An Overview of the Use of Rheology for Adhesive Manufacturers

Philip Rolfe and Joanne Langridge, Malvern Instruments,

11 Harts Lane, East Brunswick, NJ. 08816. USA

Email: prolfe@bohlinusa.com

ABSTRACT
Several aspects of the rheology of adhesives are important for their application
and end use. Coating an adhesive requires the viscosity to fall between
maximum and minimum viscosity limits and these depend upon the type of
coating process that is to be used. Viscosity data is shown for the maximum
and minimum acceptable limits for several types of coater. The modification of
the flow properties by changing the solids loading and particle size distribution
is discussed.

After application and drying, an adhesives viscoelastic properties can be used

to assess its performance. Oscillation frequency sweep data can show whether
the adhesive will have good tack properties, whether it will form an immobilized
joint and if it will peel off easily (if needed) after use. Data is shown for good
and bad diaper adhesives

Keywords: rheology, adhesive, tack, filler, particle size, viscoelastic, coating.

Introduction
Rheometer measurements can be used to evaluate the shelf life and coatability of a wet
adhesive or the tack and cohesive properties once dry.
The main areas of interest in rheology are:

1) For the development technician: optimizing a formulation to give the best flow
properties whilst minimizing the cost is important. As many additives are expensive,
using the correct amount is important to ensure that the final product is economical.
Also, using the correct particle size distribution in the latexes and any additives can
give a lower viscosity, which in turn allows higher solids loading. Increasing the solids
content of an adhesive coating is preferable as it reduces the drying time and saves
energy. Laboratory rheometers allow small test samples to be evaluated (each test
using as little as 2 mls of sample), minimizing development costs and enabling easy
experimentation.
2) For the chemical engineer: dispersion and processing conditions can be calculated
and optimised so that a small dose of an additive is effective. Additives such as fumed
silica are notoriously difficult to disperse properly and their effectiveness is severely
reduced when they are not fully dispersed.
3) For the adhesive companys storage and wholesale facilities: shelf life and
sedimentation are key issues. Rheometers can be used to predict in a short test
whether sedimentation of a filler would take place in a month or two. Therefore, such a
test can be implemented as part of the QC check for an adhesive to ensure that the
product has been manufactured in the expected way and eliminate bad batches of
adhesive.
 4) For the end user of a pressure sensitive adhesive (PSA): the tackiness, holding
strength and peelability are important. The dry properties of an adhesive can be
assessed by using oscillatory and other dynamic measurements. The tackiness of a
PSA will be proportional to how well the material can mould to the surface (ie wet the
surface) that it is applied to. The holding strength and peelability are also quantifiable
via the elasticity of the adhesive.

Obtaining a Coatable Flow Profile

High-speed roller coatings need to fulfill several rheological requirements. Their low shear
viscosity will affect the storage, roller pick-up and levelling characteristics of the coating. An
increased low shear rate viscosity will reduce the sedimentation rate of any particles and make
the roller pick up a heavier mass of the adhesive. Conversely, this product may not level so
easily after coating. Whereas the viscosity at high
shear rates should indicate the transfer properties
and mechanical stability. If the viscosity is lower
at high shear rates, this usually indicates better Electronic gap
Adjustment system
transfer properties between rollers. With coating
speeds getting faster, the rheology needs to be both High shear low inertia
well defined and stable to avoid catastrophic shear Motor and air bearing
thickening. The goal is to have high shear
coatability whilst keeping the drying time to a
minimum, thus reducing cost.

Cone measuring system

Rheometer Equipment Used for the

Analysis Peltier Plate temperature
control

With a Bohlin rotational rheometer, the viscosity

can be characterized across a wide shear rate range
to simulate the shears seen in most coating Figure 1 A Bohlin Rheometer with a Cone and Plate system
processes. The flow curve gives valuable
information to adhesive formulators and enables Piston
more reliable production of adhesive coatings.
There are two instruments that are commonly used
to measure the shear viscosity of adhesive
dispersions: rotational and capillary rheometers. In
a rotational rheometer, the sample can be measured
using a tapered plug, cone and plate or parallel
plates to determine its viscosity up to 105 s-1. In Platinum
this case the sample is placed between the Resistance
Thermometer
measuring plates and a shear rate is applied to the
upper measuring system. The shear stress is then
calculated from the torque applied to the motor and
the dimensions of the measuring system. A flow
curve is obtained by using a table of shear rates Adhesive
Sample
where the equilibrium viscosity is taken at each
shear rate. With the cone and plate and parallel Pressure
plate systems, the sample is contained in the transducer

measuring gap by its surface tension. At the

Die
highest shear rates, the sample can fracture and
come out of the gap. To obtain higher shears a

Figure 2 A Rosand capillary rheometer set-up

smaller gap is generally used, having the dual effect of giving a higher interfacial tension
between the plates and also reducing the rotational speed required giving the same shear rate.

For very high shear rates (105 s-1 - 2 x 106 s-1), a capillary rheometer is usually used. In this
apparatus the sample is
squeezed at a controlled rate
Acceptable min to max
through a very narrow die (eg. 2 Reverse Roll
viscosity of various coatings

0.30 mm). The pressure drop

across the die can be used to 1.5
calculate the shear stress, Meyer Bar

whereas the piston speed and (Pas)

die dimensions give the shear
rate.
Reverse Gravure
0.5
Slot Die
Acceptable viscosity ranges for
typical coating processes are 0
shown in Figure 3i. If the
0.1 1 10 100 1,000 10,000
adhesives viscosity profile (s-1)
needs to be changed, this can be
Figure 3- Acceptable viscosity ranges for various coating processes
carried out by adjustment of the
particle size distribution and
loading of the latex, or by addition of fillers or ultra fine additives such as fumed silica. This is
discussed in more detail below.

Optimizing a PSA formulation for loading and particle size

With a correctly defined size range of the polymer particles it is possible to get near Newtonian
flow behavior even at solids contents above 60% up to shear rates of around 106 s-1.
Commercially available latexes are usually in the size range of 150 800 nm. The upper size
being limited by the time in which aqueous dispersions can be polymerized economically. At
very small sizes, the colloidal interactions between the particles contribute considerably to the
viscosity.

The Krieger-Dougherty equation (Eq 1, below)ii shows that for hard spheres the rheological
properties are directly related to the volume occupied by the particles.

- Viscosity of the suspension,

0 viscosity of the medium,
- Volume fraction of solids in the suspension,
m maximum packing fraction (63% for random close
packing).
[] Intrinsic viscosity (2.5 for spheres)

Equation 1 The Krieger-Dougherty equation

The volume fraction of solids () is related to the solids loading by the densities of the particles
(p) and continuous phase (c). Therefore the viscosity of a sample increases markedly with
increased particle loading, especially as we tend toward the maximum packing fraction. At
solids loadings of 50 60% samples will be shear thinning from low to moderate shears,
however at high shear rates (above 104 s-1) shear thickening occurs due to the formation of
clusters of particles called hydroclustersiii. These have the effect of immobilizing part of the
aqueous phase and hence the viscosity increases. The shear thickening effect is usually
reversible when the molecules are separated by a lubricating solvent layer, however can be
permanent where shear induced aggregation occursiv. The relationship for viscosity vs. particle
loading is shown in Figure 4iii.

If we use a fixed solids loading

102 45% solids
55% solids
and see the effect of varying 61% solids
only the particle size for a
monomodal latex we find that 101
small particles infer more low
shear viscosity, however at high 100
shear rates the viscosities are (Pas)
asymptotic. This is due to a 10-1
reduced mean particle separation
for small particles, giving rise to
10-2
stronger colloidal repulsive
forces resisting the shear flow at 10-1 100 101 102 103 104 105 106
low shear. At higher shears the (s )-1

flow properties are dominated by Figure 4 - Effect of solids loading on viscosity. Graph shows viscosity vs. shear rate
for a monomodal polymer dispersion of 280 nm.
hydrodynamic interactions,
which are independent of the
particle size.

It is common knowledge that making a suspension with a bimodal particle size distribution
reduces the viscosity. This is due to the smaller particles filling the voids between the larger
particles giving the system better packingv,vi.

For studies carried out using glass beads, packing improves as the size ratio becomes larger
(large particle size: small particle size)vi. However, this is not the case for polymer
suspensions; as the colloidal repulsive forces have a greater effect as the particle size reducesvii.
With a colloidal dispersion, the closest that neighboring particles can come is larger than that
of twice the radii due to the repulsive forces, and despite reducing the size (and hence mass) of
the particle, we barely reduce the effective space that it takes up below a certain size.
Therefore the use of high solids
latexes below 100 nm diameter
1.2
is likely to be inefficient. The
optimum size ratio for a low 1.0
viscosity dispersion is around
0.8
4.5, where we would use say, /small
175 nm latex in 750-800 nm 0.6
latex. The relationship between
0.4
viscosity and particle size ratio
at low shear is shown in Figure 0.2
5iii. It can also be shown that
0.0
widening the particle size
distribution further reduces the 0 10 20 30 40 50 60 70 80 90 100
overall viscosity of the Volume fraction of small particles (%)
suspension. Figure 5- Effect of particle size ratio on viscosity of 61% solids bimodal suspensions at1 s-1.

If sedimentation is a problem, or
the coating doesnt pick up sufficiently thickly on the roller, this can be addressed by
increasing the very low shear viscosity. In this case the addition of fumed silica or other ultra
fine additives is usually effective. Fumed silicon dioxides such as Aerosil 200 from Degussa
or similar products from Cabot thicken materials due to their highly negatively charged
surface. The ultra fine nature of these additives make them hard to disperse, however a high
shear mill, ball mill or even a Cowles blade can be used. Additions of up to 1% of fumed silica
are normally sufficient if well dispersed.

Characterizing the Dry Adhesive for Tack, Holding Strength and Peelability

A dry film of adhesive can be evaluated rheologically by using dynamic oscillation tests. Due
to the high modulus of most dry adhesives, small parallel plate geometries are the most
commonly used.

An adhesive film of 100 m or thicker is first prepared by drying the dispersion onto release
paper. The dry film can then be applied directly to one of the rheometer plates and the gap is
closed with a set normal force. The normal force should be controlled during the oscillation
tests, as compression of the sample can give a falsely high elastic modulus value. The film can
then be trimmed to exactly fit the plate diameter, or always left larger that the plates giving just
comparative results.

In the test, microscopic

106 Good Adhesive
oscillations are applied to the Bad Adhesive

sample to measure the

viscoelastic structure without 105
destroying it. An oscillation Increased Tack
strain is chosen within the linear 104
viscoelastic region by first G (Pa)
running an amplitude sweep. 103
Using this strain, the oscillation Peel Resistance
frequency is varied to measure
102
the viscoelastic properties of the
adhesive. The data shows how 10-1 100 101 102
the sample will respond over (rad/s)
different timescales. Figure 6 Figure 6 Good and bad diaper adhesives. The low frequency response shows tack and
shows how the elastic modulus holding strength, whereas the high frequency response shows peelability
may vary for good and bad
diaper tab adhesives. The low frequency response indicates the samples behaviour over longer
time scales. In this case a lower elastic modulus implies that the dry adhesive will mould into
(wet) the mating surface better and so give more tackiness. This low frequency response for
good tack needs to be balanced against the need for the adhesive to have a good holding
strength. If the material is too viscous it will slowly pull apart, eventually giving a cohesive
failure. The higher frequency response shows the materials behaviour under higher shears and
in this example, indicates the peel resistance. In peeling off a PSA, the elastic modulus should
not be too high, as we want the material to fail cohesively. Therefore it can be seen that the red
curve shows a PSA with good tack but poor holding strength, and which will be more difficult
to peel off. The blue curve (dashed) is the better PSA for this application.

The modulus values for pressure sensitive tape and label adhesives are often higher and typical
data is shown in Figure 7 for two samples. The phase angle of the adhesive film at low
frequencies is also an indicator whether the adhesive will have good tack or not. A higher
phase angle indicates that the
sample is more viscously natured
and will therefore mold to the
surface more easily than one that
has a low phase angle. A higher
modulus indicates that the
adhesive is stiffer and will
therefore have more cohesive
strength. Therefore it could be
expected that a higher modulus
and lower phase angle will create a
peelable PST, whereas a higher
Figure 7 Typical modulus values for PS tape adhesives. The modulus (red) shows how
phase angle and lower modulus stiff the sample is and the phase angle can be related to tackiness
will give a very sticky adhesive
with less cohesive strength.

Conclusion
The storage stability, coatability and dry properties of an adhesive can be quantified easily
using a Bohlin rotational rheometer. The very high shear coating properties (gravure and
reverse gravure) can be simulated using a Rosand capillary rheometer fitted with a narrow die.
The use of a rheometer during formulation development allows experimentation with low
volume samples, so that the particle size distribution, loading and additive dosing can be
optimised quickly. The solids loading directly affects the viscosity of the adhesive, with
viscosity increasing rapidly as you tend towards the maximum packing fraction. The particle
size distribution affects the minimum viscosity that can be attained, with a broad particle size
distribution giving the lowest viscosity profile. The average particle size determines whether
the sample will be highly shear thinning (as with ultra fine colloidal particles) or more
Newtonian (as with coarser materials). The dry adhesive films tackiness, holding strength and
peelability can also be determined from dynamic oscillation measurements.

For more information or any questions please contact the authors at Bohlin Instruments, 11
Harts Lane, East Brunswick, NJ. 08816. Or email: prolfe@bohlinusa.com or
joanne.langridge@malvern.co.uk.

Acknowledgements

Thanks is given to Dr N Willenbacher at BASF for use of some data from his paper, to B
Ludbrook at HB Fuller for samples and informative discussions and DRT (Les Drivs
Rsiniques et Terpniques) and NND were also of great help providing samples. Thanks are
also given to Dr R Marsh, Bohlin Instruments for the PSA data for diaper closures.

i
Courtesy of H.B. Fuller Company Ltd, UK.
ii
I M Krieger and T Dougherty, Trans. Soc. Rheol. 3, 137 (1959).
iii
N Willenbacher, L Borger, D Urban, Tailoring PSA-Dispersion Rheology for High Speed Coating, Cover story.
www.adhesivesmag.com, Nov, 2003.
iv
Husband J. C. , Adams J. M. , Shear-induced aggregation of carboxylated polymer latexes , Colloid Polym. Sci.
, 270 (12) , (1992) 1194 1207
v
R J Farris, Trans Soc Rheol. 12, 281-301 (1960)
vi
J S Chong, E B Christiansen, A D Baer. J App Polym Sci. 15, 2007-2021 (1971)
vii
B Dames, B R Morrisson, N Willenbacker. Rheol Acta. 40, 434-440 (2001)
Philip Rolfe is Rheology Applications Manager and based out of the Malvern
offices in the US. He began work for Bohlin Instruments in the UK laboratory
in 1998, after attaining a First class honours BSc at Plymouth University
and working for blue chip companies such as GlaxoSmithkline and Tyco
Electronics. He has extensive analytical experience with all Bohlin and some
competitor rheometer systems. He is particularly knowledgeable in the fields
of polymers, adhesives and sealants rheology.