Application Note AN-10-001

Revision: 0 Key Words: Thermal paste

Issue Date: 2010-03-30

Prepared by: Dieter Esau

THERMAL PASTE APPLICATION

Designated use of thermal interface material (TIM) .................................................................................................. 1

Procedures for thermal paste application .................................................................................................................. 3
Quality control check on thermal paste layer............................................................................................................. 4
Determining the optimum thermal paste layer
thickness .................................................................................................................................................................... 5
References................................................................................................................................................................. 6

Designated use of thermal interface

material (TIM) baseplate or DCB of power module

When in operation, power modules produce losses which Heat sink

increase the module temperature and impair module
efficiency and/or functionality. To dissipate the heat that air gaps
builds up in power modules, the power modules are
mounted onto heat sinks. The heat is then dissipated Fig. 1: Heat transfer from a power module to a heat sink
from the power module via the heat sink. Heat transfer without TIM
between the heat-dissipating surface of the power
module and the heat sink surface depends on the A suitable way of improving heat transfer is to fill up the
surface quality of the different surfaces. Both the heat air pockets with a TIM (see Fig. 2).
sink surface and the heat-dissipating surface of the
power module are uneven. As a result, air is trapped
between the two surfaces, preventing direct heat
transfer. As air is a poor thermal conductor (the specific baseplate or DCB of power module
thermal conductivity of air is λair ≈ 0.03 W/m·K), only very
little heat can be conducted to the heat sink (see Fig. 1). Heat sink

thermal interface material

Fig. 2: Heat transfer from a power module to a heat

sink using TIM

© by SEMIKRON 2010-03-30 – Rev0 1/ 6

Application Note AN-10-001

TIM normally consist of a plastic carrier material (e.g. Rthcs

silicon oil) and thermal conductive filler substances such
1
as zinc oxide, graphite or silver. They are available in
the form of pastes, adhesives, phase-change materials
and foils. TIMs conduct heat better than air and typically
have a specific thermal conductivity of λ≈ 0.5 - 6 W/m·K.
The thermal conductivity of TIM is thus approx. 20 - 200
times better than that of air. To enable the thermal
conductivity properties of TIMs to be categorised, table 1
shows the specific thermal conductivity of materials Minimum Maximum Thickness of TIM
commonly used in power modules. The thermal paste
P12 from the company Wacker has been taken by way of
example. The R(th) values shown are based on the
module-specific thermal spreading.
Fig. 3: Dependence of thermal resistance on thermal paste
Material Specific Thickness Portion of layer thickness
Thermal [µm] R(th) for SKiM
Conductivity modules Too thin a thermal paste layer results in air pockets
λ between the underside of the module and the top of the
[W/(m*K)] heat sink, causing a high thermal resistance Rthcs. Once
the optimum has been reached, the thermal resistance
Rthcs increases quickly again in line with the increase in
Chip 106 120 2.92% thermal paste layer thickness; this happens because the
Chip solder 57 70 3.65% specific thermal conductivity λ of thermal conductive
media is very low compared with other materials in a
DBC (copper) 394 300 1.94% power semiconductor module. The minimum value as
DBC (Al2O3) 24 380 32.91% shown in the diagram above is different in each system
(module on heat sink) and has to be defined in tests.
DBC (copper) 394 300 1.31%
Thermal paste 0.81 30 57.26% The thickness of the thermal paste layer is different for
(P12 from different module types. This is why the mounting
WACKER2) instructions of power modules specify the given thermal
paste layer thickness and describe the quality of the
Tab. 1: Specific thermal conductivity of materials commonly surface of the heat sink.
used in a power semiconductor module
The thermal paste used and recommended by
If the thermal conductivity of thermal paste is compared SEMIKRON is P12 thermal paste from the company
with the thermal conductivity of other components in a Wacker and is at the lower end of the specific thermal
power module (see Tab. 1), the thermal paste does not conductivity range. The following factors are the key
rate particularly well. The extent to which thermal paste arguments in favour of this thermal paste:
contributes to the overall thermal resistance R(thjs) of the
3
module amounts to around 20-65% , depending on the • R(th) tests have shown that the thermal conductivity
module with the combination of the heat sink. The of a thermal paste in actual application does not only
thermal paste layer therefore has to be as thin as depend on its specific thermal conductivity λ, but also
possible but as thick as necessary (see Fig. 3). on it its structure (see Tab. 2). The larger the filler
particles in a thermal paste are, the higher the
specific thermal conductivity. The particle size of the
filler determines the minimum layer thickness. In
other words, the thermal paste layer applied cannot
be thinner than the largest particles in the paste. After
several temperature cycles, a paste with small
particles (e.g. P12: particle size 0.04µm - 4µm) allows
1
These materials are electrically conductive. almost for metal-to-metal contact at points where the
pressure is particularly high, resulting in a substantial
2
The filler particles (ZnO) of Wacker P12 thermal paste reduction in Rthcs.
are between 0.04µm and 4µm in diameter, which is why • The paste is highly resistant to “bleeding” and “drying
this thermal paste is very good for achieving very thin out”.
thermal paste layers.
The table below shows an excerpt from a series of
3
A thermal paste with a higher specific thermal thermal conductive media which have undergone testing
conductivity will not normally result in much improvement at SEMIKRON.
here, since this would normally mean a considerable
increase in thermal paste layer thickness, too.

2/6 2010-03-30 – Rev0 © by SEMIKRON

 Application Note AN-10-001

Name; Producer Description Filled Conduc- Possible Applicable Thermal Thermal

with tive Application thickness of Resis- resistance λ
silicon process TIM in µm tance W/(m*K)
(Data sheet)
P12, Wacker Paste, Filler: Yes No Roller, stencil-/ 10-100 + 0,81
Al2O3 screen-printing

HTC, Electrolube Paste, Filler: No No Roller, stencil-/ 10-100 + 0,9

AL2O3 screen-printing
PSX-P8, Hala Phase-Changer, No No Roller, stencil-/ 10-100 + 3,4
Contec GmbH Filler: aluminium screen-printing
powder
TIC 1000A, Paste, Filler: Yes Yes Roller, stencil-/ 15-100 O 1,5
Bergquist Al2O3 screen-printing
TIC 4000, Paste, Filler: Yes Yes Roller, stencil-/ ca. 100 + 4,0
Bergquist fluid metal screen-printing
KU ALC-5, Kunze Phase-Changer; No Yes Manually ca. 76 O 220
Aluminium foil
with wax coating
KU ALF, Kunze Phase-Changer; No Yes Manually ca. 76 + 220
Aluminium foil
with wax and
graphite coating
Keratherm 86/50, Foil, Filler: boron Yes No Manually 120 - 2,9
Kerafoil nitride
Q2-Pad, Bergquist Aluminium foil Yes Yes Manually 152 - 2,5
with graphite
coating

Tab. 2: Selection of tested TIMs

Procedures for thermal paste application
Thermal paste can be applied either to the module or to
the heat sink. This is done using a roller or in printing
processes. In roller application, a rubber roller is normally
used (see Fig. 4), while the printing process is normally
silk screen printing or stencil printing.
Applying thermal paste with a rubber roller can lead to
sufficient results provided this assembly step is
performed by experienced professional staff that are
properly trained in this critical process. This process also
has disadvantages, however, for instance
inhomogeneity, poor reproducibility and the risk of
contamination.
In stencil printing, a stainless steel stencil and stainless
steel scraper are normally used. The “effective” thermal
paste layer thickness, however, is determined by the
ratio of filled area to non-filled area, as well as by the
height of the dots applied, which in turn are determined
by the thickness of the stencil itself.
In screen printing, Monolen-PET meshes and a
polyurethane scraper with a shore hardness of 75 are
used. The thickness of the yarn and the number of yarns
per unit of length determine the thickness of the thermal
paste layer.
In stencil and screen printing, far better results can be
achieved than in the roller process, provided the printing
is done automatically. Performing this process manually
can lead to considerable process fluctuations. The
development of a process with an automatic stencil
printer that features continuous process monitoring, as is
the case at SEMIKON, requires substantial investments,
however, which in economic terms only makes sense for
large production quantities.

The stencil and screen printing process exists in all

automated stages. An example of a manual screen
printing process used to print onto a heat sink is
explained below:

Fig. 4: Paste application using a rubber roller Thermal paste application using manual screen printing
(Fig. 5):

© by SEMIKRON 2010-03-30 – Rev0 3/6

Application Note AN-10-001

a) Clean the surface with a cleansing agent to remove all Quality control check on thermal paste
grease. Position the heat sink in the device. Here, it is layer
important that the screen does not touch the surface of
the heat sink. To ensure this happens, the clearance
The thickness of a thermal paste layer can be measured
between the screen/mesh and heat sink has to be 4 - 7
directly or indirectly. An indirect way of measuring the
mm.
thickness is, for example, to weigh the thermal paste by
b) The screen is “flooded” with thermal paste (low-
performing a Tara weight measurement using suitable
pressure application).
scales. An example of a direct contact-free measurement
c) The thermal paste is spread using the polyurethane
of the thermal paste layer is a measurement using an
scraper; here, sufficient pressure must be applied to
optical profilometer such as the µSCAN from Nano
ensure that the mesh is pressed onto the print surface.
Focus. Other measurement equipment that can be used
d) Visual inspection.
to measure the thermal paste layer directly includes, for
example, thickness gauges such as wet film combs or
wet film wheels; the downside of these, however, is that
they may destroy the layer in places.

The edges of the wet film comb (e.g. from Zehntner (ZND
2051) or Elcometer Instruments or BYK Gardner (PG-
3504)) have support teeth and measurement teeth which
have a defined distance to the surface. The comb is held
a b
perpendicular to the surface and run across the surface
horizontally (see Fig. 7); when this is done, some paste
residue will be left on the teeth of the comb that are
beneath the surface of the paste layer. As shown in Fig.
8, the thickness of the layer measured in this example
lies in the range of between 25µm and 30 µm. This
measurement can be very easily distorted, however,
through improper handling, surface unevenness or even
by raised points in the thermal paste surface.
c d
Measurement comb

Fig. 5: Thermal paste application using manual screen

Scrape direction Thermal paste
printing

In addition to complying with recommended layer 90°

thickness, care should be taken when applying the
thermal paste to ensure that the thermal paste layer is Heat sink
evenly and homogenously spread on the underside of
the module or the heat sink surface. An inhomogeneous
thermal paste layer (extreme case: application of one or
more thermal paste blobs) (Fig. 6) can result in fractures Fig. 7: Measuring the thermal paste layer with wet film
or breakage in the DBC ceramic substrate. This applies comb
to modules with and without a base plate alike. In
addition to this, thermal paste inhomogeneity can also
lead to local overheating resulting from the air pockets 2 25 30 µm
between the underside of the module and the upper side Thermal
of the heat sink surface.
Thermal
paste
paste
Heat sink

Screw Screw
holes holes Edges

Measurement teeth
thermal paste
blobs Fig. 8: Measuring the thermal paste layer with wet film
comb (here PG-3504 from BYK Gardner)

Fig. 6: Module underside showing problematic thermal

A wet film wheel (e.g. from Zehntner (ZWW 2100-2102),
paste layer application
see Fig. 9 or BYK Gardner) produces more accurate
results in thermal paste layer thickness tests than a wet
film comb. The wet film wheel consists of two support
discs which are positioned at the outer edges, and one
measurement disc located between the support discs.
The measurement wheel is rolled across the surface that
has been coated with thermal paste (see Fig. 9). The
thermal paste layer thickness can be read from the scale,

4/6 2010-03-30 – Rev0 © by SEMIKRON


taken from the end of the wet segment of the middle
measurement disc (see Fig. 10).
Fig. 9: Measuring the thermal paste layer with a wet film
wheel (here ZWW 2102 from Zehntner)
Area with less contact to
the measurement disk
The layer thickness can be read here
Fig. 10: Measuring the thermal paste layer with a wet film
wheel (here ZWW 2102 from Zehntner)
Determining the optimum thermal paste
layer thickness
Thermal paste layer thickness is different for different
module types. This is why the mounting instructions of
power modules specify the given thermal paste layer
thickness and describe the quality of the surface of the
heat sink. In most cases, however, the specified layer
thicknesses apply to the thermal paste Wacker P12. If
other thermal pastes are being used, we recommend
observing the following procedures:
Varying, pre-defined thermal paste layer thicknesses are
applied to the modules or the heat sink. A module can be
screwed onto a standard module or onto an aluminium
plate in accordance with the given mounting instructions.
When tightening the mounting screws, the tightening
torques specified in the given mounting instructions must
be observed. To achieve a relaxed system state, i.e. with
no mechanical load, the mounted and secured module
should undergo three thermal cycles (20°C/100°C/1h)
(see Fig. 11).
© by SEMIKRON 2010-03-30 – Rev0
Application Note AN-10-001
Temperaturzyklus
Temperature Cycle
100°C
Temperature
One Cycle
Temperatur
20°C
0 60 120 180
t [min]
Fig. 11: Thermal cycling performed to determine optimum
thermal paste layer thickness
As the module is pressed onto the heat sink/aluminium
plate and sticky thermal paste distributed in the space
between, once the screws have been undone, a module
without base plate can not be easily removed without
causing destruction. To ensure non-destructive removal,
the module should be left untouched at room
temperature for 12 hours after the screw has been
loosened or should undergo 1-2 thermal cycles. The
remaining thermal paste can then normally be easily
removed using a clean, solvent-free, lint-free cloth
(where applicable, thermal paste manufacturer
specifications are to be complied with).
Each module should be used just once, as repeated
loosening and tightening of the screws can alter the
pressure properties. For each thermal paste layer
thickness to be tested, a minimum of two modules should
be used.
When module underside surface is fully covered with
paste, the thermal paste layer application is optimum.
(see Fig. 12).
Fig. 12: Optimum thermal paste layer
thickness
If the visual inspection reveals areas on the contact
surface of the module which haven’t been filled, then the
thermal paste layer is too thin (see Fig. 13).
Fig. 13: Too little thermal paste
5/6
 Application Note AN-10-001

References

1. Freyberg, M.: Application of thermal paste for power

modules without base plate. SEMIKRON
International, 1999.
2. Kolpakov, A. I.: „SKiiP – intellektual’nye silovye IGBT
moduli SEMIKRON“ in „Komponenty i technologii“ Nr.
1, 2003
3. Kolpakov, A. I.: „SKiM – novoe pokolenie
intellektual’nych silovych modulej SEMIKRON“. In
ölektronnye komponenty, Nr. 1, 2003
4. Kolpakov, A. I.: „SEMITOP kak al’ternativa TO“. In:
Silovaja ÷lektronika, Nr. 2, 2004
5. Goldman, W. E.: An Introduction to the Art of Heat
Sinking. In: Electronic Packaging and Production,
1966
6. Strube M (2007) Wärmeleitpastenauftrag als
Dienstleistung. Elektronik Praxis Nr.10, S. 24
7. Strube M (2007) Thermal Paste Spread and Ready
for Use. Bodo's Power System May 2007, S. 24

DISCLAIMER
SEMIKRON reserves the right to make changes without further notice herein to improve reliability, function or design.
Information furnished in this document is believed to be accurate and reliable. However, no representation or warranty is
given and no liability is assumed with respect to the accuracy or use of such information. SEMIKRON does not assume any
liability arising out of the application or use of any product or circuit described herein. Furthermore, this technical information
may not be considered as an assurance of component characteristics. No warranty or guarantee expressed or implied is
made regarding delivery, performance or suitability. This document supersedes and replaces all information previously
supplied and may be superseded by updates without further notice.
SEMIKRON products are not authorized for use in life support appliances and systems without express written approval by
SEMIKRON.

SEMIKRON INTERNATIONAL GmbH

P.O. Box 820251 • 90253 Nürnberg • Deutschland • Tel: +49 911-65 59-234 • Fax: +49 911-65 59-262
sales.skd@semikron.com • www.semikron.com

6/6 2010-03-30 – Rev0 © by SEMIKRON