J. Coat. Technol. Res.

, 5 (3) 285–297, 2008

DOI 10.1007/s11998-008-9096-6

The diacetone acrylamide crosslinking reaction and its influence

on the film formation of an acrylic latex
Nicola Kessel, Derek R. Illsley, Joseph L. Keddie

Ó FSCT and OCCA 2008

Abstract Waterborne colloidal polymers (i.e. latex) (as determined by the temperature relative to the
represent a promising alternative to organic solvent- polymer glass transition temperature), and the evapo-
based systems in coatings applications. The develop- ration rate (as controlled by the relative humidity).
ment of mechanical strength and hardness is often There is evidence that surfactant exudation can be
enhanced by chemical crosslinking that creates a three- triggered by the particle deformation that occurs at
dimensional network. If extensive crosslinking occurs film formation temperatures well above the glass
within the particles prior to their coalescence, however, transition temperature.
interdiffusion will be prevented. A weaker product will
result. We have explored the inter-relationship Keywords Crosslinking, Diacetone acrylamide,
between coalescence, crosslinking, and surfactant exu- Film formation, Keto-hydrazide, Surfactant
dation in an acrylic latex containing diacetone acryl-
amide exploiting the ‘‘keto-hydrazide’’ crosslinking
reaction. The complementary use of spectroscopic Introduction
techniques on a model system determined that the
crosslinking reaction yields an imine, not an enamine A growing interest in and uses for waterborne polymer
as has been proposed in some literature. Gel fraction coatings have been driven by increasing environmental
measurements were used to probe the rate and amount pressures, especially the need to comply with legisla-
of crosslinking and identified a slower rate in larger tion limiting volatile organic compounds and emissions
particles, suggesting that the transport of crosslinking associated with the use of solventborne polymer
agent is rate-limiting. The keto-hydrazide reaction was systems.1–4 Waterborne colloidal polymers (i.e.,
found to be acid catalyzed and favored at lower water ‘‘latex’’) are used in a wide range of applications,
concentration. Measurement of the latex pH relative to including adhesives, additives for paper, paints and
the polymer mass fraction during film formation coatings, printing inks, cosmetics, synthetic rubbers,
clarified the expected point of onset for crosslinking floor polishes and waxes, sealants, and drug delivery
in relation to particle packing. Atomic force micros- systems.4 Colloidal particles may be tailored to exhibit
copy was used to follow surface leveling relative to the a desired morphology, composition, particle size dis-
competing influence of crosslinking. The rate and total tribution, surface groups, and molecular weight.5 In
amount of surfactant exudation were found to be turn, these particles can be manipulated during the film
influenced by crosslinking, particle deformability formation process in order to create coatings that meet
the desired end-use requirements. Here, we consider
2007 Roon Awards winner, presented at 2007 FutureCoat! the effects of several inter-related aspects of film
Conference, sponsored by Federation of Societies for Coatings formation4 on final film structure and properties.
Technology, October 3–5, 2007, in Toronto, Ontario, Canada. The incorporation of crosslinking chemistry in
waterborne coatings is recognized to provide a parti-
N. Kessel (&), D. R. Illsley
Sun Chemical Ltd, Orpington, Kent, UK
cularly effective means of enhancing the mechanical
e-mail: [email protected] strength, chemical stability, and solvent resistance of
the final film.6–11 Recently, a system based on the
J. L. Keddie reaction of a carbonyl pendant group on the dispersed
University of Surrey, Guildford, Surrey, UK polymer backbone with a diamine, specifically where

285
J. Coat. Technol. Res., 5 (3) 285–297, 2008

this amine is a dihydrazide, has been the subject of crosslinking once they have formed a coating on the
increased interest.12–20 This chemistry, termed the substrate. This is because molecular interdiffusion
keto-hydrazide reaction, offers the advantage of fast, between neighboring particles, which is essential for
ambient-temperature crosslinking in functionalized the generation of latex film strength,33–35 must take
acrylic latex, when the dihydrazide is incorporated in place prior to the crosslinking reaction.7,11,30–32
the aqueous phase of the latex. Anecdotal evidence15 Strongly crosslinked particles are unable to inter-
also suggests that an added benefit of keto-hydrazide diffuse.24–29 In systems having an external crosslinker
chemistry, particularly in printing ink applications, is that is dissolved in the aqueous phase, its partitioning
the enhancement of adhesion, possibly through hydro- character between the polymer and water imparts a
gen bonding at the substrate interface, or the formation further complication. If, as the film dries, this cross-
of permanent covalent bonds between the dihydrazide linker does not readily dissolve within and uniformly
and carbonyl groups at the treated polymer substrate distribute within the polymer particles, then localized
surface. crosslinking may result. Although the competing
Our system of interest consists of an acrylic latex effects of crosslinking and interdiffusion in various
containing diacetone acrylamide pendant groups waterborne systems have been studied, there are no
(Fig. 1a) on the polymer backbone that reacts with such reports on keto-hydrazide coatings, until now.
an adipic dihydrazide di-functional crosslinker (Fig. 1b). Moreover, we address the fact that pH evolves during
It is conceivable that this reaction will yield either an film formation, and it has a catalytic influence on
imine12–14,17,18 or an enamine16,21–23 or a mixture of crosslinking, thus determining when the reaction
both.16,21–23 The precise mechanism has not been develops as the film dries.
reported, and there has been speculation in the Note that the crosslinking reaction in our system can
literature that there is an enamine product.16 This lack occur between diacetone acrylamide (DAAM) groups
of clarity has motivated this present work. within the same particle or it can occur at the interface
An understanding of the fundamental reaction between particles. Intra-particle crosslinking will
mechanism is essential to optimize this coating formu- increase the stiffness and strength, but without entan-
lation for specific applications and for the development glements at particle/particle interfaces, the films will
of new materials. In this work we use model com- lack cohesion. Inter-particle crosslinking will generate
pounds (Figs. 1c and 1d) to establish the product of the more cohesion, however, even in the absence of
reaction, in order to simplify the spectroscopic analysis. entanglements and interdiffusion.
In waterborne systems the time in the film formation Through the years some strange and unusual obser-
process at which crosslinking occurs can have a vations of surfactant ‘‘islands’’ and ‘‘blobs’’ on coating
profound effect.7,11,24–32 To achieve maximum film surfaces have gone unexplained and remain mysteri-
strength, particles should remain relatively free of ous.36–41 It is known that the presence of surfactants
crosslinks in the dispersion but undergo extensive profoundly influences the mechanical strength,

(CH2) 6CH3
(a) (c) NH NH

H3C(H2C) 3 HC
O CH3 O O
CH3
H 3C(H 2 C) 4
H2 C
-
NH CH3
Option (1) Enamine
Polymer CH3
CH3

O
O O
(CH2)6CH3
H2N NH2 H 2N NH
N NH
NH (CH2 )4 NH
(CH 2) 6 CH3
H3C(H 2C)4
O
CH3
(b) (d)

Option (2) Imine

Fig. 1: Molecular structure of the model reactants. (a) Diacetone acrylamide pendant groups on the polymer backbone; (b)
adipic dihydrazide crosslinker; (c) 2-heptanone, modeling the ketone group in the diacetone acrylamide pendant group; and
(d) octanoic hydrazide, modeling the amine group in adipic dihydrazide. Reaction between 2-heptanone (c) and octanoic
hydrazide (d) could yield option (1), an enamine, and/or option (2), an imine

286
 J. Coat. Technol. Res., 5 (3) 285–297, 2008

durability, adhesion, blocking, gloss, and permeability Table 1: Molar ratios of ADH crosslinker to DAAM
of the final film.42–46 During this research it was groups in standard latex and the number of DAAM
discovered that the properties of the latex particles, groups that can be crosslinked as a function of the wt%
the stages and conditions of film formation, and crosslinker in the latex
crosslinking all influence the rate of surfactant exuda-
tion. The processes are inter-related and should not be ADH wt% Number of moles Number of crosslinkable
considered in isolation. added to ADH:1 mole DAAMs per copolymer
latex DAAM molecule

1 0.37 6
Experimental 0.5 0.18 3
0.2 0.075 1
Reagents

Butyl acrylate (BA), methyl methacrylate (MMA), (10 mL) of the ADH (10, 5, or 2 wt% aqueous
DAAM, methacrylic acid (MA), styrene (St), acrylic solution) crosslinking agent were thoroughly stirred
acid (AA) ammonium persulfate (APS), adipic dihyd- into portions of the latex. These concentrations corre-
razide (ADH), octanoic hydrazide, and 2-heptanone spond to the molar ratios summarized in Table 1.
were used as supplied from Sigma–Aldrich Chemicals. Mono-phasic latices with solids content of 40% were
Sodium lauryl sulfate (Texapon K-12) was used as prepared with differing Tgs and particle sizes. The
supplied from Henkel. Acetone, methanol (HPLC ‘‘standard’’ composition of latex had a Tg of 52.5°C
grade), water (HPLC grade), ammonium hydroxide, and was made with particle sizes of 80, 150, and
hydrochloric acid, and litmus were used as supplied 300 nm. A latex with a Tg of 118°C was made using St
from Fisher Scientific Ltd. (46.5 wt%), MMA (46.5 wt%), DAAM (5 wt%), and
MA (2 wt%) and will hereafter be referred to as the
‘‘high Tg latex’’. A latex with Tg of –2.4°C was made
Preparation of latices using BA (69.2 wt%), MMA (21.7 wt%), DAAM
(7.2 wt%), and MA (1.9 wt%) and will hereafter be
The latices were prepared by a starve-feed emulsion referred to as the ‘‘low Tg latex’’.
polymerization process using BA, MMA, DAAM,
MAA, and St monomers emulsified with sodium lauryl
sulfate as the surfactant. APS was used as the initiator. Fundamentals of keto-hydrazide crosslinking
The reaction flask (1 L) was charged with deionized
water and surfactant and heated to 80°C via a The fundamentals of the keto-hydrazide reaction were
thermostatically controlled heating mantle. Care was studied using model compounds as previously
taken to avoid direct surface contact between the flask described. Solutions of 2-heptanone (0.1 M) and octa-
and heating mantle in order to prevent scorching of the noic hydrazide (0.1 M) were prepared in methanol and
flask contents. aliquots combined. The reaction solutions were
A preprepared monomer ‘‘seed’’ mixture, contain- analyzed using Fourier transform IR spectroscopy
ing all five monomers, was added to the flask, while (Mattson Research Series), nuclear magnetic reso-
stirring, and the temperature was allowed to stabilize. nance spectroscopy (Jeol EX90), and gas chromato-
Then an initiator ‘‘seed’’ was similarly added to the graphy mass spectroscopy (Thermo Finnigan Trace) in
reaction vessel. The mixture was left for approximately order to establish the nature of the reaction product
15 min to allow the latex ‘‘seed’’ to develop, before and the catalytic influence of pH.
commencing the monomer feeds. Monomer was fed
into the vessel at a rate of 1.5 mL/min. Initiator was
split into six equal portions and then added at half- Chemical crosslinking during film formation
hour intervals during the monomer feeds. Throughout
the process the reaction temperature was maintained The change in surface pH of drying films was measured
at 80°C; however, once the feeds were completed, the using pH indicator paper at the surface as a function of
temperature was raised to 83°C and allowed to time until the touch-dry point. Gravimetric gel fraction
stabilize for 15 min. A final portion of initiator was measurements were performed on 1 lm films bar-cast
then added and the reaction temperature was main- on silicon substrates via the extraction of the soluble
tained at 83°C for a further 75 min, after which the portion using acetone. The change in percentage gel
latex was cooled. The latex was filtered through a mesh with drying time was measured as a function of particle
in order to remove large aggregates. size, film thickness, and level of crosslinker.
Particle size was controlled through the surfactant An atomic force microscope (AFM) (Veeco Dimen-
concentration. Portions (100 g) of filtered latex were sion 3100) using ultra sharp silicon tips, of resonant
decanted and the pH of the latex was adjusted to 8.5 frequency 130–250 kHz, and spring constant 48 N/m,
using 25 wt% ammonium hydroxide solution. Aliquots was used in intermittent contact (tapping mode) to

287
J. Coat. Technol. Res., 5 (3) 285–297, 2008

probe the extent and rate of particle flattening of latex

1710 cm–1

1628 cm–1
1670 cm–1
films cast on to glass microscope slides. Unless stated
otherwise, film formation was in still air at a tempera- 1.6
ture of 22°C. 1.5
1.4
1.3
1.2

Absorbance
1.1
Characterization of surface residues evolved during 1.0
latex film formation 0.9
0.8
0.7
0.6
The solubility of the surface residue on latex films was 0.5 4
established by probing its solubility by comparing 0.4
3
0.3
AFM scans before and after rinsing with water. X-ray 0.2
2
photoelectron spectroscopy (XPS) provided surface 0.1 1
sensitive chemical information. Cast films were ana- 2100 2000 1900 1800 1700 1600
lyzed in order to determine the chemical composition Wavenumbers (cm–1)
of the residue and verify the nature of the film surface.
Fig. 2: An overlay of IR spectra showing the presence of an
absorption band at 1670 cm–1 in the anticipated region for
C=N imine. Spectrum 1: 2-heptanone; spectrum 2: excess
Factors influencing surfactant exudation during
2-heptanone plus octanoic hydrazide; spectrum 3: octanoic
film formation hydrazide; and spectrum 4: excess octanoic hydrazide plus
2-heptanone
During film formation, the humidity was adjusted
through the use of saturated salt solutions in a closed
container, and through air flow above the film. Film The 13C NMR spectrum of 2-heptanone exhibits a
formation at a temperature of 9°C was achieved in a peak at around 210 ppm, due to the ketone C=O
refrigerator. Surfactant exudation was studied using group. In comparison, the amide ester carbonyl in
AFM height and phase images, and probed as a octanoic hydrazide is observed at around 175 ppm. On
function of latex Tg, temperature, crosslinking (with reaction, the peaks, due to the model component, are
and without ADH), humidity, and evaporation rate. depleted in acidic and neutral solutions as would be
expected. Two new peaks are observed at around 160
and 150 ppm. The chemical shift for C=N is expected
Results and discussion in the 165–145 ppm range,47 and so this indicates that
the product is an imine. There are no peaks present in
Fundamentals of keto-hydrazide crosslinking the 140–100 ppm region, where C=C would be
expected to be observed. There is evidence in the
FTIR analysis revealed that 2-heptanone exhibits a literature to indicate that the two peaks observed in the
strong carbonyl peak at 1710 cm–1 and octanoic NMR spectrum are most probably due to rotation
hydrazide exhibits a strong band at 1628 cm–1 attri- about the N–N bond, altering the field on the amide
butable to the amide carbonyl and N–H stretching. On carbonyl47,48 and thereby giving rise to two rotamer
reaction, it would be expected that the intensity of the forms, as illustrated in the inset of Fig. 3.
ketone carbonyl and the N–H stretching and bending Analysis of the product of the model reaction using
peaks would be reduced, as they are consumed during GC/MS Chemical Ionization highlighted the presence
crosslinking. of a molecular ion of 255 amu (M + 1), as would be
On reaction of the model components, a new peak anticipated from options (1) and (2) in Fig. 1.
was observed in the region 1670 cm–1. Figure 2 shows GC/MS analysis of alkaline solutions showed the
that with an excess of 2-heptanone (2:1 molar ratio), a presence of strong peaks for the reactants, 2-heptanone
new peak is observed adjacent to the ketone carbonyl. and octanoic hydrazide. In contrast, neutral and acidic
As the level of octanoic hydrazide is low, indicated by solutions show a dominant product peak and dimi-
the low intensities of the relevant peaks, then this new nished reactant responses. Comparison of the initial
peak is unlikely to be related to the hydrazide. For a rates of the reactions under different pH conditions
reaction mixture containing equimolar proportions of clearly shows that the chemical reaction rate increases
reactants, the ketone carbonyl is partially masked by with decreasing pH, as illustrated in Fig. 4. We
the new peak. In the presence of excess octanoic conclude then that the reaction is acid catalyzed.
hydrazide (2:1 molar ratio), the new peak appears It is known that in a carbonyl group the electronega-
adjacent to the amide peak. This would seem to rule tive oxygen withdraws electrons from the carbon
out any possibility that the new peak is the result of an bond resulting in the carbonyl carbon attaining a
amide solution shift. Critically, the new peak at relative positive charge.49 The electrophilic character
1670 cm–1 is in the correct region for a reaction of the carbon therefore significantly enhances its
product containing C=N (1690–1630 cm–1).47 susceptibility to nucleophilic attack by electron-rich

288
 J. Coat. Technol. Res., 5 (3) 285–297, 2008

R2
NH R2
Rotation
H3C NH N

N O O
R1 CH3
R1
No methyl group Methyl group influences
influence on carbonyl the field of carbonyl
R1 = CH3(CH2)4

R 2 = CH 3 (CH 2) 6

Amide C=O group C=N group

Depleting
ketone C= O

250 200 150 100 50 0 PPM

13
Fig. 3: Carbon nuclear magnetic resonance spectrum for reaction product between 2-heptanone and octanoic hydrazide
under acidic conditions. The spectrum shows the depleted C=O peak at 210 ppm and four peaks in the 150–180 ppm region
attributed to the amide C=O group and the C=N group. Two peaks appear for each of these groups due to the possibility of
two rotamer forms (structures shown inset)

1.00 The susceptibility of the carbonyl group to nucleo-

Octanoic hydrazide
0.90 2-Heptanone philic attack is reduced by its attachment to electron-
0.80
releasing alkyl groups that reduce the degree of
positive charge on the carbon.21,49 Therefore, as ketone
Initial rate (micro moles/s)

0.70
structures comprise two alkyl groups attached to the
0.60 carbon, compared to alkyl and hydrogen in an alde-
0.50 hyde molecule, then ketones would be expected to be
less reactive than aldehydes. Indeed, the presence of
0.40
two alkyl groups would also be expected to incur some
0.30 steric hindrance to attack on the carbon. It has been
0.20 reported elsewhere that primary amines produce
0.10
imines on reaction with aldehydes and ketones,
although ketones react more slowly than aldehydes,
0.00
3 4 5 6 7 8 9 10 sometimes requiring higher temperatures and longer
pH reaction times.51,52 In addition, the progression of the
reaction is inhibited by the formation of water, and the
Fig. 4: Graph showing the relationship between solution literature identifies that, in reaction studies, the water
pH and initial reaction rate for octanoic hydrazide and must be removed either by distillation or with a drying
2-heptanone agent.51 This latter behavior is of particular interest in
the case of the waterborne latex. It would seem
reasonable to propose then that the presence of water
reagents.21,23,49–52 Moreover, the part of the molecule will ‘‘block’’ the crosslinking reaction until after
immediately surrounding the carbon is flat, with sufficient drying has occurred. The water content of
oxygen, carbon, and the two atoms directly bonded the film reaches a critical level where the equilibrium
to the carbon lying in the same plane. Consequently, of the keto-hydrazide is shifted in favor of crosslinking.
this part of the molecule is open and relatively While primary amines and carbonyl groups will form
unhindered from attack from above or below, in the imines, it is known that enamines can be formed on the
direction perpendicular to the plane of the carbonyl reaction of aldehydes and ketones with secondary
group.22,23 As nitrogen exhibits electronegative char- amines.21–23,51 Furthermore, for carbonyl compounds
acter, and will readily donate a pair of electrons, it is with an a-proton, a tautomeric equilibrium can exist
expected that the carbonyl group will be prone to react between the imine and the enamine, but in which the
with amines. The literature maintains that amines will imine form predominates.23,53 Despite the abundance
react with both aldehydes and ketones, with the nature of literature supporting imine formation from reaction
of the product being dependent on the species involved between the carbonyl group and primary amines,
and reaction conditions.51 specific reports on the keto-hydrazide are few.

289
J. Coat. Technol. Res., 5 (3) 285–297, 2008

The literature indicates that the reaction of a Chemical crosslinking during film formation
nucleophile with a carbonyl compound is often cata-
lyzed by acid.21,23,51 Protonation of the carbonyl Gravimetric measurement of latex solids fraction as a
oxygen further increases the positive charge at the function of drying time during film formation was used
carbon, making it much more susceptible to nucleo- as a rough indicator of the point of particle close
philic attack. So, in the case of a carbonyl compound, packing. Parallel measurements of the wet film’s pH
nucleophilic addition will be favored by high acidity. with indicator paper revealed an increase in acidity
However, protonation of the amine produces a spe- over time, reaching a pH of 6 after 6 min of drying. In
cies lacking unshared electrons, resulting in the loss a complementary experiment, litmus was added to the
of nucleophilic character. Thus, with respect to the wet latex dispersion prior to film formation. This
amine compound, addition is favored by low acidity. In experiment clearly revealed that the film became
reality, an efficient reaction condition would be a significantly acidic when dried. The decreasing pH
compromise, with the exact requirements depending during film formation is attributed to the loss of
on the reactivity of the carbonyl, and the basic ammonia through evaporation. Significantly, latex pre-
character of the reagent. neutralized with nonvolatile amine did not exhibit the
Based on the experimental results of this research pH change.
work and literature evidence, we propose the mecha- Combining the pH measurement with the solids
nism shown in Fig. 5 for the keto-hydrazide reaction in fraction measurements (Fig. 6), we are then able to
our latex system. establish the solids fraction when crosslinking is
In an acid environment the process commences with favored. The figure shows that the pH decreases below
the formation of a bond between H+ and the carbonyl 7 when a solids content of approximately 68 wt% is
oxygen. The nucleophilic amine group on the hydra- achieved. Depending on the particular packing of the
zide molecule attacks the electrophilic carbonyl particles, this solids content would correspond to the
carbon, resulting in the formation of a C–N bond. point of contact between particles. (For mono-size
Proton transfer occurs from the positively charged hard particles, the maximum solids fraction at close
nitrogen to form a bond between H+ and the oxygen, packing is 74 vol%.) The gas chromatography data
yielding a carbinolamine. Cleavage of the positively clearly showed that the chemical reaction rate in-
charged H2O from the central carbon then occurs, creases substantially below a pH of 7. Therefore,
forming a resonance stabilized intermediate. The crosslinking in the film is favorable only near and after
nitrogen stabilized carbocation is the conjugate acid the point of close packing. Polymer interdiffusion
of the imine and transfer of the hydrogen atom, likewise cannot occur until particles have made
attached to the nitrogen, to water yields the imine. physical contact, of course. Thus, interdiffusion and

H Carbinolamine Oxygen protonated carbinolamine

Polymer O Polymer O+ Polymer CH3 R NH2 Polymer CH3 R NH2

+
H +
CH2 C CH2 C CH2 C NH2 CH2 C NH

CH3 CH3 +
OH OH2
..
H2N R
..
NH2
- H2O

Polymer CH3 R NH2

+
CH2 C ..
NH

Polymer CH3 R NH2 - H+ CH3

Polymer R NH2
CH2 C N
+
CH2 C NH
Imine
stabilized ca
Nitrogen stabili bocation
arbocation

Fig. 5: Proposed mechanism for the keto-hydrazide reaction within waterborne coating yielding an imine—not an enamine
as has been proposed elsewhere

290
 J. Coat. Technol. Res., 5 (3) 285–297, 2008

9 molecule. In order to investigate this observation

8 further, films subjected to vacuum drying rather than
drying in air were measured also. An increase of
7
around 2–3% (not shown) in the final gel fraction was
pH

6
t = 0.5 min found, but 100% gel was still not attained. These
5 results indicate the presence of some residual water
t = 6 min
presence in air-dried films.
4
30 40 50 60 70 80 90 100 110 Possible—but unlikely—explanations for why the
wt% Solids maximum gel fraction is around 90% are that some
macromolecules do not contain DAAM or that some
Fig. 6: Graph showing the change in pH for a thin latex film molecules only have one crosslink. Another possibility
during film formation with respect to change in wt% solids. is that there is some intra-molecular crosslinking in
(As the densities of the water and copolymer are nearly which two DAAM groups on the same molecule are
identical, wt% will be very close to vol%) reacted with an ADH crosslink molecule. Recent
simulations of gel fraction,32 which consider the effects
crosslinking will probably occur simultaneously. As of molecular weight polydispersity and the number of
already noted there is an optimum pH to achieve the crosslink units, offer the most likely explanation.
fastest reaction rate, and it is expected that the acidity Statistical variation means that some polymer mole-
in the polymer phase will be correlated with that in the cules will have greater than average crosslinks per
water phase. molecule, while others will have less than the average
The influence of the level of ADH crosslinker on the or even zero. Simulations32 show that as polydispersity
polymer gel fraction is shown in Fig. 7. Gel content increases then more crosslinked units are required to
increases rapidly over the first 15 min of film drying, achieve a given gel fraction. Thus, a lower polydisper-
after which the rate slows but never reaches 100%. The sity favors a higher gel fraction for a given number of
rate of increase in gel content and the final gel value crosslinks. The ‘‘standard’’ latex has a polydispersity of
are identical for 0.5 and 1 wt% ADH, indicating that 2.7. When 1 wt% ADH is added it is estimated that
0.5 wt% is sufficient to ensure that ADH reacts with at there can be up to six crosslinks per polymer molecule.
least one DAAM group per molecule. Hence, a film In this case it is estimated that approximately 90% gel
containing 0.2 wt% achieves a lower gel fraction, being can be achieved, as was observed in the experiments.
insufficient for the crosslinking of every polymer To achieve one crosslink per polymer molecule, one
half of a molecule of crosslinker is required for each
polymer molecule. On average each ‘‘standard’’ latex
100 polymer molecule contains 16 DAAM groups. Using
90 the previously calculated molar ratio values (Table 1),
80 the number of DAAM groups that can actually be
70 reacted may be calculated. For 1 wt% ADH added to
% Gel fraction

60 ‘‘standard’’ latex, there are 0.37 moles of ADH per

50 0.5% ADH ~ 1 µm film
1 mole of DAAM. It follows then that 37% of the
40 0.5% ADH ~ 15 µm film
available DAAM groups can be reacted, correspond-
30 0.5% ADH ~ 30 µm film
ing to 6 crosslinks per polymer molecule. In the case of
20 1% ADH ~ 1µm film
0.5 and 0.2 wt% ADH ‘‘standard’’ latex, this reduces
10 0.2% ADH ~ 1 µm film to 3 and 1 crosslinks per polymer molecule, respec-
0
00 28 57 86 115 144
tively. Clearly this indicates that there is insufficient
Time after casting (min)
ADH crosslinker present to react with all of the
available DAAM groups, and so the maximum cross-
100 link density cannot be achieved in these formulations.
90 The limited solubility of ADH in water at higher
80 concentrations also inhibits the potential to improve
% Gel fraction

70
the level of crosslink density.
60
50 Gel fraction measurements (Fig. 7) also indicate
40 that the latex particle size influences the rate of
30 80-100 nm increase of the gel content. The gel fraction of latex
20 150-200 nm comprising 80 nm particles reaches a plateau value
10 250-300 nm after around 15 min. However, the rate of gelation
0
00 28 57 86 115 144
becomes slower with increasing latex particle size (150
Time after casting (min)
and 300 nm). In order to achieve the maximum gel
fraction, the ADH must diffuse a distance on the order
Fig. 7: Gel fraction graphs showing the influence of level of the particle radius. The dependence of gelation rate
of crosslinker (top), film thickness (top), and particle size on the particle size indicates that the reaction is limited
(lower) on the rate of gel evolution in the latex film by the diffusion of ADH into the latex particles. Our

291
J. Coat. Technol. Res., 5 (3) 285–297, 2008

results indicate that the ADH is partitioned mainly in been discussed previously in this work that the keto-
the aqueous phase initially and enters the polymer hydrazide reaction is inhibited by the presence of
phase during the later stages of film formation. water. Therefore, that keto-hydrazide crosslinking in a
The film thickness of the latex dispersion was varied latex film is not expected to occur unless a sufficient
in order to probe the influence of water content on fraction of the water component has evaporated. By
the progression of the crosslinking reaction. It has studying films with different thickness, the quantity of

Increasing drying time

(a) (b)

µm µm
1.2 1.2
1.0 1.0
0.8 0.8
0.6 0.6
0.4 0.4
0.2 0.2

(c) (d)

µm µm
1.2 1.2
1.0 1.0
0.8 0.8
0.6 0.6
0.4 0.4
0.2 0.2

(e) (f)

µm µm
1.2 1.2
1.0 1.0
0.8 0.8
0.6 0.6
0.4 0.4
0.2 0.2

Fig. 8: AFM topographic images showing the dependence of film structure on polymer Tg, crosslinking, and film formation
time. The left column shows the structure after 1 h and the right column shows the structure after 24 h. (a, b) High Tg latex
plus ADH crosslinker; (c, d) low Tg latex plus ADH crosslinker; and (e, f) as-prepared low Tg latex. All images are
1.5 · 1.5 lm

292
 J. Coat. Technol. Res., 5 (3) 285–297, 2008

water per unit area required to evaporate was therefore 5.00 5.00

varied. It was observed that the onset of crosslinking is (a) (b)

delayed in thicker films, presumably due to the presence
of more water per unit area. This effect is particularly
evident in the first 20 min of drying. Furthermore, the 2.50 2.50
rate of crosslinking in thicker films appears to be slower,
which could be attributed to ‘‘skin formation’’ in the
thicker film slowing down water evaporation and pH
changes in the later stages of drying. 0 0
0 2.50 5.00 0 2.50 5.00
The AFM topographic images in Fig. 8 show that in
µm µm
films from the high Tg latex with ADH crosslinker
(Figs. 8a and 8b), particles retain their particle identity. 5.00 5.00

The surface does not exhibit significant particle flat- (c) (d)
tening, and particle boundaries remain. In the absence
of ADH, the film structure is the same (images not
shown). Hence, in this case, it is the particle rigidity 2.50 2.50
itself, and not the crosslinking, that is resisting particle
flattening and interdiffusion. In contrast, the low Tg
latex with ADH crosslinker (Figs. 8c and 8d) rapidly
forms a significantly flatter film. Interestingly, a honey- 0 0
0 2.50 5.00 0 2.50 5.00
comb structure was retained at the surface with
µm µm
boundaries apparent, as shown in Fig. 8d. This struc-
ture is not observed in crosslinker-free low Tg latex 5.00 5.00

films (Figs. 8e and 8f), which exhibited essentially (e) (f)

instantaneous flattening of the surface particles and in
which no particle boundaries are apparent. These
results show that crosslinking inhibits particle flatten- 2.50 2.50
ing and retards interdiffusion.
Film formation of the low Tg latex in the absence of
crosslinker is accompanied by the development of a
surface residue of 5 nm in thickness, which spreads 0 0
0 2.50 5.00 0 2.50 5.00
laterally as drying time increases. Figures 9a–d show
µm µm
the growth of the residues with time. At later times this
residue appeared to aggregate and reticulate into 5.00 5.00

isolated ‘‘blobs.’’ This observation prompted further (g) (h)

investigation.
Films were subjected to XPS analysis that esta-
blished the chemical composition of the residue to be 2.50 2.50
consistent with surfactant. After rinsing the film
surface with water, sodium and sulfur, attributable to
sodium lauryl sulfate, were not detected with XPS.
This is expected for water-soluble surfactant. 0 0
0 2.50 5.00 0 2.50 5.00
Figures 9e and 9f show the corresponding surface
µm µm
structure after rinsing, in which no residue is observed.
When crosslinker is added to the low Tg latex, XPS and Fig. 9: AFM micrographs (topographic images on the left
AFM analysis find no evidence for surfactant excess at and phase images on the right) of low Tg latex without
the film surface. Hence, crosslinking inhibits the crosslinker showing the evolution of surface residue
exudation of surfactant. After rinsing, surfactant con- (attributed to sodium lauryl sulfate surfactant). (a, b) Latex
tinued to exude to the surface of the low Tg latex, as film 1 h after film casting; (c, d) latex film 6 h after casting;
revealed in Figs. 9g and 9h. As surfactant was not (e, f) the same latex as c and d after rinsing; and (g, h) the
detected at the surface of the high Tg latex, particle rinsed latex after aging for a further 24 h. All AFM images
deformation is likely to be a key factor, prompting are 5 · 5 lm
further investigation discussed below.

films were subjected to drying at T < Tg of the polymer

Factors influencing surfactant exudation during prior to AFM imaging at room temperature. The
film formation topography and phase AFM images in Figs. 10a–c
show that surfactant exudation is inhibited in compar-
In order to explore the influence of particle deforma- ison to films dried at room temperature (T > Tg),
tion on the rate of surfactant exudation, low Tg latex shown already in Figs. 9a–d. When the cooled films

293
J. Coat. Technol. Res., 5 (3) 285–297, 2008

Increasing Time
5.00 5.00 5.00
(a) (b) (c)

2.50 2.50 2.50

0 0 0
0 2.50 5.00 0 2.50 5.00 0 2.50 5.00
µm µm µm
5.00 5.00 5.00

(d) (e) (f)

2.50 2.50 2.50

0 0 0
0 2.50 5.00 0 2.50 5.00 0 2.50 5.00
µm µm µm
5.00 5.00 5.00
(g) (h) (i)

2.50 2.50 2.50

0 0 0
0 2.50 5.00 0 2.50 5.00 0 2.50 5.00
µm µm µm
Increasing Humidity

5.00 5.00 5.00

(j) (k) (l)

2.50 2.50 2.50

0 0 0
0 2.50 5.00 0 2.50 5.00 0 2.50 5.00
µm µm µm

Fig. 10: AFM phase images showing the evolution of SLS surfactant to the surface of a low Tg latex with time as a function
of temperature and humidity. The age of the films in the first column is 90 min; second column 6 h; third column 3 days.
(a–c) Film formed at 9
C and 51% relative humidity; (d–f) film formed at 22
C and 13% relative humidity; (g–i) film formed at
22
C and 39% relative humidity; and (j–l) film formed at 22
C and 80% relative humidity. All AFM images are 5 · 5 lm

were raised to room temperature, the evolution of been formed at room temperature. Figure 11 shows
surfactant domains were then observed, although at a some quantitative analysis of the surfactant exudation
lower surface coverage compared to the film that had phenomenon in which the surface area of the surfac-

294
 J. Coat. Technol. Res., 5 (3) 285–297, 2008

tant regions is estimated from AFM images. The particles that readily coalesce. Holl and co-workers54
surface area of the surfactant increases approximately reached a similar conclusion when they compared
at a constant rate in all cases, but the rate is heavily surfactant exudation from two latices having different
dependent on the particle deformability. Harder Tg values. In their experiment, the differences in
particles, as obtained here when T < Tg, have a lower surfactant exudation might also have been explained
rate of surfactant exudation in comparison to softer by differences in chemical composition. In our experi-
ments, on the other hand, the same latex is being used,
0.16
so the variation in exudation rate cannot be attributed
22°C and 39% humidity
to chemical disparity. Our results could be explained
0.14 22°C and 13% humidity by a co-operative mechanism between particle defor-
0.12
mation and surfactant exudation. When there is
flattening at particle/particle boundaries, surfactant
0.1 might be pushed away from that surface.
Area (µm2)

Further experiments were conducted in which the

0.08
water evaporation rate during film formation was
0.06 adjusted through the relative humidity. Recent
modeling55 of surfactant distribution has predicted
0.04 that slower evaporation rates will lead to a more
0.02
uniform distribution of surfactant in the vertical
direction. The AFM images in Figs. 10d–l show
0 visually that low Tg latex films exhibit rates of
00 1440 2880 4320 5760 7200 8640 10080 11520
surfactant exudation that are dependent on the evap-
Time after casting (min)
oration rate. In a dry atmosphere (9% relative
0.09 humidity), wherein the water evaporation rate is
0.08 9°C and 51% humidity fastest, surfactant exudation occurs very slowly and
22°C and 39% humidty
0.07 appears as small beads on the film surface. In
comparison, a film dried at an intermediate humidity
Area (µm2)

0.06
(39%) shows a higher rate of exudation and the small
0.05
domains rapidly aggregate to form larger ‘‘blobs.’’
0.04
Finally, at high humidity (80%) and slowest evapora-
0.03 tion, the surfactant rapidly exudes to the surface,
0.02 forming extensive coverage. Figure 11 shows quanti-
0.01 tatively how the humidity affects the surfactant exu-
0 dation. This observed trend in surfactant coverage is
00 720 1440 2160 2880 3600 4320 5040 5760 6480 the exact opposite of what is expected from the
Time after casting (min) model55 that considers the surfactant transport in the
aqueous phase. However, in these experiments not
Fig. 11: The time-dependence of the surface coverage only is the evaporation rate varied (as considered in the
(in lm2) of the surfactant measured from analysis of AFM
images. Films were stored at two different relative humidi-
model), but also the hydrophilicity at the film/air
ties (13% and 39%) at 22
C (at top) and at two different tem- interface is increased with increasing relative humidity.
peratures (9 and 22
C) but similar humidities (at bottom). The surfactant could be favored at a more hydrophilic
After 5760 min, the humidity was increased from 13% to interface. Furthermore, water condensation in the
39% (top graph) surfactant layer could impart mobility to the surfactant.

5.00 5.00 5.00 5.00

(a) (b) (c) (d)

2.50 2.50 2.50 2.50

0 0 0 0
0 2.50 5.00 0 2.50 5.00 0 2.50 5.00 0 2.50 5.00
µm µm µm µm

Fig. 12: AFM phase images showing the influence of evaporation conditions on the exudation of SLS surfactant to a low Tg
latex film surface (17 h old). (a) After film formation with fast evaporation under an air stream; (b) film formation under
normal evaporation conditions (39% RH); (c) after film formation with slower evaporation obtained with a partly covered
enclosure; and (d) after film formation with the slowest evaporation, obtained within a closed container. All AFM images are
5 · 5 lm

295
J. Coat. Technol. Res., 5 (3) 285–297, 2008

In order to explore this phenomenon further, low Tg References

latex films were prepared under different drying con-
ditions. Films subjected to an air stream, keeping the 1. Kerhsaw, Y, ‘‘VOC Legislation on the Threshold of
humidity low immediately above the film, were found Change.’’ Eur. Coat. J., 4/98 230–232, 234, 235 (1998)
to exhibit little surfactant exudation (Fig. 12a). On 2. Steward, PA, Hearn, J, Wilkinson, MC, ‘‘An Overview of
removal of the air stream, surfactant domains evolved Polymer Latex Film Formation and Properties.’’ Adv.
as for a film dried under normal conditions. Films dried Colloid Interface Sci., 86 195 (2000)
3. Provder, T, Urban, MW, Film Formation in Coatings,
in partly and completely enclosed conditions (Figs. 12c
Mechanisms, Properties and Morphology. ACS Symposium
and 12d), effectively creating a higher humidity imme- Series 790, Oxford University Press, vii (2001)
diately above the film, were found to exhibit rapid and 4. Keddie, JL, ‘‘Film Formation of Latex.’’ Mater. Sci. Eng.
extensive surfactant exudation and coverage that Rep., R21 (3) (1997)
increased in area as the evaporation rate decreased. 5. El-Aasser, M, Sudol, ED, In: Lovell, PA, El-Aasser, MS
(eds.) Emulsion Polymerisation and Emulsion Polymers,
Chapter 2. Wiley, Chichester (1997)
6. Hellgren, AC, Wallin, M, Weissenborn, PK, McDonald, PJ,
Conclusions Glover, PM, Keddie, JL, ‘‘New Techniques for the Deter-
mining the Extent of Crosslinking in Coatings.’’ Prog. Org.
Fundamentals of the keto-hydrazide reaction have Coat., 43 85 (2001)
been studied utilizing a model system. The comple- 7. Aradian, A, Raphael, E, De Gennes, PG, ‘‘Strengthening of
mentary use of Fourier transform IR spectroscopy, a Polymer Interface: Interdiffusion and Crosslinking.’’ Mac-
NMR spectroscopy, and GC/MS has provided support- romolecules, 33 9444 (2000)
ing evidence to clarify—for the first time—that the 8. Athey, RD, ‘‘Additives for Waterborne Coatings Part 7:
reaction yields an imine. No data support the presence Curatives.’’ Eur. Coat. J., 816 (1996)
9. Buckmann, F, Overbeek, A, Nabuurs, T, ‘‘Self-Crosslinking
of an enamine in the product. NMR provides addi-
Surfactant Free Acrylic Dispersions.’’ Eur. Coat. J., 6 53 (2001)
tional verification suggesting rotamer behavior and 10. Pinenq, P, Winnik, MA, ‘‘Polymer Diffusion and Mechanical
hydrogen bonding of the product in solution. More- Properties of Films Prepared from Crosslinked Latex Parti-
over, the crosslinking reaction is acid catalyzed and the cles.’’ J. Coat. Technol., 72 (903) 45 (2000)
reaction rate increases as pH decreases. Hence, in our 11. Winnik, MA, ‘‘Interdiffusion and Crosslinking in Thermoset
ammonia neutralized latex the crosslinking reaction is Latex Films.’’ J. Coat. Technol., 74 (925) 49 (2002)
favored by the loss of water during drying and the 12. Hanrahan, B, Hakko, K, Murayama, T, Hotta, I, FSCT
simultaneous decrease in pH arising from the evapo- Midyear Symposium: Crosslinking for Coatings – Meeting the
ration of ammonia. Crosslinking is certainly possible Challenges. Self-Crosslinking Emulsions for Improved Chem-
once the latex particles are close packed, and there is a ical Resistance Based on the Reaction of Diacetone Acryl-
amide with Dihydrazides, Orlando (2004)
rapid increase in gel content in a short period.
13. Bakker, P, Mestach, D, ‘‘Self-Crosslinking Acrylic Disper-
The crosslinking influences the later stages of film sions for the Wood Industry.’’ Surf. Coat. Int. B: Coat. Trans.,
formation. The flattening of latex particles and inter- 84 (B4) 243 (2001)
diffusion leading to the blurring of particle/particle 14. Lee, SB, Billiani, J, Pfohl, WF, Lee, KI, ‘‘Waterborne 1K
interfaces, as observed with AFM analysis, were both Self Crosslinkable Acrylic Core-Shell Polymers.’’ Proceed-
inhibited by crosslinking. The amount of particle ings of 30th International Waterborne, High Solids and
deformability was identified as a key factor in the Powder Coatings Symposium, New Orleans (2003)
amount of surfactant exudation. Factors that increase 15. Illsley, DR, Sun Chemical Ltd, Products & New Materials
the amount of particle deformation and coalescence Confidential Laboratory Report: Current and Emerging
(e.g., higher film formation temperature, lower latex Themes in Polymer Science (2003)
16. Esser, RJ, Devona, JE, Setzke, DE, Wagemans, L, ‘‘Water-
Tg, and no crosslinking) promote surfactant exudation.
based Crosslinkable Surface Coatings.’’ Prog. Org. Coat., 36
Conversely, surfactant exudation is inhibited when par- 45 (1999)
ticles are less deformable and slower to coalesce. In 17. Mestach, D, FATIPEC Congress 25th, One Pack Crosslink-
addition, a more hydrophilic atmosphere (high humidity) ing Acrylic Dispersions for Industrial Wood Coatings, 2, 347
encourages surfactant segregation at the interface with (2000)
the film. Exudation to a ‘‘clean surface’’ can be triggered 18. Mestach, D, Ahmed M, ‘‘One Pack Crosslinking Acrylic
by raising the temperature (thus, increasing particle Dispersions for Industrial Wood Coatings.’’ Proceedings of
deformation) or by raising the humidity (thus, raising the 28th International Waterborne, High Solids and Powder
hydrophilicity at the interface). Coatings Symposium, 195 (2001)
19. Fischer, GC, Fioravanti, LC, Frazza, MS, Rohm and Haas
Company, US Patent US006,512,042B1 (2003)
Acknowledgments We acknowledge Dr Ann
20. Helmer, BJ, Murray, DL, Foster, CH, Eastman Chemical
Woollins for NMR support, Mr John Humphrey for Company, Patent International Publication no: WO99/58608
support with GC/MS, Mr Steven O’Flynn and Mr Daniel (1999)
Azzian for their practical help, Miss Junko Sano 21. Howarth, J, Core Organic Chemistry. Wiley, Chichester
(Dainippon Ink and Chemical) for the XPS analysis, (1998)
and Professor Stephen Davidson (University of Kent) 22. Patrick, G, Organic Chemistry, 2nd edn. Taylor & Francis
for his useful comments on NMR and Imine chemistry. Group, London (2004)

296
 J. Coat. Technol. Res., 5 (3) 285–297, 2008

23. Morrison, FT, Boyd, RN, Organic Chemistry, 4th edn. Allyn 39. Du Chesne, A, Gerharz, B, Lieser, G, ‘‘The Segregation of
& Bacon Inc., MA (1983) Surfactant upon Film Formation of Latex Dispersions: an
24. Zosel, A, Ley, G, ‘‘Influence of Crosslinking on Structure, Investigation by Energy Filtering Transmission Electron
Mechanical Properties, and Strength of Latex Films.’’ Mac- Microscopy.’’ Polym. Int., 43 187 (1997)
romolecules, 26 2222 (1993) 40. Tzitzinou, A, Jenneson, PM, Clough, AS, Keddie, JL, Lu,
25. Winnik, MA, ‘‘Crosslinking and Polymer Interdiffusion in JR, Zhdan, P, Treacher, KE, Satguru, R, ‘‘Surfactant
Latex Films.’’ Polym. Preprints, 44 100 (2003) Concentration and Morphology at the Surfaces of Acrylic
26. Zosel, A, Ley, G, ‘‘Film Formation of Latices with Cross- Latex Films.’’ Prog. Org. Coat., 35 89 (1999)
linked Particles.’’ Polym. Bull., 27 459 (1992) 41. Belaroui, F, Hirn, MP, Grohens, Y, Marie, P, Holl, Y,
27. Tronc, F, Liu, R, Winnik, MA, Eckersley, ST, Rose, GD, ‘‘Distribution of Water-Soluble and Surface-Active Low
Weishuhn, JM, Meunier, DM, ‘‘Epoxy Functionalised Low Molecular Weight Species in Acrylic Latex Films.’’ J. Colloid
Glass Transition Temperature Latex. 1. Synthesis, Charac- Interface Sci., 261 336 (2003)
terisations, and Polymer Interdiffusion.’’ J. Polym. Sci. A: 42. Hellgren, AC, Weissenborn, P, Holmberg, K, ‘‘Surfactants in
Polym. Chem., 40 2609 (2000) Waterborne Paints.’’ Prog. Org. Coat., 35 79 (1999)
28. Tamai, T, Pinenq, P, Winnik, MA, ‘‘Effect of Crosslinking on 43. Butler, LN, Fellows, CM, Gilbert, RG, ‘‘Effect of Surfactant
Polymer Diffusion in Poly(Butyl Methacrylate-co-Butyl Systems on the Water Sensitivity of Latex Films.’’ J. Appl.
Acrylate) Latex Films.’’ Macromolecules, 32 6102 (1999) Polym. Sci., 92 1813 (2004)
29. Daniels, ES, Klein, A, ‘‘Development of Cohesive Strength 44. Butler, LN, Fellows, CM, Gilbert, RG, ‘‘Effect of Surfactants
in Polymer Films from Latices: Effect of Polymer Chain used for Binder Synthesis on the Properties of Latex Paints.’’
Interdiffusion and Crosslinking.’’ Prog. Org. Coat., 19 359 Prog. Org. Coat., 53 112 (2005)
(1991) 45. Aramendia, E, Barandiaran, MJ, Grade, J, Blease, T, Asua,
30. Brown, WT, ‘‘Effect of Crosslinker Reaction Rate on Film JM, ‘‘Improving Water Sensitivity in Acrylic Films using
Properties for Thermoset Coatings.’’ J. Coat. Technol., 72 Surfmers.’’ Langmuir, 21 1428 (2005)
(904) 63 (2000) 46. Charmeu, JY, Kientz, E, Holl, Y, ‘‘Adhesion of Latex Films:
31. Winnik, MA, Pinenq, P, Kruger, C, Zhang, J, Yaneff, PV, Influence of Surfactants.’’ Prog. Org. Coat., 27 87 (1996)
‘‘Crosslinking vs. Interdiffusion Rates in Melamine-Formal- 47. Williams, DH, Fleming, I, Spectroscopic Methods in Organic
dehyde Cured Latex Coatings: A model for Waterborne Chemistry (3rd ed.). McGraw-Hill Book Company Limited,
Automotive Basecoat.’’ J. Coat. Technol., 71 (892) 47 (1999) UK (1980)
32. Taylor, JW, Winnik, MA, ‘‘Functional Latex and Thermoset 48. Davidson, S, Woollins, A, Sun Chemical Confidential Inter-
Latex Films.’’ J. Coat. Technol. Res., 1 (3) 163 (2004) nal Memorandum (2003)
33. Richard, J, Wong, K, ‘‘Interdiffusion of Polymer Chains and 49. Murray, PRS, Advanced Chemistry. Pan Books Ltd., London
Molecular Dyamics in Dried Latex Films.’’ J Polym. Sci. B: (1987)
Polym. Phys., 33 1395 (1995) 50. Patai, S, The Chemistry of the Carbonyl Group, Chapter 1.
34. Yoo, JN, Sperling, LH, Glinka, CJ, Klein, A, ‘‘Character- Interscience Publishers, NY (1966)
isation of Film Formation from Polystyrene Latex Particles 51. Patai, S, The Chemistry of the Carbon-Nitrogen Double
via SANS. 1. Moderate Molecular Weight.’’ Macromole- Bond, Chapter 2. Interscience Publishers, NY (1970)
cules, 23 3962 (1990) 52. Graham Solomons, TW, Organic Chemistry, 3rd edn. Wiley,
35. Sperling, LH, Klein, A, Sabisivam, M, Kim, KD, ‘‘Molecular Chichester (1993)
Basis of Healing and Fracture at Polymer Interfaces.’’ 53. Furniss, BS, Hannaford, AJ, Smith, PWG, Tatchell, AR,
Polym. Adv. Technol., 5 453 (1994) Vogel’s Textbook of Practical Organic Chemistry, Chapter 5,
36. Bradford, EB, Vanderhoff, VW, ‘‘Morphological Changes in 5th edn. Longman Scientific & Technical, London (1989)
Latex Films.’’ J. Macomol. Chem., 1 (2) 335 (1966) 54. Belaroui, F, Cabane, B, Dorget, A, Grohens, Y, Marie, P,
37. Bradford, EB, Vanderhoff, VW, ‘‘Additional Studies of Holl, Y, ‘‘Small Angle Neutron Scattering of Particle
Morphological Changes in Latex Films.’’ J. Macomol. Sci. Coalescence and SDS Desoprtion during Film Formation
Phys., B6 (4) 671 (1972) from Carboxylated Acrylic Latices.’’ J. Colloid Interface Sci.,
38. Juhue, D, Wang, Y, Lang, J, Leung, OM, Goh, CM, Winnik, 262 409 (2003)
MA, ‘‘Surfactant Exudation in the Presence of a Coalescing 55. Gundabala, VR, Zimmerman, WB, Routh, AF, ‘‘A Model
Aid in Latex Films Studied by Atomic Force Microscopy.’’ J for Surfactant Distribution in Latex Coatings.’’ Langmuir, 20
Polym. Sci. B: Polym. Sci., 33 1123 (1995) 8721 (2004)

297