal treatment, in the adopted conditions, determines a strong decrease

of adhesion on untreated glass surfaces, while good adhesion properties are still

displayed by the treated samples.

These results, although they are performed in very different conditions, are in

agreement with those obtained using the microbond testing.

Tab. 4.7: Adhesion results on treated or untreated glass sheets.

*sampled immersed in water.

4.4 Conclusions

In this chapter properties of micro- and nano-composites have been investigated.

Nano-composites, made with epoxy matrices and grafted or ungrafted nanosilica as

inorganic reinforcing agent, were characterized using dynamic-mechanical analyses.

The results show that nanocomposites present lower Tg values with respects to the

pure monomers (CE and DGE). The decrease in the Tg values indicate that a reduction

of the crosslinking density of the matrix is obtained in presence of silica, according to

the analyses of reaction kinetics, that indicate a decrease of the curing rate when silica is

added to the system.

The microbond technique was used to measure the interfacial adhesion for

microcomposites.

Adhesion

monomer glass

treatment

25°C

24 hours

60°C

6 hours*

60°C

24 hours*

100°C

3 hours*

untreated 100% 40% 0% 0%

CETS 100% 60% 20% 0% CE

C3 0% - -

untreated 100% 0% DGE

GPTS 100% 100% 60% 40%

untreated 0% SOA

MEMO 70% 40%



108

The data obtained for all the systems, expressed in terms of IFSS at interface,

evidenced that the best results are reached when a very low silane agent concentration is

used, i.e. as the grafted layer tends to be a monolayer.

Epoxy matrices display good adhesion even on untreated glass fibres, according to

literature6,12, because of the polar interactions. In agreement with these conclusions,

interfacial adhesion on C3 treated glass surface is very low.

CE-treated glass fibres systems show an increase of adhesion after a thermal

treatment. This can be explained by admitting the formation of a strong covalent bond

between the inorganic surface and the silane agent.

Moreover the thermal curing induces a better crosslinking of the silane film leading

to the formation of a highly branched siloxane chains15. The silanol groups of small

siloxane chains react with other of small, longer or branched chains, to form a polymer

network. This crosslinked silane layer can explain the adhesion improvement. In fact, if

the polysiloxane film is not completely condensed, it can display elastomer

characteristics, the so-called “rubber bumper” behaviour16; in these conditions the stress

transfer is less efficient as the shear stress is applied on a layer with lower modulus. The

thermal treatment leads to a post-condensation that induces stiffness in the polysiloxane

layer, thus more effective stress-transfer properties (expressed in term of higher values

of IFSS).

Moreover, the not completely condensed siloxane layer is still rich in SiOH groups

that can bond water molecules during a hydrothermal ageing process; these molecules

could act as plasticizer on the polymer matrix modifying the composite properties.

Adhesion decreases after hydro-thermal ageing; treated glass fibres show better

resistance than the untreated ones because they form stable covalent bonds with the

polymeric matrix.

Experiments done on micro-composites using SOA matrix highlight that the sizing

procedure is necessary to assure interfacial adhesion. Experimental results indicate that

also in this case a thermal curing of silane film is necessary to achieve better

performance.

Similar conclusions are obtained from the adhesion measurements performed on the

films UV-cured on treated or untreated glass sheets.

CHAP. 5 CONCLUSIONS

Conclusions

In this work the preparation of polymeric composites through cationic

photopolymerization and the properties of the obtained products were investigated.

First it has been set up and optimized the reaction to modify the surface properties of

inorganic fillers (nanosilica and glass fibers), in order to improve their interaction with

the polymeric matrix. This procedure was applied to nanosilica and glass fibers

reinforcing agents and its effectiveness was confirmed by TGA analyses and surface

properties evaluation.

The experimental results evidenced the modification of the surface properties of both

the fillers used.

Afterwards their influence on the composite preparation reaction was analyzed; the

photopolymerization reactions kinetics were evaluated, as well as the total conversion of

the photopolymerization reaction by means also of “real-time” techniques.

The results obtained using treated or untreated fillers were compared to those

attained with the pure resins. This second part of the work evidences that the

photopolymerizable active species interact with nanosilica during the UV-reaction,

reducing both the kinetics and total conversion of the reactive groups.

CHAP. 5 CONCLUSIONS


110

Finally the characteristics of the interfacial region were studied through adhesion

measurements performed with the microbond technique. Experimental results evidence

the presence of a correlation between adhesion and interphase thickness; as a

consequence of the applied treatment, adhesion between the two phases has been

improved leading to enhanced mechanical properties even in hydrolytic conditions.

The following developments of this research could be of interest:

to deepen the study of the interaction between the photopolymerization active

species and the inorganic surface. This could be useful in order to control them or to

reduce the lowering of the reaction kinetics.

Adhesion measurements results suggest that it could be useful to study

interphase reactions when a thermal post-curing treatment is applied, in order to

understand which part of the interphase is interested and to optimize the post-curing

procedure.

Experimental techniques

SPECTROSCOPIC ANALYSIS

FT-IR

A Fourier transform infrared spectrophotometer Mattson Genesis II was used.

Reaction kinetics were calculated measuring the variations during UV-curing of the

infrared bands characteristics of the monomer used. Conversion x after a given time t

can be calculated from the relationship

[ ][ ]0

1λ

λ

AA

x t−=

[Aλ]0 = integrated area under the infrared peak characteristic of the specific

functional group of the monomer, at the very beginning of UV reaction.

[Aλ]t = integrated area under the infrared peak characteristic of the specific

functional group of the monomer, at t time.

Samples were put between two KBr disks; analyses were made in absorbance

accumulating 64 scans, resolution 2 cm-1. All the results have been calculated using an

internal standard, the integrated area related to the stretching of the C=O group (it does

not participate to UV reaction) at 1730 cm-1. The FT-IR measurements done in this

work were made after finite intervals of irradiation time, realized using a UV lamp

(Helios Italquartz) which is separate from the FT-IR spectrophotometer.

EXPERIMENTAL TECHNIQUES


112

UV-vis

A UNICAM UV-vis spectrometer was used.

SURFACE ANALYSIS

Kruss goniometer

Measurements using a Kruss goniometer DSA 10 were carried out in air at room

temperature by the sessile drop technique bidistilled; bidistilled water was used. The

values of contact angles are obtained automatically through the evaluation of a

digitalized video image. In Fig. 1 a scheme of the apparatus used is presented.

Fig. 1: scheme of a Kruss goniometer.

The contact angle measuring instrument G10 applies the so-called sessile drop

method to determine contact angles.

If a pendant drop is in hydrometrical equilibrium then the analysis of its contour can

be used for the determination of the surface or interfacial tension. The contact angle of a

sessile drop on a solid surface is measured. The drop environment can be a gas or a

liquid. The contact angle measuring instrument offers the following opportunities:

measurements of contact angles of single drops of test liquids on a solid surface;

measurements of advancing and receding contact angles by volume control of

sample drop.



113

The surface of the solid to be measured should be horizontal. A drop of liquid is

placed on the solid. The contact angle is not independent of time effects: it can change

rapidly within seconds or minutes, depending on the liquid used and on the nature of the

solid. If the solid dissolves into liquid or the liquid changes its composition due to

different vapor pressure of different components, the contact angle will decrease

rapidly.

Dynamic contact angle describes the properties between solid-liquid surfaces during

the wetting procedure. It can be distinguished between advancing and receding contact

angle.

Advancing contact angle: an advancing contact angle can be measured with an

increasing liquid drop on a solid surface (Fig. 2). This can be done with a syringe that

remains in the drop during the measurement. A liquid drop is placed on the solid surface

and enlarged by pushing more liquid through the needle. While the volume increases

the border line between liquid and solid moves. The contact angle has to be measured

while the drop volume is increasing. Only the physical interaction is measured.

Receding contact angle: the receding contact angle provides information on the

macroscopic roughness of the surface. A relatively large drop is placed onto the surface.

The syringe remains in the drop during the measurement. The drop volume is decreased

by sucking liquid back into the syringe (Fig. 3). When the border line liquid-solid starts

moving, the contact angle has to be measured.

Fig.2: advancing contact angle produced with a syringe.



114

Fig. 3: receding contact angle produced with a syringe.

Parameters that influence contact angles are:

Temperature, for this reason it is necessary to keep a standard temperature for

the measurements.

Time, the contact angle can change during time (Fig. 4), caused by evaporation

of the liquid or by the forces between surface and liquid. The quality of the surface itself

is time-dependent; this is especially valid for polymers.

Volume, density and gravitation: the influence of gravitation on the contact

angle is closely connected to the density of liquid used and to the drop volume.

Drop size.

Fig. 4: time dependence of equilibrium contact angles.

Drop environment, it concerns the surrounding gas phase and the solid surface

that should be examined. The evaporation can influence the contact angle, as well as

adsorption processes of the test liquid at the solid surface.

Surface roughness.

Dosing rate.



115

Chan balance

The instrument used was a DCA 322 Cahn balance. Measurements with bidistilled

water were performed by means of the Wilhelmy technique with an immersion rate of

20 µm/s.

CALORIMETRIC TECHNIQUES

Photo-calorimeter (photo-DSC)

The instrument used was a calorimeter DSC 7 Perkin Elmer coupled with a UV-lamp

Perkin-Elmer DPA 7 XBO 450 W (Fig. 5).

The program used for the analyses is described:

36 minutes of total run

lamp on after 2 min

lamp off after 33 min

isothermal conditions (30°C)

nitrogen atmosphere.

The lamp intensity was adjusted to have I = 0.58 µW/cm2 at the selected wavelength.

Each sample pan was treated with a sulfo-chromic mixture in order to remove any

surface coating and to assure a good wettability to the sample.

The lamp is equipped with a monochromator to select the optimal wavelength. The

lamp intensity for each wavelength was measured with a radiometer Solatell 2000 and

the obtained emission spectrum is reported in Fig. 6.

Fig. 5: photo-DSC apparatus.



116

Fig. 6: photo-DSC UV-lamp intensity.

The DSC instrument registers the variations in the heat flow between sample and

reference; the measurements are done in a temperature-controlled environment, under

inert atmosphere (nitrogen).

Temperature and enthalpy are calibrated starting from the measure of the same

values of known materials (ex. indium). An example of obtained thermogram is

reported in Fig. 7.

Fig. 7: photo-DSC thermogram.

19

19,5

20

20,5

0 10 20 30 40time (min)

Hea

t flo

w (e

ndo

up)



117

Thermogravimetric analysis (TGA)

A TGA Mettler-Toledo Star System and a LECO TGA-601 were used.

Analyses were performed following the indicated programs:

Dynamo-mechanical analysis (DMTA)

A Rheometric Scientific MKIII apparatus was used.

Samples were tested in bending configuration, single cantilever; the temperature

range was from 0° to +250°C for the CE/silica systems and from -50° to +80°C for the

DGE/silica systems. Measurements were carried out at 1Hz frequency.

An apparatus for dynamo-mechanical analysis is represented in Fig. 8.

Fig. 8: DMA apparatus scheme.

Mettler-Toledo apparatus

sample weight = 10 mg

heating rate = 10°C/min

nitrogen atmosphere

temperature range = 50°-700°C

LECO TGA-601 apparatus

sample weight = 500 mg

heating rate = 1°C/min

temperature range = 50°-900°C

air atmosphere.



118

It is constituted by:

an electrical engine, to apply a cyclic stress that vary in a sinusoidal way;

a transducer, to measure the amplitude of the sample deformation;

an oven, to operate in an inert atmosphere, with a programmed temperature;

a thermocouple, to measures the effective sample temperature.

The instrument measures directly the following items:

the transducer position, in order to calculate the sample deformation, related to

its geometry;

the sample temperature;

the phase.

The measure can be done in a range of temperature or frequency; anyway to the

sample under a constant charge is superimposed a charge that varies according to the

following relationship:

( )tt ωσσ sin0=

Polymers present viscoelastic characteristics, so that a out of phase, δ, is observed

between the sinusoidal function that express the stress variation and the function

representing the correspondent strain, as shown in Fig. 9. The given energy is stored in

an elastic way, but also partially dissipated due to internal friction. If δ = 0, the material

presents elastic characteristics, while if δ = π/2, the material presents the characteristics

of a Newtonian Fluid. If 0 < δ < π/2 the sample behavior is viscoelastic and the quantity

of dissipated energy is proportional to the value of δ.

Fig. 9: viscoelastic deformation when a sinusoidal stress is applied.



119

The complex modulus can be defined as: '''* iEEE +=

Where E’ and E’’ are the storage and loss modulus respectively. In particular:

δ

δ

sincos

*''

*'

EEEE

=

=

A measure of the relationship between the energy loss as heat and the energy stored

as elastic deformation is given from the following equation:

δδδ tg

EE

EE

==cossin

*

*

'

''

MICROSCOPY

Atomic Force Measurement (AFM)

A Digital Instrument Nanoscope IIIa was used to carry out the measurements on

silicon wafer and glass fibers surfaces using the tapping mode.

This microscope (Fig. 10) is based on interaction forces created between the

cantilever and the sample surface, like repulsion forces, Van Der Waals attractions or

magnetic interactions.

Fig. 10: AFM image.



120

The tapping mode imaging is implemented by oscillating the cantilever assembly at

or near the cantilever’s resonant frequency using a piezoelectric crystal. The piezo

motion causes the cantilever to oscillate with high amplitude when the tip is not in

contact with the surface. The oscillating tip is then moved toward the surface until it

begins to lightly touch or “tap” the surface. During scanning the vertically oscillating tip

alternately contacts the surface and lifts off. As the oscillating cantilever begins to

intermittently contact the surface, the cantilever oscillation is necessarily reduced due to

the energy loss caused by the tip contacting the surface. This reduction in oscillating

amplitude is used to identify and measure surface features.

In Fig. 11 the cantilever oscillation amplitude in free air and during the measurement

is illustrated.

Fig. 11: cantilever oscillation in the tapping mode measurement.

Scanning Electron Microscopy (SEM)

A Hitachi S800 microscope was used during this work to investigate the surface

morphology of treated and untreated glass fibers. Samples were analyzed at 15 kV

tension.

In this type of analysis the sample is hit by high energy (300-600 kV) electrons

generated by a specific source. The most common sources are:

thermo-ionic source (W filament);

LaB6 catod source;

field emission source.



121

The interaction between the sample and the electron beam creates a great number of

events so a great number of signals. They are elaborated by different types of detectors.

Analyses performed with this technique present some advantages:

medium/high vacuum degree (10-6 torr);

high precision in the focalization of the electron beam on the sample;

high resolution dues to the short wavelengths present in the high energy electron

beam.

In Fig. 12 a scheme of a typical SEM apparatus is presented. It is composed by:

an electron source;

one or more magnetic lens to reduce the beam dimension to 5 nm;

a device to control the astigmatism;

a detector.

Fig. 12: SEM scheme.

The obtained electron beam, called “spot”, moves under lines on the sample; the

higher the numbers of lines generated (slow scan), the better the resulting image is.

In Fig. 13 the main areas of signals emission for a given sample are schematically

indicated. The secondary electrons (SE) give information on the morphology of the

material; the backscattered electrons (BSE) give qualitative information on sample

composition; the characteristics X-rays give elemental analyses of the sample (type,

distribution and quantity of the elements present in the sample).



122

Fig. 13: main signals (SE, BSE, X-rays) emission areas.

Samples for SEM analyses should be prepared in order to guarantee an electric

continuum between the beam, the sample and the sample holder. For samples with

insulating characteristics, for example ceramics and polymers, it is necessary to

metallize the surface before the analysis.

MECHANICAL ANALYSES

Microbond technique

A dynamometer Adamel Lhomargy DY 25 was used. For microbond measurements

a 10 N cell load was used; samples were pulled at 0.1 mm/min speed. The inferior

clamp is made of two razor blades; their position can be adjusted by using two

micrometric screws. The test is followed using a MICAM X video camera equipped

with a NAVITAR macro zoom. The apparatus is presented in Fig. 14.

Fig. 14: microbond technique apparatus.



123

All mechanical measurements were done at 20°C and 50% HR.

UV-LAMP DEVICES

Different UV-lamps have been employed in this work.

Fusion lamp: an industrial UV-lamp with a standard mercury-filled bulb,

equipped with a conveyor belt (Fig. 15); by adjusting the conveyor belt speed it is

possible to change the sample exposition time. The lamp total intensity (measured with

a Solatell 2000 radiometer) is 370 mW/cm2. Its emission spectrum is reported in Fig.

16.

Fig. 15: FusionUV- lamp equipment.

Fig. 16: Fusion UV- lamp emission spectrum.

Helios Italquartz: is a laboratory UV-lamp with a standard mercury-filled bulb,

equipped with a support which allows changing the distance of the sample from the

lamp; there is possibility of operating in nitrogen atmosphere. The lamp intensity,

related to the different distances of the support used, varies from 10 to 50 mW/cm2.

REFERENCES

PART 1

CHAP. 1

[1] MATTHEWS F.L., RAWLINGS R.D. Composite Materials: Engineering and

Science, London: Chapman&Hall, 1993.

[2] KELLY A. Concise Encyclopedia of composites materials, New York, USA:

Pergamon Press, 1995.

[3] SEYMOUR R. The role of Fillers and Reinforcements in Plastic Chemistry. In:

DEANIN R.D., SCHOTT N.R. Fillers and Reinforcements for Plastic, Washington,

D.C., USA: ACS, 1974, pp. 1-6 (Advances in Chemistry Series, Vol. 134).

[4] PIGGOT M.R. The Effect of the Interface/Interphase on Fiber Composite

Properties. Polymer Composites, 1987, vol. 8, n. 5, pp. 291-287.

[5] PIGGOT M.R., SANADI A., CHUA P.S., ANDISON D. Mechanical

Interactions in the Interphasial Region of Fibre Reinforced Thermosets. In: ISHIDA H.,

KOENIG J.L. Composite Interfaces, New York, USA: North-Holland, 1986, pp. 109-

121.

REFERENCES

Photopolymerized micro- and nano-composites: interface chemistry and its role on interface adhesion

125

[6] PLUEDDEMANN E.P. Silane Coupling Agents New York and London: Plenum

Press, 1982.

[7] HOH K., ISHIDA H., KOENIG J.L. The diffusion of Epoxy Resin into a Silane

Coupling Agent Interphase. In: ISHIDA H., KOENIG J.L. Composite Interfaces, New

York, USA: North-Holland, 1986, pp.251-263.

[8] KIM J.K. Engineered interfaces in fibre reinforced composites. Paris Elsevier,

1998, pp. 171-182.

[9] KOENIG J.L. FT-IR Studies of Interfaces. In: LEYDEN D.E. Silanes, Surfaces,

and Interfaces. Colorado USA: Gordon and Breach Science Publishers, 1985 pp. 43-57

(Vol. 1).

[10] ISHIDA H., KOENIG J. L. Fourier Transform Infrared Spectroscopic Study of

the Silane Coupling Agent/Porous Silica Interface. Journal of Colloid and Interface

Science, 1978, vol. 64, n. 3, pp. 555-564.


the Structure of Silane Coupling Agent on E-Glass Fiber. Journal of Colloid and

Interface Science, 1978, vol. 64, n. 3, pp. 565-576.

[12] ISHIDA H., KOENIG J. L. An Investigation of the Coupling Agent/Matrix

Interface of Fiberglass Reinforced Plastics by Fourier Transform Infrared Spectroscopy.

Journal of Polymer Science: Polymer Physics Edition, 1979, vol. 17, pp. 615-626.

[13] KOENIG J.L., SHIH P.T.K. Raman Studies of the Glass Fiber-Silane-Resin

Interface. Journal of Colloid and Interface Science, 1971, vol. 36, n. 2, pp. 247-253.

[14] PHILLIPS L.V., HERCULES D.M. Surface Spectroscopic Studies of

Organosilane Layers on Glass. In: LEYDEN D.E. Silanes, Surfaces, and Interfaces.

Colorado USA: Gordon and Breach Science Publishers, 1985, pp. 235-264 (Vol. 1).

REFERENCES


126

[15] KIM J., HODZIC A. Nanoscale characterization of thickness and properties of

interphase in polymer matrix composites. Journal Of Adhesion, 2003, n. 79, pp. 383-

414.

[16] MAI K., MÄDER E., MΫHLE M. Interphase characterization in composites

with non-destructive methods. Composites: Part A, 1998, n. 29A, pp. 1111-1119.

CHAP. 2

[1] MEHNERT R., PINCUS A., JANORSKY I., STOWE R., BEREJKA A. UV &

EB Curing Technology & Equipment London: John Wiley & sons ,1998, Vol. I

[2] DECKER C. Photopolymerization and Ultraviolet Curing of Multifunctional

Monomers. In: Process of Polymers, Weinheim, Germany: H.E.H.Meijer, 1997, pp.

617-635.

[3] LEDWITH A., Photochemistry and Materials: Progress and Opportunity. In:

KELLY J.M., McARDLE C.B., MAUNDER M.J. Photochemistry and Polymeric

Systems, Cambridge: Royal Society of Chemistry, 1193, pp. 1-14.

[4] DECKER C. Photoinitiated crosslinked polymerization. Progress in Polymer

Science, 1996, vol.21, pp. 593-650.

[5] DECKER C., Photo-induced Polymerization-Kinetic Study of Ultrafast Processes

Induced by UV Radiations or Lasers. In: KELLY J.M., McARDLE C.B., MAUNDER

M.J. Photochemistry and Polymeric Systems, Cambridge: Royal Society of Chemistry,

1193, pp. 32-46.

REFERENCES


127

[6] CRIVELLO J.V., LAM J.H.W., The Photoinitiated Cationic Polymerization of

Epoxy Resins. In: BAUER R.S. Epoxy Resin Chemistry, Washington DC: ACS, 1979,

pp. 1-15 (ACS Symposium Series, Vol. 114).

[7] DECKER C., MOUSSA K. Kinetic Study of the Cationic Photopolymerization of

Epoxy Monomers. Journal of Polymer Science: Part A: Polymer Chemistry, 1990, vol.

28, pp. 3429-3443.

[8] CRIVELLO J.V. Synergistic Free Radical Effects in Photoinitiated Cationic

Polymerization. In Photoinitiated Polymerization, Washington D.C.: ACS, 2003, pp.

178-186 (ACS Symposium Series Vol. 847).

[9] CRIVELLO J.V. The Discovery and Development of Onium Salt Cationic

Photoinitiators. Journal of Polymer Science: Part A: Polymer Chemistry, 1999, vol. 37,

pp. 4241-4254.

CHAP. 3

[1] SHI W., RÅNBY B. UV Curing of Composites Based on Modified Unsatured

Polyester. Journal of Applied Polymer Science, 1994, n. 51, pp.1129-1139.

[2] SHI W., RÅNBY B. Photopolymerization of Dendritic Methacrylated Polyesters.

III. FRP Composites. Journal of Applied Polymer Science, 1996, n. 59, pp. 1951-1956.

[3] DECKER C., ZAHOUILY K., KELLER L., BENFARHI S., BENDAIKHA T.

Ultrafast synthesis of bentonite-acrylate nanocomposite materials by UV-radiation

curing. Journal of Material Science, 2002, n. 37, pp. 4831-4838.

[4] CRIVELLO J.V., NARAYAN R., STERNSTEIN S.S. Fabrication and

Mechanical Characterization of Glass Fiber Reinforced UV-Cured Composites from

REFERENCES


128

Epoxidized Vegetable Oils. Journal of Applied Polymer Science, 1997, n. 64, pp. 2073-

2087.

[5] GLAUSER T., JOHANSSON M., HULT A. A comparison of radiation and

thermal curing of thick composites. Macromolecular Material Engineering, 2000, pp.

274, 25-30.

PART 2

CHAP.1

[1] VANSANT E.F., VAN DER VOORT P., VRANCKEN K.C. Characterization

and chemical modification of the silica surface, Amsterdam: Ed. Elsevier.

[2] BARTHEL H., HEINEMANN M., STINTZ M., WESSELY B. Particle Sizes of

Fumed Silica. Chemical Engineering and Technology, 1998, n. 21, pp. 9.

[3] HAIR M.L. Silica Surfaces. In: LEYDEN D.E. Silanes, Surfaces, and Interfaces.

Colorado USA: Gordon and Breach Science Publishers, 1985 pp. 25-41 (Vol.1).

[4] MATTHEWS F.L., RAWLINGS R.D. Composite Materials: Engineering and

Science, London: Chapman&Hall, 1993.

[5] SIMS G.D., BROUGHTON W.R. Glass Fiber Reinforced Plastic-properties.

Elsevier, 2000, pp. 151-197 (Comprehensive Composites Materials, Vol. 2).

[6] ZINCK P. De la caractérisation micromécanique du vieillissement

hydrothermique des interphases polyepoxyde-fibre de verre au comportement du

composite unidirectionnel. Relations entre les échelles micro et macro Thèse du

doctorat, Lyon : INSA-Lyon, 1999.

REFERENCES


129

[7] PLUEDDEMANN E.P. Silane Coupling Agents York and London: Plenum

Press, 1982.

[8] DIBENEDETTO A.T. Evaluation of fibre surface treatments in composite

materials Pure & Applied Chemistry, 1985, vol. 57, n. 11, 1659-1665.

[9] CLARK H.A., PLEUDDEMANN E.P. Modern Plastic, 1963, vol. 40 n. 6, pp.

133.


the Structure of Silane Coupling Agent on E-Glass Fiber. Journal of Colloid and

Interface Science, 1978, vol. 64, n. 3, 565-576.

[11] ISHIDA H., KOENIG J. L. An Investigation of the Coupling Agent/Matrix

Interface of Fiberglass Reinforced Plastics by Fourier Transform Infrared Spectroscopy.

Journal of Polymer Science: Polymer Physics Edition, 1979, vol. 17, 615-626.

[12] Cyracure Cycloaliphatic Epoxides (Technical informations).

[13] PEARCE C.I., GUTHRIE J.T. The development and characterisation of coating

formulations from selected epoxidised sunflower oil derivatives. Surface Coatings

International Part B: Coatings Transactions, 2002 vol. 85, B3, pp. 169-142.

[14] CIBA Photoinitiators for UV Curing-A Formulator’s Guide (Technical

informations).

[15] MILLER A., GOU L., NARAYANAN V., SCRANTON A.B. Modeling of

photobleaching for the Photoinitiation Thick Polymerization Systems. Journal of

Polymer Science Part A: Polymer Chemistry, 2002, vol. 40, pp. 793-808.

REFERENCES


130

CHAP. 2

[1] LIU Q., DING J., CHAMBERS D.E., DEBNATH S., WUNDER S., BARAN

G.R. Filler-coupling agent-matrix interactions in silica/polymethylmethacrylate

composites. Journal of Biomedical Materials Reserch, 2001, n. 57, pp. 384-393.

[2] SUZUKI K., SIDDIQUI S., CHAPPELL C., SIDDIQUI J. A., OTTENBRITE R.

M., Modification of Porous Silica Particles With Poly(acrylic acid), Polymers for

Advanced Technology, 2000, vol. 11, pp. 92-97.

[3] PLUEDDEMANN E.P. Silane Coupling Agents. New York and London: Plenum

Press, 1982.

[4] JUN L., SIDDIQUI J.A., OTTENBRITE R.M. Surface Modification of

Inorganic Oxide Particles with Silane Coupling Agent and Organic Dyes. Polymers for

Advanced Technology, 2001, n. 12, pp. 285-292.

[5] PLUEDDEMANN E.P. Silane Compounds for Silylating Surfaces. In: LEYDEN

D.E. Silanes, Surfaces, and Interfaces. Colorado USA: Gordon and Breach Science

Publishers, 1985 pp. 1-24 (Vol. 1).

[6] MAI K., MÄDER E., MΫHLE M. Interphase characterization in composites with

non-destructive methods. Composites: Part A, 1998, n. 29A, pp. 1111-1119.

[7] NAKAMURA Y., NAGATA K., TOBITA Y., YOKOUCHI N., NIGO H., IIDA

T. Surface structure of silane-treated glass beads and its influence on the mechanical

properties of filled composites. Journal of Adhesion Science and Technology, 2002, vol.

16, n. 5, 523-542.

REFERENCES


131

CHAP. 3

[1] PLUEDDEMANN E.P. Catalytic Effects in Bonding Thermosetting Resins to

Silane-Treated Fillers. In: DEANIN R.D., SCHOTT N.R., Fillers and Reinforcements

for Plastic, Washington, D.C., USA: ACS, 1974, pp. 86-94, (Advanced in Chemistry

Series, Vol. 134).

[2] DECKER C., MOUSSA K. Kinetic Study of the Cationic Photopolymerization of

Epoxy Monomers. Journal of Polymer Science: Part A: Polymer Chemistry, 1990, vol.

28, pp. 3429-3443.

[3] MALUCELLI G., CHEN Y., PRIOLA A., BONGIOVANNI R.,

SANGERMANO M. Preparation of silica-or fiber glass-reinforced composites through

cationic photopolymerization technique. Materials Engineering, 2002, vol. 13, n.1 pp.

99-108.

[4] MALUCELLI G., SANGERMANO M., PEDITTO F., BONGIOVANNI R.,

PRIOLA A. Silica or fibre glass reinforced composites via photopolymerisation of

acrylate systems. Plastic, Rubbers and Composites, 2003, vol.32, no.3, pp. 93-97.

CHAP. 4

[1] AMDOUNI N., SAUTEREAU H., GERARD J.F. Epoxy Composites Based on

Glass Beads. I. Viscoleastic Properties. Journal of Applied Polymer Science, 1992, vol.

45, pp.1799-1810.

[2] GAUR U., MILLER B. Microbond Method for Determination of the Shear

Strength of a Fiber/Resin Interface: Evaluation of Experimental Parameters. Composites

Science and Technology, 1989, n. 34, pp. 35-51.

REFERENCES


132

[3] MILLER B., GAUR U., HIRT D.E. Measurement and Mechanical Aspects of the

Microbond Pull-Out Technique for Obtaining Fiber/Resin Interfacial Shear Strength.

Composites Science and Technology, 1991, vol. 42, pp. 207-219.

[4] PIGGOTT M.R. The single-fiber pull-out method: its advantages, interpretation

and experimental realization. Composite Interfaces, 1993, vol. 1, n. 3, pp. 211-223.

[5] ZINCK P., WAGNER H.D, SALMON L., GERARD J.F. Are microcomposites

realistic models of the fiber/matrix interface? I. Micromechanical modeling. Polymer,

2001, n. 42, pp. 5401-5413.

[6] MAI K., MÄDER E., MΫHLE M. Interphase characterization in composites with

non-destructive methods. Composites: Part A, 1998, n. 29A, pp. 1111-1119.

[7] ZINCK P. De la caractérisation micromécanique du vieillissement

hydrothermique des interphases polyepoxyde-fibre de verre au comportement du

composite unidirectionnel. Relations entre les échelles micro et macro Thèse du

doctorat, Lyon, INSA-Lyon, 1999.

[8] EAKINS W.J. Effect of water on glass fiber-resin bonds: Interfaces in

Composites Materials, ASTM STP 452, USA: American Society for Testing and

Materials, 1969, pp. 137-148.

[9] BANKS L., ELLIS B. The glass transition of an epoxy resin and the effect of

adsorbed water. Polymer Bulletin, 1979, vol. 33, pp. 377-382.

[10] FULLER R.T., FORNES R.E., Memory J.D. NMR study of water absorbed by

epoxy resin. Journal of Applied Polymer Science, 1979, vol. 23, pp. 1871-874.

[11] MOY P., KARASZ F.E. Epoxy-water interactions. Polymer Engineering

Science, 1980, vol. 20, n. 4, pp. 315-319.

REFERENCES


133

[12] KEUSCH S., HAESSLER R. Influence of Surface Treatment of Glass Fibers

on the Dynamic Mechanical Properties of Epoxy Resin Composites. Composites: Part

A, 1999, vol. 30, pp. 997-1002.

[13] CHATEAUMINOIS A. Comportement viscoélastique et tenue en fatigue

statique de composites verre/époxy. Influence du vieillissement hygrothermique, Thése

de Doctorat, Lyon, UCB Lyon I, 1991, p. 209.

[14] WU H.F., DWIGHT D.W., HUFF N.T. Effects of Silane Coupling Agents and

Performance of Glass-Fibers-Reinforced Polymer Composites. Composites Science and

Technology, 1997, n. 57, pp. 975-983.

[15] FRANQUET A., TERRYN H., VEREECKEN J. IRSE Study on effect of

thermal curing on the chemistry and thickness of organosilane films coated on

aluminium. Applied Surface Science, 2003, n. 211, pp. 259-269.

[16] GERARD J.F., AHLSTROM C., ROUBY D. Introduction of Elastomeric

Interlayers Between the Fiber and The Epoxy Matrix. Study of Monofilament and

Unidirectional Composites. In: VERPOEST I., JONES F. Interfacial Phenomena in

Composite Materials ’91, Oxford: Butterworth and Heinemann Ltd., 1991, pp. 57-60.

photopymerized micro- and nano-composites - les thèses de...

Documents