Polymer encapsulation of inorganic submicron particles inaqueous dispersionCitation for published version (APA):Caris, C. H. M. (1990). Polymer encapsulation of inorganic submicron particles in aqueous dispersion.Technische Universiteit Eindhoven. https://doi.org/10.6100/IR332570

DOI:10.6100/IR332570

Document status and date:Published: 01/01/1990

Document Version:Publisher’s PDF, also known as Version of Record (includes final page, issue and volume numbers)

Please check the document version of this publication:

• A submitted manuscript is the version of the article upon submission and before peer-review. There can beimportant differences between the submitted version and the official published version of record. Peopleinterested in the research are advised to contact the author for the final version of the publication, or visit theDOI to the publisher's website.• The final author version and the galley proof are versions of the publication after peer review.• The final published version features the final layout of the paper including the volume, issue and pagenumbers.Link to publication

General rightsCopyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright ownersand it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.

• Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying the publication in the public portal.

If the publication is distributed under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license above, pleasefollow below link for the End User Agreement:www.tue.nl/taverne

Take down policyIf you believe that this document breaches copyright please contact us at:openaccess@tue.nlproviding details and we will investigate your claim.

Download date: 04. Jun. 2021

https://doi.org/10.6100/IR332570https://doi.org/10.6100/IR332570https://research.tue.nl/en/publications/polymer-encapsulation-of-inorganic-submicron-particles-in-aqueous-dispersion(916abef9-0efe-4629-9ccf-2420c187dfef).html

POLYMERENCAPSULATION OF INORGANIC SUBMICRON PARTICLES

IN AQUEOUS DISPERSION

C.H.M. CARIS

POLYMERENCAPSULATION OF INORGANIC SUBMICRON PARTICLES

IN AQUEOUS DISPERSION

PROEFSCHRIFT

ter verkrijging van de graad van doctor aan de Technische Universiteit Eindhoven, op gezag van de Rector Magnificus, prof. ir. M. Tels, voor een commissie àangewezen door het College van Dekanen in het openbaar te verdedigen op vrijdag 1 juni 1990

te 16.00uur

door

CAROIA HUBERTA MARIA (ROBERTA) CARIS geboren te Weert

Druk: Bock en Offsetdrukkerij Letru. Helmond. 0492Q-J7797

Dit proefschrift is goedgekeurd door de promotoren prof. dr. ir. A.L. Oerman en prof. dr. B.H. Bijsterbosch en de copromotor dr. A.M. van Herk

This investigation was financially supported by the 'OSV' (i.e. The Netherlands Organization for the Actvancement of Paint and Coating Research).

Gutta cavat lapidem,

non vi sed saepe cadendo

(Latijns Spreekwoord)

aan mijn ouders,

aan Jan

Table of Contents

1 Introduetion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1

1.1 Applications of polymer-encapsulated particles . . . . . . . . . . . .1

1.2 Lirerature survey on the encapsulation of inorganic particles

1.2.1

1.2.2

1.3

1.4

1.5

with polymer . . . . . . . . . . . . . . . . . . . . . . . . . . . .2

Polymer adsorption at the inorganic surface . . . . . . . . . .2

Polymerization at the inorganic surface . . . . . . . . . . . . . . . .4

Aim of the present thesis . . . . . . . . . . . . . . . . . . . . . . . .9

Scope of this thesis . . . . . . . . . . . . . . . . . . . . 1 0

Outline of thesis . 10

2 Experimental ................................ 13

3

2.1 Synthesis and characterization of inorganic particles . . . . . . . . 13

2.2 Modification of particles with titanates ............... 14

2.3 Initiator formation at the partiele surface .............. 15

2.4 Surfactant adsorption at (modified) inorganic particles . . . .. 18

2.5 Polymerizations ........................... 18

Modification of inorganic particles with titanates .

3.1 Introduetion . . . . . . . . . . . . . . . . . .

3.2 The inorganic particles . . . . . . . . . . . . .

3.3 Chemica! structure and stability of titanates

3.4 Surfactant adsorption at the partiele surface ..

. . . . . . . . . . . 23

... 23

. .. 24

... 28

. 39

4 Polymerizations at the surface of hydrophobic Ti02 particles . 45 4.1 Introduetion . . . . . . . . . . . . . . . . . . . . . . . . . . 45

4.2 Polymerization k:inetics . . . . . . . . . . . . . . . . . . . . . . . 46

4.2.1 Effect of hydrapbobic particles on polymerization k:inetics . . . . . 46

4.2.2 Effect of surfactant concentration . . . . . . . . . . . . . . . 53

4.2.3 Effect of mixing conditions ..................... 55

4.2.4 Effect of Ti02 content • . . • • . • • . . . . • • • . • . . • • . • . 57

4.2.5 The nature of the monomer . . . . • • . • • • • . • • . . . • . . . 57

4.2.6 Effectofinitiatorconcentration ••.•.••....•....•.. 61

4.2.7 Oppositely charged initiator and surfactant species • . . . . • . • • 63

4.3 Concluding Remarks . • • • • • . . • . • . . . • . • . . . . . • • • 65

S Copolymerization at the surface ofTi02 •..... · ... · ......... 83

5.1 Introduetion ......•..•...••.•••...•.•.•.. 83

5.2 Monoroer conversion roeasurements • • • • • . . . . . . . . • . • 83

5.3 Copolymerization kinetics . • • • • • • . • • • • • . . • . • . • • • 87

5.3.1 Effect of stirring condinons .•..•.••••••......•.• 87

5.3.2 Effect of surfactant concentradon • . . • • • . . • . . . . . . . • . 89

5.3.3 Effect ofinitiator concentration ..••..••......••••. 91

5.3.4 Effect of Ti02 content . . . . . . • . • . . • . . . . . . . . . . . • 92

5.3.5 Effect of the nature of the modification • • • . • . . . . . . . • . • 92

5 .. 3.6 Polymerizations at "low" monomer concentranon . . . • . . . . . 93

5.4 Concluding Remarks • • . . • . • • • • • . • . . . . . . . • • • . . 94

6 Initiation atthe partiele surface . • . . . . . . . • . . . . . . . . . . . . 97

6.1 Introduetion . • . . • . . • . . • • . • . . . . . . . . . . . . • . . 97

6.2 The photoactivity ofTi02 .......•.........•••.• 97

6.3 Chemically bound azo initiators at the partiele surface . . . . . . 10 l

6.4 Polymerization with an initiator adsorbed at the partiele surface . 110

6.5 Concluding Remarks . . . . • . . . • . • • . . . . • • . . . . . . 111

7 Polymerization Produels . . . . . . . . . . , . . . . . . . . . . . . . . 117

7.1 Introduetion • . . . . . . . • . . . . . • . . . . . . . • • . . . . 117

7.2 Polymerizations at high monoroer concentration • • . • . • . . . 118

7.3 Polymerizations at low monoroer concentrations. • . • • • . . . . 127

7.4 Coneludi~gRemarks ....••••......•..••..... 134

8 Final Conclusions and Suggestions for forther Research . . . . . . . 137

9 References ..•...........••...•............. 143

Summary .. 155

Samenvatting 159

Curriculum Vitae 165

Dankwoord ................................... 166

Chapter 1

Introduetion

1.1 Applications of polymer-encapsulated particles

In literature several processes have been described to obtain particles that consist

of an inorganic core and a polymer shell. Polymer encapsulated particles offer

interesting prospects in applications requiring a good coupling between polymer

matrix and inorganic partiel es. Mechanica! properties, like tear strength of e.g. SBR

rubbers, will improve when an interfacial bond between the rubber and the CaC03

filler exists [1]. Composite sheets ofpolymer coated carbon powders give excellent

thermal radiation sheets for devices such as transistors, diodes, integrated circuits etc.,

and can be used as electroconductive addidves to plastics [2]. Besides, composite

matenals can also be applied as electricallyresistive harriers [3]. Because oftheir high

specific modulus and Ioss tangent, expanded graphite-PMMA composites form excel-

lent diaphragm matenals for high fidelity loudspeakers [2]. Polymer encapsulated

inorganic matenals cannot only improve the properties of products, but may also offer

many advantages in processing, because of their good dispersability in organic media

[4], and for instanee because they are directly moldable. These particles also offer

prospects as carriers for catalysts, as diagnostic matenals in human and animal health

applications [5], toners in electro-photographic applications, moldable magnetic pow-

ders, and pigmentsin paint and ink formulations [6]. The latter application is of special

interest in water based paints. These are becoming more and more important, as they

cause less air pollution upon drying than the conventional oil based paints. Besides,

there are more advantages to water based paints, such as the ease of application and

1

Chapter 1

the easy clean-up of brushes and rollers with soap and water. However, one of the probieros in latex paint technology is the agglomeration of untreated pigment and filler particles during film formation [2]. The opacifying capability or hiding power depends on three factors: light absarpdon due to the inherent coloration of the pigment particles, light reflectance (which is fixed for any given combination of binder and pigment),

and light reflection and dispersion as a result of scattering by the surfaces of pigment particles in the solidified paint film. Particularly in a white paint the net hiding power is essentially determined by the scattering effectiveness of the dispersed pigment layer, which in turn is highly influenced by the regular positioning of the pigment particles in the film, as well as the regularity of the partiele size distribution. Light scattering power and hiding coverage of the p;ûnt will be optimal if the pigment particles are

uniformly spaeed apart an ideal distance. If the partiele spacing is irregular and if pigment agglomerates are present that deviate considerably from the desired uniform pigment size, the light scattering will be degraded. as will be the hiding power. Encapsulating the hydrophilic pigment particles with polymer will improve their compatibility with the organophilic polymer binder, and agglomeration. can be prevented. Thus film appearance and performance might be improved: higher gloss and better scrub resistance, stability, weather resistance, water ( vapour) permeability,

modulus, colour stability and durability can be obtained [6]. Since the pigment concentration, as aresult of the optimum use of all pigment particles, might be reduced, this may also result in better economics.

1.2 Llterature survey on the encapsulation of lnorganlc particles wlth polymer

The processes described in literature to obtain polymer encapsulated inorganic

particles can be divided into two main groups: already existing polymer ebains are coupled to the inorganic surface, or polymerization is carried out at the partiele surface.

12.1 Polymer adsorption at the inorganic surface

Polymer adsorption at the inorganic surface was described by Princen et al. [7] and by Biglieri c.s. [8]. The interaction between polymer and inorganic surface depends

on the polymer composition and the nature of the surface. The presence of certain

oxides at the partiele surface may enhance adsorption. Makioen [9] found that the

2

Introduetion

A) +'Vllll-

B} + Y"VW-

()

Figure 1.1 Reaction of polymers with partiele surface by activadon of surface

and/or polymer. A) Activation of partiele surface; B) Activation of polymer; C) Ac-

ti vation of partiele surface and polymer

amount of adsorbed sodium poly(acrylate) could be largely increased by using Ti02

with Al203 at the surface insteadof pure Ti02. Thus, the arnount of adsorbed polymer

can be influenced by activaring the partiele surface. It is also possible to activate the

polymer insteadof the inorganic partiele. Hamann et al. [10-12] activated polymers

by attaching silane endgroups to the polymer chains, that can react with Si02 surfaces.

The amountofpolymerchemisorbed at the partiele surface depends u pon the molecular

weight of the polymers, the specific surface area of the particles, and the number of

reactive sites on the Si02 surface. During the frrst part of the reaction adsorption takes

place very rapidly, but eventually the polymer diffusion rate will decrease, as aresult

of steric bindrance by the layer already present at the partiele surface.

It was found that pigment Sedimentarlon in paints can only beprevented by polymer

adsorption at a sufficient layer thickness (thus providing steric stabilization) [ 13]. The

layer thickness required depends on the molecular weight of the polymer (very short

ebains provide poor dispersion stability, but very long chains result in entanglement

and thus also in flocculation [14)). In case of a copolymer, it also depends on its

composition: in general block copolymers proved to be more effective than random

copolymers [15]. The solvent can also play a very important role [16]: a poor solvent

enhances polymer adsorption, whereas a good solvent shows the opposite effect and

3

Chapter 1

can even enhance desorption. Meguro et al. [17 ,18] used another metbod of enhancing the adsorption of polymers at inorganic surfaces. They synthesized polystyrene by means of an emulsion polymerization process with an amphoteric surfactant (N,N-dimethyl-n-laurylbetaine ). Dispersion, flocculation and redispersion of ( encapsulated)

Ti02 or Si02 particles then were smdied as a function of pH and polymer concentra-tion. They also studied the adsorption of oligomeric surfactants on iron oxides, in relation totheir chemica! composition and pH [19].

In all these cases adsorption strongly dependedon polymer composition, nature of

the inorganic surface, and conditions lik:e pH and the presence of electrolytes. Accord-ing to Dietz and Hamann [20] the presence of po lar additives can result in partlal or even total desorption of the polymer layer (depending on the interaction between polymer and solvent on one hand and additive, solvent and surface on the other hand, and on the molecular weight of the polymer). Besides, the surface of inorganic particles is mosüy covered with small polymer particles, resulting in uneven coatings and the formation of polymer-bonded, coated agglomerates of mineral particles [13]. Thus, in most cases non-uniform polymer layers at the partiele surface will be obtained. Processes to obtain adsorption or chemisorption of polymers on inorganic surfaces are schematically shown in figure 1.1.

1.2.2 Polymerizarion at the inorganic surface

There are many ways to carry out a polymerization at the surface of inorganic particles. For instanee Fukano and Kageyama [21,22] describe the radiation-induced bulk polymerization of styrene and methyl methacrylate at the surface of inorganic

particles lik:e several types of Si02, zeolite and A1203. They found that the presence of aluminium at the surface enhanced grafting of the formed polymer. The molecular weight distribution of the grafted polymer appeared to depend on pore-size, partiele

diameter, number of hydroxylic groups and amount of water adsorbed at the surface.

Iler and Lipscomb [23] describe a metbod by which a radical initiator is adsorbed at the partiele surf ace, and monoroer vapour is admitted to the particles. Degtyarev et al. [24] also describe the polymerization of MMA vapour adsorbed at the surface of

various roetal oxides. The polymerization is initiated by radicals, formed in the oxide by irradiating the mixmre. Solomon et al. [25-30] studied the solution polymerization

of styrene and other vinyl monomers, initiated by compounds adsorbed at the surface

of clay minerats and Ti()z. Decomposition of different types of initiators is highly

4

Introduetion

Al -x·+ M - -x~· B) + M" -() + ·vvv- +Y

Figure 1.2 Polymerizations at the surface of inorganic panicles in salution or bulk

processes. A) Initiation at the paniele surface; B) Copolymerization.at the paniele

surface; C) Termination at the paniele surface

influenced by surface acidity: acidic groups appeared to catalyse cationic polymeriza-

tions and the heterocyclic decomposition of peroxides, but inhibited radical initiation.

Hamann et al. [31-35] extensively studied bulk and salution polymerizations in

the presence of panieles like Ti02, Si02 and Ah03. Grafring occurred by a reaction

with adsorbed (co)monomers (like methacrylic acid) or initiators [31]. It is also

possible to use chemisorbed (co)monomers or initiators [32]. They describe several

methods to obtain a covalent bond between initiator and Si02 surface [33-35]. In this

way more polymer can be grafted to the inorganic substrate than by using an initiator

that has only been adsorbed at the surface. Laible and Hamann [10] also give a survey

of the methods available to obtain a chemical bond between polymer and substrate. A

covalently bound monomer gives less grafring than a covalently bound radkal or

anionic initiator.

Grafting can also be obtained in an anionic polymerization by termination ofliving

polymers by hydroxylic groups at the inorganic surface. The ra te determining step in

this process is transport of macromolecules through the polymer layer already formed

at the partiele surface. In bulk polymerization processes more grafting will occur than

in a solution polymerization, because of the higher monomer concentration and the

gel-effect that may occur at the paniele surface.

5

Chapter 1

Synthesis of isotactic poly(methyl methacrylate) by means of an anionic

polymerization at the surface of Si()z, modified with an initiator attached to a silane

group, was studied by E. Schomaker et al. [36]. Nakatsuka [37] described the

suspension polymerization of butyl acrylate initiated by a radical initiator in a silane

group at the Si()z surface. He also studied solution polymerizations with initiators or

monoroer rooieties coupled to CaC03 particles by means of phosphate groups

[1,38,39]. This metbod appeared to give better results than modifying the CaC03

surface with long alkyl chains, although also in that case wetting toward the polymer and dispersion in the polymer matrix were improved. The molecular weight of the grafted polymer was higher than that of the free polymer formed (homopolymer), as

radicals at the partiele surfaces are restricted in their freedom of conformational changes, resulting in a slower recombination and thus in a gel effect [40]. Bulk and

solution polymerizations at the surface of inorganic particles are schematically shown

in figure 1.2. By means of emulsion polymerization processes several types of polymerparticles

with .a core-shell morphology can be obtained. Similar techniques can be applied to

synthesize particles with an inorganic core and a polymer shell. Yamaguchi et al.

[41-47] have described detergent-free emulsion polymerizations in the presence of

several inorganic partic les. They used oxides (like Fez03, CuO, Co2.03, Ni203, C!203,

MnOz, TiOz, ZnO, SiQz, Al203), metal powders (Al, Fe, Cu, Ag), blast furnace slag,

glass, graphite and CaS03. Initiadon was caused by radicals formed by oxidation of

H2S03. As the inorganic surface takes an active part in the oxidation of H2S03, polymerization rate and monoroer conversion strongly depend on the type of substrate

(in the case of TiOz the maximum conversion was only 6%, whereas for CuO 85%

was found). According to Scanning Electron Microscopy the surface is covered with

polymer particles instead of with a homogeneons layer. The polymer adheres to the

inorganic surface because of an electrostatle interaction between the anionic endgroups

of the polymer and the positively charged surface. Thus, the nature of the inorganic

surface appeared to play a major role, not only in the initiadon process, but also in the

adhesion of the polymer to the partiele surface. The latter will presumably depend on

parameters like point of zero charge and pH, which strongly affect partiele surface charge densities.

During polymerization most of the oligomer and polymer molecules will be

adsorbed at the partiele surface. As monomer will be adsorbed in this polymer layer,

polymerization will predominantly take place at the inorganic surface, resulting in a

6

Introduetion

Al

MONOMER •

MONOMER •

MONOMER •

Figure 1.3 "Emulsion" polymerization at the surface of inorganic particles. A)

Soapless emulsion polymeryzation; B) Emulsion polymerization at low surfactant

concentration.

gel effect. However, the major part of the polymer thus attached to the surface desorbs

upon a soxhlet extraction with benzene, and only a small amount of polymer will

remain adhered to the particles. A sirnilar process was described by Ono [2], for the

encapsulation of carbon powders like graphite and diamond.

Yamaguchi and Ono c.s. [48,49] also studied the polymerization ofMMA with a water soluble anionic or cationic radical initiator (K2S20s and 2,2'-azo-bis-(-2-

arnidino propane) hydrochloride (AIBA.2HC1), respectively), in relation to specific

charge effects (as controlled by Zeta-potential measurements) of inorganic particles

like BaS04, a-Al203, Si02, FeS, CuS, Fe203, CdS, Cu and HgS. Opposite charges

appeared to enhance the adsorption of initiator molecules and oligomers, resulting in

a better encapsulation of the particles with polymer.

Dekking [50,51] studied a process slightly differing from the process studied by

Y amaguchi and Ono. He used opposite electrical charges to adsorb a radical initiator

at the surface of several types of clay. U pon decomposition of the initiator part of the

radicals adhered to the surface, while the other part was free to enter the continuons

phase. Their relative activity appeared todependon the nature of the substrate.

Hergeth and Schmutzler c.s. [52-54] applied a theory of Schmutzler [55,56] descrihing partiele formation in detergent-free emulsion polymerization, to the deter-

7

Chapter 1

H~H Ti02

H OH

Figure 1.4 Polymerization at the surface ofTi02, modified with titanates.

gent-free polymerization of vinyl acetate in the preserree of Si02, and calculated the

minimum amount of particles required to prevent the formation of homopolymer.

During the frrst part of the reaction oligomers are adsorbed at the partiele surf ace, where

they forma layer in which further polymerization takes place.

Araiet al. [57-61] describe a mathematica! model for the detergent-free emulsion

polymerization of MMA. Predictions based on this model were compared with the

detergent-free emulsion polymerization ofMMA in the preserree ofBaS04 and CaS03

[62,63]. Monomertransfer from dropiets to the loci at the partiele surface, affected by

the impeller speed, appeared to play a very important role in the polymerization at the

surface ofBaS04. Conversion-time curves of polymerizations on the surface of CaS03

had a shape similar to those obtained in a normal emulsion polymerization.

According to Hasegawa c.s. [64] the partiele surface itself can also play an active part in grafting of polymer in a detergent-free emulsion polymerization. More polymer

adhered to the surface of freshly ground limestone, especially at new corners, defects

and disturbances in the crystal lattice. They also noticed that in a detergent-free

emulsion polymerization surface active oligomers are adsorbed at inorganic particles

lik:e CaC03 and BaS04, forming a hydrophobic layer in which further polymerization

can take place. Thus polymerization at the partiele surface is enhanced by the presence

of a hydrophobic layer at the partiele surface. A similar effect can be obtained when a

layer of adsorbed surfactant molecules is used [65-67]. In the absence of a surfactant,

8

Introduetion

polymer particles of the same size as the inorganic particles are formed. However, at

a low surfactant concentranon ( well below cmc) a thin bilayer is formed at the partiele

surface, in which polymerizarion can take place, like in the polymerizarions studied

by Hasegawa c.s .. This results in the formation of a uniform polymer layer. For this purpose anionic or carionic surfactauts could be used. It is worthwhile to norlee that

contrarily to what was reported by Martin [68], a nonionic surfactant proved to be less

effecrive, because of the relarively small amount that can be adsorbed at the partiele

surface, due to its low cmc. Similar methods were described by Hemmerleb c.s. [69] and by J. Solc [6). In this relation the adsorption of several types of surfactauts at the

surface of inorganic particles, and their effect on the dispersion, flocculation, and

redispersion behaviour of the particles was extensively studied by K. Meguro et al.

[19,70-75]. They also used an organophilic double layer of surfactauts at the surface

of pigments to carry out a polymerization of styrene at the pigment surface [3,76].

Furusawa et al. [77] used dispersions of Si02 particles with a dense layer of

hydroxypropyl cellulose in a surfactant solution. Polymerizarion took place in the

hydrophobic layer, formed by the hydroxypropyl cellulose and adsorbed surfactant

molecules. At surfactant concentrations above cmc a lot of free polymer was formed,

with a lower molecular weight than the polymer formed at the partiele surface. The

homopolymer tended to coagulate with the encapsulated Si02 particles, resulting in

particles of a "raspberry shape".

The characteristics of a bilayer at the surface of an inorganic partiele strongly

depend on conditions like ionic strength and pH. As a result polymer formation at the

partiele surface is largely influenced by these conditions. Even afterpolymerization is

completed ionic strength and pH remain very important, as desorption can take place

when these condinons are changed.

Theemulsion processes described to encapsulate inorganic particles with polymer

are schemarically shown in figure 1.3.

1.3 Aim of the present thesis

The goal of this investigation is to provide a metbod that can be used to encapsulate

inorganic particles with a polymer layer. Thus, the comparibility of e.g. Ti02 pigments

and the binder in latex paints can be improved. However, this technique can also be

used for the polymer encapsulation of other types of inorganic particles, like for

example Si02 and Al203.

9

Chapterl

1.4 Scope of thls thesis

In this thesis a metbod is described, in whlch .. emulsion-like" polymerizations are carried out in a bilayer at tbe surface of inorganic particles. Inorganic submicron particles, mainly Ti02 pigments, were modified·with titanate coupling agents, and subsequently dispersed in an aqueous surfactant solution. Polymerization was carried out in the bilayer fonned by the titanate ebains and the adsorbed surfactant molecules. A physical bond between polymer and titanate may be formed because of entangle-ments witb the titanate chains, on the other hand, by using a titanate containing a copolymerizable group (C=C) or an initiating moiety (N=N), also a chemica! bond can be obtained (figure 1.4)

1.5 Outllne of thesis

In chapter 2 experimental metbods and procedures are described. Chapter 3 deals with the chemica! and physical properties of the inorganic particles used in this work. The chemica! structure and stability of some titanates, botb in solution and at the partiele surface, and their effect on surfactant adsorption are studied. In chapter 4 the

polymerization kinetics of methyl metbacrylate (MMA}, styrene and methyl acrylate (MA) in the presence of Ti02 particles, modified with hydrophobic titanates, are described. In these polymerizations a physical bond between polymer and inorganic partiele is formed. The effect of several reaction parameters on conversion-time curves is studied by means of gas chromatography, densitometry, and electron microscopy. A qualitative model is proposed to explain the experimental results. Chapter S deals with copolymerizations of MMA and a copolymerizable titanate at the Ti02 surface, resulting in a chemica! bond between polymer and inorganic particle. In chapter 6

polymerizations are described, in which the radical initiator is bound to the partiele surface instead of dissolved in the continuons phase. In chapter 7 polymerization products are described, and a strategy is presented to obtain encapsulated particles preventing agglomeration of the particles during polymerization. Chapter 8 contains the conclusions of this thesis and some suggestions for further research.

Parts of this work have been publisbed or will sbortly be published: part of chapters 3 and 4 in references [167 -171], part of chapters 4 and S in reference [ 172], and chapter 7 in references [173, 174]. Chapter6 was partly publisbed in reference [175].

10

Chapter 2

Experimental

2.1 Synthesis and characterization of inorganic particles

Experiments were carried out with different kinds of inorganic particles: amor-

phous Ti02 , commercial titania pigments, crystalline Al203 and amorphous Si02.

Amorphous Ti02 was synthesized by adding water to a solution of a tetra alkoxy

onhotitanate in the corresponding alcohol, dried by means of molecular sieves. For

this purpose 5.13 g tetraethyl onhotitanate was dissolved in 70.3 ml dry ethanol. At

room temperature a mixture of 2. 7 mi H20 and ethanol (total volume 75 rol), was added

tothe solution. Theresultingmixture (0.15 mol titanate/1 and l.OmoiHzO/l)was stirred

with a magnetic stirrer. After 15 s TiOz precipitated. The precipitate was isolated,

wasbed with ethanol (three times), centrifuged, and driedat room temperature. A

similar procedure was used for the synthesis ofTiOz prepared from tetrabutyl titanate

and tetraisopropyl onhotitanate, to study the effect of the type of titan a te on the product

formed. The tetra alkoxy titanates were supplied by Merck (p.a.) and used without

funher purification. The surrounding atmosphere had to be kept free of moisture and

dust, to prevent seed formation in the titanate solution before the addition of water.

Experiments were carried out in a nitrogen atmosphere.

The commercial pigments were supplied by Kronos (Anatase: AD; Rutile: RLK,

RLP2, 2073,2160,2190, R1053, B87/1185 with ZnO), Tiofine (R60 andR80, treated

with different amounts of Si02 and/or Al203), Degussa (P25) and Merck (808). Also

n-Al203 (0.3 j!m, Buehler, Micropolish 11) and SiOz (kindly supplied by A.J.G. v.

13

Chapt~~r2

~iemen, Labaratory of Colloid Chemistry and Thermodynamics, Eindhoven Univer-

Sity ofTechnology [78]) were used.

The characteristics of the synthesized particles were compared with those of the commercial pigments (sometimes kindly moditiedon request). Partiele diameters were

determined with a Malvem Autosizer 2c (dynamic light scattering). The specific

surface area of some of the samples was determined according to the one-point BET

method, using a Stroehlein Areameter. Some measurements were carried out by

Kronos GmbH in Leverkusen, and some by mr. A. Korteweg of the Agricultural University iri Wageningen, The Netherlands. The latter used a technique by which the

adsorption curves of nitrogen over the complete range of relative pressures are

measured. Pore sizes were determinetl by means of mercury porosimetry (Carlo Erba

Instruments). The chemical surface composition of (modified) particles was studied

by means ofESCA (Electron Scattering for Chemical Analysis), using an instrument

of Physical Electtonics Industries Inc .. Most experiments were carried out with pure rutile pigment (Kronos RLK). This

materlal was wasbed with water, in order to remove some K2S04 adsorbed at the

surface, and driedunder vacuum at 130"C before use.

2.2 Modlflcation of particles with tltanates

Titanates KR TTS, KR 7, KR 212 and KR 26S of Kenrich Petrochemieals Inc.

were used without further purification. The titanate content, chemical structure and

stability against solvolysis of KR TTS and KR 7 were determined by means of 1H and 13C NMB. in deuterated chloroform, using a 60 MHz Hitachi Perkin Elmer High Resolution NMR Spectrometer R-24B and a 200 MHz Bruker AC 200 instrument. Further details will be provided in chapter 3 .

. Modification of TI02 was carried out in isopropanol (Merck. p.a.), diethylether

(Merck, p.a.) or dichloromethane (Merck, p.a.). In the beginning Ti02 was dispersed

in the solvent by a magnetic stirrer, but later the dispersion metbod was improved: 30

g Ti02 and 30 g glass pearls (diameter 2 mm) were added toa flask containing a titanate

in tlie appropriate solvent, in general in a concentradon of 1.5 - 4.5 gil (0.5 to 1.5 wt%

with re gard to TI02), and the mixture was shaken vigorously forabout two hours. Then

the gtass pearls were removed by tiltration and the modified TIÓ2 was isolated by

centrifugation. The product was wasbed three times with solvent, and then dried at

room temperature under vacuum. The amount of titanate at the surface was determined

14

Experimental

by elemental analysis (TNO, Zeist, The Netherlands), and by measuring the weight

loss after heating forabout one hourat 800'C. Stability against solvolysis was studied

by means of elemental analysis of titania modified with titanates in isopropanol or in

diethyl ether. By means of UV spectroscopy (using a Hewlett-Packard 8451A Diode

Array Spectrophotometer) titanate solutions in isopropanol were studied after the

addition of Ti02 or smal! amounts of water. By means of conductornetTic titrations

(using a Radiometer CDM80 conductometer) the adsorption of sodium dodecylsul-

phate was measured as a function of the titan a te content of the inorganic material, both

at room temperature and after heating modified Ti02 in water for one hourat 60'C.

Thus information about the stability of titanates at the TiO! surface against hydrolysis

was obtained. The structure of titanates at the surface was determined by FfiR (diffuse

reflection and transmission in a KBr pellet).

2.3 Initiator formation at the partiele surface

Titanate KR 26S contains an aromatic amine group, which can be used for funher reacrions at the partiele surface. Diazotation of Tî0z!KR26S was carried out with

hydrochloric acid and NaNO!. First, 4.00 g ofTi()zfKR26S was dispersed in 30 mi of

water for 30 s by means of an Ystral type X 1020 high shear stirrer. This dispersion

was transferred to an erlenmeyer flask and placed on a magnetic stirrer in an ice bath.

Subsequently 6.3 mi of concentrated hydrochloric acid (37%) was added dropwise,

after which the mixture was stirred for 5 more rninutes at O'C. Then a fresh solution

of 2.1 g of NaNO! (Merck, p.a.) in 25 mi of distilled water was added dropwise. When

all the nitrite had been added (after 30 min.) the mixture was left to react for 5 more

minutes. The solids were removed by fiJtration (through a cellulose acetate ester

membrane filter), wasbed twice with distilled ice water (40 mi) and dispersed in 10 ml

ice water, after which a solution of 2.4 g of 13-naphtalenethiol (Merck, p.a.) (0.0 15 mol) and 0.6 g NaOH (Merck,p.a.) in 25 mi of distilled water was added, which tumed the

reaction mixture yellow. After 30 min. the solids were removed by flitration through

a teflon filter, and wasbed with ice water until the yellow colour had disappeared. Then

tbe mixture was wasbed two more times with cold ethanol and twice with cold dietbyl

ether. The product was driedinair ato•c and then storedat a temperature between -20

and -10·c. Insteadof ~-naphtalenethiol also phenol ornaphtol (both Merck, p.a.) could

be used.

15

Chapter2

Another way to obtain a radical initiator at the partiele surface is by coupling

4,4'·azo-bis-(4 cyanopentanoic acid) (ACPA) (Fluka AG, purum) to KR26S at the

partiele surf ace. 1.50 g of ACP A was dispersed in 15 ml benzene (Brocacet) at o·c. Then 3.00 g PCls (Merck, p.a.) was added, after which the mixture was stirred for 15

minutes at o·c. Next the mixture was brought to room temperature and stirred for another three hours. The nearly clear solution was filtrared and the solvent evaporated in vacuum at room temperature, resulting in a yellow paste. The product was wasbed twice with 5 ml of a diethylether (Merck p.a.)/n-hexane (Merck, extra pure) (1:3)

mixture and the liquid was removed by decantation. The product was dissolved in

CH2Cl2 (Merck, p.a.) and subsequently, precipitated with hexane. It was filtrated after

which the procedure was repeated. The isolated product (4,4'-azo-bis-(4 cyanopen-

tanoic acid chloride) was driedinair at room temperature. Subsequently 0.054 g was

dissolved in 10 ml dichloromethane and added toa dispersion of20 g ofTi02/KR26S in 50 ml CH2Cl2. Glass pearls and 0.035 g triethylamine (Fluka, p.a.) were added, after

which the mixture was shaken for three hours at o·c. The modifled Ti02 was isolated by centrifuging the mixture, after which it was wasbed three times with distilled water, three times with ethanol and twice with diethyl ether. It was dried in air at room

temperature.

The sameproduct could also be obtained by reacting 20.75 g Ti02 (modified with

1.0 wt% ofKR26S) with 0.038 g dicyclohexyl carbodiimide and 0.050 g ACP A in the presence of some p-toluene sulphonic acid in 30 ml toluene (Merck, p.a.) at ts•c (30 g glass pearls had been added for dispersion). After one night the glass pearls were

removed, and the product was wasbed twice with toluene and once with isopropanoL

After isolation by means of centrifugation, it was dried at room temperature under

vacuum. The presence of the initiator at the partiele surface was shown by diffuse reflection FITR.

Initiation at the partiele surface could also be obtained with K2S20s, adsorbed at the partiele surface, for example by treating pure, wasbed Kronos RLK with a

concentrated aqueous solution ofK2S20s. In order to prevent desorption of persulphate

after the Ti02 is dispersed in water, the particles were modifled with KR TTS in

diethylether, after modiflcation with K2S20s.

16

Experimental

41

5::5

29

Figure 2.1 Reaction vessel type A.

(1) (2)

l 9.3 8.3 71

Figure 2.2 Reaction vessel type B: 1) "Half moon" type stirrer; 2) "Butterfly" type stirrer.

17

Chapter2

densitometer 15(7)

170 E9i~?(O 85 .

Figure 2.3 Reaction vessel type C; Values denoted between brackets for system with

baffles.

2.4 Surfactant adsorption at (modified) lnorganlc particles

The adsorption of sodium dodecyl sulphate (SDS; Fluka Chemie AG, 95% pure) at the surface of inorganic particles was determined by conductometric titrations. It can also be determined by dispersing particles in an aqueous solution (with a surfactant concentranon above the cmc), and subsequently removing them by centrifugation. The

remaining SDS concentranon in the liquid can be measured by a two--phase titration (water/chloroform) with hyamine, according to the metbod described by Reid et al. [79]. The indicator is a mixture of disulphine blue and elimidiurn bromide.

SDS and hexadecyl trimethyl ammoniumbromide (CTAB; Sigma, ca. 99% pure) were used without further purification, and dispersions were made using an Y stral high

shear stirrer.

2.5 Polymerizations

Polymerizations were carried out with methyl methacrylate, ethyl methacrylate, methyl acrylate, butyl acrylate, styrene (all supplied by Merck; p.a.), and butyl methacry-

late (Norsolor). These monomers were distilled at reduced pressure under nitrogen, to

remove the inhibitor. Polymerizations were carried out with a radical initiatorbasedon

18

Experimental

ACP A. Because of the limited water solubility of the acid the sodium salt was used

(prepared by reacting the acid with 2 equivalents of sodium methanolate in methanol).

As an alternative 2,2'-azo-bis (-2-amidino propane} hydrocloride (AIBA, Polyscience

Hicol) was used, without purification.

Initiation could also be accomplished by an initiator at the partiele surface (adsorbed

at the inorganic surface, or chemically bound toa titanate at the surface), or by using

radicals formed at the Ti02 surface under the influence of UV irradiation (350 nm).

Dispersions of (modified) particles in an aqueous salution of SDS (concentration

varying between 5.2 and 13.9 mmoVl) or CTAB (1.1 mmoVl) were made with an Ystral

type X 1020 high shear stirrer, and added to the reaction vessel. Then monomerwas

added and the mixture was flusbed with nitrogen fora bout 45 to 60 min. at 2o•c, in order

to remove oxygen. Subsequently, the mixture was stirred for at least 30 min., in case of

a water soluble initiator at 55 or 58·c, thus settling the equilibrium. Polymerization was

started by adding the initiator salution to the mixture, by switching on the UV light

source, or, in the case of an initiator bound to the surface, by heating the mixture to the

appropriate temperature.

Polymerizations were carried out in double walled thermostated reaction vessels,

kept under a slight excess pressure of nitrogen. Three different types were used:

(A) a 50 rnl glass reaction vessel. The mixture was stirred with a magnetic

stirrer (figure 2.1).

(B) a 250 ml glass reaction vessel, equipped with any of two different

types of stirrers (figure 2.2).

(C) a 11 steel (or glass) reaction vessel, with a turbine type stirrer (two

sizes ). The small stirrer was used in combination with four baffles, to

obtain ideal mixing conditions (figure 2.3) (see chapter4).

Vessels of type A and B were equipped with a 10 rnl addition funnel, from which

extra monomer, surfactant or initiator solutions were added dropwise to the reaction

mixture. Samples were taken at regular intervals during the entire course of the

polymerization, using a syringe through a septurn on the reaction vessel,. After each

sampling the septurn was capped, to prevent any leakage. Samples (0.2 mi) were diluted

with ca. 5 rnl distilled water, or with an aqueous salution of ca. 3 g SDS/1 in order to prevent problems with the automatic sampler of the gas chromatograph caused by

instahilities of the samples. Some hydroquinone was added to stop the polymerization.

Conversion was determined by gas chromatography, with isopropanol (approximately

3 wt%) as an internal standard in the reaction mixture (this compound was added before

19

Chapter2

dispersing the Ti02 in the surfactant solution, in order to prevent foaming). The gas

chromatograph was aHewlett-Packard S890 with an 1/8 inch poly(phenyl ether) packed column; column temperature 80.C; injection port temperature 100•c; FID temperature

1so·c; carrier gas nitrogen. Conversion-time curves were also obtained by determining

the dry solids content of the mixture during polymerization.

Reaction vessel type C was connected to an Anton Paar, Densitometer (OMA SS).

The reaction mixture was continuously circulated through the external cell (OMA 401 W) by means of a peristaltic pump (Verder UNI-V; flow ca. 4S ml/min.;

intemal diameter of the butyl rubber tubes 3 mm, extemal diameter 6 mm). Just before entering the densitometer the mixture passed through a heat exchanger inside the cell,

thermostated at exactly the same temperature as the reaction vessel, in order to compen-

sate for any possible heat losses during transport. The temperature during polymerization

did not vary by more than o.os·c; The reaction vessel was kept under a slight excess

nitrogen pressure (ca. 0.2 bar), which enabled taking samples during polymerization.

The samples were collected in a flask containing a known amount of hydroquinone to immediately stop the reaction.

All polymerizations were carried out in distilled water. Some reacrions were carried

out in a system buffered at pH = 8.00 (Titrisol: borate/HCl) or at pH = S.OO (Titrisol:

citrate/N a OH), or in a system containing 104 mol/1 HCl, Na OH, or Na Cl. Reacrions in

the presence ofTi02 modified with ZnO were carried out in an aqueous HCl solution at

pH=3.

During polymerization two types of product were formed: polymer encapsulated

inorganic particles and polymer particles formed by conventional emulsion polymeriza-

tion (homopolymer). Both products were separated by means of centrifugation and

subsequent washing of thesediment with an aqueous solution of SOS (ca. 10.4 mmol/1) and distilled water. After dispersing the precipitate in distilled water, by means of a high

shear stirrer, no free polymerparticles could be observedin the liquid phase. The polymer

content of the encapsulated particles was determined gravimetrically by hearing a sample

for about one hour at 8oo·c. In general, for PMMA the relative experimental error in

the results thus obtained appeared to be lessthan 2%, for styrene less than about 3%. In

those cases where massive coagulation had occurred, the experimental error was much

larger: up to 1S% for polymer formed in the presence of CTAB and an anionic water

soluble initiator. Polymerization products were studied by dark field microscopy (Zeiss)

and scanning and transmission electron microscopy (using a Cambridge and Jeol

20

Experiment al

instrument, respectively). Electron micrographs were kindly prepared by mr. H.C.B.

Ladan.

21

Chapter 3

Modification of inorganic particles

with titanates

ó6o ó~ ~6v~ KaÀQÇ auvtoraoOal rpirou x~ptç oû óuvar6v· óea~ov yàp Èv ~~a~ óei rtva á~~oiv ouvay~yov ytyveaOat.

But the mere faultless joining of two things without a third is impossible; for there

has to be a binder between them that keeps those two together (Plato, Timaeus, 31 b8-c2; transl. A. Nieskens)

3.1 Introduetion

In this chapter the chemica! and physical properties of several titanias are descri bed.

Special attention is paid to their surface composition in relation to small amounts of

other inorganic oxides, often added to decrease the photoactivity of the pigments.

The chemica! structure and stability of some titanates, both in solution and at the

Ti02 surface, is discussed in paragraph 3.3. Surfactant adsorption at the modified

titania surface is also discussed. Experimental details have been described in paragraphs 2.1, 2.2 and 2.4.

23

ChapterJ

Table 3.1: Physical properties of several titanium dioxides

titan ia Partiele diameter (run) Specific surface area (m2/g)

synlhesized from 400 Ti(0Czlis)4

synlhesized from 360 Ti(OC3H7)4

synthesized from 410 Ti(OC4H9)4 KtonosRLK . 260

Kronos 1072 (=AD) 370 Merck808 211 :.1:/

DegussaP25 23 3)

37 2)

Kronos2081 (=RLP2) -Ktonos2073 -Kronos2160 . Kronos2190 -

KronosR1053 -Kronos with ZnO -i~ Detennined by Kronos GmbH in Leverlrusen. :3) Described in literature [89].

344.

104

64

7.7 8.3 l) 8.5 4)

10.4 7.4 :./:)

49 42.5 Z)

56~5 4> 10.2 l) 6.7 I)

11,1 IJ

14.9 l) 3.3 I)

6.3 I)

Total ~volume (cm3/g)

0.29

0.11

-0.58

0.77

0.56 2.69

---. --

According to Degussa. 4> According to adsorption curves over the complete range of pressures [90].

3.2 Thè lnorganlc partJeles

Most commercially available titania pigments are coated with organic and/or inorganic substances. The inorganic coating with oxides of Si, Al, Zr, Zn etc. decreases the photoactivity of pigmehts. Pure Ti02 can genera te radicals under the influence of UV irradiation, and thus cause the deterioration of the polymer binder {chalking) [13]. The coating of pigments also influences the stability of dispersions in organic or

aqueous media [80]. As in this thesis a modelprocessis described, we used pure Ti()z

24

Modification oji110rganic particles with titanates

Table 3.2: Surface composition of inorganic particles (modified with several wt% of

oxides) in atomie %, according to ESCA.

Partiele Ti 0 c Si Al K s p Zr Na N Zn From Ti(OCzHs)4 28 62 10 - - - - . . . - -Frmn Ti(OC3H7)4 16 37 48 . - . - - - - - - i From Ti(OCtH9)4 23 55 23 - - - - I - I - - - -K RLK,

Chapter3

Table 3.2: Continued

T2 19 61 16 0.4 3.0 . . - - . - -T2, S% Alz03, org. 5.5 61 17 . 16 -

coat.

. - . - I . I -T2, fJrSt S% AlzO:!. 7.5 70 5.2 3.3 14 . - . - - - -then 1.5% Si()z,org.

coat. I T2, fitst 5% Alz03. 6.3 69 6.0 6.5 12 - - - - - - -then 3% Si.Ol, org.

coat.

Si()z - 68 4.8 28 . - - . - - 0.2 -a-Alz03 - 60 7.2 . 33 - - - - - - -

K = Kronos; M= Merck; D == Degussa; T = Tiofme

particles in most experiments, to prevent any effects caused by for example a non-uniform coating of the particles.

In literature several processes have been described to synthesize (monodisperse) TiOz particles [81-85]. In order to obtain some pure material we synthesized TiQz by hydrolysing a tetra alkoxy titanate (tetraethyl orthotitanate, tetrabutyl titanate or

tetra-isopropyl orthotitanate ), according to the metbod described by E.A. Barringer et al. [86-88].

The physical and chemica! properties of the amorphous Ti02 thus obtained were compared to those of several commercial Ti02 pigments (Tables 3.1 and 3.2). The decrease in specific surface area and total pore volume with increasing size of the alkoxide molecule, can be explained by a slower hydrolysis, caused by steric bindrance by the larger alkoxide chains. This results in the formation of more dense particles. The valnes obtained forthese three types of particles are in good agreement with those described by Barringer [87]. He found that washing and sintering of the particles

resulted in a decrease in specific surface area. As the commercial pigments all have been sintered, this explains the large difference in specific surface area of these particles, as compared with the particles synthesized from a tetra alkoxytitanate. The relatively large pore volumes of the commercial pigments probably reflect the pore

volume of the powder, and not that of the particles themselves.

The large amounts of carbon observed in table 3.2 may be ascribed to non-hydrolysed alkoxide ebains (in case of Ti02 particles, synthesized from a tetra alkoxy

26

Modification ofinorganic particles with titanales

HOJ§tCIH Ti02 +

HO OH

Figure 3.1 Moditicarlon of Ti()z with a titanate coupling agent

titanate ), to an organic coating of the pigment, and/or to the presence of a smal! amount

of CO gas in the instrument. The latter is adsorbed at the partiele surface before or

during the measurements (analysis of the gas in the sample chamber revealed the

presence of a substance with a molecular weight of 28). From table 3.2 it can be

concluded that the addition of a small amount of silicium-, aluminium- or zinc oxide

results in the covering of a large part of the partiele surface by this oxide. Therefore,

these particles will not exhibit the chernical surface behaviour of titania, but of the

oxides they were coated with. When Al2Ü3 and Si02 are added subsequently, the

second oxide will mainly precipitate on top of the frrst oxide, due to their opposite

surface charges at a pH of about 6. Kronos RLK has not been modified with any

inorganic or organic substances, but according to ESCA some K2S04 is present at the

partiele surface (probably as a re sult of the synthesis of the pigment from the rutile ore,

by means of the sulphate process [13]). This salt could easily be removed by washing

the pigment with distilled water. This pigment was used in the major part of the

experiments described in this thesis. In those polymerizations where the photoactivity

of this uncoated pigment caused some problems, Tiüz doped with a smalt amount of

ZnO [13] was used. According to Table 3.2 the ZnO is preferentially present at the

partiele surface, but it can be removed by washing the pigment with distilled water.

Therefore, this pigment was used without any other purification. Some polymeriza-

tions were carried out under exposure to daylight, but in order to prevent any effects

caused by the photoactivity of certain (modified) particles, all others were carried out

routinely in the dark.

27

Chapter3

fz

j, ~ ~ ~ ~ ~ ~

'MM!Iergth ~ -Figure 3.2 Hydrolysis of titanates, as observed by UV spectroscopy. --KR 26S in isopropanol; ---- after addition of water; - after additi.on of Ti02 particles.

3.3 Chemica! structure and stablllty of titanates

Recently, many articles have been publisbed on titanate coupling agents [91-99]. These are organic compounds which consist of a central titanium atom, one or two

small hydrolysable groups, and two or three long organic ebains or functional groups. A reacti.on with surface hydrox:ylic species can take place, in 'which an alcohol is formed, teaving the titanate covalently bourid to the pigment surface (Figure 3.1).

As compared with silanes, titanates have the advantage of forming strong honds with the surfaces of nonsiliceous minerals. Also, they are less likely to form bilayers by polymerizing in themselves. The principal difference between silanes and alkoxide modifying agents, like titanates,lies in the nature of the bond between metal ( or silicon)

eentres and the functional organic substituents that interact with the organic matrix. The functional organic moieties of the chemisorbed silanes are attached by hydrolysis-resistant Si-C bonds; on the contrary, those of the alkoxide based reagents contain

hydrolysable Ti-0-C honds. According to Sugerman and Monte [91-97] for many titanates, modification can

best be carried out in isopropanol, as the same alcohol is very often formed in these reactions. However, we found that titanates are very sensitive to aleoholysis and

hydrolysis. Moreover, hydralysis by water adsorbed at the partiele surface is probably

even catalysed by the surface (Figure 3.2).

28

Modîfication ofinorganic particles with titanales

Figure 3.3 60 MHZ 1H NMR spectrum oftetraisopropyltitanate in DCCI3.

H3Ç ~ [ 2HÇ-O~-Ti-LOC-C17 H3slz H3C.

_.....,_-9 8 7 6oomS 4

Figure 3.4 60 MHz 1H NMR spectrum of KR ITS in DCCI3.

29

Chapter3

1 Sppil 4 3

Figure 3.5 60 MHz 1H NMR spectrum of KR TIS in DCC13, after purposely ad-

ding another equivalent of isopropanol.

Figure 3.3 shows the 1.H NMR spectrum of tetraisopropyltitanate in deuterated chloroform. The septet at' 4.47 ppm can be attributed to Ha. The ratio Ha:Hb = 1:6, as was expected.

Figure 3.4 shows the 1H NMR spectrum of KR TIS, which, according to Kenrich [100], consistsof 95 wt% isopropyl triisostearoyl titanate and 5% isopropanol as a

solvent. As the molecular weight of isopropyl triisostearoyl titanate is 957.4 amu, it

can be calculated that 5% of isopropanol is exacdy one equivalent. However, from the

spectrum it can beseen that the signal ofisopropanol at3.8 ppm fails to appear. Instead

the ratio Ha:other protons is 1:62. Thus it can be concluded that KR TIS consists of 72% diisopropyl diisostearoyl titanate and 28% isostearoic acid. This result was

confinned by 50 MHz 13c NMR. After purposely adding another equivalent of isopropanol to a solution of KR TIS

in chloroform, further solvolysis occurred almost immediately, resulting in a mixture

of diisopropyl diisostearoyl titanate, triisopropyl isostearoyl titanate, isostearoic acid

and only a negligible amount of isopropanol (Figure 3.5).

30

M odiflcatifm of irwrganic particles with titanales

Figure 3.6 200 MHz 1H NMR spectrum of KR7 in DCCI3.

0,3 0 in dittlhyltllw 0 ... ~propanal

0

0,8 1.2 wt% KR7

\6

Figure 3.7 Carbon analysis ofTi02 modified with various amounts of KR7. Dis-

persion by conventional stirring.

Similar results were obtained for KR7, which, according to Kenrich [1 0 1], consists

of 88% isopropyl dimethacryl isostearoyl titanate, 10% isopropanol (solvent) and 2%

methacrylic acid. Here too, the alleged amount of isopropanol is one equivalent, and

the amount of methacrylic acid 0.15 equivalents. From figure 3.6 it can beseen that in

KR7 there are twodoublets (at5.95 and at 5.80ppm) in a ratio 1.16:1.00. The doublet

31

Chapter3

R, o oo3 ' I Ti

R2o' 'rr.i+

w =R2 = H)C-CH-CH3 ~.~tt. -Kc17~ -1 _, I -~ q g ~H3 ~

Modification ofinorganic particles with titanates

o.os .----.------.--.,

2928 1416 '549

2856

Figure 3.9 Diffuse Reileetion FfiR spectra of the isostearate group in KR TIS at

the surface of Ti02 (absorbance vs wavenumber).

o.s.--.....,-----r-...,

0'2

2800 3000 - Wl!venumbers cm~

01 .----..---or----,

1423 A

O.zt---,1400~---.:y,oo±=-----' -~s cm-1

Figure 3.10 Diffuse Reflection FfiR spectra of Ti02, modified with KR 7 (absor-

bance vs wavenumber): A) signals of the isostearate group; B) signalof the

methacrylic group.

0;7...------r---.

2600

Woverumbers ~cm-1)

Figure 3.11 Diffuse Reflection FfiR spectra ofTi02, modified with KR 212(ab-

sorbance vs wavenumber): signals of -CsH17.

33

Chapter3

0,2.-----.----, o.1s . .--....-----,29:-:::

300"!""T-..,

B (

Figure 3.12 Transmission and diffuse reflection FTIR spectra ofTi02, modified with KR 26S(absorbance vs wavenumber): A) transmission signals of SOl; B) dif-

fuse reflectance signals of the fenyl group; C) diffuse reflectance signals of the

Ct2H25 group.

Table 3.3: Surface composition of {modified) Ti02 according to ESCA (atomie%).

Titania Ti 0 c p TiOz(RLK) 18 56 22 -

TiOz+ l%KR26S 9.9 33 47 -Ti0z+l%KR7 16 44 36 -

Tiüz + 1% KR212 9.5 46 44 -

The titanates used in this work are shown in figure 3.8. So far we have notbeen

able todetermine the real chemica! structure of KR 212 and KR 26S, although it is

certainly different from the structures described in literature [111-112]. KR 26S

probably consists of a mixture of several types of titanates. However, the nett amount

of titanium and of the specific groups (like the aromatic amine groups) is correct. For

practical reasons, in this thesis the structure and composition of these titanates as

described by Kenrich are used.

The presence of titanate at the Ti02 surface was established not only byelemental

analysis, but also by diffuse reflection FTIR (figures 3.9, 3.10, 3.11, and 3.12 Band

C) and by transmission FTIR (figure 3.12A).

34

Modification ofinorganic particles with titanates

Figure 3.13 Thermogravimetrical analysis ofTi02, modified with various amounts of titanate: • KR TTS;•KR 7; *KR 26S.

35

Chapter3

l

M odification of inorganic particles with titanotes

Table 3.4: BET-analysis of (modified) Ti02

Titania Specific surface area (m''/g)

Pure Ti()z (RLK) 7.7

8.5 I)

RLK + 1% KR 212 8.0

RLK+ l%KR TTS 8.4

7.0 I)

RLK + 1% KR 26S 8.2

Merck 808 + 2% KR TTS 6.0 I)

Pure Ti()z (Degussa P25) 56.5 -1}

Degussa P25 + 11% KR TTS 45.8 l)

Degussa P25 + 11% KR 26S 38.8 l)

l) According toa total adsorption curve insteadof a one-point measurement [90).

monolayer, the excess of titanate molecules (i.e. those not chemically bound to the

surface) will be adsorbed by the titanate layer. This might eau se all kinds of (negative)

effects [97]. The intercept at the ordinate reflects the amount of water adsorbed at the

Ti02 surface.

From figure 3.14 it can be concluded that, especially at titanate contents below 1

wt%, there is a fairly good agreement between the theoretica! carbon content of the

particles (calculated from the chemica! structure of the titanates) and the experimen-

tally determined carbon content.

According to one-point BET-analysis the surface modification with titanates

hardly affects the specific surface area of the Ti02, as shown in Table 3.4. However,

by measuring total adsorption curves it was shown that the partiele surface area

decreases by about 18% because ofthe modification with KR TTS. The second metbod

is believed to result in a better accuracy [90].

Since the titanate-modified particles have to be dispersed in water at ca. 60'C for

the "emulsion polymerization" process, the sensitivity of the titanate bonding towards

hydralysis had to be investigated. Therefore, the amount of SDS adsorbed at the

partiele surface was determined. The frrst metbod used was based on the following

idea. As surfactant is adsorbed by partiel es, more surfactant must be added toa mixture

to reach the cmc [113]. Thus, Ti02 modified with various amounts of KR TTS, was

37

Chapter3

Figure 3.15 Adsorption of SOS a:t the parpcle surface, as a function of the titanate content of the partic les, determined by conductometric titrations ( 66,67 g TiOl/1).

- 20 °e; --- 60 °e, 1 h.

dispersed in water, and in a conductometric ritration SOS was added. It appeared (figure 3.15) that SOS is adsorbed at the partiele surface, while its amount is a function ofthe titanate content of the particles. Through the adsorption of SDS at the modified,

hydrophobic, Ti02 surface the dispersion was stabilized against coagulation. The ritration was repeated with dispersions, that had been heated at oo·c for 1 hour. Only at low titanate contents, hydrolysis was found to have occurred, resulting in a "negative" adsorption, obviously caused by solubilization of the hydrolysedisostearoic

acid. At titanate contents over 0.7 wt%, the two curves in tigure 3.15 coincide, indicating a selfprotecting effect of the titanates against hydrolysis. Obviously, under those conditions surface groups inducing hydrolysis (like e.g. adsorbed water mole-cules or acidic OH groups) are absent or sufficiently shielded, while the hydrophobic

mantie bas sufficient density to repel water effecri vely from the hydrolysable rooieties near the partiele surface. Similar results were obtained by using the second metbod

(however without a "negative adsorption", as the amount of SOS is measured). In this case the modified Ti02 was dispersed in an aqueous surfactant solution ( concentradon

above the cmc), and subsequently removed by centrifugation. The equilibrium surfac-tant concentration in the liquid phase was determined by ritration [79] (Figure 3.16). The amount of surfactant adsorbed at the partiele surface then was calculated from the

38

Modification ofinorganic particles wilh titanmes

20 ,..-----------.,

o'-----~-~---' 0.0 1.0 2.0

%KR TTS

Figure 3.16 Adsorption of SDS at the partiele surface, as a function of the titanate

content ofthe particles, determined by depletion measurements (ca. 140 g Ti0211;

original SDS conc. 8.3 mmol/1). o 20 °e; • 60 °e, 1 h.

difference between the original and the equilibrium SDS concentration in the volume

in which the TiQz had been dispersed. Underthe conditions used maximum adsorbance

was obtained. The total amount of surfactant in Figure 3.15 was somewhat lower than

the amount observed in figure 3.16, probably because of a less efficient dispersion of

the hydrapbobic particles in the water phase at the beginning of the_ measurement.

From these experiments it can be concluded, that a hemimicelle [114-116] can be

formedat the partiele surface. The hydrapbobic titanate layerwillremain stabie against

hydralysis under polymerization conditions, if the TiQz bas been modified with more

than 0. 7 wt% of titanate in an aprotic solvent, thus showing a "selfprotecting effect".

3.4 Surfactant adsorption at the partiele surface

In the previous section it was concluded that a surfactant (SDS) is adsorbed at the

modified partiele surface, forming a hemimicelle. On a non-modified, hydrophilic

surface these hemimicelles may consist of a double layer of surfactant molecules, in

equilibrium with the surfactant, monoroers and micelles in the aqueous phase (Figure

3.17).

39

Chapter3

Table 3.5: SDS adsorption at the surface of (modified) inorganic particles

Inorganic partiele * 1if' mol SDS per gram inorganic material RLK 0.85

RLK + 1% KR lTS 2.05 l) 2.02 Z)

RLK+l%KR26S 1.57 RLK+0.49% KR 7 +0.51% KR lTS ;,J 1.40 RLK + 0.50% KR 7 + 0.50% KR lTS '*1 2.61 RLK + 0.67% KR 7 + 0.33% KR lTS J) 1.64 RLK+0.76% KR 7 +024% KR lTS J) 1.21 RLK + 0.75% KR 7 + 0.25% KR lTS 4> 1.06 RLK + 0.80% KR 7 + 0.20% KR lTS 4> 0.53

RLK + 1.0% KR 7 1.29 TiQz!ZnO; pH=6.0 1.94 TiOuZnO; pH=3.1 4.32

TiOuZnO + 1% KR26S; pH=6.0 0.95 TiQz!ZnO + 1% KR26S; pH=3.1 1.83

Merck808 0.10 Merck 808 + 2% KR lTS 1.79

DegussaP25 6.37 DegussaP25 + 11.5% KR lTS 12.1

Degussa P25 + 11% KR26S 7.62 Degussa P25 + 12% KR 7 2.69

a-Ah03 3.69 a-Ah03 + 1% KR 26S 4.94

1) 3.333 g Ti021KR TIS was dispersed in 100 ml with SDS concentradon of 8.32 mmoliL 2) 6.662 g Ti02/KR TIS was dispersed in 100 ml with SDS concentradon of 8.29 mmolJI. 3) Modificalion was canied out using a mixture ofboth titanates. 4) Modiflcation was canied out using KR 7 and KR TIS, subsequendy.

On a titanate -modified hydrophobic surface the hemimicelles consist of a titanate

layer, chemically bound to the surface, and an adsorbed surfactant layer, as depicted

in Figure 3.18 (the adsorption of surfactant molecules on a hydrophobic surface was

described by Esumi et al. [70] and by Hunter [117].

The amount of surfactant adsorbed at the partiele surface was determined for

various types of inorganic particles and titanates (table 3.5).

40

Modification ofinorganic particles with titanales

Figurc 3.17 Surfactant adsorption at a Figure 3.18 Surfactant adsorption at a

(positively charged) hydrophilic surface hydrophobic surface.

The maximum adsorption of a surfactant at the surface of polymer particles in a

latex system can be calculated using equation 3.1 [113].

6 Es=---

dppAm (3.1)

where Es is the amount of surfactant adsorbed per weight of partiele (moVg Ti02),

dp is the partiele diameter, p is the density of the particles, and Am is the surface area

occupied by one mole of surfactant. Applying this equation to Krones RLK modified

with KR TTS gives a surface area occupied by a single molecule of SDS of about 50

À 2, which is in good agreement with the literature value of 50 À 2 for the adsorption of SDS on polymer particles [118]. So, most probably the whole partiele surface is

covered with SDS molecules.

As both Merck 808 andDegussa P25 consistof anatase (and therefore will probably

show a similar behaviour towards adsorption), the partiele diameter of Degussa P25

(modified with KR TTS) can be estimated using equation 3.1, and the data obtained

for Merck 808 modified with KR TTS. These considerations support the validity of

the value of 37 nm calculated by Janssen [89]. This also is in better agreement with

the value obtained for the specific surface area than the partiele diameter according to

Degussa (neglecting the effect of the porosity on the specifïc surface area for particles

41

Chapter3

with a diameter of about 23 nm an area of a bout 67 m2 is calculated). It is possible that the original primary particles have a diameter of about 23 nm, but have agglomerated.

Titanate KR 7 only contains one hydrophobic group, and thei:efore may be less effective in protecting itself against hydrolysis. For most polymerizations in which KR 7 was used as a (macro)comonomer, a combination of KR 7 and the more hydrophobic

KR TIS was used. Best results were obtained for equal amounts of both titanates, especially when attached to the surface subsequently, as can be concluded from Table 3.5 (most effective proteetion is obtained if KR TIS is added after the TiOz bas been modified with a small amount of KR 7). A higher amount of KR TIS results in more steric hindrance, while a higher KR 7 content is insufficiently stabie against hydrolysis.

lt is known from literature [113], that the presence of a water soluble salt results in the adsorption of a larger amount of surfactant at a hydrophobic partiele surface. This effect may account for the relatively large amount of SOS adsorbed at the surface of TiOz containing ZnO, as, according to table 3.2, part of the ZnO will be dissolved upon dispersing the pigment in water. Most experiments with this pigment were carried out at pH=3 (HCI solution), to obtain a stabie dispersion (at higher pH severe flocculation was found to occur). Obviously, the presence of HCl also increases the amount of surfactant that can be adsorbed at the partiele surface, as expected from the above. At pH=6 the modified pigment can only adsorb a small amount of surfactant.

From these experiments it can be concluded, that it is possible to form a hemimicelle at the surface of inorganic particles, consisting of titanate chains, chemi-cally bound to the partiele surface, and an adsorbed surfactantlayer. This system was used to carry out "emulsion-like" polymerizations at the partiele surface, in order tó obtain polymer encapsulated inorganic particles, as will be described in the next chapters.

42

Chapter 4

Polymerizations at the surface of

hydrophobic Ti02 particles

4.1 Introduetion

Polymerizations are carried out in dispersions of Ti02, modified with KR TTS or

KR 26S, stabilized against coagulation by a surfactant. Reacrions were started by

thermal decomposition of a water soluble initiator. As the reacrions were carried out

in a glass reaction vessel exposed to day-light, initianon by radicals generated inside

the titania crystal (by UV irradiation) may also take place. In the system Ti021KR26S

there might be a small effect of initianon by UV irradiation. However, it was found

that this effect can be neglected in the system Ti02/KRTTS when using an initiator

dissolved in the waterphase. These effects will be discussed in more detail in chapter

6. Two types of polymerization can take place simultaneously: (1) conventional

emulsion polymerization in micelles, swollen with monomer, and (2) polymerization

in the monoroer swollen hemimicelles at the partiele surfaces. Polymer formed by the

latter polymerization can be physically bound to the Ti02 by entanglements with the

hydrophobic titanate ebains and by adsorption. The effect of various important

parameters (like type and concentratien of surfactant and initiator, Ti02 content, ionic

strength, and hydrophobicity ofthe monomer) on polymerization kinetics is discussed

in the present chapter.

45

Chapter4

Figure 4.1 Effect of (modified) Ti02 on polymerization kinetics. (modified) Ti02 :

MMA = 1:3; -·-·- MMA; --- MMA + Ti02;- MMA + TiÛ2}KR TTS; ···· MMA + TiÛ2}KR 26S.

4.2 Polymerization kinetics

4.2 .1 Effect of hydrophobic particles on polymerization ldnetics

The effect ofthe presence ofTi02 on the polymerization behaviour is demonstrated

in figure 4.1. From literature [13] it is known that polymerization can take place at the

surface of unmodified Ti()z, because of adsorbed initiator or SDS molecules (see figure

7 .2). However, unmodified Ti()z (RLK) appears to have only a negligible effect on

the shape of the conversion-time curve of the emulsion polymerization of methyl

methacrylate (the increase in the inhibition period may have been due to spurious oxygen in the vessel). Unmodified Ti()z particles, however, do not become encapsu-

lated by a polymer layer, and their dispersion will not be stabie for more than about

one day. Only the presence of modified Ti()z particles during polymerization appears

to lead to the interesting phenomenon of uniform polymer encapsulation.

Figure 4.1 shows that at a certain conversion the polymerization rate suddenly

decreases, and, after a few minutes, increases again. This effect is observed for both

KR TTS and KR 26S and can be explained as follows. At the start of the polymeriza-

tion, monomer is present at four different places: a small part is dissolved in the water

46

Polymerizations at the surface of hydrophobic Ti02 particles

phase ( MMA solubility 0.15 mol/l [67]), the largest part is present in large monomer

droplets, part is adsorbed in normal micelles, and part is adsorbed in hemimicelles

containing a Ti0:2 partiele in the core. Polyrnerization takes place both at the partiele

surface and in free micelles. In order to keep polymerizations in free micelles at a

minimum, the surfactant concentranon in most experiments was kept at, or slightly

above the cmc. As the growing particles need more surfactant to remaio stabilized,

aftera short period of reaction a deficiency of surfactant may arise (more information

a bout this effect can probably be obtained by electrophoretic measurements to deter-

mine a possible change in zeta-potenrial during polymerization). Durlog this period

orthokinetic coagulation of the particles may occur, as described by Wahl and Baker

[119], and by Gregory [120,121]. Their theories are based on the model of von

Smoluchowski [122]. Whether two colliding particles form an aggregate or not

depends on for example Brownian diffusion, the interaction between the partiel es, their

surface roughness [123], and hydrodynamic effects. The coagulation of Ti()z and

polymer particles is favoured, as the surface charge density decreases as aresult of the

lower surfactant concentratien at the partiele surface. This effect might be enhanced

by the presence of a polymer layer at the Ti0:2 surface, that can play an active role in

coagulation by "bridging flocculation". Recently many extensive studies on floccula-

tion, induced by polymer adsorption at inorganic partiele surf aces, have been publisbed

[124-134]. Flocculation appeared todependon the electrical surface charge of the

inorganic particles and the ionic endgroups of the polymers, the nature of the adsorption

(reversible or irreversible), the presence of sites without adsorbed polymer available

for bridging, the concentration of inorganic and polymer particles in the continuons

phase, and the molecular weight of the polymer. Pelssers et al. [133, 134] found that

the attractive force between two polymer-covered surfaces passes through a maximum

with increasing polymer coverage, and is time-dependent. Polymer molecules at a

partiele surface can cause bridging flocculation if they extend far enough into the

solution to exceed the range of action ofthe electrical double layer. From adsorption

studies it was concluded that immediately after attachment the extension of a polymer

molecule will be of the order of the diameter of a free coil in solution, which is (for

sufficiently high molecular weights) longerthan the range of action ofthe double layer.

On the other hand, it is knowntbat adsorbed polymers tend to lie flat, unie ss the surface

is almost saturated with polymer. Therefore, an attached polymer chain of moderate

molecular weight will eventually relax to a rather flat conformation, thus loosing the

capability to form bridges. Bridging is only possible between a polymer chain on one

47

Chapter4

particle, and an "empty" site on anotherparticle. Therefore, at high polymer concentra-tions, when no free sites are avallab Ie, polymer adsorption does not cause flocculation, but quite contrarily shows a stahilizing effect (steric stabilization) [125,126,135].

Whether bridging flocculation occurs when two particles collide, depends on the surface charge density of the particle, the presence of empty sites at the partiele surface, the ionic strength and the nature of the pPlymer: molecular weight and chemical composition determine the polymer chain conformation at the partiele surface, and

thus its ability to form large flocs. Similar processes may occur during polymerization processes in the systems

described in this thesis. In a regular emulsion polymerization (i.e. "zero-one" systems) growing particles

consisring of pure polymer and absorbed monomer contain only one radical at the same time, as the entrance of a second radical results in immediate termination with the flrst. As the TIOz core cannot be penetrated byprimary or oligomeric radicals, Ti()z particles with a thin polymer shell are likely to contain more than one radical at the same time. Thus the average life-time of radicals at the inorganic surface will be longer than the average life-time of radicals in regular polymer particles, resulting in a high intrinsic reaction rate at the partiele surf ace. So, afterlarge coagulates of the latter particles have been formed, the polymerization rateis bound to decrease, because many radicals may become trapped inside. It is known from literature [136] that severe occlusion within tighdy packed particles, which are not swollen by monomer, can shield free radicals from reaction with monomer, so that occlusion in fact favours the occurrence of termination. In other words, partiele coagulation reduces the available partiele surface

area and may effectively terminate free radicals within the particles by confl.ning them

toa limited volume in close prox.imity, leading to reaction with one another. These effects also have been observed for polymerizations in which the polymer is insoluble in its own monomer, like the polymerization of vinylidene chloride [137-141] and

vinyl chloride [142,143]. Because we use a water soluble initiator in the aqueous phase new radicals will be generated, causing an increase in polymerization rate aftersome minutes. Thus, in the second part of the reaction polymerization will continue mainly in free polymer particles, as well as in newly · formed polymer particles, generated

because surfactant molecules have become available as aresult of the limited coagula-tion. Polymerization also continues at the surface of the agglomerates, although this polymerization does not play a very important role anymore, as the total surface area

of the agglomerates is Ie ss than the total surface area of the polymer particles or that

48

Polymerizations at the surface of hydrophobic Ti02particles

" ~ .... .. . . .. .. • ~ " ... ..

·~ .... ·~

'! (t

l< . .;-• -. i i$ 0 °/o 10°/o 50°/o 90°/o

CONVERSION

Figure 4.2 Coagulation of polymer and Ti02 particles during polymerization,

studied by dark field microscopy.

~80 0 -§00 ëii '-4 ~ c: 0 u

0 60 70

Figure 4.3 Polymerization of MMA in the presence of Ti02/l% KR TIS at the

cmc. Ti02: MMA = 1:3.

49

Chapter4

of the original Ti02 particles. The course of the polymerization was stuclied by means of dark field and electron microscopy. From figure 4.2 it can be concluded that some coagulation also occurs at the early stages of the polymerization. Coagulation of polymer particles during that period also is frequently observed in regular emulsion polymerizations [144-146], and therefore agglomeration of polymer and (encapsu-lated) Ti02 particles doesnotseem unlikely. In figure 4.2 separate encapsulated Ti02 particles remain distinguishable in the agglomerates, indicating that the encapsulation

by PMMA prevents the massive agglomeration observed when using unmodified Ti02 particles.

The polymerization productsof which the conversion-time curve is shown in figure 4.3, were stuclied both with Scanning and Transmission Electron Microscopy (see appendix I, page 66). From the scanning electron micrographs it can be concluded that at the early stages of polymerization also pure PMMA particles are formed (for example at 5.4% of conversion). PMMA particles are notably smaller than Ti02 particles and tend to decompose under the electron beam. At a conversion of about 12% some agglomeration of Ti02 and PMMA can be observed, although this effect

doesnotseem to be very important up to 27% conversion. Then suddenly, within two minutes (at the moment When a sudden, small deercase in polymerization rate can be observed in the conversion-time curve), severe flocculation appears to occur, and the reaction rate drops. From micrographs of samples taken at higher conversions it can

be concluded that only part of the particles coagulate, while indeed new polymer particles are formed. During the rest of the polymerization micrographs show more or less similar images. In other experiments it also was found that severe flocculation

occurs at the moment when the reaction rate suddenly drops. Transmission electron micrographs also show the formation of PMMA particles at 5% of conversion, and

some coagulation of Ti02 particles (the sharp lines observed for pure Ti02 (figure 1.11) are reflections caused by the titania lattice). From these micrographs it can be concluded that "bridging flocculation" indeed plays a role in the coagulation

mechanism, as the encapsulated particles seem to be connected by small polymer bridges. The Ti02 of sample 9 ( conversion 11.6%) is covered by a PMMA layer of

approximately 10 nm. In sample 11 (at a conversion of 27.7 %) large agglomerates can be observed. In sample 17 ( conversion 51.8%) two types of PMMA particles are found: large ones, formed during in the early stages of polymerization, and small ones, formed after limited coagulation had occurred.

50

Polymerizations at the surface of hydroplwbic Ti02 particles

100 e.t~~c:c 12

~ ,. _..-- 10 80 l z 8 < 0 60 :E

ën 6 :E a: a. w 40 4 ..,. > z

20 0 2 (,)

0 0 0 30 60 90 120 150 180

TIME (min)

0 CONVERSION • AMOUNT PMMA

Figure 4.4 Polymer content of Ti()z during polymerization. MMA : Ti0211% KR

TIS = 3:1; SDS conc.= 9.34 mmoliL 0 monomer conversion; • wt% PMMA in the

encapsulated product.

Figure 4.4 shows the increase in PMMA content of the Ti()z particles during

polymerization. From this figure it can be concluded that the temporary slow-down is

most distinct in polymer formation at the partiele surface, which nearly stops during

this period. The decrease in surface area accounts for the relatively small amount of

polymer formed at the Ti()z surface after the agglomeration.

In the case of Ti02 modified with KR 26S the "plateau" in the conversion-time

curves occurs earlier. This might be due to initiation by radicals formed at the Ti02

surface. As described in chapter 6, radical formation at a rutile surface by UV

irradiation is more likely to occur on KR 26S modified surfaces as compared with KR

TIS modified surfaces. Thus, relatively more polymer will be bound to the surface,

advancing the effect of bridging flocculation. This accounts for the temporary slow-

down at a lower conversion.

The temporary decrease in reaction rate seems to be primarily due to bridging

flocculation. In terros of surface charge the ionic strength of the solution can play a

role in these polymerizations by influencing surfactant adsorption (chapter 3). How-

ever, this appears to be less important here, as is indicated by Figure 4.5: polymeriza-

tions in buffered systems at high ionic strength (pH = 5.00 or 8.00), show kinetic

behaviour similar to that of unbuffered systems, to which no extra ions have been

added.

51

Chapter4

Figure 4.6 Effect of ionic strength on polymerization kinetics (MMA:Ti02 = 3:1, SDS conc. = 8.31 mmol/1). --pH= 5.00; ----pH= 8.00; ~··~pH= 6.5 (unbuf-

fered).

80

20

10 20 30 40 50 60

time (mln)

Figure 4.5 Polymerization of MMA in the presence of Ti02/l% KR TIS, at a sur-

factant concentradon slightly above cmc (9. 71 mmol/1).

52

Polymerizalions at