Successfully Linking Leadframe Roughening Processes - Atotech

Authors:
Dr. J. Barthelmes (Author to whom correspondance should be adressed)
mailto: juergen.barthelmes@atotech.com
C. Wunderlich
M. Danker
D.-G. Neoh
Atotech Germany
Erasmusstrasse 20
D-10553 Berlin
Germany
Successfully Linking Leadframe Roughening Processes for Adhesion
Improvement to the Ultra Thin Ni/Pd/Au Technology
Abstract
The introduction of plating ultra thin Ni/Pd/Au onto leadframes as a lead-free
alternative brought about an adhesion deterioration between the metal alloy
substrate and the molding compound. This fact can be explained by the IEPS
(Isoelectric Point of the Surface) concept and consequently the reduced MSL level
performance calls for improvement measures.
Roughening the leadframe prior to plating of Ni constitutes a feasible path to strongly
improve the adhesion and there are several technologies available to achieve this
target. However it is observed that some of the roughness created is lost during the
PPF (preplated leadframe) process, attributed to the levelling property of the 0.5-1
µm thick Ni layer.
In order to demonstrate the influence of the main factors contributing to the levelling a
novel concept to characterize the surface roughness, the relative surface area
increase (RSAI), will be introduced. This becomes necessary, as conventional
roughness properties like Ra, Rrms or Rmax fail to adequately correlate to the adhesion
improvement found.
The current publication explains the leadframe´s final (after roughening and PPF
plating) RSAI as a function of its most influential parameters, such as the posttreatment
of the roughening process, the pre-treatment prior to Nickel plating and the
Ni Reverse Pulse and DC plating conditions. Process differences for the alloy
materials C194 and C7025 will be demonstrated and accounted for.
The information presented will lead to a total process giving the applicator maximum
benefit when combining Ultrathin Ni/Pd/Au with a leadframe roughening technology.
Introduction
With growing importance of environmental concerns, legislative measures like RoHS
and WEEE by the EU, are being imposed restricting the usage of lead for many
applications in the electronic industry. Pure Tin layers and other Tin alloys constitute
an popular alternative in the connector business. However, due to the risk of whisker
growth, there is considerable uneasiness with respect to these options in the
IC/Leadframe (LF) industry, especially for the production of fine pitch products.
With the relatively reasonable precious metal market prices for Gold and especially
Palladium many manufacturers of integrated circuits choose the solderable and
bondable Ni/Pd/Au finish, thereby replacing Ag plating on inner and SnPb plating on
outer leads.

Besides the economical risk of fluctuating precious metal costs the technical
disadvantage of the noble finish is its high IEPS [1] (isoelectric point of the surface),
which decreases the adhesion between metal and molding compound. In general a
lower thermal reliability and a reduced MSL (moisture sensitivity level) performance is
found for a Ni/Pd/Au or preplated leadframe (PPF). The trend towards thinner
packages/new package designs and the higher reflow temperature further deteriorate
the MSL rating.
Leadframe roughening through immersion in an intergranular etch solution increases
the LF surface area and thereby the mechanical adhesion between metal and
molding compound. This technology is already applied by LF and IC manufacturers
mainly for their Ag spot plating products, such as QFN. This paper will discuss
various important aspects of applying the LF roughening technology in combination
with plating Ni/Pd/Au finishes.
The Leadframe Roughening Process
The process sequence consists of a pre-treatment (mild etch to activation) to clean
the Cu alloy off surface impurities from preceding operations and to minimize the
drag in of contaminants into the main bath. The following etching electrolyte is a
mixture of hydrogen peroxide, sulphuric acid and proprietary organic additives. It
attacks the LF, roughens it through intergranular etching effected by the organic
components and creates an organo-metallic coating of Cu(II) on the LF surface. The
sequence is concluded by an alkaline PostDip which removes most of the coating. In
conventional Ag spot plating technology its purpose is cleaning the Ag surface for
best wire bondability, in combination with PPF it strips off the organic film to optimize
the adhesion to the Ni layer, which will be deposited subsequently. Table 1 gives a
summary of process conditions of the MoldPrep LF roughening process for
conveyorized and rack application.
Process Step T (°C)
Dwell Time (sec)
Conveyor Rack
Mild Etch 25-30 30 60
Cleaning 45-55 30 240
Activator 30-40 15 120
LF Etching 30-40 60 90
PostDip 40-50 20 60
Table 1: Summary of Dwell Times and Temperatures for the Individual Steps of the
MoldPrep LF Process as a Function of Application
Typical etch depths in production range between 0.7 and 1.3 µm.
The Cu Alloy Surface Morphology – The RSAI Concept
During the etching process the Cu surface roughness is altered and strongly
increased. The conventional parameters to quantify this property are the average
roughness Ra or the root mean square roughness Rrms.
In Table 2 the change in Ra and Rrms with roughening treatment is shown. On the
average Ra and Rrms increase by 100 %. The values were obtained by Scanning
Force Microscopy (SFM). The area scanned was 100µmx100µm.

Condition Ra [µm] % Rrms [µm] %
Untreated LF 0,088 100% 0,113 100%
After Roughening 0,195 222% 0,250 221%
Table 2: Ra and Rrms as measured by SFM (100µmx100µm) of a C194 LF surface
before and after roughening
An optical impression of a surface morphology can be obtained by Scanning Electron
microscopy (SEM). Figure 1 depicts the common leadframe alloys C194 and C7025
after treatment with MoldPrep LF.
MoldPrep LF – Morphology of Alloy Surfaces
Alloy C194
Alloy C7025
3000 X 5000 X
Figure 1: SEM Images of Alloys C194 and C7025 after LF Roughening at 3000-fold
and 5000-fold Magnification
However when attempting to quantify roughness to precisely assess and describe the
etching processes’ result with respect to adhesion, neither Ra nor Rrms are meaningful
parameters, as they do not relate to the increase of the interfacial area between
molding compound and leadframe.
In a first approximation the adhesion strength between two different materials is a
function of the joint contact interface. Three dimensional data from SFM and
Interference Microscopy (IM) can be used to calculate a morphology property called
Relative Surface Area Increase (RSAI). The RSAI relates a computed alloy surface
area to an ideally flat one.
Figure 2 displays a series of images taken after each step of the LF roughening
process MoldPrep LF. It shows the base material C194 as seen by SFM. Ra ,Rrms
and RSAI of these morphologies are summarized in Table 3.

MoldPrep LF – Morphology and
Relative Surface Area Increase (RSAI)
Morphology Change with Process Step (by SFM) – Example C194
Before
After Descale
After Cleaner After Activator
After MoldPrep
Figure 2: The C194 Alloy Surface after Each Step of the LF Roughening Process (10
µm x10 µm) as seen by Scanning Force Microscopy
The value of the various parameters is a function of the surface area scanned. As the
resolution increases more information about a given morphology is obtained,
therefore the RSAI increases with decreasing image size. At the same time Ra and
Rrms become smaller. This is another clear indication for the fact that these
parameters should not be assessed to quantify and relate to the expected adhesion.
Sample Image Size R rms [nm] R a [nm] RSAI [%]
100 µm x 100 µm 113 88 0,6
62 µm x 62 µm 89 71 0,8
Before
after
Cleaner
after
Descale
after
Activator
after
MoldPrep
20 µm x 20 µm 83 69 3,3
10 µm x 10 µm 60 48 4,7
100 µm x 100 µm 118 88 0,6
62 µm x 62 µm 109 85 1,3
20 µm x 20 µm 83 64 3,8
10 µm x 10 µm 55 45 5,9
100 µm x 100 µm 132 104 1,5
62 µm x 62 µm 112 88 2,8
20 µm x 20 µm 85 67 7,8
10 µm x 10 µm 72 56 10,6
100 µm x 100 µm 130 103 1,4
62 µm x 62 µm 120 96 2,9
20 µm x 20 µm 87 69 7,3
10 µm x 10 µm 73 57 10,9
100 µm x 100 µm 250 195 30,9
62 µm x 62 µm 295 240 66,7
20 µm x 20 µm 250 202 78,3
10 µm x 10 µm 239 193 92,9
Table 3: Comparison of Different Roughness Values obtained by SFM,
Characterising the C194 Surface after Each Step of the LF Roughening Process;
Function of Image Size scanned
RSAI Development with Post-Treatment of the Roughening Process
As mentioned above after roughening the LFs will be immersed in a postdip to either
clean the Ag surface for bonding or to prepare for optimized adhesion of the Ni (Ag)
layer to be plated.
Different alloy materials thereby exhibit different results with respect to the roughness
created during the etching process. Table 4 compares typical results obtained for
alloys C7025 and C194 for an etch depth of 1 µm.

Alloy R rms [µm] R a [µm] R max [µm] RSAI %
C7025 0,290 0,232 2,320 54,1
C194 0.330 0.263 2,443 79.0
Table 4: Roughness Characteristics of C7025 and C194 after Treatment with the
MoldPrep LF Roughening Process
Acidic Postdip
The relative surface area increase for the Si containing alloy C7025 is lower than for
C194. More important however is the formation of a Si smut on the C7025 LF surface
during the etching operation. For further processing of this material it is imperative to
remove this smut within the subsequent post-treatment. This can only be done by an
acidic type of postdip, which is slightly oxidising and which will remove some of the
surface Cu layer.
The following data explains the influence of the smut removing acidic postdip on the
RSAI of the LF morphology. It is found that some of the extra surface area created is
removed by this type of post-treatment.
Alloy PD Condition RSAI
C7025 No PostDip 54%
C7025 Acidic PD, low conc. 44%
C7025 Acidic PD, high conc. 38%
Table 5: RSAI Development of a C7025 LF Surface after Treatment with the Acidic
Postdip for Si Smut Removal
After immersion of C7025 into the acidic postdip the LF can be Ni (Ag) plated without
adjustment of the common pre-treatment for the subsequent operation.
Alkaline Postdip
An alkaline postdip is applied to all LF materials other than C7025, i.e. C194, EFTEC
64T, MF202 and C151. In general the RSAI of those base alloys will slightly increase
with alkaline post-treatment, along with that an enhancement in adhesion as
witnessed in the button shear test is seen. The increase in RSAI can be as high as
10-20 %.
Table 6: Button Shear Test Results (in N/mm 2 Alloy Alkaline PD w ithout PD
C194 17,1 13,5
C7025 16,4 10,3
EFTEC 64T 15,1 11,8
) for Various Alloys with and without
Alkaline Post Treatment
RSAI Development with Ni-Plating
When combining a leadframe roughening process with the ultrathin Ni/Pd/Au
technology it is evident that the RSAI increase obtained shall not be destroyed or
overly reduced by the subsequent pre-treatment of the PPF line. The same
consideration holds true for the following plating of Ni, Pd and Au.

As the deposits of Pd and Au are in general between 5 to 50 nm thick, thus a
noticeable levelling of the surface roughness can be expected by the Ni thickness
only. The most prominent Ni electrolyte used in production is the Sulfamate type, for
it exhibits the best ductility performance in the LF bending test.
In Figure 3 images of C194 LF surfaces are displayed as a function of Ni thickness
plated. Together with these the corresponding RSAI values are denoted. A gradual
decrease of the RSAI with Ni thickness is observed. When manufacturing PPF the Ni
layers are between 0.5 to 1.0 µm thick. Plating 0.7 µm Ni will reduce the RSAI as
effected by the LF roughening process by around 40 % of its initial value. This holds
true only for alkaline post-treatment after LF etching and proper, non-etching or
oxidising, Ni pre-treatment.
RSAI Development with Ni Thickness
Morphology Change with Ni Thickness Plated – Example C194
Images from Interference Microscope (IM)
After Etching/Alk. PD
RSAI: 95 %
After 0.6 µm Ni
RSAI: 58 %
After 0.2 µm Ni
RSAI: 80 %
After 0.8 µm Ni
RSAI: 51 %
After 0.4 µm Ni
RSAI: 68 %
After 1.0 µm Ni
RSAI: 41 %
Figure 3: Morphology Change with Ni Thickness as Witnessed by the Interference
Microscope, Base Alloy C194, Ni Thickness Range 0-1.0 µm, Ni Sulfamate Type
Figure 4 depicts the RSAI decay with Ni thickness. The Ni layer was determined by
means of XRF, which averages over the entire size of the measurement spot. As the
surface area is strongly increased, the real Ni thickness will be lower.
60
RSAI [%] 50
40
RSAI Decay with Ni Thickness
Morphology Change with Ni Thickness Plated – Example C194
Images from Interference Microscope (IM)
•Slow Exponential Decay of RSAI with Ni Thickness
•Pre (before Ni) - and Post Treatment (after MP) contain no Etch Cleaner
RSAI Decay after MoldPrep LF with Ni Thickness Plated RSAI Decay with Ni Thickness
100
90
80
70
30
20
10
0
0 0,2 0,4 0,6 0,8 1
Ni Thickness [µm]
RSAI [%]
y = 94,842e -0,8197x
100
90
80
70
60
50
40
30
20
10
0
0 0,2 0,4 0,6 0,8 1 1,2
Ni Thickness [µm]
Figure 4: RSAI Decay with Ni Thickness, Base Alloy C194, Ni Thickness Range 0-1.0
µm, Ni Sulfamate Type
Direct Current Plating Conditions

The levelling property of the Ni Sulfamate bath has been investigated as a function of
direct current density and Ni concentration. For this, the samples to be plated were
characterized in terms of RSAI before and after Ni plating. The reduction of RSAI
serves as an indicator for the levelling performance of a given parameter set. As the
absolute errors add up only substantial differences in the RSAI reduction will be
considered significant. Starting from an RSAI between 60 – 70 % a reduction for 0.7
µm Ni plated of 20-30% absolute RSAI is found. The current density was varied
between 5 and 15 Asd (A/dm 2 ), the Ni concentration between 70 and 110 g/l. Due to
the large relative error of the measurement the differences between the individual
RSAI reductions are not significant. Figure 5 summarizes the results of this
investigation. The two values displayed on the x-axis of the chart denote the current
density in Asd and the Ni concentration in g/l.
RSAI Reduction with Ni Plating Conditions
DC Plating Conditions (0.7 µm Ni, Initial RSAI at 60-70 %)
Current Density and Ni Concentration
RSAI Reduction [%]
35
30
25
20
15
10
5
0
Reduction of RSAI
by Ni Plating of 0.7 µm
5/70 15/70 5/110 10/110 15/110
DCPlating Conditions CD/Ni conc. [Asd and g/l]
Figure 5: RSAI Reduction for different DC Plating Conditions in the Ni Sulfamate
bath, 0.7 µm Ni plated, Initial RSAI at 60-70 %
Pulse Break and Reverse Pulse Plating Conditions
Reverse Pulse Plating (PP) is widely applied for plating acid Cu in the field of PCB
and IC substrate manufacturing. One well known property of these layers is that their
levelling capacity is lost with stronger reverse pulses or larger reverse charges.
Therefore basic investigations were initiated to understand the influence of PP for the
deposition of Ni on the surface morphology of roughened LFs.
In Figure 6 the RSAI decrease is shown for three different sets of PP parameters.
The average current density applied is 10 Asd, the ratio of reverse to forward current
density is 3:1, the Ni thickness plated 0.7 µm. As for the DC plating experiments the
initial RSAI was between 60-70 %.
Even though the overall RSAI reduction is smaller than in the DC experiments the
difference observed between PP and DC can not be considered significant either.
Pulse breaks (PB) or pauses are frequently used as the so called chopper plating
also in LF manufacturing. The influence of such current-time functions on the RSAI
change during the Ni deposition was investigated and is displayed in Figure 6 as
well. The commonly applied pulse durations did not exhibit different responses as
compared to those obtained in the DC or PP experiments.

RSAI Difference [%]
30
25
20
15
10
5
0
RSAI Reduction with Ni Plating Conditions
Pulse Break (PB) and Reverse Pulse Plating (PP) Conditions
(0.7 µm Ni, Initial RSAI at 65 %, 10 Asd average CD, CD rev/fwd 3:1)
PB: On and Off Time [ms]
PP: Forward and Reverse Plating Times [ms]
Reduction of RSAI by Ni Plating of 0.7 µm
Pulse Break and Reverse Pulse Plating at Average 10 Asd
10 on/10off 20on/20off 20on/40off 10/0,5 80/4 400/20
PP Reverse Plating Times fwd/rev [ms]
Figure 6: RSAI Reduction for different Pulse Break and Reverse Pulse Plating
Conditions in the Ni Sulfamate bath, 0.7 µm Ni plated, Initial RSAI at 60-70 %
Summary
Due to its higher IEPS the ultrathin Ni/Pd/Au finish exhibits deteriorated MSL
performance as compared to conventional Ag spot plated leadframes. One option to
increase the adhesion of the noble metal surface to the molding compound
constitutes the roughening of the LF prior to the Ni deposition.
Commonly used roughness parameters can not appropriately describe the surface
morphology when adhesion phenomena have to be assessed. The concept of the
relative surface area increase (RSAI) is introduced to correlate the result of the
roughening process to the expected adhesion increase. The RSAI compares the
surface area of a three dimensional morphology to an ideally flat surface.
The influence of different postdip options for the LF etching process on the RSAI was
demonstrated. An alkaline postdip generally increases the RSAI and the adhesion as
measured in the button shear test. The acidic postdip is required for smut removal
from C7025 LFs and slightly decreases the RSAI through.
During Ni plating the RSAI as generated in the roughening process is reduced. The
function of RSAI with Ni thickness was established for a Ni Sulfamate electrolyte. The
deposition of 0.7 µm Ni reduces the RSAI by 40% of its initial value.
A significant influence of DC or pulse current conditions on the RSAI decrease could
not be established.
For certain packages the roughening of the LF prior to Ni/Pd/Au effects an MSL
improvement of two levels. These packages can achieve MSL 1.
References
[1] J.C. Bolger and A.S. Michaelis “Interface Conversions for Polymer Coatings”, P.
Weiss and D. Cheever, Editors, Chapter 1, Elsevier, New York (1968)