European Polymer Journal 46 (2010) 1225–1236

Contents lists available at ScienceDirect

European Polymer Journal

journal homepage: www.elsevier.com/locate/europolj

Manipulation of chain transfer agent and cross-linker concentration

to modify latex micro-structure for pressure-sensitive adhesives
Lili Qie, Marc A. Dubé *
Department of Chemical and Biological Engineering, University of Ottawa, 161 Louis Pasteur Pvt., Ottawa, Ontario, Canada K1N 6N5

a r t i c l e i n f o a b s t r a c t

Article history: N-Dodecyl mercaptan (NDM) chain transfer agent and allyl methacrylate (AMA) cross-lin-
Received 9 December 2009 ker were used to manipulate latex properties in a starved seeded semi-batch emulsion
Received in revised form 16 February 2010 polymerization of butyl acrylate (BA) and methyl methacrylate (MMA) or with a third
Accepted 22 February 2010
monomer, acrylic acid (AA). Latexes with higher gel content and lower sol polymer molec-
Available online 24 February 2010
ular weight (Mw) were produced by adding only AMA. On the other hand, latexes with
lower gel content and Mw were produced by adding only NDM. In addition, at a constant
Keywords:
AMA concentration (0.2 phm), the addition of NDM (0.2 phm) decreased gel content,
Pressure sensitive adhesive
Latex polymer
increased molecular weight between cross-linking points (Mc), and decreased Mw. Adding
Emulsion polymerization more NDM (to a total of 0.4 phm) further decreased the gel content, while decreasing the
Chain transfer agent tested Mc and increasing Mw. It was also found that using higher concentrations of both
Cross-limker AMA and NDM could produce latex with similar gel content, but smaller Mc and Mw, com-
Polymer microstructure pared to the latex produced at lower concentrations of both NDM and AMA. Regarding the
influence of AA, gel content was increased and Mw was significantly decreased with an
increase in AA concentration and a decrease in MMA concentration. The performance of
the latexes was evaluated for application as a pressure-sensitive adhesive (PSA).
Ó 2010 Elsevier Ltd. All rights reserved.

1. Introduction search has been performed on manipulating polymer prop-

erties by using different reaction conditions such as
Emulsion polymers are becoming increasingly impor- initiator, chain transfer agent (CTA) and cross-linker con-
tant due to their environmental compliance [1]. They are centrations [10–15], initiator and monomer feed rates
widely used in many applications such as pressure-sensi- [10], monomer composition [11,16], seed properties, [17]
tive adhesives (PSA) [2]; coatings [3,4], paints and binders. etc.
The performance of these products greatly depends on the Among the above mentioned polymerization condi-
latex polymer properties such as composition, glass transi- tions, CTA and cross-linker concentrations are the two
tion temperatures (Tg), gel content and structure, molecu- most effective and commonly used factors for tailoring
lar weight between cross-linking points (Mc), molecular emulsion polymer micro-structural properties. It is well
weight of sol polymers (Mw), and molecular weight be- known that adding CTA can decrease Mw in the case of
tween entanglement points (Me) [5–9]. gel-free latex, or in the case of gel-containing latex, de-
Acrylic monomers are extensively used to produce la- crease gel content [11,14,18,19]. Adding cross-linker can
tex; and usually two or more monomers are copolymer- increase gel content, but decrease Mc and Mw [12]. In any
ized via a seeded semi-batch approach to provide latexes event, it is not possible to manipulate the gel content inde-
with desirable properties. With this approach, much re- pendently of the Mw. However, the effect of combining CTA
and cross-linker is not well studied [15,20]. Chauvet et al.
* Corresponding author. Tel.: +1 613 562 5920; fax: +1 613 562 5174. reported some unique phenomena caused by combining
E-mail address: [email protected] (M.A. Dubé). CTA and cross-linker in a seeded semi-batch emulsion

0014-3057/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.eurpolymj.2010.02.014
1226 L. Qie, M.A. Dubé / European Polymer Journal 46 (2010) 1225–1236

polymerization of butyl acrylate (BA) and acrylic acid (AA) tained from Sigma–Aldrich and were used as supplied. Dis-
[15]. They showed how gel content and Mw could be in- tilled deionized water (DI-H2O) was used throughout the
creased simultaneously at the same cross-linker concen- study. Ammonia (30% by weight in H2O) was obtained
tration by decreasing the concentration of CTA. However, from British Drug House. All solvents used in the polymer
the latexes produced with both CTA and cross-linker cov- characterization such as tetrahydrofuran (THF, HPLC grade,
ered a narrow gel content range from 71 to 88%. Kajtna EMD Chemicals) and toluene (99.8%, Fisher Scientific),
et al. also studied the effect of combining CTA and cross- were also used as supplied by the manufacturer. Nitrogen
linker but in a batch suspension polymerization of ethyl- gas (Linde Canada) was used to purge the reactor. PTFE
hexyl acrylate and ethyl acrylate [20]. They reported that porous membranes with pore size of 0.2 lm, for use in
at the same cross-linker concentration, the addition of gel content measurements, were purchased from Cole-
CTA could lead to a lower gel content and higher Mw. The Parmer Canada.
trend in Mw is contradictory to what was reported by
Chauvet et al. [15]. In addition, they found the polymer 2.2. Polymer latex preparation
molecular weight distribution (MWD) of the sol polymer
was very broad when CTA and cross-linker were com- All runs were performed as seeded semi-batch emul-
bined; i.e., the polydispersity ranged from 21 to 27. Mc, sion polymerizations in three stages. The first stage was a
which has a crucial influence on the properties of high batch seed production step, the second was a continuous
gel content polymer, was not reported in either study. stage involving feeding a monomer mixture and initiator
In the present study, the combined influence of CTA and stock solution, and the third stage was a batch finishing
cross-linker on various polymer micro-structural proper- stage to increase the monomer conversion to completion.
ties was studied via starved seeded semi-batch emulsion For the initiator stock solution, 0.8996 g KPS was dis-
polymerizations of BA/methyl methacrylate (MMA) and solved into 90 g DI-H2O. The amount of KPS fed to the reac-
BA/MMA/AA. Our goal was to explore a broad gel content tion was 0.26 parts per hundreds parts monomer (phm) for
range, measure the Mc, and test bulk polymer properties all runs. The monomer emulsion stock solution consisted
using dynamic mechanical analysis (DMA) in order to pro- of 6.7 g SDS dissolved into 89 g DI-H2O, along with
duce latexes with similar gel contents but differing Mc and 304.6 g BA and 33.8 g MMA. Varying amounts of the chain
Mw. The influence of AA on the polymer micro-structural transfer agent, NDM, and cross-linker, AMA, were added to
properties was also studied. the monomer emulsion stock solution prior to the reaction.
The BA/MMA or BA/(MMA + AA) weight ratio was set at A homogeneous aqueous solution containing 202 g DI-
90/10 in order to produce latexes for use as PSA. PSA are H2O, 0.048 g NaHCO3, and 0.45 g SDS was charged into a
viscoelastic materials and can stick to a surface even under one-litre Mettler-Toledo LabMax™ reactor. The stainless
very light pressure. They are usually polymerized from two steel reaction vessel was equipped with an anchor stirring
or more monomers, one with a low Tg (e.g., BA) to provide blade and various feed and sampling ports. The stirring
tackiness and the other with a high Tg (e.g., MMA and AA) speed was maintained at 250 rpm throughout the runs.
to provide cohesive strength. The performance of PSA is The reactor was purged with N2 and the reactor tempera-
usually evaluated by tack, peel strength and shear ture was increased to 70 °C within 30 min. An initiator
strength. Tack measures how easily a PSA can stick to a solution containing 0.45 g KPS and 15 g DI-H2O and a
substrate, peel strength tests how difficult a PSA can be re- monomer mixture containing 10.8 g BA and 1.2 g MMA
moved from a substrate, and shear strength reflects the were charged to the reactor to form the seed latex. The
cohesive strength of the PSA and is defined as the time temperature was then raised to 75 °C within 5 min and
needed for a PSA to fall from a testing panel under a certain the seed polymerization was continued for an additional
vertical shear force. 10 min. At that point, the initiator and monomer emulsion
Up to now, many studies have been focused on the stock solutions were added slowly to the reactor using two
influence of polymer micro-structural properties on PSA separate metering pumps (Model: Prominent gamma G/
performance [5–9]. Armed with a better understanding of 4b; Prominent Fluid Controls Inc.). The feeding times for
the combined effects of CTA and cross-linker on polymer the initiator and the monomer emulsion stock solutions
micro-structural properties, in this study, we attempted were 3.5 and 3.0 h, respectively. After the feed was com-
to relate the PSA performance to these properties. pleted, the reaction was allowed to proceed for an addi-
tional 50 min to increase monomer conversion. The latex
was then cooled to below 30 °C, and NH4OH (30 wt.%)
2. Experimental methods was added to increase the pH to about 3.5 in order to en-
hance the shelf stability of the latex.
2.1. Materials
2.3. Polymer latex characterization
The monomers including BA (reagent grade), MMA (re-
agent grade), AA (reagent grade), and allyl methacrylate Particle sizes were measured using a Dynamic Light
(AMA, reagent grade), the chain transfer agent, 1-dode- Scattering (DLS) instrument (Malvern NanoS Zetasizer)
canethiol (NDM, reagent grade), the surfactant, sodium with an angle of 176°. One drop of the latex was diluted
dodecyl sulphate (SDS, GC grade), the buffer, sodium bicar- with 2 mL DI-H2O in a 4 mL polystyrene cuvette. The re-
bonate (NaHCO3), the initiator, potassium persulfate ported diameter is an intensity-weighted average particle
(K2S2O8) and the inhibitor hydroquinone (HQ) were ob- size. The polydispersity index values (PDI) given by the
 L. Qie, M.A. Dubé / European Polymer Journal 46 (2010) 1225–1236 1227

instrument reflect a narrower distribution with PDI values To measure Mc, a new method was established. About
closer to 0.01. Latexes having a PDI smaller than 0.4 were 0.03 g of dry gel polymer was weighed and put in a
considered to have a narrow particle size distribution. 15 mL glass vial with a diameter of 10 mm, and then
The detection range of the instrument was 0.6 nm to 6 lm. 10 mL toluene was added. The bottle was then tightly
A standard gravimetric method was used to calculate sealed with a plastic cap and put on a shaker for a range
the monomer conversion based on dried polymer. Around of times (0.5 h, 1 h, 5 h, and 20 h). Next the toluene was re-
1 g of latex sample was weighed and put in an aluminum moved using a glass pipette, and the solvent-swelled gel
dish, then the dish was dried in a fume hood at room tem- was weighed. It can be seen from Fig. 1 that the amount
perature for a week and then in an vacuum oven at 30 °C of toluene absorbed by the dry gel did not change signifi-
for 2 days. cantly with swelling time. Thus, the shortest swelling time,
For the measurement of gel content, a modification of 0.5 h, was chosen for future Mc tests. To calculate Mc, the
the method by Tobing and Klein [6] was used. Around following equation was used:
0.03 g dry polymer was weighed and sealed in a PTFE n h io
coated membrane pouch. Then the membrane pouch was V 1 qp /1=2  /=2
Mc ¼    ð2Þ
immersed into 25 mL THF in a 100 mL glass bottle. The  lnð1  /Þ þ / þ v/2
bottle was then tightly capped and put on a shaker for
24 h. The membrane pouch was then removed and dried where V1 is the molar volume of toluene, 106.3 cm3/mol
in a fume hood until it reached a constant weight. The [6]; and qp is the density of polymer (1.06 g/cm3 and
weight of the remaining dry gel was taken and the gel con- 1.19 g/cm3 for PBA and PMMA, respectively [1], 1.073 g/
tent was calculated using: cm3 for the BA/MMA (90/10 by weight) copolymer). / is
the volume fraction of gel polymer in the swollen gel and
Gel content can be calculated as follows:
 
¼ mass of the dry gel=mass of initial dry polymer W p qs
/¼ ð3Þ
ð1Þ W p qs þ W s qp

The THF solution remaining from the gel content test where Wp and Ws are the weight fractions of the gel poly-
was concentrated, if necessary, and analyzed for sol poly- mer and solvent (toluene) in the swollen gel, respectively.
mer molecular weight. The solution was filtered using a qs and qp are the densities of solvent (0.8669 g/cm3) and
0.45 lm syringe PTFE filter (Pall Corporation) and then polymer, respectively. v is the polymer and solvent inter-
analyzed with a Waters Gel Permeation Chromatography action parameter, and can be calculated as:
(GPC) instrument equipped with a Differential Refractive fV 1 ðd1  d2 Þg
Index detector, a manual injector and three Waters Styra- v ¼ 0:34 þ ð4Þ
fRT g
gel columns (HR6, HR4, and HR3) in series. THF was used
as the eluent and the flow rate was set at 0.3 mL/min. d1 is the solubility parameter of the BA/MMA copolymer.
The internal temperature was set at 37 °C. The data were As the solubility parameter value of poly(butyl acrylate
analyzed using Empower 2 software from Waters. The cal- (hereafter referred to as PBA) is 8.9 (cal/cm3)1/2 [6], that
ibration curve included a set of 12 polystyrene standards of poly(methyl methacrylate) (hereafter referred to as
(EasiCal from Polymer Laboratories) with a range of 162– PMMA) is 9.2 (cal/cm3)1/2 [1], d1 should be 8.93 (cal/
6,035,000 g/mol. The Mark–Houwink parameters, K and cm3)1/2 for the BA/MMA (90/10 by weight) copolymer;
a, for BA and MMA are 1.1  104 dL/g and 0.708, and and d2 is the solubility parameter for toluene, 8.9 (cal/
1.28  104 dL/g and 0.690, respectively. cm3)1/2 [6].

35000 8.5
Gel swelling (g toluene/g dry gel)

30000 8.0
Mc (g/mol)

25000 7.5

20000 7.0
0 5 10 15 20 25
Swelling time (h)
Mc
Mc vs. swelling time Gel swelling vs. time

Fig. 1. Gel swelling and Mc changes with swelling time (gel samples from latex 8; see Table 1).
1228 L. Qie, M.A. Dubé / European Polymer Journal 46 (2010) 1225–1236

2.4. Glass transition temperature roll coater. The roll coater was passed through the film
front to back twice (i.e., along the length of the film). The
Two methods were used to test glass transition temper- dwell time did not exceed one minute. The substrate and
atures (Tg): Differential Scanning Calorimetry (DSC) and the strip were inserted into the grips and the upper grip
dynamic mechanical analysis (DMA), the latter to be de- was set to move upward at a speed of 300 mm/min. The
scribed in the DMA test section below. average force per m required to peel the strip from the
A Differential Scanning Calorimeter (DSC) Model Q1000 substrate was recorded and reported as peel strength.
from TA Instruments was used. It was equipped with an Lastly, PSTC 7 measures shear strength. A specimen of
auto sampler, a refrigerated cooling system and nitrogen 25.4 mm  152.4 mm of the film was cut. The strip was
as the purge gas. To do the test, about 0.04 g of dry polymer laminated onto a stainless steel substrate with a contact
was weighed into standard DSC hermetic aluminum sam- area of 25.4 mm  25.4 mm and then placed in the
ple pans. The analysis was performed using a modulated home-built shear tester using a C-clamp. A 500 g weight
DSC method with modulation amplitude of ±1 °C every was suspended at the end of the strip. The time to failure
60 s. The sample was cooled to 80 °C and followed by a was recorded automatically using Labview™ software.
heating ramp of 3 °C/min until the sample reached
150 °C. The Tg was calculated from the inflection point in 2.6. Dynamic mechanical analysis
the Reversed Heat Flow curve using the software provided.
The viscoelastic properties of the PSA were character-
ized via a RDA III rheometer (TA Instruments) with
2.5. PSA testing
25 mm parallel plate geometry. The sample thickness
was 1.7 ± 0.2 mm. The DMA measurements were done by
Tack, peel strength and shear strength were measured
heating the sample from 40 °C to 150 °C at a rate of
according to the Pressure Sensitive Tape Council standards
3 °C/min and under a shear frequency of 1.591 Hz. The Tg
PSTC-6, PSTC-1, and PSTC-7, respectively [21]. The latex
was also obtained from the maximum of the tan d curve.
was filtered using glass wool and cast onto a 50 lm MylarÒ
sheet with a 30# Meyer rod. The cast film was dried at
room temperature and conditioned for 24 h at standard 3. Results and discussion
conditions of temperature and humidity (23 ± 1 °C and
50% ± 5 Relative Humidity). A Universal Instron tester 3.1. Experimental design
was used to evaluate loop tack and peel strength. For tack,
a strip of 25.4 mm  177.8 mm was cut from the film and In this study, BA, MMA, and in some cases, AA, were
was formed into a loop with the adhesive side facing out- used as the monomers. For the BA/MMA system, a 32 facto-
ward. Approximately 25.4 mm at both ends of the strip rial design consisting of 10 runs (one replicates) with two
was masked with tape and inserted into the upper grip. factors (NDM and AMA concentrations) and three levels
The instrument moved the upper grip downward at a (0, 0.2, and 0.4 phm) were carried out (see Table 1). For
speed of 300 m/min until an area of 25.4 mm2 came into all these runs, the BA/MMA weight ratio was 90/10. An
contact with the stainless steel substrate mounted into empirical model relating gel content with NDM and AMA
the lower grip. Next, the tester moved the upper grip up- concentrations was established. Table 1 summarizes the
wards at the same speed while recording the force needed experimental design and final latex properties (i.e., gel con-
to de-bond the loop from the substrate. The maximum tent, Mw, Mn, and Mc) for 13 runs, which also includes three
force per metre necessary to remove the adhesive was re- additional runs. Run 11 was carried out in order to validate
ported as loop tack. PSTC 1 Test Method A evaluates peel the model’s predictive power and also produce a latex with
strength at a peel angle of 180°. A PSA film specimen of similar gel content as that from Run 1 of the factorial de-
25.4 mm  304.8 mm was cut. The strip was laminated sign; Runs 10 and 12 were performed to expand the range
onto a stainless steel substrate with the help of a 2040 g of factorial design. They used 0.5 phm and 0.11 phm NDM,

Table 1
Experimental design and latex properties (BA/MMA weight ratio: 90/10).

Latex ID NDM (phm) AMA (phm) Gel content (wt.%) Mw (103) (g/mol) Mn (103) (g/mol) Mc (103) (g/mol) Tg (°C, DSC)
1 0.40 0.20 40 165 42 13.0 –
2 0.40 0 0 117 35 – 39.2
3 0 0 64 296 90 65.0 38.0
3R 0 0 67 291 71 – –
4 0 0.20 91 107 51 32.3 38.20
5 0.20 0.20 74 89 28 45.7 –
6 0.40 0.40 75 72 27 38.5 –
7 0.20 0.40 86 86 35 31.7 –
8 0 0.40 96 – – 24.1 37.0
9 0.20 0 0 275 54 – –
10 0.50 0 0 88 35 – –
11 0.2 0.10 42 230 73 15.5 –
12 0.11 0 0 586 73 – –
 L. Qie, M.A. Dubé / European Polymer Journal 46 (2010) 1225–1236 1229

respectively, and resulted in two additional gel-free Table 3

latexes. Experimental data for developing Model A.

A second experimental design included AA in the mono- Latex ID x1 x2 x1 x2 x21 x22 y (%)
mer system. A total of six runs involving three CTA levels
1 1 0 0 1 0 40
(0, 0.2, and 0.4 phm) and two AA levels (2 and 4 phm) were 2 1 1 1 1 1 0
completed. For all these runs, the total weight of monomer 3 1 1 1 1 1 64
including AA and also the weight ratio of BA to the mono- 3R 1 1 1 1 1 67
mer mixture was fixed at the levels for the initial experi- 4 1 0 0 1 0 91
5 0 0 0 0 0 74
mental design: the BA/(MMA + AA) weight ratio was set 6 1 1 1 1 1 75
at 90/10, and the amount of AA was based on the total 7 0 1 0 0 1 86
amount of monomers. The results from these runs are 8 1 1 1 1 1 96
shown in Table 2. 9 0 1 0 0 1 0

3.2. Empirical model building

3.3. Influence of CTA and cross-linker concentrations on latex
An empirical quadratic model (Model A) was estab- particle size
lished based on the factorial design from Table 1 to relate
the gel content (y) of the latexes to the concentrations of During seeded semi-batch emulsion polymerization,
NDM (x1) and AMA (x2) used in the runs. x1 and x2 are secondary nucleation and particle coagulation may occur
the coded values (see Table 3), and their relations to the ac- depending on polymerization conditions. These two phe-
tual concentrations of NDM (X 1 ) and AMA (X 2 ) are shown nomena may not only affect the polymerization rate but
in Eqs. (5) and (6). The units of X 1 and X 2 are phm (per hun- also the final latex properties such as composition, Mw
dreds parts of monomers). and gel content [22]. In this study, we wanted to avoid sec-
ondary nucleation and particle coagulation, and therefore
ðX 1  0:2Þ we monitored the particle growth during the polymeriza-
x1 ¼ ð5Þ
0:2 tion process. It is known that secondary nucleation can
ðX 2  0:2Þ be caused by several factors, but it is in particular, gov-
x2 ¼ ð6Þ
0:2 erned by the emulsifier concentration [23,1]. In the event
Model A: of desorption of radicals from particles due to transfer to
CTA reactions, new particles can also be nucleated. Such
y ¼ 0:6263  0:2356x1 þ 0:3127x2 þ 0:1209x1 x2 an effect is uncommon when NDM is used as CTA in the
range used in this study [24,11,25]. Elevated levels of
þ 0:0856x21  0:1394x22 ð7Þ
cross-linker, well beyond those used in this work, have also
After statistical testing, it was found that Model A (Eq. been reported to induce secondary particle nucleation [26].
(7)) was adequate and also significant at a = 0.05, but the Particle coagulation can be caused by low emulsifier con-
coefficients of the last three terms in the model were not centrations, and can be influenced by CTA and cross-linker
significant. Therefore, a simplified model (Model B, Eq. concentrations. Fig. 2 shows a typical progression in the
(8)) was developed by removing these terms. Model B particle size distribution as demonstrated for Run 1.
was also found to be adequate and significant and was an From Fig. 2, we can see that the particle size distribu-
improvement to Model A. tions do not give any indication of secondary nucleation
Model B: or particle coagulation. The increase in particle size with
monomer feeding time (cf. 20 min, 60 min, 120 min, and
y ¼ 0:5893  0:2475x1 þ 0:3008x2 ð8Þ 180 min) is also evident from the figure. After a reaction
Using Model B, a new run (Run 11) was designed in or- time of 180 min, monomer feeding was stopped and only
der to produce a latex with gel content similar to the latex the initiator solution was fed to the reactor until
from Run 1, but using less cross-linker. The NDM and AMA 210 min. At this point, the reaction mixture was allowed
concentrations for Run 11 were chosen as 0.2 and 0.1 phm, to consume the remaining monomer for an additional
respectively. The gel content of the latex from Run 11 was 50 min. Beyond 180 min, the particle size distribution re-
42%, which compared well to the predicted gel content of mained unchanged. This is consistent with the fact that
43.5% using Model B. at 180 min, the monomer conversion was already very

Table 2
Influence of AA on polymer properties.

Latex ID AA (wt.%) CTA (phm) Gel content (wt.%) Mw (103 g/mol) Mn (103 g/mol) Tg (°C)
13 2 0.4 0 120 34 38.0
14 4 0.4 0 126 38 37.5
15 2 0.2 0 305 57 37.7
16 4 0.2 0 349 68 36.8
17 2 0 69 267 87 –
18 4 0 72 223 57 –
1230 L. Qie, M.A. Dubé / European Polymer Journal 46 (2010) 1225–1236

30 1

Instantaneous monomer conversion

0.8
20
Intensity (%)

0.6

10 0.4

0.2
0
10 100 1000
z-average particle diameter (nm) 0
0 0.2 0.4 0.6 0.8 1
0 min 20 min 60 min 120 min 180 min Overall monomer conversion

Fig. 2. Particle size distribution change during polymerization (Run 1). Run 1 Run 2 Run 3 Run 8 Run 9

high and the particle size would not be expected to change Fig. 3. Evolution of instantaneous monomer conversion.

further. All of the latexes produced in this study had aver- et al. [11] observed, the growth in particle size likely led
age particle sizes around 148 ± 3 nm and narrow particle to a decrease in the rate of radical exit from the particles,
size distributions (PDI < 0.04). Thus, as expected, the CTA and accordingly, the number of radicals per particle in-
and cross-linker did not affect the number of particles un- creased, so the instantaneous conversion increased as well.
der all of these conditions.

3.4. Influence of CTA and cross-linker concentration on 3.5. Latex polymer composition
monomer conversion
In a conventional batch polymerization, due to the differ-
Both the instantaneous and overall monomer conver- ent monomer reactivity ratios (r1 = 0.507 and r2 = 2.375,
sions were calculated and are defined as: where 1 represents BA and 2 represents MMA) [27], MMA
will tend to polymerize faster thereby resulting in polymer
Instantaneous monomer conversion composition drift. So in this study, we used a starved seeded
¼ weight of dry polymer=weight of added monomers semi-batch approach in order to achieve a homogeneous
copolymer composition. The Tg was measured by DSC for
ð9Þ
several of the latexes. All the DSC results indicated the pres-
ence of a single Tg close to that calculated by the Fox-equa-
Overall monomer conversion tion [28]. Fig. 4 gives an example of a DSC thermogram for
¼ weight of dry polymer=weight of total monomers latexes 2 and 9. Only slight differences in Tg were measured
ð10Þ from run to run (38.5 ± 1.5 °C). DMA measurements pro-
vided an even narrower range in Tg (±0.25 °C). This confirms
The instantaneous monomer conversion can be affected that the copolymer composition was indeed homogeneous.
by many factors, such as the amounts of initiator and sur-
factant, monomer to initiator ratio, monomer mixture and
initiator solution feeding rates, as well as the CTA and 3.6. Influence of CTA and cross-linker on polymer latex
cross-linker concentrations. In this study, all factors, ex- properties
cept for the CTA and cross-linker concentrations, were
fixed. Fig. 3 shows the instantaneous and total monomer By comparing latex 3 to latexes 12, 9, 2, and 10 in
conversions measured by gravimetry for Runs 1, 2, 3, 8, Table 1, it can be seen that adding the CTA (i.e., NDM) de-
and 9, which covers the entire range of CTA and cross-lin- 0.0
ker concentrations. All runs in our factorial design gave
nearly identical conversion profiles. Thus, the CTA and Latex 2
-0.1 Tg: -39.2 C
cross-linker concentrations did not appear to influence
Rev heat flow (mV)

monomer conversion. This is consistent with the findings

of Chauvet et al. [15]. From Fig. 3, it can also be seen that -0.2
Latex 9
the instantaneous monomer conversion was high (greater Tg: -38.2 C
than 90%) for a large part of the time, which is typical of -0.3
monomer-starved conditions, and the final overall mono-
mer conversion was between 99.2 and 99.8%, which gave
-0.4
the latexes a solid content of 47%. In addition, the instan-
taneous conversion was seen to slightly increase during
the monomer feed period (i.e., from 20 to 180 min). As -0.5
-100 -50 0 50 100 150
there was no secondary particle nucleation, as concluded
Temperature (°C)
from the particle size measurements, this increase was
not caused by secondary nucleation. Rather, as Plessis Fig. 4. DSC thermograms for latexes from Runs 2 and 9.
 L. Qie, M.A. Dubé / European Polymer Journal 46 (2010) 1225–1236 1231

creased the gel content to zero. In addition, with the in- whole polymerization process even though AMA has two
crease in NDM concentration, Mw greatly decreased for non-symmetrical double bonds. Hence, we can divide the
the gel-free latexes. Fully expected, these phenomena re- polymerization process into two sub-processes: (1) the for-
sulted from an increase in chain transfer events, and there- mation of polymers with kinetic chain length, which are
fore the kinetic chain length was significantly decreased. here defined as ‘‘primary” polymers; and (2) the further
By comparing latexes 3, 4, and 8 in Table 1, it can be growth of some of the primary polymer chains, either at
seen that when no CTA was used, the gel content increased the pendent tertiary hydrogen points of the BA units via
with AMA concentration, while the Mc of the gel polymers chain transfer to polymer (a.k.a. back-biting) or at the pen-
and the Mw of the sol polymers decreased. This phenome- dent double bonds of the cross-linker units. When cross-
non is well known for the emulsion polymerization of BA linker is used, the latter growth mechanism dominates.
and BA-dominated comonomer systems [12]. When Larger primary polymer chains have a higher probability
cross-linker was not used, gel formed by either inter- or in- to engage in further chain growth; and once their sizes be-
tra-molecular chain transfer to polymer via back-biting, come sufficiently large after further growth, they become
plus termination by combination [29,10]. When cross-lin- gel; otherwise they will remain as sol polymers.
ker was used, gel also formed via another mechanism, Based on the above mechanism, one can surmise that at
propagation to the pendent double bond of the cross-linker constant AMA concentration (0.2 phm), an increase in
unit in the polymer chain. In this case, the latter mecha- NDM from 0 to 0.2 phm resulted in the formation of much
nism dominated the gel formation. shorter primary polymer chains, so Mw decreased as ex-
From Table 1, it can be seen that when the same pected. At the same time, the gel content decreased be-
amount of cross-linker (0.2 phm) was used, an increase cause the shortened primary polymer chains decreased
in NDM concentration from 0 (latex 4) to 0.2 (latex 5) to the probability for primary polymers to transfer to gel by
0.4 (latex 1) phm resulted in a decrease in gel content from further growth. Finally, the formation of a looser gel net-
91 to 74 to 40 wt.%, respectively. The tested Mc first in- work was because NDM (CTA) can also terminate some
creased from 32.3 to 45.7 kg/mol, and then decreased to of the chain growth at the cross-linker points.
13.0 kg/mol. At the same time, the sol polymer molecular, When the NDM was further increased to 0.4 phm, the
Mw, first decreased from 107 to 89 kg/mol, and finally in- primary polymer chains should become much smaller. As
creased to 165 kg/mol as shown in Fig. 5. It can be seen expected, this resulted in a decrease in the gel content,
in Fig. 5 that the MWD became broader with the increase due to the reasons outlined earlier. However, the Mw in-
in NDM concentration. Similar trends were found for la- creased, which contradicted our expectations. It is likely
texes 7 and 6 as shown in Table 1. that small primary polymer chains, possessing cross-linker
The seemingly contradictory changes in Mc and Mw in their backbones, reacted further to form highly branched
warrant further discussion. Chauvet et al. reported on the polymers. However, some of these polymers did not reach
evolution of gel content for the emulsion copolymerization the size of gel, so they were not gel but rather, sol poly-
of BA and AA under starved seeded semi-batch emulsion mers. This, in turn, resulted in higher than expected Mw.
conditions (i.e., conditions similar to that in this study) As for the change in Mc, this was possibly due to the imper-
[15]. They demonstrated that the gel contents were almost fect structure of the gel formed under this condition. The
constant for most of the polymerization process (from formed gel in this case had many branches and very few
20% overall monomer conversion to almost complete cross-linking points, which implies that the actual Mc
conversion of monomer) and only slightly increased during was very large. The formation of such a gel structure was
the cooking period when AMA and NDM were used sepa- possible because although chain growth could still occur
rately or together. In addition, the gel content was in- at the pendent double bond points of the cross-linker, most
creased by adding AMA. This implies that NDM and AMA of the growing side chains were terminated by CTA and
can simultaneously affect the latex properties during the therefore became branches. The mechanical strength of
this kind of gel is very weak, and it could easily break
and collapse during swelling. The collapse causes signifi-
1.2 cant solvent loss, which results in a lower swelling ratio.
Latex 4 Thus, if calculated using Eq. (2), the Mc would be smaller.
1.0 CTA: 0 phm In this case, the calculated Mc from the gel swelling model
Mw: 107 kg/mol
Mn: 54 kg/mol cannot describe the gel structure.
0.8 PDI: 2.0 We can further confirm the explained Mw change by
dwt/d(LogM)

Latex 5 comparing latex 1 to latex 2. Latex 2 was produced with

0.6 CTA: 0.2 phm the same amount of NDM (0.4 phm) as Run 1, but without
Mw: 89 kg/mol Latex 1
Mn: 28 kg/mol
AMA. Its Mw was 117 kg/mol, which was significantly low-
CTA: 0.4 phm
0.4 PDI: 3.2 Mw: 165 kg/mol er than that of latex 1, 165 kg/mol. When no CTA was used,
Mn: 42 kg/mol with the addition of cross-linker, we would expect the gel
PDI: 3.9
0.2 content to increase, but Mw to decrease. However, when
0.4 phm NDM was used, the gel content increased from 0
0.0 to 40%, but the Mw increased when the cross-linker was in-
2 3 4 5 6 7 8
creased from 0 to 0.2 phm. This is because at 0.4 phm
LogM (g/mol)
NDM, the chain lengths of the primary polymer were very
Fig. 5. Mw changes for latexes 1, 4, and 5 (0.02 phm cross-linker). small, so after further growth at the cross-linker points,
1232 L. Qie, M.A. Dubé / European Polymer Journal 46 (2010) 1225–1236

some of them became gel and the others became branched crease back-biting by lowering the rate of formation of six-
sol polymers with bigger Mw. membered transition rings, and therefore chain transfer to
By comparing latex 5 with 6, and latex 1 with 11 (see polymer is greatly lowered and accordingly, leads to a low-
Table 1), we can see that latexes with similar gel contents er gel content [30]. As the pendent group of AA is much
but different Mc and Mw could be produced by simulta- smaller than that of MMA, it is likely that AA would not de-
neously increasing the concentration of NDM and AMA. crease back-biting as much as MMA. In addition, as AA has
At higher levels, the Mw and Mc were significantly lower, a tertiary hydrogen similar to that of BA, increasing the
as shown in Fig. 6 and Table 1 (note: As we mentioned ear- concentration of AA while decreasing that of MMA and
lier, the tested Mc could not describe the gel structure for keeping that of BA constant will also enhance inter-molec-
latex 1 and here, also for latex 11 due to its imperfect gel ular chain transfer through back-biting, and accordingly,
structure. Therefore, here we only compared the Mc of la- increase the gel content. Therefore, the gel content was in-
texes 5 and 6). creased with the increase in AA concentration and de-
The polymer property changes can also be explained via crease in MMA concentration. Furthermore, as the pH of
the polymerization process analysis mentioned earlier. At the reaction mixture was around 2.5 for all runs, and also
higher levels of NDM, the chain length of the primary poly- the pKa of AA and poly(acrylic acid) are about 4 [31,32],
mer would be smaller, and correspondingly the Mw will be it is unlikely that AA or poly(acrylic acid) underwent disso-
smaller; at higher levels of AMA, there would be more ciation during the polymerization and caused the gel con-
cross-linking points, so the Mc became smaller. Moreover, tent and Mw changes.
as the effect of increase CTA concentration on gel content
can counterbalance that of increase cross-linker concentra- 3.8. Bulk properties of PSA film made from latexes 1 and 11
tion, so the gel content could be stay unchanged. (hereafter referred to as PSA1 and 11)
From the above results, we can see that to achieve sim-
ilar gel contents with different Mw, and also very different DMA temperature sweep curves for PSA1 and 11, which
Mc, we require latexes at fairly high gel contents (70%). In have similar gel contents of 40%, are shown in Figs. 7 and
addition, in order to study the influence of Mc and Mw on 8. Figs. 7 and 8 show that the storage (G0 ) and loss (G00 ) mod-
polymer mechanical properties, it is better to make the dif- uli of these two PSA were similar at temperatures below
ference in Mc and Mw more significant. 10 °C; and as the temperature increased, the G0 and G00 of
PSA11 became much larger than those of PSA1. The signifi-
3.7. Influence of AA on latex polymer properties cant increases in G0 and G00 were consistent with the much
larger Mw for latex 11. The modulus increase also indicates
To study the influence of AA on latex polymer proper- that at high temperatures, PSA11 had a much higher cohe-
ties, we compared latexes from Table 1 to those in Table 2; sive strength than PSA1. In this study, we did not report
specifically latex 2 with latexes 13 and 14, latex 9 with la- the DMA data for the PSA made from latex 17 and 18, which
texes 15 and 16, and latex 3 with latexes 17 and 18. We can had similar gel contents of 75%, as the moduli of these two
see that changing the BA/MMM/AA weight ratio from 90/ PSA were beyond the testing limit of our DMA instrument,
10/0 to 90/8/2 to 90/6/4, at 0.4 and 0.2 phm NDM, the especially at the low temperature range.
gel content was 0 while the Mw and Tg slightly increased.
At 0 phm NDM, the gel content slightly increased, Mw de- 3.9. PSA performance
creased and Tg slightly increased. The increase in gel con-
tent and decrease in Mw with the increase in AA Henceforward the PSA will be named after their latexes
concentration and decrease of MMA concentration was ID for the purpose of simplification. For example, the PSA
likely caused by an increase in back-biting reactions. It is made from latex 1 was referred to as PSA1.
known that when MMA copolymerizes with BA, it can de- Table 4 shows the Mw and Me values for the BA/MMA
gel-free latexes, and Fig. 9 shows the performance of the
PSA films made from these latexes.
1
1.E+10
Latex 1 Latex 11
CTA: 0.4 phm CTA: 0.2 phm
0.8 1.E+09
Cross-linker: 0.2 phm Cross-linker: 0.1 phm
Mw: 165 kg/mol Mw: 230 kg/mol
Mn: 42 kg/mol Mn: 73 kg/mol 1.E+08
G' (dyn/cm 2 )
dwt/d(LogM)

0.6 PDI: 3.9 PDI: 3.2

1.E+07

0.4 1.E+06

1.E+05
0.2
1.E+04
-50 0 50 100 150
0
2 4 6 8 Temperature (°C)
LogM (g/mol) G' of latex 1 G' of latex 11

Fig. 6. Mw of latexes with similar gel content (40%). Fig. 7. Storage modulus (G0 ) of PSA1 and 11.
 L. Qie, M.A. Dubé / European Polymer Journal 46 (2010) 1225–1236 1233

1.E+09 M e;monodisperse
M e;polydisperse ¼ ð13Þ
/2:3
1.E+08
G" (dyne/cm2)

where / is the weight fraction of polymers having Mw

1.E+07 greater than the Me of the monodisperse sample.
Fig. 9 shows that tack greatly increased when the Mw/
1.E+06 Me ratio increased from 3.7 to 5.2, then greatly decreased
when the Mw/Me ratio changed to 13.0 and 27.5. Peel
1.E+05 strength greatly increased when the Mw/Me ratio increased
from 3.7 to 5.2 to 13.0, and then greatly decreased when
1.E+04 the Mw/Me further increased to 27.5. Shear strength in-
-50 0 50 100 150
Temperature (°C) creased with Mw/Me. The change in PSA performance can
G" of latex 1 G" of latex11
be explained as follows. For gel-free PSA, with the increase
in Mw or Mw/Me ratio, the deformability of the PSA would
Fig. 8. Loss modulus (G00 ) of PSA1 and 11. decrease, and at the same time, the cohesive strength
would increase. As shear strength is mainly related to the
cohesive strength of the PSA, it increased with the increase
Table 4 in the Mw/Me ratio. As for tack and peel strength, it is
Polymer properties of gel-free latexes (BA/MMA weight ratio: 90/10).
known that their tests involve both a bonding and debond-
Latex CTA Mw (103 Mn (103 Me (103 Mw/Me ing procedure [9,35–38]. For a PSA film to have high tack
ID (phm) g/mol) g/mol) g/mol) and peel strength, it must stick well to the testing panel
10 0.50 88 35 23.7 3.7 during the bonding process via deformation and flow,
2 0.40 117 35 22.6 5.2 and also dissipate a lot of energy during the debonding
9 0.20 275 54 21.2 13.0
process. Accordingly to Zosel, only when Me is larger than
12 0.11 586 73 21.3 27.5
10 kg/mol, can fibrils be formed during the debonding pro-
cess and therefore dissipate a significant amount of energy
[9]. In addition, Lackout et al. reported that when Mw was
700 4 low, fibrils only formed in the lateral direction during deb-
onding, and in this case, only a small amount of energy was
Tack and peel strength (N/m)

600
Shear strength (1”x1”, h)

dissipated; however, with the growth in Mw, fibrils formed

3
500 not only in this direction but also in the elongation direc-
400 tion, so in this case, a lot of energy was dissipated [39].
2 Based on this, it is likely that at a certain Mw/Me ratio, an
300
optimum condition can be reached under which the PSA
200 film could adhere well to the testing panel during bonding
1
and also a lot of energy could be dissipated during debond-
100
ing. In Fig. 9, one can see that the maximum tack occurred
0 0 at a lower Mw/Me ratio than the maximum peel strength.
0 5 10 15 20 25 30
This is because the loop tack test uses a shorter bonding
Mw/Me ratio
time and smaller contact force than the peel strength test.
Tack Peel strength Shear strength
Therefore, in order to bond well during the loop tack test,
Fig. 9. Tack, peel strength and shear strength vs. Mw/Me ratio for gel-free the PSA must deform and flow better, which corresponds
latexes (BA/MMA weight ratio: 90/10). to a lower Mw/Me value.
Given the information displayed in Fig. 9 and the anal-
ysis regarding the deformability and cohesive strength of
The values of Me in Table 4 were obtained as follows: (1) PSA, we can conclude that any PSA having lower cohesive
Calculate the Me of monodisperse copolymer of BA/MMA strength than that of PSA10 would give lower tack, peel
with weight ratio of 90/10 with Eq. (11). strength and shear strength than those of PSA10. This
M e;copolymer;monodisperse ¼ W BA  Me;PBA;monodisperse was confirmed with a PSA reported by Tobing and Klein,
which we named PSA10TK [6]. The Mw and Me of PSA10TK
þ W PMMA  Me;PMMA;mnodisperse ð11Þ were 97 and 24 kg/mol, respectively, which are very simi-
lar to that of our PSA10. But PSA10TK was produced using
where the Me of monodisperse poly(butyl acrylate) (re-
BA/AA (97.5/2.5) and its Tg was 50 °C, which was much
ferred to as PBA) is 20,773 g/mol [6], Me of poly(methyl
lower than that of PSA10 (38 °C). Therefore, the cohe-
methacrylate) (referred to as PMMA) is 10,013 g/mol
sive strength of PSA10TK should be much lower than that
[33], so
of PSA10. Fig. 10 shows that PSA10TK did have lower tack,
M e;copolymer;monodisperse ¼ 0:9  20; 773 þ 0:1  10; 013 peel strength and shear strength than PSA10. One should
note that the shear strength of PSA10TK was reported to
¼ 19; 697 g=mol ð12Þ
be ‘‘<<0.1 min” with a contacting area of 1/200  1/200 . In this
(2) Calculate the Me for our polydisperse polymers study, all the shear strength measurements were com-
using the following plastic model [34]: pleted with a contact area of 100  100 , so the reported shear
1234 L. Qie, M.A. Dubé / European Polymer Journal 46 (2010) 1225–1236

600 10 places these PSA in the Mw range between PSA2 and the
PSA9 on Fig. 9. According to Fig. 11, the effect of increasing
Tack and peel strength (N/m)

Shear strength (1”x1”, min)

500 AA concentration resulted in the same PSA performance
8
trends as in Fig. 9 where the Mw/Me ratio increased from
400
6
5.2 (PSA2) to 13.0 (PSA9). The AA concentration effect is
due to the formation of strong hydrogen bonds, which in-
300
crease the cohesive strength of the PSA. This is similar as
4
200 the effect of increasing Mw/Me on the cohesive strength
of the BA/MMA PSA. The PSA performance shown in
2 Fig. 12 differs only from that in Fig. 11 with respect to
100
the peel strength. The peel strength decreased with
0
1
0 increasing AA concentration in Fig. 12. This effect is also
Latex 10 Latex210-Re
due to the formation of hydrogen bonds with AA, but as
Tack Peel strength Shear strength the Mw of PSA9, 15, and 16 were 275, 305 and 349 kg/
Fig. 10. Performance of PSA10 and PSA10TK. mol, respectively, this places these PSA in the range of
Mw/Me ratio from 13.0 (PSA9) to 27.5 (PSA12) (see Fig. 9).
Thus, from PSA9 to 15 to 16, the peel strength decreased,
strength of PSA10TK was converted to a corresponding va- which is the same trend as changing from PSA9 to 12.
lue at 100  100 contact area using the following equation Table 5 shows the performance of PSA made from the
[40]. gel-containing latexes produced in this study. Combining
n o the polymer property data from Table 1 with the PSA per-
L2 W g formance data from Table 5, a number of observations can
T¼ ð14Þ be made. First, comparing the PSA made from latexes 3, 4,
f2tMg g
and 8 (CTA concentration = 0; AMA concentration = 0, 0.2,
where T is the shear strength, L is the length of contact area and 0.4 phm, respectively), we can see that with the in-
between PSA and testing panel, W is the width of the con- crease in gel content, the shear strength greatly increased,
tact area, g is the zero shear viscosity of the PSA, t is the and tack and peel strength slightly decreased. This is due to
thickness of the PSA film, M is load for testing shear the increase in gel content which served to increase the
strength and g is the gravitational constant. When convert- cohesive strength of the PSA. It should be noted, however,
ing the shear strength from 1/200  1/200 to 100  100 using Eq. that the elevated gel contents of PSA4 and PSA8 at a Tg of
(14), only L and W doubled, and all the other factors re- 38 °C make them impractical for use as a PSA. Second,
mained unchanged, so the shear strength at 100  100 should comparing the PSA produced from latexes 4, 5, and 1
be: (AMA concentration = 0.2 phm; CTA concentration = 0,
0.2, and 0.4 phm, respectively), we observed that an in-
T 100 100 ¼ 22  2  T 1=200 1=200  8  0:1 min ¼ 0:8 min ð15Þ crease in gel content resulted in an increase in shear
strength and a decrease in tack and peel strength. The
Figs. 11 and 12 describe the influence of AA on the per- PSA from latexes 4 and 5 cannot be used as PSA due to their
formance of PSA made from latexes produced with 0.4 and high gel contents at a Tg of about 38 °C. In this case, the
0.2 phm NDM, respectively. Fig. 11 shows a decrease in influence of gel content on PSA performance overshadows
tack and increases in peel strength and shear strength with any effects due to Mw or Mc. Thus, to study the influence of
increases in AA concentration. This behaviour can best be Mc and Mw on PSA performance, it will be necessary to
explained by considering the PSA performance changes study PSA with similar gel contents. Finally, comparing
with Mw /Me as shown in Fig. 9. PSA2, 13, and 14 had sim- PSA with similar gel contents and similar Mc, those from
ilar Mw of 117, 120, and 126 kg/mol, respectively; this

600 8
600 0.5
Tack and peel strength (N/m)

7
Shear strength (1”x1”, h)
Tack and peel strength (N/m)

500
Shear strength (1”x1”, h)

500 6
0.4
400
400 5
0.3
300 4
300
0.2 3
200
200
2
0.1 100
100 1

0 0 0 0
No AA
1 2
2 wt.% AA 4 wt.%
3
AA No1AA 2 wt.%
2 AA 4 wt.%
3 AA
Tack Peel strength Shear strength Tack Peel strength Shear strength

Fig. 11. Influence of AA on PSA performance at a CTA concentration of 0.4 Fig. 12. Influence of AA on PSA performance at a CTA concentration of 0.2
phm NDM. phm NDM.
 L. Qie, M.A. Dubé / European Polymer Journal 46 (2010) 1225–1236 1235

Table 5 and cross-linker levels had a lower Mw, and a similar Mc,
PSA performance of latexes containing gel (BA/AA weight ratio: 90/10). which was due to the imperfect gel structure. The PSA per-
Latex ID Gel content Tack Peel strength Shear strength formance of films cast from these latexes correlated well
(wt.%) (N/m) (N/m) with DMA data. Under similar conditions (i.e., higher CTA
1 40 187 91 4.5 h and cross-linker levels) but at higher gel contents (75%),
11 42 172 80 5.8 h the latex had both a lower Mw and Mc. The higher gel con-
3 64 61 47 101.6 h tent latexes could not be employed as PSA due to their high
4 91 30 20 >1 month
5 74 43 29 >1 month
gel content at a Tg of 38 °C. Thus, a modification of the
6 75 38 26 >1 month monomer mixture composition in order to lower the Tg
7 86 40 30 >1 month should enable us to generate adequate PSA at this higher
8 96 28 15 >1 month gel content.

latexes 1 and 11, we observe a slight decrease in tack and Acknowledgements

peel strength and a significant increase in shear strength
with an increase in Mw. This is consistent with the change We thank the Natural Science and Engineering Research
in G0 and G00 moduli as shown in Figs. 10 and 11. These find- Council (NSERC) of Canada and Omnova Solutions Inc. for
ings concur with the report that the shear strength is re- financial support.
lated to the modulus at high temperature (127 °C), and
the bigger the modulus, the higher the shear strength References
[41]. The fact that the increases in tack and peel strength
are small is because tack and peel strength are propor- [1] Lovell PA, El-Aasser MS. Emulsion polymerization and emulsion
tional to the G00 at debonding frequency/G0 at bonding fre- polymers. New York: John Wiley & Sons; 1997.
[2] Jovanovic R, Dubé MA. Emulsion-based pressure-sensitive
quency [38], and Figs. 7 and 8 clearly show both G0 and G00 adhesives: a review. J Macromol Sci-Polym Rev 2004;C44(1):1–51.
increased with Mw at the same magnitude. Latex 5 has sim- [3] El-Aasser MS, Tang JS, Wang XR, Daniels ES, Dimonie VL, Sudol ED.
ilar gel content as latex 6, but larger Mc and Mw values (see Advances in emulsion polymerization for coatings applications:
latex blends and reactive surfactants. J Coatings Technol 2001;
Table 1), and from Table 5 we can see that the influence of
73(920):51–63.
Mc and Mw on PSA performance are not observable. This is [4] Nakayama Y. Development of novel aqueous coatings which meet
because the polymer material from these two latexes no the requirements of ecology-conscious society: novel cross-linking
system based on the carbonyl-hydrazide reaction and its
longer satisfied the Dahlquist contact criterion [42] due
applications. Prog Org Coatings 2004;51(4):280–99.
to their high gel content at a Tg of about 38 °C. In a future [5] Tobing S, Klein A, Sperling LH, Petrasko B. Effect of network
study, we will modify the polymer composition to lower morphology on adhesive performance in emulsion blends of acrylic
the Tg, and thus be able to use higher gel content latexes pressure sensitive adhesives. J Appl Polym Sci 2001;81(9):2109–17.
[6] Tobing SD, Klein A. Molecular parameters and their relation to the
as PSA. adhesive performance of acrylic pressure-sensitive adhesives. J Appl
Polym Sci 2001;79(12):2230–44.
[7] Tobing SD, Klein A. Molecular parameters and their relation to the
adhesive performance of emulsion acrylic pressure-sensitive
4. Conclusion
adhesives. II. Effect of crosslinking. J Appl Polym Sci 2001;79(14):
2558–64.
The combined influence of CTA and cross-linker on var- [8] Zosel A, Ley G. Influence of cross-linking on structure,
mechanical-properties, and strength of latex films. Macromolecules
ious polymer micro-structural properties, such as gel con-
1993;26(9):2222–7.
tent, Mc and Mw, was studied via starved seeded semi- [9] Zosel A. The effect of fibrilation on the tack of pressure sensitive
batch emulsion polymerizations of BA/MMA. We showed adhesives. Int J Adhes Adhes 1998;18(4):265–71.
that by combining CTA and cross-linker, latex polymers [10] Plessis C, Arzamendi G, Leiza JR, Schoonbrood HAS, Charmot D, Asua
JM. Seeded semibatch emulsion polymerization of n-butyl acrylate.
with unique micro-structural properties could be pro- Kinetics and structural properties. Macromolecules 2000;33(14):
duced. For example: at the same cross-linker concentra- 5041–7.
tion, adding CTA led to a latex polymer with lower gel [11] Plessis C, Arzamendi G, Leiza JR, Alberdi JM, Schoonbrood HAS,
Charmot D, et al. Seeded semibatch emulsion polymerization of
content, higher Mc and lower Mw; adding even more CTA butyl acrylate: effect of the chain-transfer agent on the kinetics and
resulted in a latex polymer with highly branched sol poly- structural properties. J Polym Sci Part A-Polym Chem 2001;39(7):
mers having higher Mw and gel polymer with an imperfect 1106–19.
[12] Bouvier-Fontes L, Pirri R, Asua JM, Leiza JR. Cross-linking emulsion
network structure. The performance of the PSA produced copolymerization of butyl acrylate with diallyl maleate. J Polym Sci
from these latexes was greatly affected by the gel content, Part A-Polym Chem 2005;43(20):4684–94.
which overshadowed the effects of Mw and Mc. Outside of [13] Bouvier-Fontes L, Pirri R, Magnet S, Asua JM, Leiza JR. Effect of the
diacrylate ester size on the semicontinuous cross-linking emulsion
PSA applications, these latexes may be very useful for a
copolymerization of BA. Macromolecules 2005;38(7):2722–8.
broad range of uses due to their unique micro-structure. [14] Shen HZ, Zhang JY, Liu SJ, Liu GD, Zhang LQ, Qu XW. Effect of the
For example, these highly branched latex polymers might chain-transfer-agent content on the emulsion polymerization
process and adhesive properties of poly(n-butyl acrylate-co-acrylic
be added to linear polymers to improve their rheological
acid) latexes. J Appl Polym Sci 2008;107(3):1793–802.
behaviour [43]. [15] Chauvet J, Asua JM, Leiza JR. Independent control of sol molar mass
Latexes with similar gel contents were produced by and gel content in acrylate polymer/latexes. Polymer 2005;
simultaneously increasing the CTA and cross-linker con- 46(23):9555–61.
[16] Elizalde O, Arzamendi G, Leiza JR, Asua JM. Seeded semibatch
centrations according to an empirical model. At a moderate emulsion copolymerization of n-butyl acrylate and methyl
gel content range (40%), the latex produced at higher CTA methacrylate. Ind Eng Chem Res 2004;43(23):7401–9.
1236 L. Qie, M.A. Dubé / European Polymer Journal 46 (2010) 1225–1236

[17] Plessis C, Arzamendi G, Agnely M, Leiza JR, Asua JM. Seeded [29] Plessis C, Arzamendi G, Leiza JR, Schoonbrood HAS, Charmot D, Asua
semibatch emulsion polymerization of n-butyl acrylate: effect of JM. A decrease in effective acrylate propagation rate constants
the seed properties. J Polym Sci Part A-Polym Chem 2002;40(16): caused by intramolecular chain transfer. Macromolecules
2878–83. 2000;33(1):4–7.
[18] Barudio I, Guillot J, Fevotte G. Efficiency of mercaptan chain transfer [30] Gonzalez I, Asua JM, Leiza JR. The role of methyl methacrylate on
agents in emulsion copolymerizations. I. Influence on kinetics and branching and gel formation in the emulsion copolymerization of
microstructure. Modeling of radical desorption. J Polym Sci Part A- BA/MMA. Polymer 2007;48(9):2542–7.
Polym Chem 1998;36(1):157–68. [31] Perrin DD, Dempsey B, Serjeant EP. pKa predictions for organic acids
[19] Former C, Castro J, Fellows CM, Tanner RI, Gilbert RG. Effect of and bases. London: Chapman & Hall; 1981.
branching and molecular weight on the viscoelastic properties of [32] Lee JW, Kim SY, Kim SS, Lee YM, Lee KH, Kim SJ. Synthesis and
poly(butyl acrylate). J Polym Sci Part A-Polym Chem 2002;39(20): characteristics of interpenetrating polymer network hydrogel
3335–49. composed of chitosan and poly(acrylic acid). J Appl Polym Sci
[20] Kajtna J, Golob J, Krajnc M. The effect of polymer molecular weight 1999;73(1):113–20.
and crosslinking reactions on the adhesion properties of [33] Fetters LJ, Lohse DJ, Richter D, Witten TA, Zirkel A. Connection
microsphere water-based acrylic pressure-sensitive adhesives. Int J between polymer molecular-weight, density, chain dimensions, and
Adhes Adhes 2009;29(2):186–94. melt viscoelastic properties. Macromolecules 1994;27(17):4639–47.
[21] Pressure Sensitive Tape Council. Test methods for pressure sensitive [34] Wool RP. Polymer interfaces, structure, and strength. Cincinnati:
adhesive tapes. Northbrook, Illinois: Pressure Sensitive Tape Hanser Gardner Publications; 1995.
Council; 2004. [35] Chang EP. Viscoelastic windows of pressure-sensitive adhesives. J
[22] Odian G. Principles of polymerization. Hoboken, NJ: Willey; 2004. Adhes 1991;34(1–4):189–200.
[23] Tang JS, Ding TH, Daniels ES, Dimonie VL, Klein A, El-Aasser MS. [36] Chang EP. Viscoelastic properties of pressure-sensitive adhesives. J
Synthesis of well-defined, functionalized polymer latex particles Adhes 1997;60(1–4):233–48.
through semicontinuous emulsion polymerization processes. J Appl [37] Tse MF, Jacob L. Pressure sensitive adhesives based on VectorR SIS
Polym Sci 2003;88(1):30–41. polymers I. Rheological model and adhesive design pathways. J
[24] Maxwell IA, Morrison BR, Napper DH, Gilbert RG. The effect of chain Adhes 1996;56(1–4):79–95.
transfer agent on the entry of free-radicals in emulsion [38] Yang HWH, Chang EP. The role of viscoelastic properties in the design
polymerization. Makromol Chem (Macromol Chem Phys) of pressure-sensitive adhesives. Trends Polym Sci 1997;5(11):380–4.
1992;193(2):303–13. [39] Lakrout H, Creton C, Ahn DC, Shull KR. Influence of molecular
[25] Sayer C, Lima EL, Pinto JC, Arzamendi G, Asua JM. Kinetics of the features on the tackiness of acrylic polymer melts. Macromolecules
seeded semicontinuous emulsion copolymerization of methyl 2001;34(21):7448–58.
methacrylate and butyl acrylate. J Polym Sci Part A-Polym Chem [40] Dahlquist CA. Handbook of pressure sensitive adhesives. New
2000;38(2):367–75. York: Van Nostrand Reinhold; 1989.
[26] Bajpai S, Rai JSP, Nigam I. Swelling behavior of poly(MMA-co-BA-co- [41] Dale WC, Paster MD, Haynes JK. Mechanical-properties of acrylic
PPGDA) polymers. J Appl Polym Sci 2009;112(4):2374–82. pressure sensitive adhesives and their relationships to industry
[27] Dubé MA, Penlidis A. A systematic-approach to the study of standard testing. J Adhes 1989;31(1):1–20.
multicomponent polymerization kinetics – the butyl acrylate [42] Satas D. Handbook of pressure sensitive adhesive technology. New
methyl-methacrylate vinyl-acetate Example. 1. Bulk York: Van Nostrand Reinhold; 1989.
copolymerization. Polymer 1995;36(3):587–98. [43] Brigitte IV, Albena L. Hyperbranched and highly branched polymer
[28] Robeson L. Polymer blends: a comprehensive review. Munich, architectures – synthetic strategies and major characterization
Germany: Hanser; 2007. aspects. Chem Rev 2009;109:5924–73.