www.epotek.

com

A Meridian Adhesives Group Company

EPOXY ADHESIVE
TEST
MEASUREMENT
GUIDE
 ABOUT THIS GUIDE
This guide is an educational tool designed to aid adhesive users in gaining a more
thorough understanding of adhesive properties and testing methods. It was compiled
through the combined efforts of multiple departments at Epoxy Technology, including
Quality Control, Research & Development, and Technical Service.

Although we have based our information on the most up-to-date data, advances in test
methods and materials are constantly evolving. Please use this reference booklet as a
guide only, and follow any recommendations listed on the data sheets, along with any
technical information that may be included with your adhesive product shipment.

We hope that the information in this guide will be helpful and valuable to you in choosing
the best adhesive for your specific application.

ABOUT EPOXY TECHNOLOGY

Epoxy Technology, A Meridian Adhesives Group Company, offers an extensive line of
specialty adhesives including Electrically Conductive, Thermally Conductive, Optical
Grade Adhesives, UV Epoxy and UV Hybrid Adhesives.

All of our products are tested thoroughly and consistently in our state-of-the-art
laboratories to ensure product reliability. Epoxy Technology is very proud of its
recognized quality program, including comprehensive ISO 9001 and MIL-STD 883/5011
certifications, as well as RoHS Compliance and Green Partnerships.

Our EPO-TEK® brand adhesives have been recognized worldwide for quality and
reliability since 1966.

www.epotek.com

2
Table of Contents

1.0 Viscosity/Rheology 4

2.0 Pot Life and Working Life 8

3.0 Adhesive Curing 10

3.1 Kinetics of Cure 12
3.2 Glass Transition Temperature (Tg) 16

4.0 Thermogravimetric Analysis (TGA) 20

4.1 Degradation Temperature 21
4.2 Outgassing 22

5.0 Lap Shear and Die Shear 25

6.0 Dynamic Mechanical Analysis (DMA) 28

7.0 Thermomechanical Analysis (TMA) 39

8.0 Electrical Conductivity (Volume Resistivity) 42

9.0 Thermal Conductivity 45

3
 1.0 Viscosity/Rheology
Viscosity is a term and value used to describe the internal friction of a fluid. The degree of
friction is observed when a force is applied to the material. The more friction detected, the
more force required to move the material or to "shear" it. Shearing occurs when the fluid is
actually moved. A low viscosity product like water requires less force to flow than a high
viscosity material, like peanut butter. The force required to shear a material can be
measured and reported using a Brookfield® Viscometer. One type is a cone and plate
viscometer shown in figure 1.

The cone and plate viscometer precisely measures the torque established by the
resistance in a fluid, sandwiched between a rotating cone and stationary plate. As torque is
directly proportional to the shear stress in the material, it can easily be converted to
absolute centipoise (cPs) units through known geometric constants of the cone, the rate
of rotation, and the stress related torque (or from pre-calculated range charts). The plate
portion of the assembly is the sample cup (shown in figure 1) which holds the fluid and
remains stationary. It is jacketed so that a constant temperature is maintained throughout
the test. The cup is locked onto the viscometer so that the material comes into contact
with the cone.

figure 1

Newton defined viscosity by the model in figure 2 below. Two parallel planes of fluid of
equal area "A" are separated by a distance "dx", moving in the same direction, but at
different velocities (V1 and V2 ). He assumed that the force to maintain this difference in
speed was proportional to the difference in speed throughout the liquid. Also called the
velocity gradient. To express this, Newton said that "η" is a constant for a given material and
is called its "viscosity." The velocity gradient, dv/dx, is a measure of the change in speed at
which the layers move with respect to one another. This describes the shearing a liquid
goes through and is referred to as the "shear rate (S)" which is reported in "reciprocal
second" (sec-1). The term F/A indicates the force per unit area required to produce the
shearing action. It is referred to as "shear stress" and will be symbolized by "F́ ". Its unit of
measurement is "dynes per square centimeter" (dynes/cm 2 ).

figure 2

4 Viscosity/Rheology
Using these simplified terms, viscosity may be defined mathematically by this formula:

η = Viscosity = F/S = shear stress/shear rate

ASTM Test Method D2196, "Standard Test Method for High-Shear Viscosity Using a Cone
and Plate Viscometer", is based on the above definition and is followed for viscosity
measurements.

The fundamental unit of viscosity is the "poise." If a material requires a shear stress of one
dyne per square centimeter to produce a shear rate of one reciprocal second, it would
produce a viscosity of one poise, or 100 centipoise. Viscosity can also be expressed in
"milli-Pascal-seconds" (mPa∙s). One milli-Pascal-second is equal to one centipoise. Newton
assumed that all materials have, at a given temperature, a viscosity that is independent of
the shear rate. In other words, twice the force would move the fluid twice as fast. This holds
true for Newtonian fluids only. The following is a schematic (figure 3) of a cone and plate
viscometer in action:

figure 3

The type of flow described above is true for Newtonian fluids only. The following graphs
show how this type of product behaves. Figure 4 shows the straight line relationship
between shear rate (F') and shear stress (S). Figure 5 shows how the viscosity remains
constant (assuming temperature is held constant) with varying shear rates.

figure 4 figure 5

But not all fluids are Newtonian in their flow characteristics. In fact, there are several types
of flow behavior. These products are classified as Non-Newtonian. There are three
subgroups within the Non-Newtonian classification: Pseudoplastic, Dilatent and Plastic.

www.epotek.com Viscosity/Rheology 5
 PSEUDOPLASTIC materials will decrease in viscosity with an increase in shear rate. This is
sometimes referred to as shear thinning. Figures 6 & 7 depict this:

figure 6 figure 7

A DILATENT type of behavior as seen in figures 8 and 9 below, causes products to

increase in viscosity with an increasing shear rate. This is also called shear thickening and is
commonly seen with deflocculated solids such as clay.

figure 8 figure 9

The third and final type of flow is PLASTIC. When a material exhibits Plastic behavior similar
to figures 10 and 11 below, it will remain in a "solid" state until a certain amount of force (yield
value) is applied to it before it will flow. A great example of this is ketchup.

figure 10 figure 11

6 Viscosity/Rheology
Epoxies can exhibit any or all of these flow
behaviors which is important to consider
when choosing an adhesive for a specific
application. Thixotropy is another parameter
that is related to viscosity and can be
measured using a viscometer. A thixotropic
fluid decreases in viscosity with time, while it
is subjected to constant shearing as seen in
figure 12. This type of behavior can occur in
combination with any of the above flow figure 12
types.

When talking thixotropy, often we refer to Thixotropic Index (TI). This is a value reported
by taking the ratio of two separate readings at different speeds on a viscometer. For
example, if a material produces a reading of 15,000 cPs at 1 RPM and 10,000cPs at 10
RPM, the thixotropic index is equal to 15,000/10,000 or 1.5. Usually, TI is calculated from
viscosities measured at RPMs that are a decade apart (i.e between 1 and 10 RPMs or
between 10 and 100 RPMs). This number is also key in choosing the proper type of
material for a particular dispensing technique.

All of the preceding terminology and definitions are also "rheological parameters".
Rheology is defined as the science of the deformation and flow behavior of materials.
Flow behavior is important in many industrial processes and the success or failure of an
application is widely based on these properties for a specific adhesive. Coatings, molded
plastics, adhesives, personal care items, cosmetics, inks, cement, solder pastes and
medicines are examples of the range of commercially available products whose viability
depends on having the proper rheology. For all of the above materials, the necessary
rheological properties must be examined before processing and application.

The following graph (figure 13) shows how a user may want to balance viscosity and
thixotropy in order to determine the best application method for a given adhesive:

Thixotropy vs. Viscosity

figure 13

www.epotek.com Viscosity/Rheology 7
 2.0 Pot Life and Working Life
Pot life is a metric used to define the length of time that a material can be handled before
its viscosity changes substantially. Generally, pot life is defined as the amount of time it
takes for the initial viscosity to double. Figure 1 below shows the increase in viscosity over
time for a typical low temperature cure epoxy system. You will notice that the viscosity
builds slowly at first as the chemical reaction begins. However, the reaction itself gives off
heat which, in turn, increases the rate of reaction. This process causes the viscosity to
increase in an exponential manner.

figure 1

As shown in figure 2, this material has about 3 hours of pot life, i.e. it takes three hours
before the viscosity doubles (100% change).

figure 2

Often, knowing the pot life alone is not sufficient to determine how long a given material
can be used for a given manufacturing process. For example, figure 3 shows the
normalized (% change) viscosity versus time for two different products. Both the blue and
green products exhibit the same percent change in viscosity versus time. Since both
formulations double their viscosity after 3 hours (shown as a 100% change), they both
would have 3 hours of pot life.

8 Pot Life and Working Life

However, as figure 4 shows, the two materials have very different starting viscosity values.
The starting viscosity of the green material is ten times greater than that of the blue
material. As a result, the magnitude of viscosity change over a given period of time for the
green product is much higher than the blue product. In these cases, the process limits for a
specific application should be used to define the working time for a given material in a given
process.

figure 3

figure 4

In the example above, customer A may be able to work with the blue material as long as the
viscosity is lower than 3,000 cPs. This would give customer A, 5 hours of working life (2
hours longer than the pot life). On the other hand, customer B may only be able to work
with the green material if the viscosity is less than 12,500 cPs. In this case, the material
would have a working life of only 2 hours in customer B’s process.

Thus, pot life is a data sheet value designed to give a first order description of the speed of
viscosity increase over time. However, the actual working time of the product will be
application specific and should be determined and defined based on the individual process.

www.epotek.com Pot Life and Working Life 9

 3.0 Adhesive Curing
The word polymer comes from the Greek words poly (meaning many) and meros (meaning
parts). As such, polymers are long molecule chains composed of many thousands of small
units (monomers) joined end to end. An example of linear polyethylene, the simplest
hydrocarbon polymer, is shown in figure 1:

Linear
Polyethylene

figure 1

Polymers fall into two main categories: thermoplastics and thermosets.Thermoplastic

materials are generally composed of linear polymer chains with little or no branching or side
groups similar to the thermoplastic polyethylene polymer in the diagram above. As a result,
these chains are able to easily slide past each other when heated. In this way,
thermoplastics can be heated and formed, then re-heated and re-formed repeatedly
without damage.

On the other hand, thermosets undergo a chemical reaction when initially heated and cured
to form a three-dimensional cross-linked network. Once these materials are cured, this
structure is locked in place and the material cannot be reformed or reprocessed. An
example of the chemical crosslink structure for a common type of thermosetting polymer,
an epoxy, is shown in figure 2.

10 Adhesive Curing
Example of Thermoset Cross-Linking Reaction

figure 2

Epoxies get their name from the three membered "epoxide" rings located on the starting
epoxide monomer. In the case above, the starting monomer is actually a diepoxide, which
contains two epoxide groups one at each end of the monomer. The hardener in the above
system is a diamine, which contains a reactive amine group at either end of the molecule.
Each of the epoxide rings can be opened by an active hydrogen on the end of the diamine
to produce chemical bonds between the epoxide monomer and the hardener.

This particular diamine hardener provides four active hydrogens that are capable of
reacting with four different epoxide monomers to chemically cross-link them together.
Typically, the higher the functionality (number of reactive sites) of the hardener, the more
tightly cross-linked the resulting cured thermoset will be. In general, highly cross-linked
systems will have higher hardness and more thermal, chemical and moisture resistance.

www.epotek.com Adhesive Curing 11

 3.1 Kinetics of Cure
The cure kinetics rate for thermosetting adhesives can be monitored using Differential
Scanning Calorimetry (DSC). DSC measures the energy absorption of a sample as it is
subjected to a specified thermal profile. The amount of heat contained in a material at any
given time is known as its enthalpy. As additional thermal energy is supplied to the material,
its enthalpy will increase. The level of the resulting temperature change in this material will
be dictated by its specific heat. At a given temperature, specific heat is the amount of
thermal energy needed to change the temperature of that material by one unit.

Figure 3 below shows the cross-section of a typical DSC sample holder. A small amount of
uncured sample (~5-10mg of resin content) is placed in a small aluminum sample pan with
lid. The sample pan is placed on top of an individual chromel heating disc. An identical
empty aluminum pan is placed on a second chromel heating disc to serve as a reference for
the experiment. Energy is supplied separately to the sample pan and the reference pan in
order for each of them to exactly match the heating rate of a predetermined temperature
profile. The rate of energy supplied to the sample to heat it at a controlled rate is
proportional to the specific heat of the sample.

The specific heat of a material will change only slightly with temperature as long as that
material remains in the same physical state. However, when the thermal profile imposed on
a material causes a transition to a new state, (i.e. melting, decomposition, curing, etc.) this
will cause a significant change in the specific heat. The rapid change in specific heat at a
transition, then causes a change in the amount of energy needed to be delivered to the
sample in order to maintain the desired heating or cooling profile.

Standard
DSC Cell*

figure 3
*Courtesy of TA Instruments

12 Adhesive Curing
Changes in state that absorb energy are known as endothermic changes. Melting is an
example of an endothermic state change. Heat is put into the system to allow molecules
the mobility to begin to move. Eventually, enough heat is put into the system to cause
overall flow of the melted material. At this point, the heat energy becomes converted to
the kinetic energy of the liquid. On the other hand, changes in state that actually evolve
energy are exothermic transitions. Heat of reaction (curing) and crystallization are
examples of exothermic transitions. Figure 4 shows some typical endothermic and
exothermic transitions as they would appear in a DSC scan.

figure 4

As mentioned earlier in this section, DSC can be used to monitor the cure of a
thermosetting material. This is accomplished by subjecting the uncured thermoset to a
controlled temperature ramp and analyzing the location, size and shape of the resulting
exothermic cure peak.

Figure 5 shows a DSC kinetic curing profile for a typical epoxy system. The peak in the
sample’s heat flow response is located above the baseline, representing an exothermic
event. The peak of this curing exotherm is located at approximately 150˚C, indicating that
the fastest cure for this product will occur at this temperature. The onset of the cure peak
for this product is 140˚C.

www.epotek.com Adhesive Curing 13

 However, this does not mean that the product cannot cure below 140˚C. As the enlarged
portion of the kinetic curve reveals, there is actually a small amount of exotherm starting as
low as 80˚C. This particular product will cure at 80˚C, but it will take around four hours. On
the other hand, when this product is cured at 150˚C, the peak cure temperature, it will only
take about 10 minutes.

DSC Kinetic Cure

figure 5

The total change in enthalpy that the sample undergoes during cure can be calculated by
integrating the area under the exotherm peak. For the material in the diagram above, the
total heat of reaction is 450J/g. As specified in ASTM D3418, "Standard Test Method for
Transition Temperatures and Enthalpies of Fusion and Crystallization of Polymers by
Differential Scanning Calorimetry", this total heat of reaction can then be used to determine
the extent of reaction of the material after it is cured according to a desired cure condition.
The cured material is analyzed in the DSC using the same thermal profile used to
measure the uncured material. Figure 6 shows the original exotherm of the uncured
material above, along with that of the same material cured at 150˚C for 10 minutes.

14 Adhesive Curing
 DSC to Calculate Residual Cure

figure 6

The material cured at 150˚C for 10 minutes exhibits only a very small residual exotherm
peak. This indicates that most of the theoretical cross-linking reactions were completed
during the cure. The enlarged view of the cured sample exotherm shows that only 18.5J/g
of reactivity remains of the original 450 J/g heat of reaction for this material. Thus, only
4.1% of the material is left unreacted by the imposed cure. Epoxies do not need to achieve
a full 100% reaction in order to perform well as adhesives. Generally, systems that have
been cured to at least 90% conversion will exhibit mechanical and physical properties that
do not differ significantly from their fully reacted state. In addition, even lower degrees of
conversion may produce properties that are perfectly sufficient for a given application.

DSC can also be used to determine how long a particular material will need to be cured at a
given temperature in order to reach the maximum degree of conversion at that
temperature. This is achieved by measuring the exotherm curing peak of a material
exposed to a constant temperature over time or an isothermal scan.

www.epotek.com Adhesive Curing 15

 Figure 7 shows the different exotherm peaks for a typical epoxy cured at 80˚C, 90˚C,
100˚C and 110˚C. If the material below is cured at 80˚C, it requires close to an hour to
reach the full conversion level at this temperature. However, if it is possible to increase the
cure temperature to 110˚C, the material will need less than 15 minutes to reach final
conversion.

figure 7

One important note is that the faster a thermoset is cured, the more stress is imparted to
it. As a result, it may not always be desirable to cure quickly at the highest temperatures.
For some applications, it may be necessary to ramp the temperature to the desired cure
condition, or start with a lower temperature precure to yield lower stress bond lines.

3.2 Glass Transition Temperature (Tg)

As discussed in the beginning of this section, the chemical cross-links that form during the
cure of thermosetting materials do not allow the cured material to melt or flow when
re-heated. However, cross-linked materials do undergo a very slight softening at elevated
temperatures at a point known as the Glass Transition Temperature (Tg).

The Tg is not a discrete thermodynamic transition, but is actually a temperature range over
which the mobility of the polymer chains increases significantly and the bulk material
changes from a rigid/glassy state to more of a rubbery/ compliant state. The temperature
range over which this transition takes place is very dependent on the level of chemical
cross-links and physical chain entanglements in the system, as well as the degree of
flexibility of the polymer chains themselves.

16 Adhesive Curing
Polymers with rigid backbone segments, bulky side groups and a high crosslink density will
require very large amounts of thermal energy to provide the system with enough mobility
to transition to a rubbery state. Such systems, like that shown in figure 8b below, will have a
high Tg. On the other hand, systems like that shown in figure 8a, which have long sections
of flexible polymer chains between cross-links, will shift to a rubbery state at significantly
lower temperatures, resulting in a lower Tg.

Tg is an endothermic state change where heat energy must be supplied to the material in
order to transition to a rubbery state. This transition can also be characterized using DSC in
accordance with ASTM E1356, "Standard Test Method for Assignment of the Glass
Transition Temperatures by Differential Scanning Calorimetry". Samples are prepared and
tested in the same manner used for measuring curing kinetics, with the exception that the
sample is already cured. The sample can either be cured outside the DSC using an
application specific cure profile (ie. box oven, tunnel oven, heated tooling, etc.), or can be
cured in the DSC during a previous kinetic or isothermal scan.

figure 8

www.epotek.com Adhesive Curing 17

 The format of a Tg scan is very similar to that of the kinetic scan. Temperature is plotted on
the X axis and the heat flow response on the Y axis. However, unlike the large exothermic
peaks evolved during the kinetic cure, Tg is exhibited as a small endothermic step change
in the heat flow. Figure 9 below shows a typical Tg for an epoxy. The Tg is actually a
transition region, rather than a specific temperature. In this case, the transition begins
around 60˚C and is complete by about 100˚C. However, the convention is to report a
single temperature defined as midpoint of the transition region as it is bounded by
tangents to the two flat portions of the heat flow curve. For this material, the Tg would be
reported as 75˚C.

figure 9

The reason that the Tg spans a temperature range rather than occurring at a specific
temperature, is that the cross-linked polymer chains take time to increase their mobility in
response to applied thermal energy. With just a little bit of applied energy, some of the
more flexible chain segments and side groups can begin to move a little bit. This small-scale
movement allows an increase in free volume that in turn allows the freedom for some of the
larger and more rigid segments to begin to move once higher levels of energy are applied.

Thus, even at constant temperature, the transition to the rubber-like state will take some
time to evolve. However, a material's Tg is measured by subjecting it to some sort of
temperature ramp. The rate of this scan will play a large role in determining the
temperature location of the Tg. Slower scans will allow more time for the polymer chains to
orient in response to the applied energy and will result in a lower temperature
measurement of the Tg.

On the other hand, a very quick scan will cause the temperature to increase more quickly
than the chains are able to respond to the supplied energy, and the resulting Tg response
region will be pushed to higher temperatures. Thus, it is very important to check the test
conditions under which Tg has been reported. Ramp rates of 10˚C/min and 20˚C/min are
typical testing conditions used in the industry. The 20˚C/min condition is frequently used
as it cuts testing time in half. However, it will result in a higher reported value for Tg.

18 Adhesive Curing
The Tg of a material will also depend on its degree of cure. Generally, the reported Tg for a
material will be based on 100% cure/conversion. However, depending on the cure schedule
used for a given application, a lower degree of conversion may be attained. This may have
less cross-linking, and therefore, will result in a lower Tg.

A high Tg is not necessarily required for good performance of an adhesive. In many cases,
it is actually desirable for a material to operate above its Tg. Since the system is chemically
cross-linked, it will not degrade in performance above the Tg, it will just be in a softer state.
Softer/tougher materials are often desirable for applications in which resistance to thermal
or mechanical shock is required.

High Tg materials are generally desirable for applications which require low movement in a
given temperature range or very high temperature resistance.

Figure 10 shows Tg scans for some typical low, medium and high Tg epoxies.

figure 10

www.epotek.com Adhesive Curing 19

 4.0 Thermogravimetric Analysis (TGA)
TGA is a method for measuring very small weight changes in a material as a function of
temperature. This technique is used to determine thermal degradation temperature. It can
also be used to measure moisture content and level of outgassing. Additionally, it is used as
a method to check the content of inorganic fillers (such as silver flake) in formulations for
QC lot acceptance. ASTM D3850, "Standard Test Method for Rapid Thermal Degradation
of Solid Electrical Insulating Materials by Thermogravimetric Method (TGA)", governs this
testing.

As shown in the schematic in figure 1, a TGA is essentially a very high precision

micro-balance in a controlled atmosphere. A sample basket is suspended from one arm of
the balance while a weighted tare pan balances the other side. The system is designed to
compensate for any changes in sample weight to maintain a balanced position.

As a sample is heated, it will normally begin to lose mass. At low temperatures, this is
generally due to either evaporation of moisture or low molecular weight species; while at
higher temperatures, larger weight losses due to thermal degradation will eventually occur.
The mass loss will cause the sample arm of the balance to tip up slightly. An infrared LED
light is mounted in the middle of the balance arm. When the balance tips, the light from the
LED moves off center and a photodetector notes a difference in the measured intensity.
Current is then supplied to a meter which rotates the balance until the light from the LED
is again centered on the photodetector. The amount of current required to return the
balance to its zero position is directly proportional to the displacement of the sample
balance arm and, hence, the mass change of the sample.

TGA 2950
Schematic*

figure 1
*Courtesy of TA Instruments®

20 Thermogravimetric Analysis (TGA)

 4.1 Degradation Temperature
The output of a TGA scan is generally a plot of percent sample weight versus temperature
(as shown figure 2). The degradation temperature of a cured material can be calculated in
two ways. The first approach is the onset degradation temperature. This temperature is
calculated based on the intersection of the tangent to the maximum change in slope of the
weight loss curve with the tangent to the baseline.

The second approach is the 10% weight loss temperature. This calculation is very straight
forward for unfilled systems. In this case, it is simply the temperature at which the material
has lost 10% of its starting weight. For filled systems, this weight loss is determined based
on a loss of 10% of the polymer component of the formulation. For example, if a conductive
material is 70% filled with silver flake, only 30% of the system is polymer. A loss of 10% of
the weight of the polymer component will be a 3% loss in weight of the overall system.
Generally, as seen in the system below, the two methods result in very similar values for
degradation temperature.

figure 2

As generally reported on data sheets, the suggested operating temperature windows is

calculated on the degradation temperature. The upper end of the continuous operating
window is calculated by subtracting 150˚C from the degradation temperature.

www.epotek.com Thermogravimetric Analysis (TGA) 21

 figure 3

For the example material in figure 3, the degradation temperature is approximately

420˚C. Thus, 270˚C would be calculated for the suggested operating temperature
window. Generally, we would add a safety factor so that the datasheet would actually list
250˚C as the continuous operating windows.

4.2 Outgassing (Thermal Stability)

The percent weight versus temperature curve is also used to determine weight loss at
specified temperatures. This weight loss is often referred to as outgassing or thermal
stability. Datasheets generally report outgassing levels at 200˚C, 250˚C and 300˚C.
<1.0% @ 200˚C is required by MIL-STD 883 Method 5011 (low outgas)
Figure 4 below shows an example of outgassing calculations.

figure 4

One important industry specification is the NASA outgassing requirement. Products that
meet this specification must exhibit less than 1.0% Total Mass Loss (TML) after being
exposed to 125˚C for 24 hours in a vacuum. They must also contribute less than 0.1%
Collected Volatile Condensable Materials (CVCM) during this exposure.

22 Thermogravimetric Analysis (TGA)

EPO-TEK Materials Meeting NASA Outgassing Requirements
Product TML* CVCM** Cure Time Cure Description
(<1.0%) (<0.1%) Temp

301 0.98 0.01 1 Hour 65° C Optical

301-2 0.70 0.01 3 Hours 80° C Optical

302-3M 0.70 0.01 7 Days 25˚ C Optical

314 0.77 0.00 2 Hours 120˚ C Optical

320 0.99 0.00 2 Hours 65° C Black

353ND 0.81 0.01 1 Hours 120˚ C Optical

354 0.31 0.00 3 Hours 80° C

Optical
Hours +100° C

377 0.54 0.02 1 Hours 150˚ C Optical

390 0.43 0.01 30 Min. 25˚ C Polyimide

920 0.65 0.01 45 Min. 80˚ C Alumina-Filled

930 0.49 0.00 45 Min. 80˚ C Boron Nitride-Filled

930-4 0.82 0.01 1 Hours 150˚ C Boron Nitride-Filled

E2101 0.32 0.01 1 Hours 150˚ C Silver-Filled

H20E 0.62 0.01 1 Hours 150˚ C Silver-Filled

H20E-D 0.54 0.01 1 Hours 150˚ C Silver-Filled

H20E-PFC 0.76 0.01 1 Hours 150˚ C Silver-Filled

H21D 0.39 0.00 1 Hours 150˚ C Silver-Filled

H22 0.99 0.01 20 Min. 100˚ C Silver-Filled

H27D 0.52 0.09 1 Hours 150˚ C Silver-Filled

H31 0.54 0.01 1 Hours 150˚ C Silver-Filled

H31D-LV 0.47 0.02 1 Hours 125˚ C Silver-Filled

H35-175MP 0.33 0.02 1.5 Hours 150˚ C Silver-Filled

H37-MP 0.27 0.01 1 Hours 150˚ C Silver-Filled

H44 0.27 0.00 1 Hours 150˚ C Gold-Filled

H63 0.19 0.01 1 Hours 120˚ C Alumina-Filled

H67-MP 0.37 0.00 1.5 Hours 150˚ C Alumina-Filled

H70E 0.99 0.03 12 Hours 60˚ C Alumina-Filled

H72 0.31 0.00 30 Min. 100˚ C Alumina-Filled

H73 0.43 0.01 20 Min. 100˚ C Alumina-Filled

H74 0.27 0.01 24 Hours 50˚ C Alumina-Filled

H77 0.22 0.00 1 Hours 125˚ C Alumina-Filled

H81 0.62 0.00 16 Hours 60˚ C Gold-Filled

MA-5 0.40 0.04 1 Hour 150˚ C B-Stage

OG198-54 0.53 0.00 UV Cure +

150˚ C Cationic UV
1 Hour

T7109 0.80 0.13 1 Hours 150˚ C Boron Nitride-Filled

T7110-38 0.60 0.01 7 Hours 70° C Alumina-Filled

U300-2 0.97 0.01 30 Min. 150˚ C

* Total Mass Lost ** Collected Volatile Condensable Materials

www.epotek.com Thermogravimetric Analysis (TGA) 23

 EPO-TEK Products Meeting MIL-STD 883/5011 Standards:
Electrically & Thermally Conductive Adhesives
EPO-TEK® Cure Outgassing (≤1.0%)
EK1000-MP 200°C - 60 min 0.11%
EK1000-1MP 200°C - 60 min 0.04%
H20E-MP 150°C - 60 min 0.25%
H35-175MP 150°C - 90 min 0.33%
H35-175MP-LV 180°C - 60 min 0.79%
H35-175MPT 180°C - 60 min 0.51%
H37-MP 150°C - 60 min 0.69%
H37MP-2 150°C - 60 min 0.30%
H37-MPT 150°C - 60 min 0.31%

Thermally Conductive Adhesives

EPO-TEK® Cure Outgassing (≤1.0%)
H65-175MP 165°C - 90 min 1.00%
H67-MP 150°C - 60 min 0.37%
H67MP-GB 150°C - 60 min 0.05%
H67-MP-T 150°C - 60 min 0.54%

The Fiber Optics industry also has a set of outgassing requirements. These are known as
Telcordia GR-1221. The cured adhesive is heated from 50˚C to 150˚C at 5˚C/min in the
TGA. A 0.1% weight loss or less is considered evidence of a properly cured thermal cure
system. While, 0.25% weight loss or less is considered evidence of a properly cured UV
system.

EPO-TEK Materials Meeting Telcordia Standard:

Telcordia GR-1221
EPO-TEK® Cure Weight Loss
323LP 150˚C - 60 min. 0.070%
353ND 150˚C - 30 min. 0.037%
375 150˚C - 30 min. 0.041%
383ND 150˚C - 30 min. 0.085%
OE184 150˚C - 30 min. 0.027%
OG116-31 UV - 2 min. at 320 - 500 nm 0.087%
OG198-54* UV - 2 min. at 320 - 500 nm 0.084%
* Material was post cured for 1 hour @ 150˚C

24 Thermogravimetric Analysis (TGA)

 5.0 Lap Shear and Die Shear
Lap shear strength and die shear strength tests are commonly used in the adhesive world
to characterize the bond strength of epoxies. Each measures the strength of an adhesive
on a specific substrate.

Lap shear strength is assessed using ASTM D1002, "Apparent Shear Strength of
Single-Lap-Joint Adhesively Bonded Metal Specimens by Tension Loading (Metal to
Metal Bonding)". This test allows not only the determination of shear strength of an epoxy
when bonded to aluminum, but also the examination of its mode of failure which is equally
as important.

Each lap shear specimen is created using two 1 inch wide acid etched aluminum coupons
with a 1/2 inch overlap. A side view of this assembly is shown in figure 1.

figure 1

Once the specimens have been assembled and cured according to the recommended cure
schedule on the data sheet, they are cooled to room temperature and placed one at a time
into the Instron® Test System for analysis. The two coupons are clamped vertically
between "grips" and pulled 180˚, at a constant rate set by the Operator. The pulling force
is increased on the coupons, until the adhesive joint fails resulting in a curve similar to figure
2. The tensile shear strength is then calculated from the curve using the following formula:

Lap shear strength (psi) = maximum load force (lbs.) / bond area (in2)

figure 2

www.epotek.com Lap Shear and Die Shear 25

 Typical values for lap shear strength range from 500 - 3000 psi. However, the strength
reading is not the only important parameter of the test. The mode of failure is equally, if not
more important to consider.

There are two types of failures that can occur adhesive and cohesive. When a material fails
adhesively, the adhesive completely loses its bond to the substrate. This occurs when the
adhesive strength of a bonding material is lower than its cohesive strength. With a cohesive
failure, the strength of the adhesive itself is not as great as the forces applied to it and the
adhesive pulls apart leaving portions of the adhesive bonded to both substrates.

Figure 3a & 3b show an example of both failure modes.

figure 3a figure 3b

In applications where a very high constant load is present, a high tensile strength adhesive
with good mechanical hold to a chosen substrate is recommended.

High: 2500 – 4000 psi

Average: 1200 – 1900 psi
Low: < 800 psi

26 Lap Shear and Die Shear

Die shear strength is a test common to the semiconductor industry. It is used to determine
the strength of a die bonded to a package substrate using a die attach adhesive. By
subjecting the die to a stress that is parallel to the substrate, the result is a shearing stress
at the interface between the die and die attach material. The purpose of die shear testing
is to gauge the overall quality of the bond to both die and substrate.

The following is an enlarged schematic of the test:

Die shear strength is typically reported in kilograms and can vary significantly due to die and
substrate composition as well as die size. For larger dies, a lower stress material is
recommended because of the greater surface area. Typical die shear testing is done on
2mm x 2mm gold dice on gold plated Kovar® substrates. The higher the value, the stronger
the bond.

High: >10 kg / 3400 psi

Med: 5 – 10 kg / 1700 – 3400 psi
Low: <2 – 5 kg / <1000 psi

www.epotek.com Lap Shear and Die Shear 27

 6.0 Dynamic Mechanical Analysis (DMA)
DMA measures a material’s response to an imposed oscillating stress or strain. Either the
imposed stress is controlled and the resulting strain in the sample is recorded, or the strain
is controlled and the resulting stress level is measured.

When a stress is imposed on a polymeric system, this stress will be dissipated throughout
the material and generally results in some level of deformation. Polymers are viscoelastic in
nature. As a result, only a portion of the deformation will be reversible when the stress is
released. This is the elastic portion of the response.

Recall Hooke’s Law for an ideal elastic solid:

σ = κ∗ε Where:

σ = stress
ε = strain
κ = spring constant

The above equation shows that the amount a spring will be stretched is proportional to the
amount of stress imposed on it. The proportionality is described by the spring constant, k.
The spring constant may be thought of as the stiffness of the spring. Another term for the
stiffness of a material is the modulus, E.

Thus, E = σ/ε

figure 1

Figure 1 above shows HOOKE'S LAW for the linear relationship between applied load and
elongation of an elastic spring.

28 Dynamic Mechanical Analysis (DMA)

However, as shown below (figure 2), if the load becomes too large, the spring will actually
undergo some permanent deformation and will no longer return to its original length when
the load is released.

figure 2

As shown in figures 3a and 3b below, when very low levels of stress are applied to a polymer
system, the coiled polymer chains will begin to stretch and straighten out slightly. This
results in a low level of deformation. However, like the spring, entropy will drive the majority
of chains to recoil back into their original preferred orientation as soon as the load is
released.

At higher loads, as shown in figure 3c, the chains will not only stretch and uncoil, but will
actually start to untangle and slide past each other. This deformation is not reversible. At
the extreme, the slipping of chains becomes similar to a viscous flow. If the system is not
cross-linked, the material will eventually pull apart. However, if the system is cross-linked, it
will stretch and flow until the cross-links become the only forces holding it together.
Eventually, the chains will break and the sample will fail.

20kg
5kg

figure 3

www.epotek.com Dynamic Mechanical Analysis (DMA) 29

 While HOOKE'S LAW describes the elastic component of a polymer’s stress/strain
behavior, Newton’s law of ideal fluids is often used to describe the other extreme of a
polymer’s response to stress viscous flow.

Newton’s equation shows that the amount an ideal fluid will flow in a given period of time
will be proportional to the amount of stress imposed upon it. The proportionality is
described by the fluid’s viscosity.

Often, car shock absorbers or dashpots are used to illustrate this viscous flow behavior. As
figure 4 depicts, the rate at which the liquid is forced through the holes in the cylinder is
dependant on the stress/load placed on the dashpot. Unlike the reversible deformation of
the spring, the displacement of the fluid in the dashpot is completely irreversible. Thus, all
work that has been put into the system is dissipated into the fluid.

figure 4

You will note that the viscous component of the deformation introduces time as a key
variable. As shown earlier, a polymer will respond to an applied stress first, by stretching and
uncoiling of the chains followed by disentanglement and finally slipping of the chains past
each other. These polymer rearrangements require time to occur.

30 Dynamic Mechanical Analysis (DMA)

As a result, over time, a polymer’s actual response to an applied stress looks more like
figure 5 below:

Applied Stress Strain Response

yield/ equilibrium

σ ε creep

} elastic
region
t t
figure 5

As the figure above shows, when a constant stress or load is applied to a polymer system,
the first response is elastic chain stretching/uncoiling resulting in elastic deformation. If the
polymer were a true spring, the deformation would quickly reach an equilibrium point and
no additional deformation would occur (like a weight hanging on the end of a spring).

However, with a visco-elastic system, the sustained load will cause the polymer chains to
start to untangle and slide a bit. The untangling process will also allow the polymer coils to
stretch a bit more. This slipping and additional stretching results in additional deformation
over time, called creep. Eventually, the chains will have stretched as far as they can and
further deformation will come from pure viscous chain slippage of any non-crosslinked
chains.

Many mechanical models have been developed to simulate visco-elastic behavior. One of
the most popular is the 4 element model shown in figure 6 below. The first spring governs
the initial elastic response, followed by a mixed region of spring and dashpot elements in
parallel and then, finally the purely viscous dashpot element in series with all of the above.

4 Element
Model

figure 6

www.epotek.com Dynamic Mechanical Analysis (DMA) 31

 When the stress/load is released, entropy will slowly cause the stretched polymer chains to
coil back to their original state. However, only the elastic portion of the deformation will be
recovered and the sample will remain somewhat deformed relative to its initial state
(figure 7).

figure 7

The final key element that can have an impact on the stress/strain behavior of a material is
temperature. As shown in figure 8, higher temperatures give increased mobility to polymer
chains. This results in faster and larger deformations at a given load than if the same
material was tested at lower temperatures.

figure 8

DYNAMIC TESTING As mentioned at the very beginning of this section, Dynamic

Mechanical Analysis generally measures the strain response of a material to an oscillating
applied stress. In some cases the applied strain is controlled, and the resulting stress level in
the material is measured, but this is less common. Figure 9 shows some of the many
possible test geometries for DMA.
sample

sample

test fixture
sample sampl e test
fixture

figure 9

32 Dynamic Mechanical Analysis (DMA)

Regardless of which form of stress is imparted to the sample (tensile, torsion or bending),
all DMA impart either a sinusoidal stress or strain to the given sample and measure the
resulting material response.

figure 10

Therefore, if the equation for the stress applied to material shown above is given by:

σ = σ0 sin(ωt) Where:

σ = stress at a given time

σ0 = the maximum applied stress
ω = the frequency of the oscillation
The material’s response will depend on its relative degree of elastic versus viscous
character. The more elastic the material, the closer the response will track with the applied
stress profile. However, the more viscous the material, the more time it will take to flow and
respond to the applied load and the deformation will be less reversible.

www.epotek.com Dynamic Mechanical Analysis (DMA) 33

 Therefore, the resulting strain response for the material shown in figure 10 will be out of
phase with the applied stress by an amount δ:

The strain can be expanded to the form:

This expression shows that the strain is actually composed of two terms that are 90˚ out of
phase with each other. The above equation can then be written in terms of the contribution
of real and imaginary components:

′ ′
The first term contains the real component. This represents the portion of the strain
response that is due to the elastic nature of the material. The second term is the out of
phase portion of the response, which reflects the ability of the material to dissipate energy
through viscous mechanisms. The complex modulus of the material can be calculated by
dividing the equation for stress by that for strain:

Ε∗ = σ/ε′+ σε″
Ε∗ = Ε ′+ Ε ′′ Where:

Ε′ = the real or "storage" modulus

Ε″= the imaginary or "loss" modulus
The ratio of the loss modulus to the storage modulus is known as the tan(δ):

tan(δ)= Ε″/ Ε ′

The tan(δ) is a meaningful property, as it represents the ratio of the viscous versus elastic
contributions of the material at a given condition.

While both the absolute and relative values of the storage and loss moduli at room
temperature are valuable, generally DMA is run as a function of temperature. Measuring a
material’s response to oscillating stress over a broad temperature range gives a wealth of
information about thermal transitions. For instance, DMA provides one of the highest
resolution techniques for measuring a material’s Tg.

34 Dynamic Mechanical Analysis (DMA)

 figure 11

The idealized DMA scan in figure 11 shows how changes in a material’s storage modulus with
temperature may be able to reveal a great deal about its thermal transitions. The most
commonly reported thermal transition for a thermosetting polymer is its Glass Transition
Temperature (Tg). As discussed previously, the Tg of a material is not actually a discrete
thermodynamic transition, but is actually a temperature region over which the mobility of
the polymer chains increases significantly and the bulk material changes from a rigid, glassy
state to a more rubbery/compliant state. This change is indicated by a significant drop in
the storage modulus as the material response changes from primarily elastic/ reversible
deformation to more and more viscous/irreversible chain motion. This large transition, the
alpha transition (α), is generally where the Tg is originated from.

Beyond the glass transition region is the "rubbery plateau" region where the ratio of elastic
to viscous contributions reaches an equilibrium. For a crosslinked thermosetting material,
this plateau is maintained until the degradation temperature is reached.
For non-cross-linked systems the melt temperature will eventually be reached and cause
liquid-like flow.

Figure 12 shows an example of a typical DMA plot for an epoxy resin. This plot shows both
storage and loss modulus as well as tan(δ). From this plot, the Tg region can be seen in
several ways. The first method for calculating the Tg is from the onset of the drop in the
storage modulus (E′) curve. This indicates the point where viscous motions begin to kick in.
The second method is the peak of the loss modulus (E″) curve. This indicates the point of
maximum chain slippage in a cross-linked system. Finally, the peak of the tan(δ) curve can
be used to determine Tg. This indicates the point where the ratio of viscous response to
elastic response is the highest.

www.epotek.com Dynamic Mechanical Analysis (DMA) 35

 figure 12

The peak of the tan(δ) curve is a popular measurement point for the Tg as it is usually easier
to isolate than to determine the onset of the drop in the storage modulus. However, this
case again highlights the fact that it is important to identify how a Tg was measured as each
of these three methods of determining Tg are accepted in the industry, but will give
different absolute values. Tg’s calculated from the tan(δ) peak will always be higher than
those calculated from the storage modulus. Generally, Tg’s calculated from the loss
modulus will fall somewhere in between.

The tan(δ) versus temperature curve can be used to determine much more about a system
than just its Tg. The width of the tan(δ) peak can indicate how homogeneous a system is.
Systems with very broad peaks are generally composed of a blend of different polymer
chain lengths or structures which gives rise to a broader temperature range, to initiate
significant viscous chain motions for the various components. On the other hand, systems
with narrow peaks generally have more narrow distribution of chain types and molecular
weights.

The height or amplitude of the tan(δ) curve is directly related to a material’s ability to
dissipate energy through segmental motion. Systems with tall tan(δ) peaks have higher
ratios of energy absorbing viscou s motions and are, therefore, generally tougher systems
than those with low tan(δ) amplitudes.

The final common use of the data from a DMA curve is the most basic—to determine the
modulus of a material at a given temperature for use in design calculations. Since the
modulus measures the stiffness of a material, this is often used in combination with
knowledge of the material’s CTE (which will be discussed in the next section) to determine
levels of stress that may occur during thermal stressing of bonded substrates.

36 Dynamic Mechanical Analysis (DMA)

The best way to reduce bond stresses during thermal cycling is to match the thermal
expansion character of the adhesive to that of the substrate. Unfortunately this is not
always possible. The next best choice may be to select a material with a lower modulus or
higher toughness that can absorb any stresses incurred by thermal expansion mismatch.
For these calculations, we generally use the storage modulus as we assume that the
deformations in the bond line will most likely be small enough to keep the adhesive within
its elastic limit.

Some typical ranges for Elastic Modulus are:

High: >900,000 psi (generally high Tg)

Med: 400,000 – 900,000 psi (generally med Tg)
Low: <150,000 psi

Figure 13 shows a schematic of the sample chamber of a typical DMA.

DMA7
Schematic*

Temperature Enclosure

figure 13 *Courtesy of Perkin Elmer®, Shelton, CT

www.epotek.com Dynamic Mechanical Analysis (DMA) 37

 3 Point Bend Set-up

figure 14

Figure 14 shows the sample geometry used to measure modulus at EPO-TEK.

A small, 3.0mm x 1.0mm x 18.0mm bar of material is tested in three point bending.

38 Dynamic Mechanical Analysis (DMA)

 7.0 Thermomechanical Analysis (TMA)
TMA is a technique for measuring very small expansions or contractions in a material as it is
heated or cooled. The linear expansion of a material in response to temperature is known
as the linear Coefficient of Thermal Expansion (CTE).

CTE can be measured using the same equipment as Dynamic Mechanical Analysis (DMA).
However, for CTE, the knife edge probe and three point bend fixture are replaced with a
flat bottom cylindrical probe and flat sample holder (as shown in figure 1 below).

DMA7
Schematic*

Temperature Enclosure

figure 1 *Courtesy of Perkin Elmer®, Shelton, CT

Typically, TMA measures the linear expansion (or contraction) of a material based on the
change in height (Δz) of a small bulk sample. The sample is usually a small cylinder or cube
with very flat, parallel top and bottom surfaces. The sample is placed on a flat sample
platform with very low thermal expansion—usually quartz. A flat bottom probe is lowered
until it just touches the surface of the sample. A very small load is then applied to make sure
that the probe contacts the sample in a controlled manner. This load remains constant
throughout the scan. The position of the probe is sensed using a Linear Variable
Differential Transducer (LVDT).

www.epotek.com Thermomechanical Analysis (TMA) 39

 As the sample experiences a thermal ramp, it will expand (if heated) or contract (if cooled)
under the probe. The LVDT will track this probe position versus temperature (figure 2).

figure 2

CTE testing is performed in accordance with ASTM E831, "Standard Test Method for
Linear Thermal Expansion of Solid Materials by Thermomechanical Analysis". The figure
below shows a typical TMA plot of probe position versus temperature. CTE is calculated as
the relative change in height of the sample per one degree change in temperature:

Thus, the CTE can be calculated by dividing the slope of the probe position (dh/dT) versus
temperature curve, by the initial height (ho) of the sample.

40 Thermomechanical Analysis (TMA)

 figure 9

A typical TMA plot is shown in figure 3. It will have two different slopes; a shallow slope at
lower temperatures, and a steeper slope at higher temperatures.

The shift in the slope of the probe position versus temperature plot will occur at the
material’s glass transition temperature (Tg). In the Tg region, the polymer chains will have
enough thermal energy to undergo larger scale motions which will enable a higher degree
of expansion with temperature. Thus, the TMA plot is a convenient method to determine a
material’s Tg. This is calculated by finding the point at which the tangents to the two
different slopes of the TMA plot intersect. Typically, the CTE above Tg is 3 to 4 times that
of the CTE below Tg.

From an engineering design standpoint, it is often important to match the CTE of an

adhesive as closely as possible to the substrates being bonded in order to minimize stress
at the interfaces. Thus, it is important to keep in mind that the adhesive will actually have
two different expansion rates over a broad temperature range. The temperature at which
this transition occurs will also be an important factor. If the Tg is sufficiently high relative to
the operating range of the part, only the lower CTE will come into play. On the other hand,
it is also important to keep in mind that the modulus of the adhesive will drop significantly
at temperatures above the Tg. A lower modulus or higher toughness material will be able to
better absorb and lessen the stresses incurred by thermal expansion
mismatch.

Examples of some CTE values of epoxy:

Silica Filled: Lowest ( 5-15 ppm /˚C )

Alumina Filled: Good ( 25-35 ppm /˚C )
Silver Filled: Average ( 40-55 ppm /˚C )
Typical ( 55-100 ppm /˚C )

www.epotek.com Thermomechanical Analysis (TMA) 41

 8.0 Electrical Conductivity (Volume Resistivity–VR)

Semiconductor and SMD applications require electrical contacts for the circuit to work
properly. Thus, VR is very critical to the performance of the circuit. Polymers, however, are
inherently electrically insulating materials. They can be made to conduct electricity by
adding electrically conductive filler materials. Silver and gold are the most common
conductive fillers used as gold does not oxidize and while silver does oxidize, silver oxide
remains conductive.

The most common type of electrically conductive adhesive is an Isotropic Conductive

Adhesive (ICA). In this case, the distribution of the conductive filler particles is
homogeneous throughout the adhesive, resulting in a material that is equally conductive in
all directions.

A less-common type of electrically conductive adhesive is an Anisotropic Conductive

Adhesive (ACA). As shown in figure 1 below, the amount of conductive filler particles used
in an ACA is much lower than the levels in an ICA. In addition, the particle size in an ACA is
designed to be the exact thickness of the bond line. As a result, these materials are
electrically conductive in only one direction (through the bond line), and are actually
electrically insulating in the remaining directions.

The majority of current applications are designed around ICAs.

figure 1

In order to establish electrical conductivity in an otherwise insulating resin system, enough

conductive filler must be added to create a three-dimensional network of the filler particles
for the electrical current to flow across. The size and shape of these filler particles often
dictate the loading level required. As shown in figure 2, a flake type filler geometry
(diagram 2b) will generally form a continuous conductive network at considerably lower
volumetric loading levels than spherical particles of the same material (diagram 2a).

Large flakes (diagram 2c) will typically form a conductive network at lower volumetric
loadings than smaller flakes of the same material. However, there are some cases where a
combination of large and small flakes (diagram 2d) may produce the optimum conductivity
at a given loading due to the ability for the small flakes to pack well between the interstices
of the larger flakes.

42 Electrical Conductivity (Volume Resistivity–VR)

 figure 2

The choice of flake size to impart conductivity to a resin system is also driven by the ICA’s
intended application. If the ICA is to be dispensed through a fine gage syringe needle, or
screen-printed to very tight tolerances, there may be a limit to the maximum size of flake
allowed. The required rheology of the ICA may also play a large role in flake size selection.
Smaller flakes have much higher surface area per unit volume than larger flakes. As a result,
adding smaller flakes will increase the viscosity of the ICA much more than the addition of
the same loading content of larger flakes.

The chemical compatibility of the resin system with the surface of the flake can also play a
role in determining the degree of conductivity obtained. Many conductive metal flakes are
coated with a very thin layer of organic lubricant material. This lubricant is necessary for the
manufacturing of the flake. The lubricant layer prevents the flakes from irreversibly welding
together during the mechanical process required to flatten them from the particulate
starting material to the flake geometry. The compatibility of the resin system with this
coating will determine how well the adhesive wets the flake. When there is good
compatibility between the lubricant and the resin system, the flakes will disperse well and
the resin will wet out the surface of the flake more completely.

While an uncured ICA paste may have some degree of conductivity, the optimum electrical
conductivity for a given system is not obtained until the material is fully cured. When a
thermosetting resin system cures, the chemical cross-links that form between polymer
chains bring the chains closer together and cause an overall shrinkage of the system. When
the resin shrinks, it pulls the conductive flakes closer together. This reduces the thickness
of insulating resin between the flakes and improves the electrical flow throughout the ICA.
Thus, it is very important to fully cure ICA’s in order to achieve optimum electrical
properties.

Electrical Conductivity (Volume Resistivity–VR) 43

 Volume resistivity is measured according to ASTM D2739, "Standard Test Method for
Volume Resistivity of Conductive Adhesives" and MIL-STD 883/5011. A thin stripe of the
ICA is painted onto a glass slide. Typical coating are .150 mm thick, 2.55 mm width, and the
length equals the distance between the ohmmeter probes (typically 7.5 cm). The width and
thickness of the stripe is usually controlled by using a stencil template.

Figure 3 below shows a typical process for preparing a VR sample.

figure 3

Once the sample has been prepared and cured, the two probes of a ohmmeter are applied
to the ends of the stripe to measure the resistance across the sample. Volume resistivity is
then calculated according to the following equation:

44 Electrical Conductivity (Volume Resistivity–VR)

 9.0 Thermal Conductivity
In the continued quest for miniaturization, thermal conductivity is playing an even more
important role in today’s products. More and more adhesives are relied upon to dissipate
generated heat away from active components. Thermal conductivity is a fundamental
material property that is essential for characterizing this heat transfer.

Thermal conductivity, by definition, is equal to the quantity of heat that is transferred in a

specific period of time through a known sample area when the sample’s opposite faces are
subject to an applied temperature gradient. Typical units of thermal conductivity are:

Watts
meter * Kelvin ( W
m *K (
One method, ASTM E1461, "Standard Test Method for Thermal Diffusivity by the Flash
Method", for measuring thermal conductivity is Laser Flash Diffusivity. This test method
has evolved into one of the most widely used techniques for the measurement of the
thermal diffusivity and thermal conductivity of polymeric materials in the adhesive industry.
Using this technique, the front side of a small, usually disc-shaped sample, is placed into a
horizontal fixture and heated by a short energy (laser) pulse. The resulting temperature rise
on the alternate surface is measured versus time using an IR detector. Although this is a
fairly quick test, the sample preparation and thickness are critical to the end result.

An advantage of this test is that it is non-contact and, is therefore nondestructive to the

sample.

Here is a schematic of the test apparatus:

Lens
Filter
Sample

IR
Detector Laser

Furnace

figure 1

Thermal conductivity (ThK) = Heat flow rate / (Area × Temperature gradient)

Thermal Conductivity 45
 There are several filler types that exhibit exceptional thermal conductivity: silver, boron
nitride (BN) and alumina. Any of the three will greatly aid the thermal conductivity of
epoxies which, by nature, are thermally insulative materials. Silver provides the best
thermal conductivity, but also provides electrical conductivity. Some applications require a
material to be electrically insulating so as not to affect the performance of an electronic
package. Alumina is electrically insulating and allows for good thermal conductivity. It is
commonly used in potting applications where abrasion is not critical. To reduce abrasion,
boron nitride is often used. This provides excellent thermal properties as well as reduced
abrasion for critical applications.

In general, higher particle size and higher loadings will result in a higher thermal conductivity
value for the epoxy system. Some ranges of performance for thermally filled epoxies are:

Very High: > 4.0 W/m*K

(new technologies)

High: > 2.0 W/m*K

Average: 0.7 – 2.0 W/m*K

Low: < 0.7 W/m*K

Poor: 0.1 – 0.3 W/m*K

Filler Type Thermal Conductivity of Filler Only (W/m*K)

Alumina 36
Boron Nitride 60
Silver 429

46 Thermal Conductivity
CONTACT US:
NORTH AMERICA EUROPE
Epoxy Technology, Inc. Germany, Austria, Bulgaria, Croatia,
Tel: +1 978-667-3805 Czech Republic, Greece, Hungary,
Email: [email protected] Poland, Romania, Slovakia, Slovenia
Website: www.epotek.com
Epoxy Technology Europe GmbH
Tel: +49 (0) 8 21 74 8720
Epoxies, Etc Email: [email protected]
Tel: +1 401-946-5564 Website: www.epotekeurope.com
Email: [email protected]
Website: www.epoxies.com
Switzerland, Liechtenstein, Russia,
Ukraine, Belarus
ASIA Epoxy Technology Europe AG
South East Asia Tel: +41 (0) 41-7481080
Email: [email protected]
Pacific Adhesive Systems (M) Sdn. Bhd Website: www.epotekeurope.com
Tel: +604-626-6687
Email: [email protected] United Kingdom, Ireland
Website: www.padhesive.com
Epoxy Technology Europe Ltd.
Tel: +44 1672 233104
Hong Kong
Email: [email protected]
PAS Asia Pacific LTD. Website: www.epotekeurope.com
Tel 852-2698-1727
Email: [email protected]
France
Website: www.padhesive.cn
Epoxy Technology Europe SAS
China Tel: +33 (0) 1 39 13 20 82
Email: [email protected]
PAS Trading (Shanghai) Co., Ltd. Website: www.epotekeurope.com
Tel +86-21-6631 2554
Email: [email protected]
Belgium, Luxembourg,
Website: www.padhesive.com.cn
The Netherlands

China Epoxy Technology Europe SRL

Tel: +32 (0) 2 351 1800
PAS Trading (Shenzhen) Co., Ltd.
Email: [email protected]
t: +86 755 2103 4922
Website: www.epotekeurope.com
Email: [email protected]
Website: www.padhesive.cn

For ordering, please contact us at:

Epoxy Technology, Inc.
14 Fortune Drive, Billerica, MA 01821
Tel: 978-667-3805
A Meridian Adhesives Group Company
www.epotek.com