EFFECT OF INORGANIC PIGMENTS

ON POLYMER INTERDIFFUSION

IN LOW-Tg LATEX FILMS

Mitsuru Kobayashi

A thesis submitted in conformity with the requirements

for the degree of Master of Science

Graduate Department of Chemistry

University of Toronto

O Copyright by Mitsuru Kobayashi (2000)

National Library ISfI of Canada Bibliothèque nationale du Canada

Acquisitions and Acquisitions et Bibliographie Services services bibliographiques

395 Wellington Street 395. rue Wellington Ottawa ON K I A ON4 Ottawa ON Kt A ON4 Canada Canada

The author has granted a non- exclusive licence dowing the National Library of Canada to reproduce, loan, distribute or sel1 copies of this thesis in microfom, paper or electronic formats.

The author retains ownership of the copyright in this thesis. Neither the thesis nor substantial extracts fiom it may be printed or otherwise reproduced without the author's permission.

L'auteur a accordé une licence non exclusive permettant a la Bibliothèque nationale du Canada de reproduire, prêter, distribuer ou vendre des copies de cette thèse sous la forme de microfiche/nlm, de reproduction sur papier ou sur format électronique.

L'auteur conserve la propriété du droit d'auteur qui protège cette thèse. Ni la thèse ni des extraits substantiels de celle-ci ne doivent être imprimés ou autrement reproduits sans son autorisation.

ABSTRACT

Effect of Inorganic Pigments on Polymer Interdiffusion

in Low-Tg Latex Films

Master of Science (2000)

Mitsuru Kobayashi

Department of Chemistry, University of Toronto

This thesis examines the influence of inorganic pigments on latex film formation.

We carried out fluorescence resonance energy transfer measurements on latex films for

following the extent of inter-particle polymer diffusion. We found that large calcium

carbonate particles (300 nm x 1,000 nm) had little effect on the polymer diffusion rate for

films conidning up to 80 wt % filler. whereas 25 nm silica particles had a significant

influence in poly(methy1 methacrylate-CO-2-ethylhexyl acrylate) latex films. We also

examined the influence of silica particle size on this rate in poly(buty1 methacrylate) latex

films. We found that the 12 and 25 nm SiOi particles significantly affect the extent of

mixing with increasing amounts of filler. The rate decreases as the filler size decreases.

We propose that there is a low mobility polymer layer near the filler surface than the bulk

polyrner, and that the smdl silica particles can also act as obstacles.

ACKNOWLEDGEMENTS

First 1 would like to express my deep appreciation to my research supervisor,

Professor Mitchell A. Winnik. He encouraged me to work independently, yet gave me a

lot of suggestions when needed. He also encouraged me to join a lot of meetings with

industry people, which 1 believe improved my communication and presentation skills.

I would also like to thank Professor Winnik and Professor Douglas Reeve,

Director at the Pulp and Paper Centre. University of Toronto. for inviting me to the

Surface Science Consortium and encouraging me to give presentations to those who work

in the pulp and paper industry. I also appreciate Prof. Reeve for reviewing my MSc

Thesis and giving me valuable comments.

1 would also like to thank Professor Eugenia Kumacheva for reviewing my MSc

Thesis and for giving me valuable suggestions. I also had good opportunities to discuss

my research work with her at the Surface Science Consortium.

Special thanks go to Oji Paper Co., Ltd.. for allowing me to study polymer

chemistry at the University of Toronto and for financial support.

My special thanks also go to Dr. Yahya Rharbi. Dr. Ewa Odrobina, Dr. Hung H.

Pham, Dr. Ronghua Liu, Dr. Jiangdong Tong, and Mr. Lan Cao, for valuable discussions

on my research project, and Dr. Matthew Moffitt for proofreading the thesis. it was

wonderful for me to interact with them.

I would ûlso like to thank dl other members in Prof. Winnik goup, for usehl

discussions and suggestions. and the friendly atmospheres.

Finally, I would like to thank my wife Hanimi, my son Yoshiki, my parents. sister

and brother, for supporting and encounging my graduate study and research for two

years.

TABLE OF CONTENTS

ABSTRACT

ACKNOWLEDGEMENTS

LIST OF TABLES

LIST OF FIGURES

INTRODUCTION

Background

Emulsion Polymerization

Latex Film Formation

Polymer interdiffusion in Latex Films

Fluorescence Resonance Eneqy Transfer (FRET)

References

EFFECT OF PIGMENT TYPES ON POLYMER

INTERDIFFUSION

Introduction

Experimentai

2-2- 1. Materials for this study

2-2- la. Materials for Emulsion Polymerization

2-2- 1 b. Inorganic Pigments

2-2-2. Synthesis of Dye-labeled Monorner

2-2-3. Synthesis of Poly(MMA-CO-EHA) Copolymer Latex

2 - 2 4 Characterization of Latex

2-2-4a. Particle Size Measurement

2-2-4b. Gel Permeation Chromatography (GPC)

2-2-4c. Solid Content of latex dispersions

. . 11

... 111

viii

i x

1

1

1

4

5

6

7

9

9

IO

10

10

11

12

13

14

14

15

15

2-2-5. Film Preparation

2-2-6. Energy Transfer Measurement

2-2-7. Data and Data Analysis

Results

2-3- 1. Dye Distribution Analysis by GPC

2-3-2. Polymer Interdiffusion in Latex Films Using Fluorescence

Resonance Energy Tnnsfer (FRET) Technique

2-3-3. Efficiency of Energy Tnnsfer, @&O) in Newly Formed Films

2 - 3 4 Maximum Efficiency of Eneqy Transfer a&=) in Latex

Films

2-3-5. Effect of Pigment on the Rate of Polymer Interdiffusion

Discussion

Conclusions

References

EFFECT OF SEICA PARTICLE

POLYMER INTERDIFFUSION

Introduction

Expenmen ta1

3-2- 1. Materiais for this study

3-2- la. Materials for Emulsion Polymerhtion

3-2- 1 b. Synthesis of Dye-Labeled Latex Particles

3-2- lc. Colloidal Silica

3-2-3. Characterization of Latex and Colloidal Silica

3-2-2a. Particle Size Measurements

3-2-2b. Gel Permeation Chromatography (GPC)

3-2-2c. Solid Content of latex dispersions

3-2-3. Film Prepmtion

3-2-4. Energy Transfer Measurement

3-2-5. Data and Data Analysis

3.3 Results

3-3-1. Dye Distribution of PBMA

3-3-2. Determination of Silica Particle Sizes by DLS and TEM

3-3-3. Fluorescence Resonance Energy Transfer Technique for Latex

Film Formation

3 - 3 4 Initial Efficiency of Energy Transfer, @ d o ) in Newly Formed

PBMA Latex Films.

3-3-5. Maximum Efficiency of Energy Transfer, ad-) in PBMA

Latex Films

3-3-6. Effect of Silica Particle Size on the Rate of Polymer

Interdiffusion

3-3-7. Analysis of die Diffusion Process

3-3-8. Effect of Silica Particle Size on Diffusion Coefficients

3-3-9. Dependence of the Polymer Diffusion Rate on the Volume

Fraction of Silica

3-3- 10. Fundamental Mechanism of Polymer Diffusion

3.4 Discussion

3-41 . @ d o ) in Newly Formed PBMA Latex Films in the

Presence/Absence of Silica

3-42 Effect of Silicri on Gd200 h)

3-4-3. Effect of Silica on the Polymer Diffusion Rate

3.5 Conclusions

3.6 References

4 FUTURE WORK

4.1 Effect of Modification on the Polymer Diffusion Rate of Latex

Polymers in the Presence of Minerai Fillers

4.2 Effect of Other Constituents (Water-soluble Polymen, Thickeners) on

the Polymer Difision Rate of Latex Polymers in the Presence of

Mineral Fillers

4.3 Application of FRET Technique to Styrene-butaddiene and Styrene- 105

acrylate Latex systems

4.4 References

vii

LIST OF TABLES

Table 2.1 Recipe for preparing 1 10 nm P(MMA-CO-EHA) latex.

Table 2.2 Characteristics of the latex and inorganic pigments. The particle size.

Tg, and molecular weight for both Phe-P(MMA-CO-EHA) and MeAn-

P(MMA-CO-EHA) are very sirnilar to those for MeAn-P(MMA-CU-

EHA).

Table 2.3 Final area and maximum efficiency of eneqy transfer values.

------..-----

Table 3- 1 Recipe for preparing LOO nm PBMA latex.

Table 3-2 Charactenstics of the latex and colloidal silica. The particle size. Tg.

and molecular weight for Phe-PBMA are very similar to those for An-

PBMA.

Table 3-3 The mean particle sizes and the particle size distributions for silica

used for this study.

Table 3-4 Values for M O ) and @&200 h) at 60 OC.

TabIe 3-5 QE7(2 h) at difkrent annealing temperatures.

Table 3-6 D,, and the corresponding f,(t) values for films containing different

types and amounts of silica filler.

Table 3-7 Volume content of PBMA polymers (vol %) near the filler surface,

relative to the total polymer volume in latex films. The filler content is

40 wt% for d l the calculations.

LIST OF FIGURES

Figure 1.1 Schematic representation of the initial stage of an emulsion

polyrnerization system.

Figure 1.2 Schematic representation of the latex film formation process.

Figure2.1 Chernicalsuucturesofmonomersusedforthisstudy.

Figure 2.2 Scanning electron microscopy (SEM) images for precipitated

calcium carbonate (CaC03, Le ft-hand side) and colloidal silica

(SiO2. right-hand side).

Figure 2.3 Chemical structures of dye-labeled monomers.

Figure 2.4 GPC chromatognms for (a) Phe-P(MMA-CO-EHA) and (b)

MeAn-P(bfMA-CO-EHA) latex polymers. The samples were

prepared by drying a latex dispersion. followed by dissolution of

polymer into THF solvent. Cume (1): the fluorescence signal for

two-stage polymer; curve (2): the refractive index signal for two-

stage polymer; and curve (3): the refractive index signal for seed

polymer.

Figure 2.5 Schematic representation of polymer interdifision for a latex

film. When a latex film is annealed well above its glass transition

temperature (Tg), the film has a zone of interdifision which

contains mixtures of D- and A-labeled polymer molecules.

Figure 2.6 Schematic representation of polymer interdifision at the

molecular level. When a latex film is annealed above irs Tg, D-

Figure 2.7

Figure 2.8

Figure 2.9

Figure 2. L O

Figure 2.1 1

and A-labeled polymer molecules interdiffuse across the ce11

boundary and the distances between D and A groups become

smaller.

Donor fluorescence decay profiles (a) in a P(MMA-CO-EHA) latex

film, (b) in a P(MMA-CO-EHA) latex film containing 80 wt %

CnCOj, md (c) in s P(MM.4-CO-EH.4) latex film containing 20 wt

% SiO?. Each film was annealed at 50 f 1 "C for (1) O min, (2) 60

min, and (3) 990 min. The decay profiles are integrated to obtain

the area under the curves.

Effect of CaC03 content on the initial efficiency of rnergy

transfer. We plot the initial energy efficiency of energy transfer

OET(0) vs CaC03 content in latex films (wt %).

Effect of SiO2 content on the initial efficiency of energy transfer.

We plot the initial efficiency of energy uansfer &-(O) vs Si02

content in latex films (wt %).

Comparison of P(MMA-CO-EHA) extent of mixing, f,(r) vs

annealing time for films containing different arnounts of calcium

carbonate: O wt % (O), 30 wt % (e), 80 wt % (o), and 90 wt % (a).

The weight-average molecular weight (MW) of the P(MMA-CO-

EHA) latex was ca. 240,000. Films were annealed simultaneously

at 50 f I OC.

Comparison of P(MMA-CO-EHA) extent of mixing, f,(r) vs

annealing time for films containing different amounts of silica: O

wt % (O), 20 wt % (a), 40 wt % (a), and 50 wt % (i). The weight-

average molecular weieht ( M d of the PMMA-CO-EHA) latex was

ca. 240,000. Films were anneaied simultaneously at 50 f 1 OC.

Figure 2.12 Schematic representation of the mnealing process for a latex film

containing silica particles (25 nm). The silica particles can act as

obstacles and retard polymer diffusion andor reduce the mobility

of polymer molecules near the pigment surface.

Figure 2.13 Schematic representation of the annealing process for a latex film

containing precipitated calcium carbonate. The latex particles

spread locally dong the pigment surface.

Figure 3.1 Chernical structure of butyl methacrylate used for this study.

Figure 3.7 Chemical structures of dye-labeled monomers used for this study.

Figure 3.3 GPC chrornatograms for (a) Phe-PBMA and (b) An-PBMA latex

polymers. The samples were prepared by drying a latex dispersion.

followed by dissolution of polymer into THF solvent. Curve (1):

the fluorescence signai for two-stage polymer; curve (2): the

refractive index signal for two-stage polymer; and curve (3): the

refractive index signal for seed polymer.

Figure 3.4 Transmission elecuon rnicroscopy (TEM) images for various

particle sizes of silica. (a) K-25, (b) K-50, (c) S-12. (d) S-45.

Figure 3.5 Particle size distributions of silica particles used for this study. (a)

K-25, (b) K-50, (c) S-12, (d) S-45. The particle sizes were

determined by transmission electron microscopy (TEM).

Figure 3.6 Donor fluorescence decay profiles in a PBMA latex film after

annealed for (1) 0 min, (2) 60 min, (3) 330min, (4) 12,000 min.

Figure 3.7 Donor fluorescence decay profiles in a latex film. (a) PBMA with 59

40 wt % of 25 nrn SiO-, (K-25). (b) PBMA with 40 wt % of 50 nm

Si@ (K-50), (c) PBMA with 40 wt % of 12 nm SiO- (S-13, (d)

PBMA with 40 wt % of 45 nrn SiO2 (S-45) after annealed for (1) O

min. !2) 60 min. (3) 330min. (4) 12.000 min. respectively.

Figure 3.8 Plots of the initial efficiency of enegy transfer, OET (O), vs Si02 6 1

contents in newly formed films.

Figure 3.9 Plots of the maximum efficiency of energy transfer. (200 h), 64

vs SiO- contents. Films were annealed for 200 h at 60 OC.

Figure 3.10 Plots of the maximum efficiency of rnergy uansfer, Gn(2 h). vs 65

annealing temperature. Films contain 40 wt 8 of SiO2 and were

annealed for 2 h.

Figure 3.11 Plots of the maximum efficiency of energy transfer, Qm (200 h), 66

vs the surface to volume ratio (lldsiot). Films contain 40 wt % of

SiO2, and were annealed for 200 h at 60 O C . The diarneters of S a l

were determined by transmission electron microscopy (TEM).

Figure 3.12 Plots of the extent of mixing f,(r), as a function of annealing time. 70

Films contain (a) 25 nm of SiO2 (K-25), (b) 50 nm of SiO? (K-SO),

and were annealed simultaneously for each series of sample films

at 60 OC.

Figure 3.13 Plots of the extent of mixing f,(t), as a function of annealing time. 72

Films contain (a) 12 nm of Si& (5-12), (b) 45 nm of SiOz (S-43,

and were anneaied simultaneously for each series of sample films

xii

at 60 O C .

Figure 3.14 Plots of the extent of mixing f,(t), vs the square root of annealing

time. Si02 content for al1 Si02 contained films is 40 wt %. Those

films were annealed at 60 OC.

Figure 3.15 Plors of the dope vdues in Figure 3.13. as a functinn of the

surface to volume ratio, l/dsio2. The diameters of Si02 were

determined by transmission electron microscopy (TEM).

Figure 3.16 Mean apparent diffusion coefficients, Da,,, as a function of the

extent of mixing f,(t). Films contain (a) 25 nm of SiO2 (K-25), (b)

50 nm of SiO2 (K-50). and were annealed simultaneously for each

series of sample films at 60 OC.

Figure 3.17 Mean apparent diffusion coefficients. Dapp. as a function of the

extent of mixing fm(f). Films contain (a) 12 nm of S i 0 (S-12). (b)

45 nm of Si02 (S-45). and were annealed simultaneously for each

series of sample films at 60 O C .

Figure 3.18 Plots of the extent of mixing fm(t) vs the square root of annealing

time for latex films containing different arnounts of (a) S-12 and

(b) K-25.

Figure 3.19 Plots of the dope values in Figure 3.17, as a function of the silica

volume fmction.

Figure 3.20 Plots of 4, vs fm(t) for PBMA latex films containing O wt % ( O ) ,

10 wt % (a), 20 ~t % (CI), 30 wt % (a), a d 40 ~t % (A) of S- 12.

Figure 3.21 Master curve of Dvp vs f,(t) for PBMA latex films containing O

wt % (O), 10 wt % (a), 20 wt % (o), 30 wt % (a), and 40 wt % (A)

of S-12.

Figure 3.22 Plot of {~~[D,(T,~~)/D~(T.o)]}- ' vs Qf " for PBMA latex films

containing various fractions of S- 12.

Figure 3.23 Plots of DZF, vs ,(t) for PBMA latex films containing O wt % (01.

10 wt % (a), 20 wt % (a), 30 wt % (i), and 40 wt % (A) of K-25.

Figure 3.24 Master curve of D,, vs f,(r) for PBMA latex films containing O

wt % (O), I O wt % (m), 20 wt % (o), 30 wt % (i), and 40 wt % (A)

of K-25.

Figure 3.25 Plot of ( I ~ [ D ~ ( T , ~ ~ ) / D ~ ( T , O ) ] } " vs <Dr " for PBMA latex films

containing various fractions of K-25.

Figure 3.26 Plots of ATg vs the silica volume fraction for PBMA latex films

containing S-12 (O) and K-25 (a).

Figure 3.27 Plots of ATg vs the total volume of the polymer near the silica

surface for PBMA latex films containing S-12 (O) and K-25 (a).

The diameters of silica were determined by transmission electron

rnicroscopy (TEM).

Figure 3.28 Plots of ATg vs the total volume of the polymer near the silica

surface for PBMA latex films contûining S-12 (O), K-25 (a), S-45

(a), and K-50 (i). The diameters of silica were determined by

transmission ekchon microscopy (TEM).

Figure 3.29 Plot of the g l a s transition temperature Tg2 vs thickness of the

polymer layer tipi rigidified by the silica filler for the polymer

near the silica filler surface. The thickness is expressed based on

the radius of gyration RG for the PBMA polymer with a MW =

50,000.

Figure 3.30 Plots of ( In[D,(T, ar)/Dp(T. 0.233)] }-1 vs { i-6(1.7R&D~io2/d)-'

for PBMA latex films containing S-12 (O) or K-25 (a).

Figure 3.3 1 Schematic representation of rnorphology difference for latex films

containing either smaller (left-hand side) or larger (right-hand

side) size of SiO?. The silica fillers can act as obstacles and retard

polymer interdiffusion ancilor reduce the mobility of polymer

molecules near the filler surface.

Figure 3.32 Schematic representations for pol ymer immobilization near the

filler surface and obstacle effect.

Figure 3.33 Schematic representation for polymer molecules immobilized near

the filler sul-face. We calculated the thickness of the immobilized

polymer layer based on the radius of gyration, RG, and 2RG using

RG (PBMA, MW = 50.000) = 4.7 nm.

1. INTRODUCTION

1-1. Background

Synthetic binders play a key role in the process of paper and b oard coatings. 1

These products have become more and more important since the use of latex as a coating

binder was first examined in 1946. The sheet and print quaiity of coüted papérs Ilas bem

improved for many years due to their supenor characteristics as pigment binders.

The reason that latex is widely used as a pigment binder in paper coatings is that it

has overcome a lot of problems such as binder migration, print mottle, backtnp mottle.

blister, warer retention. different coating rheology. It has also irnproved smoothness,

gloss, pnnt gloss, pnnt fidelity. press performance. reduced n w material waste. and

convenience in handling.

There is a continuing need to lower costs of coated paper production while at the

sarne time maintaining or improving paper quality. To reach this goal, one would like to

have the knowledge necessary to design the structure of the coating in a way that

optirnizes the use of latex as a pigment binder. For this purpose. one needs a deeper

understanding of the role of the latex binder in the coating. We approach this problem by

trying to understand how a large amount of pigment in the coating formulation affects the

coalescence of latex particles and the subsequent polymer diffusion that enhances the

mechanical properties.

1-2. Emulsion Polymerization

Emulsion polyrnenzation has developed into a widely used process for the

production of synthetic latexes since its fint introduction on an industrial scale in the

mid-1930s. ' Today, various kinds of synthetic polymers are prepared by emulsion

polyrnerization. Examples include synthetic rubber, high-impact polymers, latex foam,

latex paints, paper coatings, carpet backing, adhesives, binders for non-woven fabrics.

barrier coatings additives for consuuction materials such as Portland cement, mortar and

concrete, and sealants and additives. Latex particles are also used for a growing number

of specialty applications such as: diagnostic tests. immunoassays, biological cell-labeling.

drug delivery systems, chrornatographic separations, and as size calibration standards for

many instruments such as blood counters and electron microscopes.

Emulsion polymerization is defined as a free-radical-initiated chain

polymerization, in which a monomer or a mixture of monomers is dispened in an

aqueous medium, with the aid of ernulsifier molecules and polymerized by a water-

soluble kce radical initiator. ' Water is the main ingrdient in ernulsion polymerizstion. II

maintains a low viscosity and provides good heat transfer. In addition, it serves to isolate

the polymerization loci. This is one of the advantages in emulsion polymenzation, termed

compartmentalization. The water also acts as the medium through which monomer passes

from monomer droplets to the growing particles. It is the locus of water-soluble initiator

decomposition and oligomer formation. and the medium of dynamic exchange of

surfactant between the phases.

Surfactants are surface-active molecules. In emulsion polymerization, they

perform two major functions. One is to provide sites for panicle nucleation. and the other

is to provide colloidal stability to the growing particles as a result of their adsorption at

the particle-water interface. Anionic surfactants are the surfactants most commonly used

in emulsion polymenzation. Cationic surfactants are also used for special applications in

paper coatings and asphalt additives. Nonionic surfactants are also used to control the

latex particle morphology. and to enhance the colloidal stability against mechanical shear,

freezing, and added electrolytes. Reactive surfactants. which are surface active molecules

with an active vinyl group, are also used in order to bind surfactants chernically to the

latex surface. In this case, they have an advantage of reduced desorption during film

formation and reduced water sensitivity of the latex films. Water-soluble polymers such

as poly(viny1 alcohol) and hydroxyethyl cellulose can be used as non-surface-active

emulsifiers,

The most commonly used water-soluble initiaton are salts of persulfuric acid,

such as potassium persulfate and ammonium persulfate. On heating, persulfate ions

dissociate into two sulfate radical anions, which c m initiate emulsion polymerization.

Redox initiaton, typically a mixture of an oxidizing agent and a reducing agent, are used

for emulsion polymerization at a low temperature.

A variety of chah transfer agents, such as mercaptans, are used to control the

molecular weight of the emulsion polymer.

Two major mechanisms for particle formation have been proposed for emulsion

polymerization. These are micellar nucleation. and homogeneous nucleation. Both

processes are considered to proceed simultaneousiy. In micellar nucleation. initiator

radicds genented in the aqueous phase enter the monomer-swollen surfactant micelles,

as single radicals or ~lipndicals, =d initiatc potymcnzation to fom monorner-swollen

polymer particles, which grow by propagation reactions. In homogeneous nucleation

radicals genented in the aqueous phase propagate by adding monomer units to form

water-soluble oligomers until they reach the lirnit of their solubility in the aqueous phase

and precipitate out of solution. A schematic representation of the initial stage

emulsion polymenzation system is shown in Figure 1.1.

Micelle d = 5-10 nm

Monomer is solubilized.

Monomer droplet d = 1,000-10,000 nrn

Initiator + 2 Re (Radicals) , Monomer-swollen polyrner particle

Aqueous phase

Figure 1.1: Schematic representation of the initial stage of an emulsion polymerization

system.

1-3. Latex Film Formation

One of the major applications for emulsion polymers is in the area of waterbome 4 coatings. The latex onginally consists of a colloidal dispersion of particles in an aqueous

medium. Upon drying the particles are transformed into a void-free and mechanicalty

coherent polymer film. In this process. the latex particles have to be deformed in order to

fil1 space between the particles, which results in polyhednl structure. For this process to

sccur, the temperature should be above the minimum film-forming temperature (MFQ of

the system. The M F ï often corresponds to the glass transition temperature of the latex

polymer in the presence of water. Here. one must have sufficient adhesion between the

latex cells so that one can improve the mechanical and barrier properties of the film.

According to Winnik. the process of Iatex film formation can be divided into

three stages. In the first stage, water evapontes until the particles corne into close contact

to each other. In the second stage, the latex particles deform and begin to fil1 the

interstitial spaces between the individual particles at temperatures above the MFFT.

Finally, in the third stage, the particles begin to coalesce and interdiffuse, which results in

the formation of a mechanically sound film. The ease of latex film formation has a strong

influence on the properties of paper coatings. as well as paints and coatings. A schematic

representation of the latex filrn formation process is shown in Figure 1.2.

Aqueous Iatex solid content 20-50 wt %

Stage 1 b

water evaporates

Stage II b

particle deforms T2MFT Packing of deformed partides

Stage III / 1 1 1 ~ 1 / 1 / 1 / 1 / / / / / 1 / / I / l / / / l l l / l / / / / / / / l / / / / / aging and further coalesceEce t 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

T 2 T g Mechanically coherent füm

Figure 1.2: Schematic representation of the latex film formation process.

1-4. Polymer Interdiffusion in Latex Fiims

Scientists studied the surface morphology of latex films by electron microscopy

many years ago, and found that the cellular structure in newiy-formed latex films

disappeared upon aging or annealing. It seemed that this process was related to the

growth in mechanical properties of the films. Bradford and Vanderhoff referred to this

process as gradualfurther coalescence, and Voyutski ' used the term autoliesion. In his

publication, autohesion referred to the diffusion of polymer chains across the interparticle

or intercellular boundaries. Unfortunately, they did not have any procedures that allowed

one to measure this diffusion directly. Over the past 15 years, severd techniques have

allowed one to measure this difision quantitatively, and a broad body of theory has been

created to describe the diffusion process.

Polymer diffusion across interfaces, which is also termed polymer inrerdifùsion,

is an important feature in many processes such as heding of the interfaces in latex films,

polymer weiding, crack healing, powder coatings, and compression-molded polymer

powders. The tenn interdifision is used for processes of mixing, intermingling, and

homogenization at the molecular and macroscopic levels. Interdiffusion implies diffusion

among distinguishable molecules.

The theory of polyrner diffusion across interfaces has been a fascinating subject

for many years. Polyrner chains adjacent to an interface have a different distribution of

conformations to those in bulk. As they diffuse across the interface. the conformations are

iandonized md thercfore one cbtains an extra entropic dnving force for the interdiffusion

of surface polymer chains. If the polymer chains are sufficiently short, the diffusion will

follow Fick's laws. In latex films, polymer interdiffusion is recognized as an important

process in development of mechanical strength.

There are sevenl techniques to charactenze aspects of the polymer diffusion

process in latex films. Those are transmission electron microscopy (TEM), Y atornic force

microscopy (AFM). freeze-fracture TEM. 5s 10 A few other techniques allow one to

measure the rates of polymer interdiffusion. Small angle neutron scattering (SANS) and

fluorescence resonance energy transfer (FRET) are the most activety used to study the

rate of polymer interdiffusion. In this study, we use fluorescence resonance energy

vansfer (FRET) technique.

1-5. Fluorescence Resonance Energy Transfer (FRET)

Ruorescence resonance energy transfer (FRET) involves the transfer of the

excited state energy from a fluorescence donor (D) to a fluorescence acceptor (A). 11

FRET takes place without emission of a photon. and it occurs as a result of a resonant

dipole-dipole interaction between the donor and acceptor molecules. The rate of energy

transfer w(r) is expressed by the following equation:

where .rD0 is the donor fluorescence lifetime in the absence of acceptors, and Ro is the

characteristic distance (the Forster distance) over which energy transfer takes place. Sb

This rate is deterrnined by several factors. such as the extent of overlap of the emission

spectnim of the donor with the absorption spectrum of the acceptor, the relative

orientation of the donor and acceptor transition dipoles, and the distance between the

donor and acceptor molecules.

FRET was fint applied to biochernical research, '' and it has also been extensively 13-15 employed in the field of polymer science. One of the most interesting features is that

this technique can be applied to the study of polymer diffusion across the inter-particle

boundary in latex films, Sb. 16. 17 which is also important for paper coatings that contain a

large amount of minenl filtex-s. In this study Ive examine how rhose fillen affect the rate

of polymer interdiffusion in latex films.

1-6. References

Macnair, A. K., Task Group Chaiman, Syithetic Coaring Adhesives, TAPPI PRESS,

Atlanta, 1998.

Lovell, P. A.: El- Aasser. M. S ., Emiilsion Polymen'zation and Emtilsion Polymers,

Wiley, Chichester, 1997.

30' Annual Short Course "Advances in Ernulsion Polymerization and Latex

Technology" El-Aasser, M. S. ed., Lehigh University, PA, Vol. 1, 1999.

(a) Paton, T. C., Paint Flow and Pigment Technology, Wiley, New York. 1979. (b)

Turner, G. P. A., lntroduction to Paint Chemistry and Principles of Paint

Technology, 3" ed., Chapman and Hall, London. 1985.

(a) Wang, Y.; Kats, A.; Juhué, D.; Winnik, M. A. Langmoir 1992, 8, i435. (b)

Winnik, M. A. The Formation and Properties of Latex Films in Emulsion

Polymerization and Emulsion Polymers, Lovell, P. A., El-Aasser, M. S., Eds., Wiley,

New York, 1997, p 467.

Bradford, E. B.; Vanderhoff, J. W., J. Macromol. Chem., 1966,1,335.

(a) Voyutski, S. S., J. Polym. Sci.. Part A. 1958, 32, 528. (b) Voyutski, S. S.,

Autohesion and Adhesion of High Polymers, Wiley-Interscience, New York, 1963.

(a) Joanicot, M.; Wong, K.; Maquet, J.; Chevalier, Y.; Pichot, C.; Graillat, C.;

Lindner, P.; Rios, L.; Cabane, B. Prog. ColIoid Polym. Sci., 1990, 81, 175. (b)

Chevalier, Y.; Pichot, C.; Graillat, C.; Joanicot, M.; Wong, K.; Lindner, P.; Cabane,

B. Colluid Polym. Sci., 1992,270,806.

9. (a) Wang, Y.; Juhue, D.; Winnik, M. A.; Leung, O.; Goh, M. C. Lcingmzrir, 1992, 8,

760. (b) Goh, M. C.; Juhue, D.; Leung, O.; Wang, Y.; Winnik, M. A. Langmuir,

1993,9, 1319.

10. (a) Roulstone, B. J.: Wilkinson, M.; Hearn, J.; Wilson, A. J. Polyrn. Int., 1991, 24,

87. (b) Roulstone, B. J.; Wilkinson, M.; Hearn, J. J. Polym. Inr.. 1992, 27.43.

1 I . Lakowicz. J. R. Principles o f Flitorescence Spectroscop~. Plenum Press, New York,

1983.

12. Herman, B . Fluorescence Microscopy and fluorescent Probes, edited by Slavik. J..

Plenum Press. New York, 1996, p 1.

13. Rager, T.; Wegner, G.; Winnik. M. A. ~facroniolecules, 1997,30,49 1 1 .

14. Marowetz, H. and Amrani, F. Macmmolecrrles, 1978, 11,28 1 .

15. Qui, X. P.; Jiang, M. Polymer. 1995. 36,360 1 .

16. Winnik, M. A. Cwr. Opin. Colloid In., 1997.2. 192.

17. Boczar, E. M.; Dionne, B. C.: Kirk, A. B.; Lesko. P. M.; Koller, A. D.

Macronzolecrrles, 1993.26, 5772.

2. EFFECT OF PIGMENT TYPES ON POLYMER

INTERDIFFUSION

2-1. Introduction

Latex polymer is widely used as a pigment binder for various types of coated 1 papers. It is of significant importance to understand the factors that affect the structure

of the coating, especially the spatial arrangement of pigment particles and binder. These

factors are important for the control of product quality and the design of new products. ' Understanding latex particle deformation and the evolution of film properties is also

important because these are the steps that promote adhesion to the inorganic pigment and

adhesion to the cellulose fiben. The amount of binder in a paper coating is small, due to

the relatively high cost compared to the cost of conventional pigments such as clay and

calcium carbonate. Thecefore. it is essential to optirnize the characteristics of the latex,

both for maintaining or improving the quality of the paper, and for lowering the paper

cost.

It is well known that, during the drying process, latex particles deform above the

glass transition temperature (Tg) of the po1ymer to form a void-free solid cornprised of

polyhedrd cells. Over time, this cellular structure is lost as polymer diffuses across the

intercellular boundaries and it creates a continuous polymer matnx. There have been a

number of studies on wetting and adhesion of polymer latex particles to inorganic 4-6 pigments in paper coatings. These include the scanning electron microscope studies by

Scnven et al., and the recent neutron scattering study by Joanicot et ai. In spite of this

effort, it is still not well understood how the latex particles deform in the presence of

large arnounts of inoqanic pigments dunng film formation.

Our research group has developed a quantitative technique based on fluorescent

spectroscopy for studying polymer interdiffusion in latex films. In this technique, one

mixes two essentially identical types of latex particles in dispersion pnor to film

formation. Both types of latex particles are covalently labeled with about I mol % of a

fluorescent dye, one with a dye (D) that can act as a donor in a fluorescence resonance

energy transfer (FRET) experiment, and the other with a corresponding acceptor dye (A).'

In a FRET experiment, the donor dye is excited selectively with light at a wavelength

where it has a strong absorption. This excited dye can fluoresce, or, if there is a nearby

acceptor dye, the enegy can be transferred to the acceptor dye via a resonant coupling of

the transition dipoles. The energy transfer process occurs as described in eq 2-1. When a

dispersion containing donor- and accepter-labeled particles dries to form a film, some of

the boundaries separate donor- and acceptor-labeled polymer. As these polymen diffuse

across the boundaries. they bring donor and acceptor dyes into proxirnity, allowing the

extent of energy transfer to increase.

D * + A + D + A * (3- 1)

Over the past 15 years. we have used this technique to examine a varie ty of hcton

that affect the rate of polymer diffusion in latex films. Most of these experiments were

carried out on pigment-free dispersions, with the intent of understanding how the

properties of the binder phase itself evolves. "ecently we began to examine the

influence of pigments on the polymer diffusion rate in latex films. In this work, we have

been interested in modeling the effect of pigments in latex paints. which have a higher

binder content than that of paper coatings. In one study, Feng et al. found that hard

polymer filler particles such as poly(rnethy1 methacrylate) (PMMA) and RopaqueB

(Rohm & Haas) significantly retard the rate of polymer interdiffusion. In this paper, we

describe the application of FRET measurements to the study of the polymer interdiffusion

process in latex films that contain large amounts of inorganic pigments, typical of those

used in paper coatings. We show that calcium carbonate pigment and silica have very

different effects on the rate of polymer diffusion in these films.

2-2. Experimental

2-2-1. Materials for this study

2-2-la. Materials for Emulsion Polymerization

Monomers, such as methyl methacrylate (MMA, Aldrich, 99%), and (2-

ethylhexyl) acrylate (EHA, Ruka, 99%) were distilled under vacuum under a N2

aunosphere, and stored in refngerator pnor to use. Potassium persulfate (K2S2O8, KPS,

Aldrich, 99%), sodium bicarbonate (NaHC03, Caledon, 99%), sodium dodecyl sulfate

(Ci2H250S03'Na+, SDS, Aldrich. 98%) and 1-dodecanethiol (Ci2HXSH, DM, Aldrich,

98%) were used as supplied. Distilled water was further punfied through a Millipore

~ i l l i - ~ ~ system. Chernical structures of monomers used for this study are shown in

Figure 2.1.

MMA EHA

Figure 2.1: Chernical structures of monomers used for this study.

2-29 1 b. Inorganic Pigments

Precipitated caicium carbonate (TP-221GS). dispersed with sodium polyacrylate.

was supplied as a suspension by Okutûma Kogyo Co., Ltd. This pigment has an

ellipsoidal shape, termed scalenohedrd calcite, with a length of 0.7 to 1.0 Pm, and a

diameter of 0.25 to 0.3 Pm. This pigment is cornmonly used for high gloss coated paper.

ColIoidal silica (Klebosol 30R25) was supplied as a suspension by Clariant Corporation.

The filler particles are amorphous sphencd silica beads with a diameter of around 25 nm

and a narrow size distribution. 'O The particles show good dispersibility in waier. Both

pigments were used as supplied. Scanning electron microscopy (SEM) images for CaC03

and SiOl are shown in Figure 2.2. Further pigment characteristics are shown in Table 2.2.

SWc 'IL x,iTticr :-:: 4 :i . P x : c i r rue 2 5 '.'- . - 6)

- -. . -- -- - - . 'r"

Figure 2.2: Scanning electron microscopy (SEM) images for precipitated calcium

carbonate (CaC03, left-hand side) and colloidal silica (SiO?, right-hand side).

2-2-2. Synthesis of Dye-labeled Monomer

9-Phenanthrylmethyl methacrylate (Phe-MMA) were synthesized previously. ' l 9-

Methacryloxyrnethyl- 10-rnethylanthracene (MeAn-MMA) has recently been synthesized

by Liu et al., using anthraquinone as a srarting material. " We use MeAn-labeled

monomer io carry out emulsion polymerization of acrylate monomers. The chernical

structures of dye-Iabeled monomen are shown in Figure 2.3.

Phe-MMA MeAn-MIMA

Figure 2.3: Chernical structures of dye-labeled monomen.

2-2-3. Synthesis of Poly(iMMA-CO-EHA) Copolymer Latex

Poly(MMA-CO-EHA) copolymer latex samples (monomer weight ratio of 1: l),

labeled with I mol % of a fluorescent dye [either a donor, phenanthrene (Phe) or an

acceptor, anthracene (An)], were prepared by semi-continuous emulsion

copolymerization at 80°C, using KPS as the initiator, SDS as the surfactant. and L -

dodecanethiol as the chain transfer agent. The reaction conditions were similar to that

described previously. l 3 AS a donor-labeled monorner. we used (9-phenanthry1)methyl

methacrylate (Phe-MMA) to introduce the donor dye. The synthesis of this monomer is

described in ref 1 1. As an acceptor-labeled monomer. we used 9-rnethacryloxymethyl- 10-

methylanthracene (MeAn-MMA) descnbed in ref 12. The recipe for L 10 nm P(MMA-CO-

EHA) latex is shown in Table 2.1. We used the sarne seed latex particles obtained in the

first stage for preparing both the Phe- and MeAn-labeled latex particles. Their

characteristics are shown in Table 2.2. These latex particles, referred to as P(MMA-CO-

EHA), have a diameter of L 10 nm, and a Tg of 7 OC.

Table 2.1: Recipe for preparing 110 nm P(MMA-CO-EHA) latex.

MMA (g) -

EHA ($1

MeAn-MMA (g)

Phe-MMA (g)

a. 1 mole % relative to the total monomer fed in the second stage.

b. Reaction temperature.

c. Reaction time.

First stage -

(batch process)

1.75

Water (g)

KPs (g)

2-2-4. Characterization of Latex

2-2-Sa. Particle Size Measurement

The particle size and size distributions were determined by dynamic light

scattering employing a Brookhaven BI-90 particle sizer. This measurement is based oii

quasi-elastic light scattering (QELS), dso referred to as dynamic light scattering (DLS).

1.75

Second stage

(under monomer starved condition)

58.8

0.06

Phe-Iabeled

17.5

17.5

0.75 a

MeAn-labeled

17.5 1

17.5

0.78 "

27.0

0.06

27 .O

0.06

2-2-4b. Gel Permeation Chromatogaphy (GPC)

Molecular weights and molecular weight distributions were measured by gel

permeation chromatognphy (GPC), using two Ultrastyngel columns (500 + 10'' A, or 10"

A + lo5 A for very high molecular weights) with tetrahydrofuran ( T m as the eluent and

a flow rate of 0.4 mUmin. A srnall portion of a latex dispersion was dned in the oven for

2 h at 120 OC. Then the polymer was dissolved in tetrahydrofuran (THF) to give a

solution of c3. 0.5 :vt 9 polynier concentration. The solution was filtered before it was

injected into the column. Colurnns were calibrated with poly(methy1 methacrylate)

(PMMA) standards. Dual detectors (WATERS Series R-400 Differential Refnctometer

as a refractive index detector and KRATOS FS 970 Spectrofluoro Monitor as a

fluorescence detector) were used to detect the presence of the donor and acceptor dyes

and to ensure that these fluorescent dyes are randomly disuibuted in the polymer

backbone.

2-2-4c. Solid Content of latex dispersions

The solids content of latex dispersions was measured gravimeincally by

measuring the weight of a small portion of a latex dispersion (WL) and the weight of solid

after dned (Ws) as:

They were found to be ca. 30 wt %.

Table 2.2: Charactenstics of the latex and inorganic pigments. The particle size, Tg, and

molecular weight for both Phe-P(MMA-CO-EHA) and MeAn-P(MMA-CO-EHA) are very

similar to those for MeAn-P(MMA-CO-EHA).

a. The length of the pigment is 700 - 1000 nm.

' P(MMA-EH*)

Phe- + MeAn-labeled

b. Nominal molecular weights based upon linear methyl methacrylate standards.

2-2-5. Film Preparation

Latex films were prepared From dispersion mixtures with a 1: 1 number ratio of

Phe- and MeAn-labeled P(MMA-CO-EHA) latex particles and different amounts of

inorganic pigment. The final dispersions, with total solid contents from ca. 30 wt % to 60

wt %, were adjusted to pH 9, using diluted potassium hydroxide solution for the

dispersions with Si02 and those without pigment, and with diluted hydrochlonc acid

solution for the dispersions with &CO3. Film formation was carried out by the following

procedure. For each film, we took three drops from each dispersion using a Pasteur

pipette and spread them ont0 a quartz plate. Then we leveled off those dispersions using a

stainless steel blade, and the film was allowed to dry for 10 minutes in air at 23 O C ,

followed by storage in the cold room at 4 OC to minimize the amount of polymer

interdiffusion in the film. A typicd film thickness was 100 km. Films formed from latex

aione were crack-free and transparent. The films became more turbid as one increased the

diameter (nm) 110 250 - 300 "

Precipitated

Calcium carbonate Colloidd silica

amount of precipitated calcium carbonate. All films containing colloidal silica were

transparent and crack-free up to 40 wt %. but small cracks were observed in the film

containing 50 wt % of silica.

2-2-6. Energy Transfer Measurement

Al1 films were annealed at 50 i 1 O C for polymer diffusion measurements. For

each senes of samples ro be compared, the films were anneaied sirnuitanrously.

Fluorescence decay profiles were measured by the single photon-timing

technique.'" Samples were excited at 300 nm. and emission was detected at 350 nm. A

bandpass filter (350 f 5 nm) was used to minimize the scattered light and interference due

to fluorescence from directly excited acceptors. For each measurement. it took about 10

to 15 minutes to collect 5000 counts in the maximum channel. Prim to each measurement,

a film sample was placed in a quartz tube. and the tube was degassed with flowing

nitrogen gas.

Due to the high pigment content. many of the films we examine are turbid or even

opaque. For these films, the extent of light penetration into the film is limited. Due to a

relatively large extent of light scattering in opaque samples. the depth of light peneuation

may be as small as the wavelength of the excitation light (here 300 nm). One therefore

needs to dign the optics carefully to minimize the amount of scattered light reaching the

detector. Because the scattered light is at shorter wavelengths than the ernitted light,

proper use of filters can rninimize its contribution to the measured decay. We excite the

donor at 300 nm and detect the emission at 350 nm with a bandpûss filter (350 f 5 nrn).

Because scattered light has the sarne time profile as the excitation pulse, one can correct

the measured decay profile for any residual contribution due to scattered light. 15

2-2-7. Data and Data Analysis

The latex films we examined were prepared from a 1:1 mixture of donor- and

acceptor-labeled latex particles. We monitor the polymer diffusion process by meauring

changes in the extent of energy m s f e r between the donor and acceptor dyes attached to

these latex polymers. When a donor dye D is excited, it can transfer its energy to any

nearby acceptor dyes A. The important feature of this process is that the rate of energy

transfer w(r) depends sensitively on the distance r between the donor and acceptor

molecules:

where roo is the donor fluorescence lifetime in the absence of acceptors, and is the

characteristic distance (the Forster distance) over which energy transfer takes piace. ' For energy transfer for donon and acceptors randomly distributed in a three

dimensional Euclidean space, the donor decay function will have a stretched exponentiai

fom ' :

where Io is the intensity at zero time. and P is a panmeter proportional to the local

concentration of acceptors CA:

where NAv is Avogadro's number and # describes the averaged relative orientation of the

donor and acceptor di pole moments.

In our expeximents. we measure donor fluorescence decay profiles b ( t ) . To fit

each decay curve, we use the following phenomenological equation:

While this equation is sirnilar in form to eq 2-41. no meaning is ascribed to the

individual fitting panmeters. The panmeters Al , A?, and p are obtained from the fit of

each decay profile, and we use these fitting parameters to in t ep te ID ( t ) andytically,

from decay time t = zero to t = infinity.

The efficiency of energy transfer, @n(t) is defined as shown in eq 2-6:

nrrmbrr of ET events = 1 - 1; I D (4

@, (4 = nimber of photons absorbed (2-6)

J ' " I ; O dt

where J œ ID ( t)dt is the area under the fluorescence donor decay profile obtained from O

area (t) Qm(r) = 1 -

While the extent of eneqy transfer can in principle be determined by measunng

the intensities of donor and acceptor fluorescence, this experiment suffen from several

artifacts, particularly the absorption by A of light ernitted by the excited D. In Our

expenments, we avoid this problem by carrying out fluorescence decay experiments. We

measure the influence of the acceptor dye on the decay rate of the donor dye following

pulsed excitation. In the absence of acceptor, the phenanthrene chromophore we employ

as the donor dye undergoes an exponential decay with a lifetirne roo. In Our analysis of

the donor fluorescence decay data, we assume that al1 deviations from a stnctly

exponential donor decay profile are due to FRET. For the films we examine here, roo =

45 ns.

Another useful parameter is the extent of mixing,&(t), expressed in terms of the

growth in energy transfer efficiency, normalized by the maximum change associated with

complete mixing. 16

#, (r) - @, (O) - area(0) - area(t) f, (0 = -

@ ( ) - @ (O) area(0) - ares(=)

where [O&) - #do) ] represents the change in the efficiency of energy transfer between

the initially prepared film and a film annealed for time t.

2-3. Results

2-3-1. Dye Distribution Analysis by GPC

In the next section, we describe how polymer chains behave in latex films before

and during annealing. In order to carry out the FRET measurements, one needs to know if

the dyeç are randomly distributed in the polymer backbones, which affects physical

behavior of the labeled polymer. We assume that the dyes are randomly distributed dong

the polymer chnins, 3nd &si the nurnbrr of dyes is proportional to the polymer chain

length. It is important to chuacterize not only the overall content of the dye-labeled

monomer in the polymer backbones. but also the distribution of the dye attached to the

polymer.

In order to examine this. Sosnowski suggested that one determine the fluorescent

dye distribution in the polymer chains by employing Gel Penneation Chromatognphy t 7 (GPC) analysis. In this section, we describe how we analyzed the dye distribution of

different kinds of labeled polyrner latexes in a simple, qualitative way. using GPC. We

use tandem detectors: one is based upon the difference of refractive index between eluent

(here, THF) and the polymer (Ri detector), and the other is based upon the fluorescent

intensity (fluorescence or UV detector). The data are collected simultaneously, so that one

can distinguish where the fluorescence intensity cornes from.

Figure 2.4 shows a GPC chromatogram for (a) Phe-labeled P(MMA-CO-EHA) and

(b) MeAn-labeled P(MMA-CO-EHA) latex polymers. We show three curves: curve (1) for

the fluorescence signal for the two-stage polymer, curve (2) for the refractive index (RI)

signal for the two-stage polymer, and curve (3) for the Ri signai for the seed polymer. In

curve (1) and (2), the signai derived from seed polyrner is not visible because total

content of seed polymer is only 10 wt 8, and the distribution of the molecular weight is

partly overlapped with that of polymer fonned in the second stage. One can also see that

curves (1) and (2) have the sarne trend, which indicates that dyes are uniformly labeled

dong polymer backbones. The fluorescence signai is shifted slightiy to the left side of the

RI signal because the fluorescence detector is placed just in front of the Ri detector. In

addition to that, there is no fluorescence signal at long elution time, which indicates that

there is no fluorescent comonomer left unattached to the polymer backbones, and no

fluorescent oiigorner formed during the preparation of the labeled latex particles. We

prepared the labeled latex under monomer-starved condition, which enables us to obtain

polymer in which fluorescence dyes are randornly distributed dong the polymer chahs.

In Figure 2 . k one notices that this polymer has a broad molecular weight

distribution. We obtain the weight average molecular weight of ca. 240,000, and its

polydispersity index (PD1 = M W / Mn) of 4.6. The same trend can be seen in Figure 2.4b,

indicating that the broad molecular distribution is derived from the nature of the

cmnonomers (hsrc, MMX and EHX).

d 1

6 8 l O 12 1.1 16 18 20 22 2.1 16 6 8 10 17 14 16 18 10 21 2.1 26

Retention Tirne (min) Retention Tirnr (min)

Figure 2.4: GPC chromatognms For (a) Phe-P(MMA-CO-EHA) and (b) MeAn-P(MMA-

CO-EHA) latex polymers. The samples were prepared by drying a latex dispersion,

followed by dissolution of polymer into THF solvent. Curve (1): the fluorescence signal

for the two-stage polymer; curve (2): the refractive index signal for the two-stage

polymer; and curve (3): the refractive index signal for the seed polymer.

2-3-2. Polymer Interdiffusion in Latex Films Using Fluorescence Resonance Energy

Transfer (FRET) Technique

In Figure 2.5. we present a drawing of a plana. section of an idedized latex film

prepared from a mixture of D- and A-labeled latex particles in the absence of pigment.

This type of ordered structure is obtained if the particles are uniform in size and

oganized into a face-centered cubic array at a high solid content as the dispersion dries."

When the latex film is annealed or allowed to age, polyrner diffusion across the

intercellular boundmies brings the donor- and accepter-labeled polymers into proximity.

A schematic representation of polymer interdiffusion across the intercellular boundary in

a latex film is shown in Figure 2.6. When two adjacent cells are labeled, respectively.

with D and A chromophores, polymer interdiffusion will bring the D and A dyes closer

together. This process leads to a measurable increase in energy transfer.

D A Annealing A D D -+

D A T > T g

: zone of interdiffusion

Figure 2.5: Schematic representation of polymer interdiffusion for a latex film. When a

latex film is annealed well above its glass transition temperature (Tg), the film has a zone

of interdifision which contains mixtures of D- and A-labeled polymer molecules.

1 Annealing T > Tg

Figure 2.6: Schematic representation of polymer interdiffusion at the molecular level.

When a latex film is annealed above its Tg, D- and A-labeled polymer molecules

interdiffuse across the ce11 boundary and the distances between D and A groups become

smaller.

In this chapter. we examine the influence of inorganic pigments on the polyrner

diffusion rate. We use two types of inorganic pigments. One is precipitated calcium

carbonate (CaC03) with a much larger diameter than the latex panicles. The other

pigment is colloidal silica (SiO?) with a smaller diameter than the latex particles. Those

pigments may be expected to play different roles dunng the film formation process. In the

second stage of film formation, the latex puticles are likely to be located between the

CaC03 pigments, because the size of the CaC03 particles is much larger than that of the

latex particles. On the other hand, SiOl may be located between the latex paaicles,

because the size of SiOl particles is much smaller than that of the latex particles.

19 Feng et al. studied the effect of poly(methy1 rnethacrylate) (PMMA) filler

particles on the rate of polymer diffusion in films prepared from poly(buty1 methacrylate)

(PBMA). In their experiments, they examined films containing a constant fraction (35 vol

%) of PMMA particles of different sizes. They found that the difision rate decreased in

proportion to the increase in the surface area of the hard filler particles. i.e., with a

decrease in the hard particle size at constant filler volume. Based on their results, we

expect to see a larger effect on the polymer interdiffusion rate in the films containing

SiO2, due to its much larger surface area, than those containing CaC03.

Typical donor fluorescence decay profiles in a latex film are shown in Figure 2.7.

Figure ?.?a shows data obtained for a latex film containing no pigment. Figure 2.7b

shows the corresponding decay traces obtained for û film containing 80 wt % CaCO,.

Figure 2 . 7 ~ shows the corresponding decay traces obtained for a film containing 20 wt %

SiO2. When the latex film is annealed at temperatures well above Tg, polyrner

interdifision occurs. One c m see the evolution of polymer interdiffusion in the films by

Iooking at the extent of curvature of the decay profiles. As one anneals the Film for longer

times, the curvature becomes more pronounced, which indicates that polymer

interdiffusion is promoted by heat and annealing time. One can see that polymer diffusion

rate is fast for the first 60 minutes. and then slows down in Figure 2.7b. One can also see

that the extent of polymer interdiffusion for the films containing pigment is smaller than

that for the film containing no pigment.

(a) P(h1hIA-cwEHA) iI

O 50 100 150 200 250

Time (ns)

lm

* mo *œ VI e U CI

c - 100

1 O

O 50 100 150 200 250

Time (ns)

O 50 LOO 150 200 250

Time (ns)

Figure 2.7: Donor fluorescence decay profiles (a) in a P(MMA-CO-EHA) latex film. (b)

in a P(MMA-CO-EHA) latex film containhg 80 wt % CaC03, and (c) in a P(MMA-CO-

EHA) latex film cont-ng 20 wt % SiO2. Each film was annealed at 50 t 1 OC for (1) 0

min. (2) 60 min. and (3) 990 min. The decay profiles are integrated to obtain the area

under the curves.

2-33, Efficiency of Energy Transfer, M O ) in Newly Formed Films

In this section we examine the influence of pigment on the extent of energy

transfer in newly formed films. If these films are prepared at low enough temperature;

M e or no polymer interdiffusion will take place. Energy transfer will occur only across

the interface between cells formed by the D- and A-labeled latex particles. Under rhese

circumstances @&O) is a measure of the interfacial area between D- and A-labeled celis 19 in the film. The ncwly fomcd films we rxarrinsd were dlorved !O dry at room

temperature (23 OC, i.e., above the minimum film forming tempenture). But as soon as

the film appeared to be dry, it was transferred to the cold room at 4 O C for storage until

the decay profile of the cold film could be measured. Since the Tg of the matrix polyrner

is 7 OC, we imagine that minimal polymer diffusion occurs when the films are prepared in

this way. We use experimental values of area(0) and @do) to examine the effect of

pigment on the contact between D- and A-labeled latex particles in the newly formed

films.

Effect of CaCOl

In Figure 2.8 we plot @&O) vs &CO3 content (wt 8) for a series of freshly

prepared films. The values of #&O) obtained range from 0.082 to 0.095. In other

experiments on nascent films prepared from sirnilar-sized latex particles at temperatures

close to the minimum film forming temperature, @do) values on the order of 0.05 to

0.07 were obtained. 'O These results suggest that Little polymer diffusion has occurred in

the samples we have examined. Note also that meaningfui data are obtained for films

containing 80 wt % and 90 wt % CaC03. These films contain large amounts of air voids

and, as a consequence, are opaque. We obtained a flat line through al1 the data points.

Measurements were repeated several times, and the error bars were attached to each set of

data in order to examine how the data were dispened. We took the average vdue anci

showed both the minimum and the maximum vaiues as error bars. The results indicate

chat the measurements are reproducible. By carefblly aligning the optics for the

measurement, the extent of light scattering reaching the detector is minimized. Some

scattered light at the excitation wavelength is detected. but we correct its contribution to

the fluorescence signal as described in ref 15, by the software used for data analysis.

One can see in Figure 2.8 thar the magnitude of @do) is essentidly independent

of the arnount of CaC03 in the film. This result indicates that even when large amounts of

CaC03 is present, there is a comrnon extent of interfacial contact in the film between cells

formed from D- and A-labeled latex particles. One could imagine that wetting of the

pigmcnt by ille htcx pdymer rnight eïen promcte the extent of mixing in the newly

formed films. This does not occur. We conclude that CaC03 pigment does not promote

coalescence of the latex particles.

Effect of SiO2.

In Figure 2.9 we plot values of # d o ) vs Si02 (wt %) for newly fomed latex

films. In the case of latex films containing CaC03, useful films could be prepared

containing as much as 90 wt % pigment. Here we were limited to films containing up to

50 wt 8 Sioz. When we attempted to prepare films containing larger amounts of silica.

those films were so brittle after they dned that they could not be hnndled. In the films

shown in Figure 2.9, we obtained @do) values ranging from 0.060 to 0.079. The initial

values of the energy vansfer quantum efficiency are slightly lower than in the case of

CaCO,. In addition, one cm see bat, unlike CaC03, @do) values decrease slightly with

increasing amounts of Si02. As we mentioned above, @ d o ) is prirnarily a measure of the

interfacial area between D- and A-labeled cells in the film. The data in Figure 2.9 indicate

that the SiOl pigment with a diarneter of 25 nm appears to have a srnall effect either on

reducing the interfacial area between D- and A-labeled cells in the system, or on

suppressing the lirnited extent of polymer difision that occurs during film formation.

One can see that the error bars for each data point are srnail, which indicates that the

measurements are reproducible. Like CaCOi, SiOz does not pmmote coalescence of the

latex particles dunng film formation.

0.00 I I t t I 1 I I I 1

O 20 40 60 80

CaCO, content (wt %)

Figure 2.8: Effect of CaC03 content on the initial efficiency of energy transfer. We plot

the initial energy efficiency of energy transfer @do) vs CaCOj content in latex films (wt

%).

SiO, content (wt %)

Figure 2.9: EfTect of SiOl content on the initial efficiency of energy transfer. We plot the

initial efficiency of energy transfer M O ) vs SiO? content in latex films (wt %).

2-3-4. Maximum Effciency of Energy Transfer Q'nim) in Latex F i

In order to evaluate the extent of mixing. f,(t) defined by eq 2-8, we need to know

the value of #da), which corresponds to full mixing of the polymer. If the film is fully

mixed, one should have a random distribution of donors and acceptors, and the decay

profile should be descnbed by eq 2 4 . Under these circumstances, the rate and efficiency

of eneqy transfer will be determined only by &, k, and the concentration of acceptors in

tliz filrii. In order ro obtain ihis value experimcntally, one necds a mode! comsponding to

a "fully mixed sample. In the energy transfer experiments we carry out, On reaches its

maximum value when the polymers in the film diffuse a distance on the order of the

original particle radius. There are several ways to obtain a sample which will serve as a

model for a&=). First, one takes a film and anneals it at a higher temperature. Since the

polymer rate is strongly accelented by increasing temperature, (PET will norrnally

increase npidly to its maximum value. Altematively, one can dissolve a dry film sample

in an organic solvent. In solution, one expects full mixing of the polymer molecules. A

film cast from this solution is then a good model for the determination of O&=). In

pigment-Free films, both approaches give similar values of ares(-) values from which the

corresponding O&-) are caiculated. In this section. we examine the effect of pigment on

the magnitude of On(=).

For pigment-free latex films, we obtained areu(=) values from a solvent-cast film.

This film was prepared from a dry P(MMA-CO-EHA) film prepared from a 1: I mixture of

D- and A-labeled particles, which was subsequently dissolved in tetrahydrofuran o. The solution was cast onto a quartz plate and allowed to dry at room temperature for 12 h.

For these films, we obtained an ares(=) value of 18.2 ns, and a @E~(Q)) value of 0.60, and

these values did not change when the film was annealed at 80 OC for 1 h. In contrast,

when a sarnple of the latex film itself was annealed at 50 OC for 112 h, we obtained an

ares(=) value of 22.6 ns, and a @Ei(w) value of 0.50. We infer from these results, that for

prolonged anneding at 50 OC, lu11 mixing is not achieved after 112 h. As one cm see in

Table 2.3, a latex film heated at 60 O C for this length of time gave values closer to that

found for the solvent-cast film, ares(=) = 19.7 ns, and a @da) value of 0.57.

-

a. Annealing time.

b. Annealing tempenture.

c. Integnted area under the donor fluorescence decay profile.

Table 2.3: Final area and maximum efficiency of energy transfer values.

Effect of CaCOI,

When films

values of ares(=)

were prepared in the presence of pigment, we obtained different

and M m ) . This result was unexpected. As a consequence. we

i

2

3 iI

4

5

6

7

8

9

10

1 I

examined this effect in some detail, particularly for the films containing large amounts of

CaC03. As seen in Table 2.3, a film containing 90 wt % CaC03, mnealed for 112 h at 50

OC had am(=) = 30.5 ns, and @R(-) = 0.33. M e n we repeated this expenment at 60

OC, @&O) increased to 0.36, still well below the value expected for complete

interdifision. Even anneding a film containing 90 wt % CaC03 for 30 min at 120 OC

increased @dm) to only 0.48. In order to force mixing, we added THF to a latex film

containing 90 wt % CaCO,. The polymer was kept in the presence of the solvent for

t m ~ i "

min

60

6750

6750

6750

6750

30

30

30

-

6750

6750

CaC03

wt %

O

O

90

O

90

90

90

90

90

80

80

Tanneai

OC

80

50

50

60

60

100

120

140

50

60

@ET

0.60

0.50

area

ns

18.2

22.6

30.5

19.7

29.3

26.1

23.8

24.5

20.6

29.4

28 .O

solvent cast

0.33

0.57

0.36

0.42

0.48

0.46

0.55

0.36

0.39

6 +THF

several minutes, and then the solvent was allowed to evaporate. This film, after drying 12

h at room temperature. had a significantly higher extent of energy transfer, Gd-) = 0.55,

Effect of SiO?.

We also determined ares(-) values for films containing up to 50 wt % silica. For

samples annealed 112 h at 50 OC. these values ranged from 22.2 ns (no silica) to 28.7 ns

(50 wt % of SiO3, corresponding @dm) values from 0.51 to 0.37. As in the case of

calcium carbonate, the presence of pigment appears to lower the maximum efficiency of

energy transfer, @da), accessible by anneaiing the latex films. This effect is likely to be

a consequence of the pigment aff'ecting the extent of diffusion that occurs in the system.

To examine the influence of pigment on the rate of polymer diffusion, we need to

introduce values of fm(t). Since f,(t) is a measure of the extent of mixing in the system. we

need to choose values for ares(=) and that describe only the latex polymer in the

system, and are not influenced by any effect of the pigment on limiting the total extent of

interdiffusion. For this reason. we use values obtained for the pigment-free polymer

obtained by solvent casting: ares(=) = 18.2 ns. and #da) = 0.60.

2-3-5. Effect of Pigment on the Rate of Polymer Interdiffusion

In Figure 2.10, we plot values of f,(t) as a function of annealing time at 50 OC for

films containing different amounts of CaC03. In carrying out these experiments, we

assume thnt the annealing temperature of 50 O C does not affect the characteristics of

CaC03 pigment. The f,(r) values were calculated with eq 2-8 from the areas under the

fluorescent decay curves, using a common area(m) value of 18.2 ns. The most important

result in Figure 2.10 is that polyrner interdiffusion is not significantly retarded by the

presence of the CaC03 pigment. The diffusion rate is unchanged for amounts of pigment

up to 30 wt 96. Even at 80 wt %, the effect is srnail. Various experiments were repeated a

number of times, and experirnents were carried out at different tempentures. The trends

s h o w in Figure 2.10 are consistent. To cary out meaningful expenments, a set of

samples must be annealed simultaneously. It is difficult to reproduce the exact profile of

f,(r) vs t for experiments carried out at different rimes. The problem is one of reproducing

the oven temperature. The polymer diffusion rate is so sensitive to tempenture (with

apparent activation energies ranging frorn 160 ?' to 400 kllmol ") that even small

temperature differences between expenments have a noticeable effect on the rates of

polymer diffusion.

In Figure 2.1 1 we plot values of f,(t) as a function of annealing time at 50 O C for

films containing different arnounts of Sioz. Here we see that the presence of SiO2 in the

film lias a pronounseci e f k t on reducing the rate of polyner interdiffusion. Therc is

greater retardation when one increases the amount of SiO2 in the latex films.

O 60 120 180 240 300 1000

T h e (min)

Figure 2.10: Comparison of P(MMA-CO-EHA) extent of rnixing, f;.(t) vs annealing tirne

for films containing different amounts of calcium carbonate: O wt % (O), 30 wt % (a), 80

wt % (O), and 90 wt % (i). The weight-average molecular weight ( M W ) of the P(MMA-

CO-EHA) latex was ca. 240,000. Films were annealed sirnuItaneously at 50 2 1 OC.

O 60 120 180 240 300 1000 T h e (min)

Figure 2.11: Comparison of P(MMA-CO-EHA) extent of mixing. f,(t) vs anneding time

for films containing different arnounts of silica: O wt % (O), 20 wt % (a), 40 wt % (a),

and 50 wt 8 (a). The weight-average molecular weight ( M W ) of the P(MMA-CO-EHA)

latex was ca. 240,000. Films were ûnnealed simultaneously at 50 t 1 OC.

2-4. Discussion

The polymer we examine as the binder for our model paper coatings is prepared

from equal parts by weight of MMA and EHA using serni-continuous emulsion

polymerization to maintain a uniform composition throughout the polymerization. The

mole ratio of MMA / EHA components is 65 1 35. and the polymer has a Tg of 7 OC. The

seed particle used in the synthesis was prepared by batch emulsion polymerization in the

absence of iuiy shah tramfer agciit, whercas a small m o u n t of 1-dcdecxxthiol :vas

added in the second stage to limit the molecular weight. A broad molecular weight

distribution was found for these latex polyrners by gel permeation chromatography

(GPC). We know little about branching in the polymer we produce. Giving the reactivity

of the polymerizing acrylate radical and the presence of the teniary hydrogen on the EHA

side chain. it is likely that the polymer we examine contains substantial branching.

Paralle! analysis of the latex polyrner by GPC, using tandem fluorescence and refractive

index detectors. establishes that the fluorescent dyes are covalently bound to the polymer

and uniformly incorponted.

In pigment-free latex films, polymer diffusion is slow at room temperature. but

faster at 50 OC. Under these conditions. 43 O C above Tg, the initial stage of polymer

diffusion (to f,(t) = 0.4) is rapid, followed by a slower rate of increase. As mentioned

above, the polymer molecular weight is broad, and there is likely to be a distribution of

bnnching as well. The smaller and more compact structures are likely to have faster

invinsic diffusion rates. In latex films containing a mixture of species with a distribution

of diffusivities, the fastest diffusing species make the largest contribution to fm(t) at early

times. One can in principle obtain a deeper insight into the nature of the diffusion process

by assuming a model and calculating apparent diffusion coefficients. We will defer this

type of analysis for the next chapter. The tirne necessary for Full mixing is much longer

than those shown in Figure 2.10 and Figure 2.1 1. As seen in Table 2.3, even 6750 min

(1 12 h) anneaiing at Tg + 43 OC does not lead to full mixing. Here Q>rr = 0.50 whereas the

fully mixed sample, prepared by solvent casting, has @= = 0.60. It rnay be that some of

the polymers in the sample, as a consequence of long-chah bnnching, have a star-like

structure. Such polyrnen have a very slow rate of diffusion in polymer melts. "

In the presence of pigment, the polymer diffusion rate is retarded. The effect is

large for silica, which has a large surface to volume ratio. and rnuch smailer for calcium

carbonate, where the surface to volume ratio is ca. 10 tirnes smaller. In addition, the

surfaces of the pigment are very different. The Si-OH groups at the surface of the silica

are likely to be strong hydrogen bond donors toward the ester groups in the latex polymer.

The results for silica p d l e l those found by Feng et al. for PMMA particles added as

hard organic fillen to PBMA latex films.

There are two types of explanallons for the effect of filler on a reduced polyrner

diffusion rate. In the first model, the hard pigment surface serves to make the adjacent

polymer matrix more rigid. It is well known that polymer chains adjacent to a @id

surface have decreased mobility. " Tsagaropoulos and Eisenberg '5 proposed a three-

Iayer model in terms of polymer inobility. They studied changes in the glass transition

temperature associated with adding small silica particles as fillen to various bulk

polymers. As increasing amount of filler was added. a new high glass transition

tempenture (Tg) was found in addition to the Tg for the bulk polymer. As more filler was

added. they found a decrease in the magnitude of the signal from which Tg was

determined. In their model. the surface Iayer of polymer adjacent to the pigment is

suongly immobilized, but in addition, nearby polymer also has its mobility restricted. It is

this nearby polymer that contributes to the elevated Tg. From this perspective, one reason

for the decreased rate of polymer diffusion found here is that the polymer molecules near

to a pigment surface have decreased mobility.

In the case of small silica particles as the pigment, another effect can become

important. As shown in Figure 2.12, the silica particles are rnuch smaller ihan the latex

particles. When the film dnes. one c m imagine that the silica particles deconte the

interface between adjacent cells. In this way they c m serve as obstacles to the diffusion of

latex polymer. It is important to recognize that obstacles slow down diffusion without

affecting the intrinsic mobility of the difising species. The obstacles at the interface

opente in a different way to force molecules to difise along more tortuous pathways to

reach the same extent of interdiffusion. As a consequence, the time necessary for mixing

is longer.

As shown in Figure 2.13, the CaC03 particles are much larger than the cells

formed from the individual latex particles. It would be very unlikely that these large

particles could act as an obstacle to intercellular polyrner diffusion. The retardation that is

observed at very high pigment levels is likely to the ngidifying effect of the nearby

pigment surface.

Figure 2.13 also illustrates an important feature of the binder-pigment rnorphology

whan srnall amounrs of binder are used with C C O j as the pigment.

Joanicot et al. examined the binder-pigment morphology of such systems. They

spread a carboxyl functionalized styrene-butylacrylate latex dispersion on a calcium

carbonate crystal plate and observed the structure by atomic force microscopy (AFM).

Here, they took the same surface coverage as for mineral pigments in typical matte paint.

which has a pigment volume concentration (PVC) of 70 8. They have shown that the

binder dispersion spreads completely dong the crystal surface, but those latex polymer

particles in the dispersion do not form a continuous film. The latex polymer forms

patches on the surface unless they are annealed at very high temperatures. Consequently,

Joanicot et al. concluded that. in an actual coating, the CaC03 pigments can be fixed by

isolated latex particles. i.e. stuck together at many discrete points by droplets of latex

polymer. Each droplet originated from the localized coalescence of many latex particles.

In Our case, the latex particles are not acid-functionalized, but we can provide a

few further insights into this model. We know from Our measurements on newly formed

films that @ d o ) does not change as the fraction of pigment in the system increases. From

this result we infer that the interfacial area between donor- and acceptor-labeled cells

does not change significantly. even when there is 90 wt % pigment. The pigment, of

course, is denser than the latex particles, so 90 wt b pigment corresponds to about 25 vol

% binder. When these samples are mnealed, polymer difision leads to an increase in

C P ~ , but the last stages of mixing are strongly retarded. As one sees in Table 2.3. 112 h of

annealing for 90 wt % CaC03, even at 60 OC, leads to Q'n( 1 12 h) = 0.36, f,(t) = 0.53. We

interpret this result to indicate that upon annealing, the blob of latex film does not behave

exactly as shown in Figure 2.13, but rather spreads Locally dong the pigment surface. In

this way, we imagine that a significant fraction of the polymer molecuies adsorbs to the

polymer surface in such a way that they do not participate in the polymer interdifhision

process.

It is aiways possible that the chromophores in the polyrner interact differently with

the pigment surface than the polymer backbone. This effect would be very difficult to

measure, but selective adsorption would be analogous to processes involved in adsorption

chromatography. If this were the case, one could rationalize the result seen at long

annealing times in ~ h e film containhg 90 :vt 8 &CO;. If polymer diffusion were

accompanied by flow dong the pigment surface, chromophore adsorption might lower the

fraction of donor and acceptor groups able to participate in energy transfer. Another

aspect of our measurements that one should keep in mind is that our films are opaque.

Our experiments measure fluorescence in the reflectance mode from the samples. The

excitation light may penetrate into the sample to a depth not much larger than the

wavelength of light. Thus we may be observing selectively processes that occur in the

fint micrometer of ihese 100 pm thick films.

: zone of interdiffusion

: SIIica

Figure 2.12: Schematic representation of the annealing process for a latex film

containing silica particles (25 nm). The silica particles can act as obstacles and retard

polymer diffusion andor reduce the mobility of polymer molecules near the pigment

surface.

I Annealing ' D A \

: zone of interdiffusion : Calcium carbonate

Figure 2.13: Schematic representation of the annealing process for a latex film containing

precipitated calcium carbonate. The latex particles spread locally dong the pigment

surface.

2-5. Conclusions

We employed the fluorescence energy transfer technique to measure the rate of

polymer interdiffusion in the presence of large amounts of inorganic pigment. We

obtained sirnilx values of the initial efficiency of energy transfer in the presence of

different amounts of CaC03 and silica pigment. This result indicates that neither pigment

pemirbs the interfacial area between donor- and accepter-labeled particles in the newly

fomcd film, nar do they pmmote coûlescence of the tntex pxticles. CaCO; has little

effect on the polymer diffusion rate in the binder phase, but at very high solids content

(80, 90 wt %). the diffusion rate is slowed. In contrast, 25 nm diameter silica particles

have a much more pronounced effect on slowing the rate of polymer diffusion. Because

of sample brittleness, we could study films containing up to 50 wt % (3 1.2 vol %) silica.

In these films. the binder is present as the continuous phase. At 90 wt % CaC03 content.

the volume fraction of binder is about 0.25. Here the binder serves prirnarily to glue the

pigment particles together. Because of the air voids in the matrix, the films are essentially

opaque.

2-6. References

Macnair, A. K., Task Group Chairman, Synthetic Couring Adhesives, TAPPI PRESS.

Atlanta, 1998.

LePoutre. P., Prog. in Org. Coat.. 1989, 17,89.

(a) Wang, Y.; Kats, A.; Juhué, D.; Winnik, M. A., Langmuir, 1992, 8. 1435. (b)

Winnik, M. A., The Formation and Properties of Latex Films in Emdsion

Polymerization and Emiiision Polymers, Lovell, P. A., El-Aasser, M. S., Eds., Wiley,

New York, 1997, p 467. (c) Keddie, J . L., Mat. Sci Eng., 1997,21, 10 1.

(a) Granier, V.; Sartre, A., Langmuir, 1995, i l , 2 179. (b) Butt, H. J.; Gerhm, B .,

Langmuir, 1995,11,4735. (c) Unenl, W. N., Langmuir, 1998, 14,2201. (d) Sheehan,

I. G.; Whalen-Shaw, M.. Tappi J., 1990, 73, L 7 1.

5. (a) Sheehan, J. G.; Takamura, K.; Davis, H. T.; Scnven, L. E., Tappi J., 1993, 76, 93.

(b) Ming, Y .; Davis, H. T.; Scnven, L. E.; Takamura. K.; Vodnick, J. L., TAPPI 1995

Coating Conference Proceedings. TAPPI PRESS, p 39 1.

6. Joanicot. M.: Granier. V.; Wong, K., Prog. in Org. Coat., 1997,32. 109.

7. (a) Forster. Th.. Ann. Phys. (Leipzig), 1948, 2, 55. (b) Lakowicz, J. R.. Principles of

F!i<orescence Specrroscopy, Plenum Press, New York. 1983.

8. (a) Farinha, J. P. S.; Matinho, J. M. G.; Kawaguchi, S.; Yekta, A.; Winnik, M. A.,

Macromolecules, 1995. 38, 6084. (b) Kim, H.; Winnik. M. A., Macramolecdes,

1995, 28,2033. (c) Winnik. M. A.; Pinenq, P.: Krüger, C.; Zhang, J.: Yaneff. P. V.. J.

Coat. Technol., 1999, 7I,47.

9. Feng, J.; Odrobina, E.; Winnik, M. A., Macrornolecrdes, 1998,31,5290.

10. Rharbi, Y.; Cabane, B.; Vacher. A.: Joanicot, M.; Boue, F.. Ertrophys. Lett.. 1999.46.

472.

I l . Zhao, C. -L.; Wang, Y.; Hruska. 2.; Winnik. M. A., Macromoiecrries. 1990. 23.4082.

12. Liu, R.; Winnik, M. A.: Stefano, F. D.; Vanketessan. J.. in preparation, 1999.

13. Feng, J.; Winnik. M. A. Macromolecules, 1997.30.4324.

14. O'Connor, D. V.: Phillips, D., Tirne-correlated Single Photon Cotmting, Academic.

New York, 1984.

15. Martinho, I. M. G.; Egan, L. S.; Winnik, M. A.. Anal. Chern., 1987,59,86 1.

16. (a) Wang, Y.; Zhao. C.; Winnik, M. A., J. Chem. Phys., 1991. 95, 2 143. (b) Wang,

Y.; Winnik, M. A.; Haley, F., J. Coat. Technol., 1992.64, 5 1. (c) Kim, H.; Wang, Y.;

Winnik, M. A., Polymer, 1994, 35, 1779. (d) Kim, H.; Winnik, M. A.,

Macromolecules, 1994,27. 1007.

17. Sosnowski, S.; Feng, J.; Winnik, M. A., J. Polyni. Sci.: Polym. Chem., 1994. 32,

1497.

18. (a) Joanicot, M.; Wong, K.; Maquet, 1.; Chevalier, Y.; Pichot, C.; Graillat, C.;

Lindner, P.; Rios, L.; Cabane, B., Prog. Colloid Polym. Sci., 1990, 81, 175. (b)

Chevalier, Y.; Pichot, C.; Graillût, C.; Joanicot. M.; Wong, K.; Lindner, P.; Cabane,

B., Colloid Polym. Sci., 1992,270, 806.

19. Feng, J.; Odrobina, E.; Winnik, M. A. Macromolecules 1998,31,5290.

20. Pham, H. H.: Farinha, J. P. S.; Winnik, M. A., Macromolecrrles, mbmitred for

p~tblicarion.

21. Wang, Y.; Winnik, M. A., J. Phys. Chem., 1993, 97,2507.

22. Wang, Y.; Winnik, M. A., Macronzolecicles, 1993.26.3 147.

23 (a) Klein, J.; Fletcher, D., Ncznrre. 1983. 304. 526. (b) Bartels. C. R.: Crist. Jr.. B.;

Fetters, L. J.; Graessley, W. W., Macromolecriles, 1986, 19, 785. (c) Shull, K. R.;

Kramer, E. J.; Fetters, L. J., Nature, 1990, 345, 790. (d) Shull, K. R.; Dai, K. H.;

Kramer, E. J.; Fetters, L. J.; Antonietti, M.; Sillescu, H., Macromolecriles, 1991, 24,

505. (e) Gel!, C. B.: Graessley. W. W.; Efstratiadis. V.; Pitsikalis, M.. J. Poiyrn. Sci.:

Polym. Pltys.. 1997,35, 1943.

24 (a) O'Brien, J.; Cashell, E.; Wardell. G.; McBrierty, V. I., Macromolecides. 1976. 9,

653. (b) Douglass, D.; McBrierty, V. J., Polym. Eng. Sci., 1979. 19, 1054. (c) Ito. M.:

Nakamura, T.; Tanaka, K., J. Appl. Polym. Sci.. 1985, JO. 3493. (d) Dutta, N.;

Choudhury, N.; Haidar, B.; Vidal, A.; Donnet, J.; Delmotte, L.; Chezeau, J.. Polper.

1994,35,4293. (e) Pliskin, 1.; Tokita, N., J. Appl. Polym. Sci., 1972. 16,373.

25. (a) Tsagaropoulos. G.; Eisenberg, A.. Macromolecnles, 1995, 28, 396. (b)

Tsagaropoulos, G.; Eisenberg, A., Macromolecrtles, 1995.28.6067.

3. EFFECT OF SILICA PARTICLE SIZE ON POLYMER

INTERDIFFUSION

3-1. Introduction

Many polymer coatings used in a wide variety of applications such as paints.

paper coatings, contain a large amount of inorganidorganic pigments. One of the major

reasons for inuoducing these pigments is that soft polymeric matrices c m be reinforced

by hard fillers. For example. naturai or synthetic rubber is commonly reinforced by

carbon black particles. and silicone elastomers (PDMS) can be reinforced by silica. 1

Since silica is chernically inactive and optically transparent, it can be used in a wide range

of applications. nnging from paints and magnetic fluids to high-quality paper coatings. Z

In the paints and adhesives industries, vinyl acetate, acrylic ester, synthetic rubber and

other polymer latex have been used in coatings and adhesives for fribric and paper. J

Colloidd silica is used with these polymer emulsions in order to improve adhesion.

durability, and abrasion resistance. The silica also serves to prevent stickiness and

improves the washing resistance of the coatings. When colloidal silicû is used as a

constituent for paints, a superior stain resistant film can be formed due to its anti-static

characteristics. In paper coatings. this property provides better quality on sheet feeding,

and thus silica is used as an anti-static agent.

Colloidal silica is also widely used for paper coatings. For instance, high-quality

coated paper for graphics printing has a smooth layer on the top of the coating in order to

enhance gloss. ' Other inorganic pigments such as clay, calcium carbonate, talc, alurnina,

titanium dioxide. are also used with latex polymers to avoid stickiness to the stainless

steel drum. For this type of paper, one would like to make the surface of the paper more

rigid, in other words, prevent it from sticking to the hot stainless steel cirum.

Another application of (colloidal) silica for high-quality paper coatings is for

photognphic quality ink jet printing paper. ' In this case, the coating contains both silica

and a latex polymer with a relatively high glas transition temperature (Tg), as well as

water-soluble polymers such as poly(viny1 alcohol), casein, and starch. The ink receiving

Iayer is required to have a large pore volume to keep the ink jet printing ink at the surface

of the coated paper, and dry the ink dispersion as quickly as possible. Silica also serves to

prevent the ink dispersion from spreading honzontally and penetrating vertically. In this

application, the latex particles act as silica binder, but are also required to keep their

shape, which rneans that less coalescence is preferred in this system. These different

applications of silica-plus-latex in coatings require a broad understanding of the nature of

their interaction. Here we examine the influence of silica nanospheres on the rate and

extent of polymer diffusion in an ücrylate latex film.

In this chapter we examine the influence of silica as a filler on polymer

interdiffusion in poly(buty1 rnethacrylate) latex films. We carry out fluorescence

resonance energy transfer (FRET) rneasurements on latex films that allow us to follow the

extent of polymer diffusion as a function of time after the latexfpigment dispersion dries.

In this study we compare four different types of colloidal silica, and discuss how silica

fillers affect the rate of polymer interdiffusion in latex films. We focus on the effect of

silica particle size on polymer interdiffusion rate in poly(buty1 methacrylate) latex films.

3-2. Experimental

3-2-1. Materials for this study

3-2-la. Materials for Emulsion Polymerization

Butyl rnethacrylate (BMA. Aldrich. 99%) was distilled under vacuum under a N2

atmosphere, and stored in refrigerator prior to use. Potassium persulfate (K2S20s, KPS,

Aldrich. 99%), sodium bicarbonate (NaHCO,. Caiedon, 998) , sodium dodecyl sulfate

(Ci2H30S03*Na', SDS, Aldrich. 98%) and 1-dodecanethiol (Ci2HaSH, DM, Aldrich,

988) were used as supplied. Distilled water was further purified through a Millipore

~ i l l i - ~ ~ ~ system. The structure of butyl methacrylate used for this study are show in

Figure 3.1.

Figure 3.1: Chemicd structure of butyl methacrylate used for this study.

9-Phenanthrylmethyl methacrylate (Phe-MMA) were synthesized previously. 6

The synthesis of (9-anthry1)methacrylate (An-MA) was also described in ref 6b. The

structures of these dye-labeled monomers are shown in Figure 3.2.

Figure 3.2: Chernical structures of dye-labeled monomers used for this study.

3-2-1 b. Synthesis of Dye-Labeled Latex Particles

Butyl methacrylate (BMA) was distilled under vacuum pnor to use. Potassium

persulfate (KPS), sodium bicarbonate (NaHC03), sodium dodecyl sulfate (SDS) and 1-

dodecanethiol were used as supplied. Distilled water was further punfied through a

Millipore ~ i l l i - ~ ~ system. Poly(buty1 methacrylate) (PBMA) latex samples, labeled

with 1 mol 8 of a fluorescent dye [either a donor, phenanthrene (Phe) or an accepter,

anthncene (An)], were prepared by semi-continuous emulsion copolymenzation at 80 OC,

using KPS as the initiator, SDS as the surfactant, and 1-dodecanethiol as the chain 7 transfer agent. The reaction conditions were similar to those described previously. As a

donor-labeled monomer, we used (9-phenanthry1)rnethyl methacrylate (Phe-MMA) to

intmduce the donor dye. As an acceptor-labeled monomer, we used 9-mthryl

methacrylate (An-MA). The recipe for 100 nm PBMA latex is shown in Table 3.1. We

used the same seed latex particles obtained in the first stage for preparing both the Phe-

and An-labeled latex particles. Their chmcteristics are shown in Table 3.2. These latex

particles, referred to as PBMA, have a diameter of 100 nm. and a Tg of ca. 20 OC. The

solid content of the dispersion was measured gravimetrically and was found to be ca. 30

wt %.

3-2-lc, Colloidal Silica

Colloidal silica dispersions (30 wt % solids). Klebosol 30W5 (K-25) and

Klebosol 30R50 (K-50), were supplied by Clariant Corporation. The filler particles are

amorphous spherical silica beads with diameters of 25 nm and 50 nm, respectively.

Colloidal silica dispersions (20 wt % solids). Snowtex-O (S-12) and Snowtex-OL

(S-49, were supplied by Nissan Chernical Industries. These silica particles are also

amorphous sphencal beads with diameters of around 12 nrn and 45 nm, respectively.

Both samples contain acidic small particle silicasol. with a free anion content less than

300 ppm. ' The conductivity values for these dispersions are ca. 300 pS/crn, much lower

than Klebosol dispersions. Further charactenstics of these particles are shown in Table

3.2.

Table 3.1: Recipe for preparing LOO nm PBMA latex.

An-MA (g)

Phe-MMA (g)

Water (g)

l -Dodecanethiol (g)

First stage

a. 1 mole % relative to the total monomer fed in the second stage.

Second stage

(under monomer starved condition) (batch process)

3.5

57 .O

0.06

0.2 1

0.08

80

1.5

b. Reaction tempenture.

c. Reaction time.

Phe-Iabeled

35.0

0.68 "

27 .O

0.06

0.60

0.30

80

18

An-labeled

35.0

0.65 "

27 .O

0.06 ,

0.60

0.30

80

18

Table 3.2: Characteristics of the latex and colloidal silica The particle size, Tg, and

molecular weight for Phe-PBMA are very similar to those for An-PBMA.

1 1 1 Colloidal siIica 1 PBMA Klebosol 1 Klebosol

Phe- + An- / 1 30R25 30R50 labeled

Diameter (nm) i i û 0 ' i 25" i 50"

DLS 1 90.8 1 25.6 1 73.1

M w / M n 2.5 - - pH of dispersion 7 . W .4 9 .O13 .O 9.013.0

(S- 12) (S-45)

a. Determined by quasielastic light scattering (Brookhaven BI-90).

b. Obtained from the brochures.

c. Determined by dynamic light scattering (DLS).

d. Determined by transmission electron microscopy (TEM).

e. The data beforelafter ion-exchange are shown for PBMA latex and Klebosol.

Conductivity of

dispersion (pS/cm)'

3-2-2. Characterization of Latex and Colloidal Silica

3-2-Sa. Particle Size Measurements

Dynamic light scattering (DLS) experiments were carried out on a variable angle

laser light scattenng photometer from Broo khaven Instruments Corporation. A 5 m W

vertically polarized He-Ne laser from Spectra physics was the Iight source. The silica

dispersion in water was filtered through disposabie 0.45 pm Gelman filters into g las

scattenng cells. The cells were placed into the BI-ZOOSM goniometer and sat in a vat of

8001 160 2,0001300 1,2001 150

themostated toluene which matched the index of refraction of the glass cells. The angular

range of the goniorneter was 7'462'. The scattered light was detected by a

photomultiplier interfaced to the BI-2030AT digital correlator with 136 channels and

measured the correlation Function in real time. Dynamic light scattering data were 9 analyzed following the method of cumulants. The logarithm of the nonnalized intensity

autocorrelation function, g"' r, cm be fitted into a power series in terms of the delay time

(1 Ing r=-Tl?+

where rl is the first cumulant, r2 the second cumulant, and so forth. Once ri is

deterrnined, the concentration and angular dependence can be expressed as

where D, is the z-average diffusion coefficient. C is a parameter that is chancteristic of

the molecular architecture. kD is the effective interaction parameter, and q is the scattering

vector, with ifs magnitude given by

where 0 is the scattering angle, no is the refractive index of solvent, and is the

wavelength of the laser beam in a vacuum. From the difision coefficient, the z-average

hydrodynamic radius, Rh, CM be calculated frorn the Stokes-Einstein relation

where q is the solvent viscosity. For the determination of the z-average diffusion

coefficient for each sarnple, 90' scattering angle and one concentration were used.

Dynarnic light scanering data were also anaiyzed using the CONTIN 'O method to

determine the distribution of hydrodynamic size.

Transmission electron microscopy (TEM) measurements were carried out on a

Hitachi mode1 600 electron microscope. The sarnples were prepared as follows. Thin

carbon films (ca. 5 A) were grown on mica as a support. Then 25 pL of the silica

dispersion in water was added onto the carbon-coated mica surface. After the water was

evaporaied, cach carbon film :vas flonted off the micz support in water and deposited onto

a 300 mesh Gilder copper grid. The sample was air-dried before introduction into the

electron microscope. Staining of the sample was unnecessary. The diameter of the silica

was measured directly from the TEM micrograph. using "Photoshop" software. For each

simple, around 100 to 200 particles were measured and the average was taken as the final

diane ter.

3-2-2b. Ge1 Permeation Chromatography (GPC)

Molecular weight and molecular weight distributions were measured by gel

permeation chrornatography (GPC). in the s m e manner with that descnbed in Section 2-

3-4b.

3-2-2c. Soüd Content of latex dispersions

The solids content of latex dispersions was measured in the sarne manner with that

described in Section 2-2-4c.

3-2-3. Film Preparation

Before we prepared films from the PBMA latex and colloidal silica dispersions,

the latex and Klebosol dispersions were cleaned by treating the individual dispersions

with a prepurified ion-exchange resin (AG-508-X8 rnixed-bed resin, Bio-Rad) to rernove

the ionic surfactant and other ionic species. For each sample, the pH decreased from 8-9

to 2-3 after the ion-exchange. The Snowtex dispersions were used as supplied. Latex

films were prepared from dispersion mixtures with a 1:I number ratio of Phe- and An-

labeled PBMA latex particies and different amounts of silica. nie final dispersions, with

total solids content ranging from Ca. 8 wt % to 19 wt %, have pH = 3-4. Film formation

was carried out by the following procedure. For each film, using a Pasteur pipette. we

spread eight drops from each dispersion mixture onto a quartz plate. The film was

allowed to dry over 40 minutes in an oven at 3 1 f 1 OC, followed by storage in the cold

room at 4 OC to minimize the amount of polymer interdiffision in the film. A typical film

thickness was 150 Fm. Films formed from latex alone were crack-free and transparent.

-411 of the cher films were transparent and crack-free up to 31) wt % silica content. but

small cracks were observed in the films containing JO wt % filler. It was hard to work

with films containing more than 40 wt % filler.

3-2-4. Energy Transfer Measurement

Al1 films were annealed at 60 f I OC for polymer diffusion measurements. For

each series of samples to be cornpared. the films were annealed simultaneously.

Fluorescence decay profiles were measured by the single photon-timing

technique. " Samples were excited at 300 nm, and the emission was detected at 350 nm.

A bandpass filter (350 i 5 nm) was used to minimize the scattered light and interference

due to fluorescence from excited accepton. For each measurernent, it took about 10 to 15

minutes to collect 5000 counts in the maximum channel. frior to each measurement, a

film sample was placed in a quartz tube, and the tube was degassed with flowing nitrogen

gas.

3-2-5. Data and Data Analysis

Data analysis was carried out according to Section 2-3-7.

3-3. Results

3-3-1. Dye Distribution of PBMA

As we have shown in Section 2-34, it is important to know if the fluorescent dyes

are randomly distributed in the polymer backbones. In order to examine this, we canied

out Gel Permeation Chromatography (GPC) andysis.

In the case of labeled PBMA latex polymers, their molecular weight distributions

are much narrower than those of labeled P(h4MA-CO-EHA) polymers, as seen in Figure

2.4. The polydispersity index (PDI) for PBMA latex polyrners is ca. 2.5. One cm clearly

see the peak derived from seed polymer because the weight-avenged molecular weight of

seed polymer is ca. 800.000, whereas thût of second stage polymer is ca. 50,000. GPC

chromatograms for Phe-PBMA are illustrated in Figure 3.3a. and those for An-PBMA are

illustnted in Figure 3.3b.

Retention Time (min) Retention Time (min)

Figure 3.3: GPC chromatograms for (a) Phe-PBMA and (b) An-PBMA latex polyrnen.

The samples were prepared by drying a latex dispersion, followed by dissolution of

polymer into THF solvent. Curve (1): the fluorescence signal for the two-stage polyrner;

curve (2): the refnctive index signal for the two-stage polymer: and curve (3): the

refnctive index signal for the seed polymer.

3-3-2. Determination of Silica Particle Sizes by DLS and TEM

In this chapter, we examine the effect of silica panicle size on the rate of polymer

interdifision. Therefore, it is important to know the exact silica particle size. We

employed both dynamic light scattering (DLS) and transmission electron microscopy

(TEM) measurements to determine the particle size, and compared the data with those

descri bed in the product brochures. 'J' TEM images for each silica panicle are shown in

Figure 3.4. The particle size disuibutions for each siiica panicie decermined by Lhe TEM

experiments are shown in Figure 3.5. The data are summarized in Table 3.2.

When we employed DLS. the particle size data for K-25 and S-12 are very similar

to those described in the product brochures. However. the DLS diameters for K-50 and S-

45 are much larger than those described in the product brochures. DLS data correspond to

the volume average particle size (hydrodynamic radius), so the larger particle size data

are more emphasized in the calculation.

We also employed TEM measurements, and determined the silica particle size.

Approximately 100 to 200 particles were measured, and we determined the number

average particle size for each silica filler. The particle size distributions for K-25. K-50.

and S-12 are quite narrow, but that for S-45 is relatively broad, as seen in Figure 3.5d.

Consequently, the average particle sizes obtained by TEM for K-25, K-50, and S-12 are

very close to those obtained by DLS. On the other hand, the average particle size obtained

by TEM for S-45 is much smaller than that obtained by DLS. due to the relatively broad

particle size distribution.

In order to examine the silica particle size distributions in DLS measurements, we

introduce polydispersity. The first two moments of the distribution G(T) is as follows:

where q is the scattering vector and D is the average difision coefficient. Eq 3-6 shows

that p? is proportional to the variance of the "intensity" weighted diffusion coefficient

distribution. Thus it carries information on the width of the size distribution. The

magnitude and units of pz are not immediately useful for characterizing a size

distribution. In addition, distributions with the same relative width (sarne shape) may

have very different means and variances. For these reasons a relative width (reduced

second moment) is defined as follows:

Polydispenity has no units. It is close to zero for monodisperse samples. If this

value exceeds 0.080, it is considered to be a broad size distribution. The mean particle

sizes and polydispersities obtained from DLS analysis are summarized in Table 3.3. One

c m see that the polydispersity for S-12 is very large (0.283), which means that it has a

broad particle size distribution in DLS measurement. On the other hand. polydispersity of

K-50 is 0.005, indicating that it is almost monodisperse.

In the TEM analysis, we calculared the standard deviations and used those values

as a measure of particle size distributions. For S-45, the standard deviation is 13.7 nm.

indicating that it has a broad particle size distribution. The mean particle sites and the

standard deviations obtained from TEM analysis are summarized in Table 3.3.

Figure 3.4: Transmission electron rnicroscopy (TEM) images for various particle sizes of

silica. (a) K-25, (b) K-50, (c) S-12, (d) S-45.

Diameter (nm)

"9

O 20 JO 60 80 100 120 O 20 JO 60 80 100 120 Diarneter (nm) Diameter (nm)

Figure 3.5: Pûaicle size distributions of silica particles used for this study. (a) K-25. (b)

K-50, ( c ) S-12, (d) S-45. The particle sizes were determined by transmission electron

microscopy (TEM), and ûndyzed with the image analysis software "Photoshop."

Table 3.3: The mean particle sizes and the particle size distributions for silica used for

this study.

Mean diameter Polydispersity

I TEM

Mean diameter

DLS "

Determined by dynamic light scattenng (DLS).

Determined by transmission electron rnicroscopy (TEM).

Each standard deviation (SD) was calculated as follows,

1 SD = - Z ( x i - F)' , where n is the sample size and 1 is the mean. (. - I

Fluorescence Resonance Energy Transfer Technique for Latex Film

Formation

In this study. we examine the influence of colloidal silica (SiO:) as a filler on the

rate of polymer interdiffusion. We compare four different types of Si02 with dianeters

ranging from 12 nm to 50 nm.

Feng et al. l 3 studied the effect of poly(methy1 methacrylate) (PMMA) filler

particles on the rate of polymer interdiffusion in films prepared from poly(buty1

methacrylate) (PBMA). In their experimenrs, they examined films containing a constant

fnction (35 vol %) of PMMA particles of different sizes. They found that the diffusion

rate of the polymer decreased in proportion to the increase in the surface area of the hard

filler particles, i.e., with a decrease in the hard particle size at constant filler volume.

Based on their results, we expected to see a larger effect on the polymer interdifision

rate in the films containing smaller Si&, due to its much larger surface area.

In Chapter 2, we examined the effect of two inorganic pigments on polymer

interdiffusion in a low-Tg latex film. Cdcium carbonate and colloidal silica were used as

mode1 inorganic pigments, and poly(methy1 methaclylate-CO-2-ethylhexyl acrylate)

copolymer latex was used as the latex binder. We found in that system that, due to its

large pigment size, calcium carbonate did not significantly retard the rate of polymer

interdiffusion, even if the pigment content is above the critical pigment volume

concentration (CPVC). But the limiting extent of polymer interdiffusion upon long

anneaiing was significantly reduced: we found that the maximum efficiency of energy

tramfer, QC?i-) for the film containing large amounts of CaCOl (e.g 90 wt %) was much

smaller than that for the film containing no CaC03. In contrast. the presence of silica

significantly retarded the rate of polymer interdiffusion, even when the filler content is

below CPVC. In addition, the value for the film containing large amounts of SiOz

(e.g. 40 wt %) is much srnaller than that for the film containing no Sior.

Typical donor fluorescence decay profiles for latex film samples at different

stages of annealing are shown in Figure 3.6 and Figure 3.7. Figure 3.6 shows data

obtained for a latex film containing no silica filler. Figure 3.73 shows the corresponding

decay traces obtained for a film containing 40 wt % of 25 nrn diameter SiO? (Klebosol

30R25, K-25). Figure 3.7b shows the corresponding decay traces obtained for a film

containing 40 wt % of 50 nm SiO2 (Klebosol 30R50, K-50). Figure 3 . 7 ~ shows the

corresponding decay traces obtained for a film containing 40 wt O/c of 12 nm SiO2

(Snowtex-O, S-12). Figure 3.7d shows the corresponding decay traces obtained for n film

containing 40 wt % of 45 nm Si02 (Snowtex-OL, S-45). When these latex films were

annealed at 60 OC, well above the polymer Tg (20 OC), polymer interdiffusion takes place.

One can see the evolution of polymer interdiffusion in the films by looking at the extent

of curvature of the decay profiles. As one anneals the film for longer times, the curvature

becomes more pronounced, which indicates that polymer interdiffusion is promoted by

heat and annealing time.

If one compares the curves in Figure 3.7a and Figure 3 . 7 ~ with those in Figure

3.6, one sees less curvature in the decay curves at comparable annealing times (at 60 O C )

when silica is present in the sample. This result tells us that polymer diffusion is retarded

over the entire annealing time. In the decay curves themselves, one c m see that there is

greater retardation of the diKusion rate in the samples containing smaller silica particles.

For exampie, in Figure 3.7a and Figure 3 . 7 ~ ~ one can see that the extent of energy transfer

is strongly suppressed due to smailer size of SiO?. One can also see, in Figure 3.7b and

Figure 3.7d, that the rate of polymer interdiffusion is suppressed for the first several

hours, but finally reaches a similar extent of energy transfer after 200 h of annealing at 60

OC, as that shown in Figure 3.6.

PBMA latex film

O 50 100 150 200 250 Time (ns)

Figure 3.6: Donor fluorescence decay profiles in a PBMA latex film after annealed for

( 1 ) O min, (2) 60 min, (3) 330min. (4) 12,000 min.

O 50 100 150 200 250 T i e (ns)

O 50 100 150 200 250 Time (ns)

10000 10000 (d) PBMA + 40 wt % S-15

1000 1 O00 h h u U

œ .I

V1 g 100 œ

U 100 r L

I 3 CI

10 10

I 1 O 50 LOO 150 200 250 O 50 100 150 200 250

Time (ns) Time (ns )

Figure 3.7: Donor fluorescence decay profiles in a latex film. (a) PBMA with 40 wt % of

25 nm SiOr (K-25), (b) PBMA with JO wt 8 of 50 m SiO2 ( M O ) , (c) PBMA with 40

wt % of 12 nrn Si02 (S-12). (d) PBMA with 40 wt % of 45 nm Si02 (S-45) after annealed

for (1) O min, (2) 60 min. (3) 330min. (4) 12,000 min. respectively.

3-3-4. Initial Efficiency of Energy Transfer, M O ) in Newly Formed PBMA Latex

Films

In this section we examine the influence of filler on the extent of energy transfer

in newly formed films. If these films are prepared at low enough temperature, little or no

polymer interdiffusion wil1 take place. Energy transfer will occur only across the interface

between cells formed by the D- and A-labeled latex particles. Under these circumstances,

M O ) is a mesure of the interfacial area between D- and A-labeled cells in the film. 13

The newly formed films we exarnined were allowed to dry uncovered in an oven at 31 k 1

OC over 40 min, but as soon as each film appeared to be dry, it was transferred to the cold

room at 4 O C for storage until the decay profile of the cold film could be measured. Since

the Tg of the matrix polymer is Ca. 20 OC, we imagine that minimal polymer difision

occurs when the films are prepared in this way. We use expenmental values of @ d o ) to

examine the effect of filler on the contact between D- and A-labeled latex particles in the

newly formed films.

We examined four types of colloidal silica dispersions. The diameters for the

various filler particles are 25 nrn for Klebosol 30R25 (K-25), 50 nm for Klebosol 30R50

(K-5Gj. 12 nm for Snowtrx O (S-12). ancl 45 iim for Snowîex OL (S-45). notc that

two of the samples have sirnilar mean diameters (Klebosol30R50 and Snowtex OL), and

that S-45 has a particularly broad size distribution.

Effect of K-25 and K-50 on #-fi

In Figure 3.8 we plot @ d o ) vs Si02 content (wt %) for a series of freshly

prepared films containing either K-25 or K-50. Usefui films could be prepared containing

as much as 40 wt % (23.3 vol %) filler content. When we attempted to prepare films

containing larger amounts of silica. those films were so brittle after they dried that they

could not be handled. For K-25. the values of #&O) obtained nnge from 0.036 to 0.085.

For K-50. on the other hand, @do) values nnge from 0.084 to 0.093. In other

experiments on nascent films prepared from sirnila-sized latex particles at temperatures

close to the minimum film forming temperature, @ d o ) values on the order of 0.05 to

0.07 were obtained. lJ These resuits suggest that little polymer diffusion has occurred in

the samples we have examined. However, the extent of energy transfer for newly formed

films is different for films containing two different sizes of SiO2. One can clearly see that,

as one increases the amount of K-25 in the latex film, @ d o ) vaiues decrease, whereas the

magnitude of @ d o ) is essentially independent of the mount of K-50 in the film. This

result indicates that even when large amounts of K-50 are present, there is a common

extent of interfacial contact in the film between cells fonned from D- and A-labeled latex

particles. On the other hand K-25 filler appears to have a larger effect either on reducing

the interfacial area benveen D- and A-Iabeled cells in the system, or on suppressing the

limited extent of polymer diffusion that occurs during film formation than K-50 filler.

O 10 20 30 40 SiO, content (wt %)

Figure 3.8: Plots of the initial efficiency of energy transfer, Om(0), vs SiO, contents in

newly formed films.

Effect of S-12 and S-45 on @do).

In Figure 3.8 we also plot # d o ) vs SiOZ content (wt %) for a series of freshly

prepared films containing either S-12 or S-45. For latex films containing either S- 12 or S-

45, useful films could be prepared containing as much as 10 wt % filler content. For films

containing S-12, the values of #do ) obtained range from 0.020 to 0.085. For films

containing S-45. on the other hand, values range from 0.079 CO 0.086. Here we see

the same trend as the case of K-25 and K-50. These results also suggest that little polymer

diffusion has occurred in the samples we have exarnined. However, Iike K-25 and K-50,

the extent of energy transfer for newly formed films is different for films containing the

two different sizes of SiOz particles. When the films contain the smaller size of %O2, the

filler appears to reduce the interfaciai area between D- and A-labeled latex particles. As

one increases the amount of S-12 in the latex film, @ d o ) values decrease, whereas those

of S-45 are constant. This result also indicates that, even when large amounts of S-45 is

present, there is a common extent of interfacial contact in the film between cells formed

from D- and A-labeled latex particles. The magnitude of the change in M O ) is even

larger for S-12 filler than for K-25 filler. Detailed @&O) values for films containhg

different types and mounts of silica filler are surnrnarized in Table 3.4.

3-34. Maximum Efficiency of Energy Transfer, Q>n(-) in PBMA Latex Films

In order to evaluate the extent of mixing, f,(t) defined by eq 2-8, we need to know

the value of @da), which corresponds to full mixing of the polymer. If the film is fully

rnixed. one should have a random distribution of donors and acceptors, and the decay

profile should be described by eq 2 Ja . Under these circumstances. the rate and efficiency

of energy transfer will be determined only by %, 2, and the concentration of acceptors in

the film.

There are three ways to obtain a sample which will serve as a model for #d=).

First. one takes a film and anneals it for sufficiently long times. Second. one takes a film

and anneals it at higher temperatures. Since the polymer diffusion rate is strongly

accelented by increasing temperature, will norrndly increase rapidly to its maximum

value. Finally, one can dissolve a dry film sample in an organic solvent. In solution. one

expects full mixing of the polymer molecules. A film cast from this solution is then a

p o d model for the determination of # d m ) .

We have found in the past, for filler-free films. that dl these approaches give

similar vdues of ares(=) from which the corresponding @ET(=) values are calculated. '3

In this section, we examine the effect of filier on the magnitude of @d-).

In the experiments reported here, we obtained ares(=) values for filler-free latex

films, from a solvent cast film. This film was prepared from a dry PBMA film prepared

from a 1: 1 mixture of D- and A-labeled particles, which was subsequently dissolved in

tetrahydrofuran (THF). The solution was cast ont0 a quartz plate and allowed to dry at

room temperature for 12 h. For these films, we obtained an ares(=) value of 14.5 ns, and

a @da) value of 0.68, and these values did not change when the film was annealed at 80

O C for 1 h. In contrat, when a sample of the latex film itself was anneaied at 60 "C for

200 h, we obtained an are+) value of 16.1 ns, and a @dm) value of 0.65. One can see

that a latex film heated at 60 O C for this length of time gave values close to those obtained

from the solvent-cast film.

Effect of silica particle size on for a long annealing time at 60 O C .

When films were prepared in the presence of filler and annealed for 200 h at 60

OC, we obtained values of area(200 h) and #&O0 h) that depended on the amount of

filler present. We determined area(200 h) and #EI(200 h) values for films containing up

to 40 wt 8 K-25. For samples annealed for 200 h at 60 OC, these values ranged from 16.6

os (no silica) to 21.9 ns (40 wt% K-25), corresponding @A200 h) values from 0.64 to

0.52. The SiO2 filler has its most pronounced effect when it is present in amounts greater

than 20 wt %. In the case of K-25, the presence of filler appears to lower the Q>n(200 h)

values accessible by annealing the latex films.

We also determined area(î-O0 h) and #&200 h) values for films containing up to

40 wt % for K-50. For sarnple films annealed for 200 h at 60 OC. these values ranged from

16.1 ns to 17.0 ns, corresponding to #d200 h) values from 0.64 to 0.62. Unlike the case

of K-25, the presence of this filler does not Iead to a lowering of @&200 h) values.

In the case of films containing S-12 filler, we see that #E1(200 h) values decrease

with increasing amount of filler, following the same trend observed for films containing

K-25. However, the magnitude of Gd200 h) decrease for films containing S-12 is even

larger than that of S-12. We determined area(200 h) and @E.r(200 h) vaiues for films

containing up to 40 wt % for K-25. For samples annealed for 200 h at 60 OC, these values

ranged from 17.0 ns (no silica) to 31.6 ns (40 wt % S-121, corresponding to Q>n(200 h)

values from 0.63 to 0.30. Even the presence of a srnall amount of S-12 affects the

@&O0 h) values. and as the filler content is increased, it ha an even larger effect on

@d200 h) values.

We also determined area(200 h) and @d200 h) values for films containing up to

40 wt % S-45. For sarnples annealed for 200 h at 60 OC, these vdues ranged from 16.2 ns

to 16.9 ns, corresponding to QSr(2OO h) values from 0.64 to 0.63. As in the case of K-50,

the presence of filler does not lower a d 2 0 0 h) values.

In summary, the presence of 50 nm diameter silica particles (K-50 and S-45) has

almost no effect on area(200 h) values and @E7(200 h) values. The values we calculate

are almost identical to those in the filler-free film. where essentially full mixing of the

donor- and ricceptor-labeled polymer occurs. We conclude that these relatively large

particles have little effect on the extent of polymer diffusion that takes place when the

films were heated for 200 h at 60 OC. In contrat, we observe significant changes in

~ m z ( 2 O O h) and @k-l(?OO h) :.dues in latex films containing the srnaller si!ica particles

(K-25 and S- 12). The results are sumrnarized in Figure 3.9. Here the filler particles retard

polymer diffusion to such an extent that the extent of mixing after 200 h annealing is

significantly reduced. Detailed Od200 h) values for films containing different types and

amounts of silica fiIler are summarized in Table 3.4.

0e4 C l K-50 (50 nm) 0 S-12 (12 nm)

w e u

O 10 20 30 40 SiO, content (wt%)

Figure 3.9: Plots of the maximum efficiency of energy transfer, Qm (200 h), vs Si02

contents. Films were anneaIed for 200 h at 60 OC.

In Figure 3.10, we plot Od200 h) values for films containing 40 wt % of SiOz

annealed for 200 h at 60 O C , as a function of the inverse diameters of SiO2. The number

average diameten of silica particles were determined by transmission electron

microscopy (TEM). Since l/dsio2 corresponds to the surface to volume ratio of SiO2 filler,

one can see that there is a linea. relationship between Qn(200 h) values and the surface

to volume ratio.

Figure 3.10: Plots of the maximum efficiency of energy transfer. Om (200 h). vs the

surface to volume ratio (l/dsioz). Films contain 40 wt % of SiO?, and were annealed for

200 h at 60 OC. The diameters of SiOt were detennined by transmission electron

rnicroscopy (TEM).

Effect of silica oarticle size on at hioh annealine temperatures.

In the previous section, we saw that in the presence of small silica particles (S-12,

K-25) the extent of polymer diffusion that occurred in at 60 O C over 200 h was

significandy reduced. Here we examine this effect from a different perspective, in which

we mnealed films containing 40 wt % (33.3 vol I) silica for 2 h at various tempentures.

The results are shown in Figure 3.1 1. @E7(2 h) values increase with the annealing

temperature, and approach 0.6 even in the presence of S-12 and K-25. Detailed @*

values for both 10 min and 120 min are shown in Table 3.5.

These results indicate that interaction of the PBMA polymer with the surface of

the silica particles slow down the rate of polymer diffusion. On the other hand, we learn

that this interaction is not so strong as to suppress a part of the polymer rnixing process.

60 90 120 Annealing Temperature ( O C )

Figure 3.11: Plots of the maximum efficiency of eneqy transfer, am(2 h), vs annealing

temperature. Films contain 40 wt % of SiO2 and were anneded for 2 h.

Table 3.4: Values for @do) and w 2 0 0 h) at 60 OC.

PBMA + S-12

PBMA + S-45

PBMA + K-50

SiO2

content

(wt %)

Table 3.5: G d 2 h) at different annealing tempentures.

PBMA .t

40 wt % K-50

PBMA + 40 wt% $12

PBMA + 40 wt 96 S-45

10

min

10

min

120

min

10

min

Determination of areal4 and @d-).

To examine the influence of pigment on the rate of polymer interdiffusion, we

need to introduce values of ares(=) or #dm) into the calculation of f,(t). Since fm(t) is a

measure of the extent of rnixing in the PBMNsilica system, we need to determine both

are+) and ad=) values that describe only the latex polymer in the system, so that we

c m consider separately the effect of filler on the rate and the total extent of polymer

difision. For this reason, we use values obiained for the filier-free polymer obtained by

solvent casting; ares(=) = 14.5 ns, and M m ) = 0.68 in the calculation of ail f, values.

120

min

3-3-6. Effect of Silica Particle Size on the Rate of Poiymer Interdiffusion

In this section we examine the influence of silica on the rate of polymer diffusion

in PBMA latex films. To calculate f,(t) values via eq 2-8, we employ values

corresponding to each film, but #dm) values obtained as descnbed in the preceding

section.

Effect of K-25 and K-50.

In Figure 3.12a. we plot values of fm(t) as a function of annealing time at 60 OC for

films containing different amounts of K-25. The f,(t) values were caiculated with eq 2-8

from the areas under the fluorescent decay curves, using a common ares(=) value of 14.5

ns. We see that the presence of K-25 in the film has a pronounced effect on reducing the

rate of polymer interdiffusion. Even in the latex film with 10 wt % silica, it is obvious

that K-25 retards the rate of polymer interdiffusion. There is greater retardation as one

increases the amount of K-25 in latex films. When one increases the amount of K-25 in

the latex film up to 40 wt %. one can see a significant effect of retardation on the polymer

diffusion rate.

In Figure 3.12b we plot values of f,(t) as a function of annealing time at 60 OC for

films containing different arnounts of K-50. One of the interesting features in Figure

3.12b is that polymer interdiffusion is retarded by the presence of K-50 filler. but the

extent of retardation is much smaller than that of K-25. The use of K-50 filler up to 20 wt

% seems to have little or no effect on the rate of polymer interdiffusion. Even in the latex

film with 40 wt % K-50, the effect is much smdler than that of K-25.

0.8 a

0.6 e .c, w

E cw 0.4

0.2

0.0 F O 60 120 180 240 300 12000

Time (min)

(b) K-50, 60°C cwt - a18 -1

O 60 120 180 240 300 12000 Time (min)

Figure 3.12: Plots of the extent of rnixing f,(t). as a function of annealing time. Films

contain (a) 25 nrn of SiOr (K-25). (b) 50 nm of Si02 (K-50). and were annealed

simultaneously for each series of sample films at 60 OC.

Effect of S- 12 and S-45.

In Figure 3.13a, we plot values of f,(t) as a function of annealing time at 60 OC for

films containing different amounts of S42. The presence of S-12 in the film has a

significant effect on reducing the rate of polymer interdiffusion. Only 10 wt % of filler

content causes a significant retardation of the polymer interdiffusion rate. There is also a

greater retardation effect when one increases the amount of S-12 in latex films. When one

increases the amouni of S-12 iii ille latex film up :O 10 wr %, the nte of polyner diffusion

is substantially reduced. This effect is even larger than that for films containing K-25

filier.

In Figure 3.13b we plot values of f,(t) as a function of annealing time at 60 O C for

films containing different amounts of S-45. Here one notices that polyrner interdiffusion

is retarded by the presence of S J 5 filler. but the extent of retardation is much srnaller

than that of S-12. The use of S-45 filler up to 20 wt % seems little effect on the rate of

polymer interdiffusion, as one can see for films containing K-50 filler. Even in the latex

film with 40 wt % S-45, the effect is much srnaller than that of S-12.

1.0

0.8

0.6

0.4 I

0.2 I

3

0.0 O 60 120 180 240 300 12000

Time (min)

(b) S-45,60°C (wt %)O

O 60 120 180 240 300 12000 Time (min)

Figure 3.13: Plots of the extent of mVting f,(t), as a function of annealing tirne. Films

contain (a) 12 nm of SiO2 (S-12), (b) 45 nm of Si02 ( S A S ) , and were anneded

simultaneously for each senes of sample films at 60 OC.

3-3-7. Analysis of the Diffusion Process

In this section we analyze the f,(t) data more deeply to try to understand how

mineral fillers affect the polymer molecules as they dif ise across the latex ceil

boundaries. If the diffusion follows Fick's laws, one expects that the extent of rnixiog

f,(t) will be proportional to the square root of time. l5 We plot fm(t) as a function of t'" in

Figure 3.14. These plots are linear for values of f,(t) up to 0.7 for the film without silica.

In addition, the films conrainine different sizes of SiO? filler also eive reasonable linear

plots. We fit each senes of data to a straight line. and note that they al1 have a small

positive intercept. In the analysis described below. we only consider the slopes obtained

by the least-squares best fits to the lines shown in Figure 3.14. and ignore the intercepts at

t = O (with f,(r) S 0.06).

Figure 3.14: Plots of the extent of rnixing f,(t), vs the square root of annealing time. Si02

content for al1 SiOl contained films is 40 wt %. Those films were annealed at 60 O C .

The dope of each line is a measure of the polymer mobility in the film. To

examine how the surface-to-volume ratio of the silica particles affects this rnobility. we

plot the value of the slope &(t) 1 tl") against l/dsio2 in Figure 3.15. The data fa11 on a

smooth curve, but not a straight line.

When Feng et al. " examined the influence of PMMA particles as fillers on the

rate of polymer diffusion in latex films. they found a linear dependence of (f,(t) I tl") on

l/dsioz. In their experiments, the filler content was kept constant at 35 vol %.

Figure 3.15: Plots of the slope values in Figure 3.14. as a function of the surface to

volume ratio, l/dsioz for 40 wt% (23.3 vol %) silica filler. The mean diameters of SiOi

were detemiined by transmission electron microscopy (TEM).

3-3-8. Effect of Silica Particle Size on Diffusion Coef'fïcients

Another measure of the polymer mobility is the apparent diffusion coefficient

Da,, which descnbes the rate of movement of the polymer molecules across the polymer-

polymer interface between adjacent cells in the film. l w 8 We calculate the diffusion

coefficient of the polymer by fitting the extent of mixing f,(t), obtained from the energy

transfer measurements, to a spherical diffusion mode1 which satisfies Fick's laws of

diffusion, where D, is the true difision coefficient. 19

This mode1 assumes that the diffusing substance is initially distributed

in a sphere of radius R with an initial concentration Co. 'O At rime r we have

75

(3-8)

uniformly

Dapp values are calculated by equating f, with the fractional mass f, which has

diffused ûcross the interface, f, = M t I Mm. where M, = ( 4 3 ) s r ~ ~ ~ o . and carry out a

numencal integration to find the best D,, value which satisfies the equation

Simulations have shown '' that for the particle size and concentration of acceptor

employed here, f,(t) increases more npidly thanf,(t). Thus values calculated for Dapp are

larger than those for D,. The simulations also show that Dapp is proportional to D, for

values of f,(t) up to 0.7, and in this range the two D values differ by a factor of 3.

Experiments on latex films involve polymers with a distribution of chain lengths

and a corresponding distribution of Ds values. The 4, values we obtained are apparent

mean diffusion coefficients avenged over d l the chain lengths in the sample latex 13

polymer and the annealing history of the film. I 3 Previous experience - has shown that

molecular weight polydispersity leads to Dapp values that decrease with increasing fm(t). l7

The short chain polymers dominate the initial polymer diffusion. At longer times, the

growth in @ d t ) and f,(t) is due to the diffusion of longer chah polymers. Because D,,

values Vary with f,(t), values of Dapp from separate expenments should be compared at

similar extents of mixing.

The data for Da,, and the corresponding fm(t) values are surnmarized in Table 3.6.

Table 3.6: D,, and the corresponding f,(t) values for films containhg different types and

arnounts of silica filler.

PBMA

+

K-25

L

PBMA

+

S-15

0.13

0.36

0.50

0.60

0.77

0.9 Z

0.093

0.081

0.083

0.064

0.050

0.040

O. 103

0.081

0.078

0.063

0.036

0.005

0.06 1

0.067

0.060

PBMA

+

0.069

0.036

0.05.)

0.043

0.033

0.033

0.23

0.37

0.47

0.60

0.76

0.93

0.073

0.084

0.082

0.70

0.25

0.38

0.50

0.67

0.9 1

0.26

0.39

053

0.64

0.79

0.93

0.20

0.36

0.17

K-50

0.104

0.086

0.08 1

0.026

0.029

0.030

0.024

0.020

0.037

0.080

0.070

0.059

0.054

0.04 1

0.003

0.056 ---

0.029

0.W

0.26

0.40

0.53

0.16

0.2 1

0.28

0.35

0.49

0.87

0.08 1

0.063

0.056

0.016

0.034

0.003

0.24

0.36

0.49

0.22

0.40

0.54

0.066 0.065 0.64 - PP

0.053 0.8 1 0.017 0.79

0.004 0.94 0.003 0.9 1

0.56

0.69

0.94

0.080

0.054

0.054

0.12

0.22

0.31

0.39

0.56

0.92 1

0.1 1

0.32

0.42

0.52

0.67

0.085

0.066

0.065

0.23

0.33

0.45

0.042

0.028

0.024

0.019

0.015

0.025

0.030

0.0 16

0.0 13

0,012

0.0 10

0.0 12

0.73

0.35

0.45

0.56

0.73

0.92

0.063 O 5 7

0.13

O. 17

0.21

0.29

0.42

0.71

0.064

0.05 1

0.038

0.038

0.027

O.Oû3

0.053

0.030

0.027

0.025

0.092

0.042

0.00s

0.19

0.25

0.3 3

0.43

0.62

0.9 I 0.92

o n 0.94

0.003

0.042

0.003

0.77

0.92

Effect of K-25 and K-50.

In Figure 3.16a, we plot values of Da,, as a function of f,(t) for films containing

different amounts of K-25. Here we see that the presence of 20 wt % K-25 in the film has

a pronounced effect on reducing the rate of polymer interdiffision. We compare the Dapp

values at f,(t) = 0.35. One sees that the Dapp value decreases from 0.08 1 (nmls') for the

film containing no silica, to 0.027 (nm/s2) for the film containing 20 wt % K-25 and to

0.01 1 (nm/s2j for the îïini cuntainiiig 10 wî B K-25. When one i n m a e s the amount of

K-25 in the latex film up to 40 wt %, one can see a significant efTect of retardation.

However, the difference in Dapp values appears to be greater between the film with O and

20 wt % filler than those between 20 and 40 wt %.

In Figure 3.16b we plot values of Dapp as a function of fm(t) for films containing

different amounts of K-50. Unlike the case of K-25. Dapp values do not change

significantly with increasing amount of K-50 filler. We compare the Da,, values at fm(t) =

0.50. The Dapp value decreases from 0.08 1 (nm/s2) for the film containing no silica. to

0.065 (nrn/s2) for the film containing 20 wt % K-50 and to 0.050 (nrn/s2) for the film

containing 40 wt % K-50.

Effect of S- 12 and S-45,

In Figure 3.17a. we plot values of Da, as a function of f,(r) for films containing

different arnounts of S- 12. As in the case of K-25, we see that the presence of S- 12 in the

film has a significant effect on reducing the rate of polymer interdiffusion. We compare

the Rpp vaiues at fm(t) = 0.3 1. The Dapp value decreases from 0.10 (nmls') for the film

containing no silica, to 0.016 (nm/s2) for the film containing 20 wt % S-12 and to 0.004

(nrn/s2) for the film containing 40 wt % S-12.

In Figure 3.17b we plot a,, values as a function of f,(t) for films containing

different amounts of S-45. Polymer interdiffusion is retarded by the presence of S-45

filler, but the extent of retardation is much sinaller than that of S-12. We compare the D,,

values at f,(t) = 0.50. The D,, value decreases from 0.078 (nmls') for the film containing

no silica, to 0.030 (nrn/s2) for the film containing 20 wt 8 S-45 and to 0.023 (nds') for

the film containing 40 wt % S 4 5 .

n nni 1 I 1 l 1 l 1 1 i 1 I

Figure 3.16: Mean apparent diffusion coefficients, D,,, as a function of the exrent of

rnixing f,(r). Films contain (a) 25 nm of SiO? (K-25), (b) 50 nm of SiO2 (K-50), and were

annealed simultaneousIy for each senes of sample films at 60 O C .

0.001 1 1 I 1 i 1 I 1 I 1

0.0 0.2 0.4 0.6 0.8 1 .O

f&)

Figure 3.17: Mean apparent diffusion coefficients, Da,,, as a function of the extent of

mixing f,(t). Films contain (a) 12 nm of SiOz (S-12), (b) 45 nm of SiO2 (S-45), and were

annealed simultaneously for each series of sample films at 60 OC.

3-3-9. Dependence of the Polymer Diffusion Rate on the Volume Fraction of Silica

In Section 3-3-7 we described the dependence of polyrner diffusion rate on the

size of silica particles. and showed that smaller sized silica had a larger effect on the rate

of polymer diffusion. In this section we describe how the amount of silica affects the

polymer diffusion nie.

- In Figure 3.18 we plot f,(t) as a function of rl" for films containing (a) S-12 and

(b) K-25. As rve described in Section 3-3-7. al! cf the data give reasonable linear plots,

indicating that polymer diffusion follows the Fick's laws even in the presence of silica

particles. The slopes decrease as one increûses the amount of silica, and the same trend

c m be seen for films containing different sizes of S Q . We plot those slope values as a

function of silica volume fraction. We assume that the density of silica is 2.2 g/crn3. One

can see in Figure 3.19 that when the slope vaiues are plotted ûgainst silica volume

fraction, we obtain a smooth curve but not a straight line.

a::: 1 /.

(a) S-12,60 O C c

Figure 3.18: Plots of the extent of mixing f,(t) vs the square root of anneaiing time for

latex films containing different amounts of (a) S- 12 and (b) K-25.

0.00 0.05 0.10 0.15 0.20 0.25

SiO, volume fraction

Figure 3.19: Plots of the siope values in Figure 3.18. as a function of the silica volume

fraction.

3-3-10. Fundamental Mechanisrn of Polymer Diffusion

Viscoelasticity in polymers is one of the most important characteristics that lead to

a strong dependence of polymer diffusion on factors such as temperature, polymer

molecular weight, and distribution, the amount of added low molecular weight diluents,

and the degree of cross-linking. '' We try to understand the effects of such factors

obtained by different techniques and elucidate the fundamental mechanism of polymer

diffusion and its correlation with polymer propenies derived from other types of

viscoelastic measurements.

Viscoelastic properties of polymers can be well described by the Williams-

Landel-Ferry (WLF) equation. '' This equation describes temperature effects above the

Tg of the polymer on dynamic properties of polymers related to backbone motions in

tenns of changes in the free volume in the system. This equation is based on the idea of

time-temperature superposition. Events which occur on one time scale at a given

temperature To occur on a faster time scaie at higher temperature. In the traditional WLF

analysis, one defines a shift factor a ~ = thr, in terms of the shift dong the time axis needed

to bring two curves, representing measurements at different T. into correspondence at a

reference tempenture T,.

log a~ =

where Ci and C2 are parameters characteristic of a particular polymer. and 4 is the

diffusion coefficient determined at the reference temperature. In a study of the creep

cornpliance of PBMA, Ferry and coworkers '3 were able to fit the data at T, = 373 K, Ci =

14.5 and Cz = 255 K. Those values were found to be independent of molecular weight for

PBMA sarnples of MW > 6.0 x IO".

In our approach to data analysis, we shift plots of the apparent diffusion

coefficient Da,, vs. the extent of mixing fm(t) dong the D-mis at constant T to bnng the

curves into correspondence at zero silica content (mf = O). Thus we can define a shift

factor bT as

This shift factor is similar in form to that employed in the Fujita-Doolittle

expression, even though the assumptions of this mode1 do not entirely fit Our system.

Through the shift factor h, we introduce free volume theory, assurning that rnolecular

uanspon as descnbed by Dqp is regulated by the availability of free volume in the

system. 2.5

The specific occupied volume of a Iiquid Vo is defined as the volume of the

equilibrium liquid at O K. Therefore the specific Free volume VF is given by

where V is the specific volume of the liquid structure at any tempenrure T. As the

temperature is increased from O K, the increase in volume is accompanied by

homogeneous expansion of the matenal due to increasing amplitude of vibrations with

temperature, and also by formation of holes which are distributed discontinuously

throughout the material at any instant.

The self-diffusion coefficient of one component system derived from free volume

theory is expressed below:

where Dl is the self-diffusion coefficient. Doi is a pre-exponential factor. Vm is

the average hole free volume per molecule in the liquid (or per gram of the polymeric

liquid), V is the cntical local hole free volume required for a molecule (or jumping unit in

case of polymers) to jump to a new position, E is the critical energy which a molecule

must obtain in order to overcome the attractive forces holding it to its neighbors. T is the

absolute temperature and k is the Boltzman constant.

In the case of two components. one assumes that the change in volume is only due

to the change in the available hole free volume. One also assumes there is no volume

change during mixing, and that the rnolecular weight of the solvent is equal to the

molecuiar weight of a jumping unit of the polyrner chain. Fujita '6 denved an expression

for the diffusion coefficient D of a polymer in the presence of small molecular species:

where f is a total free volume of the system, B is the minimum hole size or jump size

required for the diffusion of a given molecule or rnolecular segment, A is a

proportionality constant that depends on the size and shape of a jumping unit of the

polymer chain, T is the absolute temperature, and R is the gas constant.

The free volume of a two cornponent system is in fact die sum of the hctional

fkee volume contributed by individual components. In our case those are polymer and

silica filler particles, whereas in Fujita's derivation Qf is the volume fraction of

plas ticizer.

where 0, is the volume fraction of polymer with fiactional free volume fp, af is the

volume fraction of silica filler particles with fractional free volurnefi, and P =fr -Ji is the

difference between the fractional free volume of silica filler particles and polymer. When

Of refers to a small molecule miscible with the polymer, one can obtain the Fujita-

Doolittle equation from eq 3-15 and 3-16. The Fujita-Doolittle equation describes the

influence of the volume fraction of the additive on the polymer diffusion coefficient at

given temperature as expressed belocv 26 :

where D(T, Qf) is the diffusion coefficient of polymer at temperature T in the presence of

QI volume fraction of plasticizer, D(T. O) is the diffusion coefficient of polymer at

temperature T in the absence of plasticizer. The terms fp(T, 0) and P(T) are defined in eq

3-16, at given temperatures.

Silica particles act to rigidify the surrounding matrix and are not solutes in the

traditional sense. " If for the sake of argument we assume that silica particles act as

"antiplasticizer." one may obtain any mathematical relationship between the ratio of

diffusion coefficients and the volume fraction of silica fiIIers. In order to examine this

idea, we first calculate the magnitude of the term (ln[Dp(T,O)lDp(T,<Dr)]}'l for each set of

data, with Qi ranging from 0.048 to 0.233. We assume that silica fillers have a density of

2.2 @m3. We equate Dp with D,,. Since the plots of D,, vs f,(t) in Figure 3.20 are

overdl paralle1 for al1 sets of data, the value of {ln[D,(T,O)/Dp(T,~f)] 1'' is considered to

be almost constant at different f,(r) values. Consequently al1 the data can be

superimposed. and we obtain a single master curve of a,, vs f,(t) for PBMA latex films

containing different amounts of S-12 in Figure 3.2 1.

In Figure 3.22 we plot values of {ln[D,(T,O)/Dp(T,@I)l}'l vs of-' for PBMA latex

films containing various fractions of S-12. We obtain a stnight line which indicates that

the shift factor is a function of the silica volume friction. If we plot

(ln[D,(T,O)/D,(T,~f)]}" vs Or -'? as in eq 3-17, we also obtain a straight line, widi a

negative slope. The plot also has a negative intercept. Too strict adherence to the Fujita-

Doolittle mode1 would lead to the strange conclusion that fp(T,O), the free volume in the

polymer in the absence of additive. is negative.

Figure 3.20: Plots of Da,, vs f,(r) for PBMA latex films containing O wt % (O), 10 wt %

(a), 20 wt % (O), 30 wt % (a), and 40 wt % (A) of S-12.

Figure 3.21: Master curve of Da,, vs fm(r) for PBMA latex films containing the S-12

volume fraction Of = O ( O ) , 0.048 (a), 0.102 (a), 0.163 (a), and 0.233 (A).

Figure 3.22: PLot of (ln[D,(T,O)/D,(T, af)]}-' vs -' for PBMA latex films containing

various fractions of S- 12.

We also examine this behavior with a different size of silica filler K-25. The same

trends can be seen in plots of &,, vs f,(r), master curve, and a plot of

(ln[D,(T,O)ID,(T,~f)] )- ' vs aI ? Those data are shown in Figure 3.23, Figure 3.24, and

Figure 3.25.

Note that ihe intercept is 0.19 and the dope is 0.035 in Figure 3.22 for the PBMA

latex films containing S-12. For the PBMA latex films in Figure 3.25 containing K-35.

the intmept is 0.36 and thc slopc is 0.05 1.

Figure 3.23: Plots of a,, vs f,(t) for PBMA latex films containing O wt % (O), 10 tvt %

(a), 20 wt % (O), 30 wt % (a), and 40 wt % (A) of K-25.

Figure 3.24: Master curve of D,,, vs f,(r) for PBMA latex films containing the K-25

volume fraction <Pr = O (O), 0.048 (a), 0.102 (O), 0.163 (m), and 0.233 (A).

Figure 3.25: Plot of {In~,(T,O)ID,(T, Of)]}-' vs Of-' for PBMA latex films containing

various frictions of K-25.

One notices that our system containing both PBMA latex polymer and silica looks

very similar to that containing polymer and a plasticizer (or an antiplasticizer). However,

there is an important difference between the silica-filled polymer and one of the key

assumptions made in denving the Fujita-Doolittle equation. In a plasticized polymer, the

presence of the plasticizer adds iree volume to the system and lowers Tg. in the silica-

filled system, the filler particles rigidify the matrix surrounding each particle. and reduce

the free volume of the systern. Since the silica particles do not dissolve in the polymer.

one cannot sirnply apply our system to the mode1 proposed by Fujita and Doolittle.

WLF analvsis.

Another way of analyzing this data is that one can correlate [hem to the Williûrns-

Landel-Feny (WLF) equation, as seen in eq 3-1 1. Here we assume that PBMA latex

polyrner consists of two different regions with different glass uansition tempentures

(Tg), and chat the polymer near the silica tiller surface has a higher Tg. " Thus one can

i nrroduce

where ATg is the Tg difference between the polymer near the silica filler surface and the

bulk polymer. T,' is an arbitrary chosen reference temperature. Eq 3- 1 1 can be rewritten

as

In Our analysis, the shift factor is descnbed in eq 3-12. We equate ln a~ = ln b,

and calculate ATg. Plots of ATg vs the silica volume fraction are shown in Figure 3.26.

The data fall on a srnooth curve for both series of films, but not a stmight line.

We normalize those data to a constant surface-to-volume ratio. Here the diameters

of silica particles were determined by transmission electron microscopy (TEM). The data

can be supenmposed onto a single curve as shown in Figure 3.27.

Figure 3.26: Plots of ATg vs the silica volume fraction for PBMA latex films containing

S-12 (O) and K-25 (a).

Figure 3.27: Plots of ATg vs the total volume of the polymer near the silica surface for

PBMA latex films containing S-12 (O) and K-25 (e). The diameters o f silica were

determined by transmission electron rnicroscopy (TEM).

To pursue this analysis further, we assume that the PBMA polymer c m be divided

into two types of domains. The polymer near the silica surface is assumed to have a

higher Tg, and the bulk polymer has its normal Tg. The Tg for the polymer nea. the silica

surface can be calculated by the following equation.

where TgaPP is the apparent Tg for the totd PBMA polymer (= 293 + ATg), 0, and Q2 are

the volume fractions of the domains of bulk polymer and the polymer near the filler

surface, respectively, and Tgl and Tg2 are the glass transition temperatures for the bulk

polymer and the polymer near the filler surface. TgI of the bulk polymer is 293 K. Thus

one can simulate the Tg2 values by making assurnptions about the magnitude of 0,. We

assume (0, values based on the idea that the thickness of the rigidified layer 8pl is on the

order of the radius of gyntion Rc of the polymer in the matrix. In our system for PBMA

with a MW = 50.000, Rc = 4.7 nm. '"n Figure 3.28, we present a plot of ATg vs the totd

volume of the polymer near the silica surface for PBMA latex films coniaining various

sizes of silica particles. The arnount of rigidified polymer is calculated assurning Spi =

RG. We have a situation with one measunble (bT) and two unknowns (Tgr, tipi). Various

pairs of Tgz and 6,i values will explain Our results. For example, if 6,i = 4.7 nm (Ro),

Tgz = 344 K, whereas if 6,[ = 1.2RG, Tgz = 335 K. Both of these values are above the

annealing temperature of the polymer diffusion expenments. A plot showing the

relationship between ATg and Ôpa& is presented in Figure 3.29.

Figure 3.28: Plots of ATg vs the total volume of the polymer near the silica surface for

PBMA latex films containing S-12 (O), K-25 (a), S-45 (O), and K-50 (m). The diameters

of silica were determined by transmission electron microscopy (TEM).

Figure 3.29: Plot of the glass transition temperature Tgz vs thickness of the polymer layer

6pi rigidified by the silica filler for the polymer near the silica filier surface. The

thickness is expressed based on the radius of gyration RG for the PBMA polymer with a

A third model for understanding the influence of silica on polymer diffusion is

based on the observation that the films become brittIe when Qsio2 exceeds 0.233. In this

model we make the assumption that when QSio2 = 0.233, al1 the polymer in the film has

become rigidified. Taking into account the diameters of the silica particles, we Rnd that

this result corresponds to a layer of rigid polymer of the thickness 8.0 nm surrounding

each particle for both S-12 and K-25. Since RG for the PBMA is estimated to be 4.7 nm.

i1iè rigidificd layer corresponds to 1.7 Rû. In order to proceed. we define a new chift

factor

In Figure 3.30. we plot the reciprocal value of this shift factor against the

reciprocal of the total volume of non-rigidified polymer in the system. The volume of

rigidified polymer is calculated in the following way: the term (6ld) refers to the surface

to volume ratio, and (6/d)(L.7Rc@sio2) refers to the total volume fraction of the polyrner

rigidified by the silica particles. Both plots for films containing S-12 and K-25 give

reasonable straight lines.

Figure 3.30: Plots of ( ln[D,(T, û+)/D,(T. 0.233)] 1'' vs ( 1-6( 1 *' for PBMA

latex films containing S- 12 (O) or K-25 (a).

3-4. Discussion

3-1-1. M O ) in Newly Formed PBMA Latex Films in the PresencdAbsence oPSilica

The initial efficiency of energy transfer. WO), in latex films provides a measure

of the ratio of intedacial area between D- and A-IabeIed latex cells to the volume of the

D-labeled phase. l 3 As one can see in Figure 3.8. @ d o ) values decrease with increasing

arnount of K-25 and S-12, whereas those of K-50 and S-45 are constant. The effect of K-

50 and S-45 on reducing the DIA interfacial area is very small, or almost negligible even

if their filler content is larger. On the other hand. it seems that K-25 and S-12

significantiy reduce the DIA interfacial area.

One explanation to ihis is that the film morphology in the presence of K-25 or S-

12 is different from that in the presence of K-50 or S-45. Due to its small size, filler up to

d = 25 nrn c m form a network iocated between D- and A-labeled latex ceils. The

presence of srnaIl particles in the interface reduces the interfacial area and the subsequent

efficiency of energy transfer. Those filler particles may even surround the latex particles

and prevent the polymer molecules from dif ishg across the ce11 boundaries. On the

other hand, K-50 and S-15 may not cover the whole latex panicles, and ailow polymer

interdiffusion to occur more easily. But at this point we do not know how film

morphoiogy changes in the presence of different sizes of filler, especially those particles

smaller than the latex particles.

There is evidence in the litenture ''JO that when one adds small hard particles to a

dispersion of larger soft latex particles, the small particles form a percolation network in

the film fornid upun dqing. WC depict ihis gpe of structure in the h w i n g in Figure

3.3 1. In this two-dimensional cross-section of the newly formed film, the small particles

form a connected network in the space between adjacent cells. By occupying a significant

fraction of the area between the faces of adjacent cells. the particles act to separate D-

labeled polymer from A-labeled polymer in the newly formed film.

Figure 3.31: Schematic representation of morphology difference for latex films

containing either smaller (left-hand side) or larger (right-hand side) size of Sioz. The

silica fillers can act as obstacles and retard polymer interdiffusion andor reduce the

mobility of polymer molecules near the filler surface.

We have no specific evidence about the rnorphoiogy of the films containing the

larger (45 nm, 50 nm diameter) silica particles. We can conclude from their lack of

influence on OEI(O) values that these particles do not occupy a significant fraction of the

interfacial area between adjacent D- and A-labeled cells in the nascent film.

3-4-2. Effect of Silica on w 2 0 0 h)

In the absence of silica, when latex films are annealed for long penods of time,

polymer diffusion leads to complete mixing of D- and A-labeled polymer. In the presence

of silica, we see a pronounced reduction in the extent of mixing in films containing 20 to

40 wt % of the srnall silica particles S-12 and K-25. As one sees in Figure 3.9. after 200 h

annealing at 60 OC, substantial amounts of polymer remain unmixed, leading to decreased

values of #&Ml lij. To explain ihis efkci, we imagine that the polymer adjacent to the

particle surface is adsorbed to the silica. The Si-OH groups at the surface of the silica are

likely to be strong hydrogen bond donors toward rhe ester groups of the PBMA.

Adsorbed polymer would require either very long times or elevated temperature to

desorb. In accord with this idea, we found, Figure 3.1 1. that after 2 h anneaiing at 120 O C ,

QR. values for films containing 40 wt % S-12 or K-25 approached the value expected for

complete interdiffusion of the latex polymer. The larger particles S 4 5 and K-50 had a

much smaller influence on the magnitude of a d 2 0 0 h). The plot in Figure 3.10 suggests

that the magnitude of this effect is related almost entirely to the difference in surface-to-

volume ratio for these filler particles.

3-4-3. EfFect of Silica on the Polymer Diffusion Rate

As we have shown above, the polymer diffusion rate is retarded in the presence of

silica as a filler. The effect of retardation is larger for the smaller size silica particles,

which have a kger surface to volume ratio.

There are two types of explmations for the effect of filler particles on reducing the

polymer diffusion rate as seen in Figure 3.14 to Figure 3.17. In the first model, the hard

pigment surface serves to make the adjacent polymer matrix more rigid. This is the

traditional "filler effect" in which filler particles increase the modulus of elastomers. It is

well known that polymer chahs adjacent to a rigid surface have decreased rnobility. '' On

the left-hand side of Figure 3.32, we present a drawing indicating how each filler particle

acts to reduce the mobility of polymer near its surface. Our explanation of the behavior of

the shift factor described above is based upon this idea.

Tsagaropoulos and Eisenberg " proposed a three-layer mode1 in terms of polymer

mobility. They determined Tg values from the maximum in tan6 From dynamic

mechanical (DMA) measurements. They studied changes in the glass transition

temperatures associated with adding small silica particles as fillers to various bulk

polymers. As increasing amounts of filler were added, a new high glass transition

temperature (Tg) was found in addition to the Tg for the bulk polymer. As more filler was

added, rhey found s decreze In the magnitude of the DMA sienal from which Tg was

determined. In their model, the surface layer of polymer adjacent to the filler is strongly

imrnobilized, but in addition, nearby polymer also has its mobility restricted. It is this

nearby polymer that contributes to the elevated Tg. From this perspective, one reason for

the decreased rate of polymer diffusion found here is that the polymer molecules near the

filler surface have decreased mo bility .

Immobilized polymer Obstacle effect layer at the fffler surface

Figure 3.32: Schematic representations for polymer immobilizaiion near the filler surface

and obstacle effect.

Based upon the filler-effect model, we can try to estimate what fraction of the

polymer in the system would have i ts mobili ty reduced in the presence of 40 wt % (23.3

vol 8) siiica of each type examined here. We depict a dnwing in Figure 3.33 and

descnbe that a polymer Iayer, which has a thickness of the radius of gyration RG or 2Rû,

surrounds the whole silica surface. We begin by noting that RG = 4.7 nm for BPMA of

MW = 50,000. '* For the various silica particles with diameters of 12, 25 45 and 50 nm,

their number ratio at 40 wt % to the LOO nrn PBMA panides is 176: 19:3.3:2.4: 1. This

calculation assumes a density of 2.2 @m3 for silica. As shown in Table 3.7, when we

consider polyrners within a distance RG h m the particle surface and sum over the surface

area of al1 of the silica particles. we find that for K-50 that only 20 % of the volume of the

PBMA phase is affected by the filler. For S-12. al1 of the PBMA polymer is afTected. If

we go a step further and assume that polymer molecules within 2RG are influenced by the

presence of the silica surface, we find that al1 of the polymer would be affected in the case

of S-12 and K-25. It is important to remember that aggegation of the silica particles

would expose less area to the polymer film, and lower the influence as predicted in Table

3.7. Nevertheless, we know from the uansparency of the films that even in the presence

of 23.3 vol % silica, there is not sufficient aggregation to create voids that would scatter

light.

Figure 3.33: Schematic representation for polymer molecules immobilized near the filler

surface. We calculated the ihickness of the immobilized polymer layer based on the

radius of gyration, b, and 2RG using RG (PBMA, MW = 50,000) = 4.7 nm.

An alternative model also predicts diat the presence of particles would retard the

diffusion rate. In the obstacle model, the particies act as inert obstacles, and polymers in

the system must diffuse around thern. By increasing the tortuosity of the diffusion path,

obstacles increase the time required for the polymers to mix. A drawing depicting this

model is shown on the right-hand side of Figure 3.32.

The results we obtain on the influence of silica particles on On(200 h) values

indicate to us that polymer adsorption ont0 the surface of the silica plays an important

role in aHecting the difision rate of the polyrner in the filled systems. We suspect,

however, that the filler effect is not the whole story. When we examine how plots of Dqp

vs. f,(r) are affected by an increase in the arnount of silica present, we see a striking

difference between the behavior for the films containing S-12 or K-25, compared to that

in films contaiaining S-45 or K-50. In Figure 3.1% and Figure 7.l?b, we cee thnt the

retardation effect of the larger silica particles is ver)) sirnilar throughout the interdiffusion

process. In conuast, we find that the smaller particles exhibit a much larger influence on

the early-tirne diffusion of the polymer. Thus we suspect that the obstacle effect of the

small silica particles also contributes to the reduction of the polymer diffusion rate, and

that this effect is most pronounced at srnall extents of mixing.

Table 3.7. Volume content of PBMA polymers (vol %) near the filler surface. relative to

the total polymer volume in latex films. The filler content is 40 wt % for al1 the

calculations.

Polymers within RG

(4.7 nm thickness)

(d = S-12 12 nm) I

Polymers within ZRG

(9.3 nm thickness)

K-25

(d = 25 nm)

K-50

(d = 50 nm)

48-8

20.4

100.0

48.2

3-5. Conclusions

We employed the fluorescence resonance energy transfer (FRET) technique to

measure the rate of polymer interdifision in the presence of colloidal silica, which has a

large surface area to volume ratio. The initial efficiency of energy transfer is constant in

the presence of different amounts of 50 nm SiO?, but it decreases with increasing arnounts

of 12 nm and 25 nm SiO?. This resuIt indicates that 12 nm and 25 nm SiO2 reduces the

intedaciai ÿrea ortwren donor- arid acczptor-tabeled particles in the newly fomed film

with increasing amount of filler, and they prevent the latex particles from the initiai

codescence. The extent of retardation in latex films with 50 nm SiO? is srnall, whereas

that of 12 nm and 25 nm SiO2 is much more pronounced on slowing the rate of polymer

interdifision. The mean apparent diffusion coefficient (D,,) of the latex film with 12 nm

SiOl (40 wt%) is 10 times smaller than the latex film without filler. Because of sample

brittleness. we could study films containing up to 40 wt % SiO? content. In these films,

the binder is present as the continuous phase.

3-6. References - -

1. (a) Kraus, G. Rrtbber Chem. Tech., 1978.51, 297. (b) Rharbi, Y.; Cabane, B.; Vacher,

A.; Joanicot, M.; Boue. F. Eitrophys. Lett. 1999,46,472.

2. Bourgeat-Lami. E.; Lang, J. J. Colloid interface Sci. 1998, 197,293.

3. "SNOWTEX" Brochure; Nissan Chemical Industries, 1997.

4. Imai, T.; Miyake, J.; Nojima, K. U.S. Patent 5,275,846, 1994.

5. (a) Asano, S.; Ohashi, H.; Kondo, H.; Nojima, K.; Imabeppu, K.; Sakaki, M.; Suzuki,

E. U.S. Patent 5,952,051, 1999. (b) Ogawa, S.; Senoh, H.; Andoh, M.; Nomun, H.

U.S. Patent 5,750,200, 1998. (c) Imabeppu, K.; Asano, S.; Ohashi, H.; Nojima, K.;

Suzuki, E.; Sakaki, M. U.S. Patent 5,741,584, 1998. (d) Asano, S.; Ohashi, H.; Kondo,

H.; Nojima, K.; habeppu, K.; Sakaki, M.; Suzuki, E. U.S. Patent 5,670,242, 1997. (e)

Ogawa, S.; Senoh, H.; Andoh, M.; Nomura, H. U.S. Patent 5,576,088, 1996.

6. (a) Ng, D.; Guillet, J. E. Macrarnolec~~les 1982, 15, 728. (b) Zhao, C. -L.; Wang, Y.;

Hniska, 2.; Winnik, M. A., Macromolecules, l990,23,4082.

7. Feng, J.; Winnik, M. A. Macromolecules 1997,30,4324.

-

8. Brandmp, J.; Immergut, E. H. Polymer Handbook, 3rd ed; Wiley, New York, 1989, p

VV396.

9. Koppel, D. E . J. Cheni. Phys. 1972.57.4814.

10. Provencher, S. W. J. Chem. Phys. 1976,64,2772.

11. O'Connor, D. V., Phillips, D., Time-correlated Single Photon Coiinting, Acridemic.

3 e w York, 1984.

12. "KIebosol" Brochure; Clziant, 1999.

13. Feng, J.; Odrobina, E.; Winnik, M. A. Macromolecriles 1998.31, 5290.

14. Pham, H. H.; Farinha, J. P, S.; Winnik, M. A. Itlncrornulecriles, submitted for

piiblication.

15. Whitlow, J.; Wool, R. P. Macromolecriles 1991,24,5926.

16. Dhinojwaia, A.: Torkelson, J. Macromoleciiles 1994,27,48 17.

17. (a) Liu, Y.; Feng, J.; Winnik, M. A. J. Chem. Phys. 1994, 101, 9096. (b) Liu. Y.;

Winnik. M. A. Macroniol. Chem. Symp. 1995,92,321.

18. Yekta, A.; Duhamel, J.; Winnik, M. A. Chem. Phys. Lett. 1995, 235. 1 19.

19. Crank, J. The Marhemarics of Difitsion: Clarendon: Oxford. U . K., 1974.

20. Zhao, C. -L.; Wang, Y.; Hruska, 2.; Winnik, M. A. Macromolecules 1990.23.4082.

2 1 . Farinha, I. P. S.; Martinho, J. M. G.; Yekta, A.: Winnik, M. A. Macrontoleciîles 1995.

28,6084.

22. (a) Wang, Y.; Zhao, C.; Winnik, M. A. J. Chem. Phys. 1991, 95, 2143. (b) Wang, Y.;

Winnik, M. A.; Haley, F. J. Coat. Technol. 1992, 64, 5 1. (c) Kim, H.; Wang, Y.;

Winnik, M. A. Polymer 1994, 35, 1779. (d) Kim. H.: Winnik, M. A. Macromolecriles

1994,27, 1007.

23. Wang, Y.; Winnik, M. A. 1. Phys. Chern. 1993,97,2507.

24. Ferry, J. D. Viscoelastic Proper~ies of Polymers, Wiley, New York, 1980.

25. Odrobina, E. Ph. D niesis, University of Toronto, Toronto, Canada, 2000.

26. Fujita, H. Fonschr. Hochpoiyrn.-Forsch. 1961.3. 1 .

27. (a) Tsagaropoulos, G.; Eisenberg, A. Macromolecules 1995, 28, 396. (b)

Tsagaropoulos, G.; Eisenberg, A. Macromolecriles 1995,28,6067.

28. Brandrup, J.; Immergut, E. H. PoZymer Handbook, 3rd ed; Wiley, New York, 1989, p

VIV35.

29. Eckenley, S. T.; Helmer, B. J . J. Coat. Technol. 1997.69.97.

30. (a) Dufresne. A.: Cavaille J. Y.; Helbert, W. Polymer Composites 1997, 18, 198. (b)

Hajji, P.; Cavaille J. Y.; Favier. V.; Gauthier. C.; Vigier, G. Paiymer Composites

1996, 27,612.

31. (a) O'Brien. J.; Cashell. E.; Wardell, G.; McBrierty. V. J. Macronzolecrtles 1976, 9.

653. (b) Douglass, D.; McBrierty, V. J. PoZynt. Eng. Sci 1979, 19. 1054. (c) Ito. M.;

Nakamura, T.; Tanaka, K. J. Appl. Polynz. Sei. 1985. 30. 3493. (d) Dutta. N.:

Choudhury, N.: Haidar, B.; Vidal, A.; Donnet. J.; Delmotte, L.; Chezeau. I. Poiymer

1994,35.4293. (e) Pliskin. 1.: Tokita. N. J. Appl. Po ly~ . Sci. 1972, 16.473.

4. FUTURE WORK

4-1. Effect of Modification on the Polymer Diffusion Rate of Latex Polymen in the

Presence of Mineral Fiilers

We have shown in this study that minera1 fillers affect the polymer diffusion rate

in latex films. In the actuai paper ioatings, most of the latex polyrners arc cxboxylzted

due to their high stability in dispersions. Carboxyl groups also play an important role of

binding mineral fillers. Thus it is interesting to examine the effect of carboxyl group on

the polyrner diffusion rate in the presence of various types of rnineral fillen.

Kim and Winnik ' have exarnined the effect of carboxyl group on the polymer

diffusion rate in poly(buty1 methacrylate) latex films using the FRET technique. They

found that even the carboxyl content increases up to 6 mole %. they observed that

polymer diffusion stiIl takes place near the late.u/latex interface.

The latex polymer that we used for this study is expected to have a substantial

branching as descnbed in Chapter 2. but the gel content is quite low ( l e s than 10 wt %).

The latex used for paper coatings has a higher gel content, depending on the types of

coated paper and the types of the printing method. Therefore it is important to know how

the degree of cross-linking affects the polymer diffusion rate in the presence of rnineral

fillers.

Tarnai et al. exarnined the effect of cross-linking on the polymer diffusion rate.

Surprisingly, polymer difision takes place even in the latex film with a 100 wt 76 gel

content. Based on their results, we expect that polymer diffusion occurs in the

commercial coating process.

4-2. Effect of Other Constituents (Water-soluble Polymen, Thickeners) on the

Polymer Diffusion Rate of Latex Polymers in the Presence of Mineral Fiilers

Although latex polymer is extensively used as a pigment binder for various types

of coated papers, the amount of latex used is very small. Other constituents such as water-

soluble polymen [starch, casein, poly(viny1 alcohol)] and thickeners (carboxymethyl

cellulose, associative thickeners) are also used for the actual paper coatings. It is

interesting to examine the effect of those constituents durhg latex film formation.

4-3. Application of FRET Technique to Styrene-butadiene and Styrene-acrylate

Latex systems

We have used acrylic latex polyrners throughout this study. In paper coatings.

styrene-butadiene latex is die most commoniy used syntlietic bindzr due m high

performance for commercial printing. Styrene-acrylate copolymer latex is dso used for

paper coatings. Both synthetic latex polymen have styrene in polymer backbones.

There is a problern that styrene has a tluorescence background thût affects energy

transfer from phenanthrene to anthracene molecules. It is dificuli to remove the

contribution of styrene to analyze the energy uansfer data. We hope that a new pair of

fluorescence donorhcceptor pair will be developed in the hiture,

4-4. References

1. (a) Kim, H-B.; Wmg. Y.; Winnik, M. A. Pofymer 1994, 35, 1779. (b) Kim, H-B.;

Winnik, M. A. Mucrurnolecules 1994, 27. 1007. (c) Kim, H-B.; Winnik, M . A.

Macrontalecrtles 1995,28,2033.

2. Tamai, T.; Pinenq, P.: Winnik, M. A. Macrumolecriles 1999.32. 6102.

3. Thayer, A. M . Chem. Eng. News 1993,71,28.