Texas Instruments | Ni-Pd-Au Component Lead Finish and Its Potential for Solder-Joint Embrittlement | Application notes | Texas Instruments Ni-Pd-Au Component Lead Finish and Its Potential for Solder-Joint Embrittlement Application notes

Texas Instruments Ni-Pd-Au Component Lead Finish and Its Potential for Solder-Joint Embrittlement Application notes
Application Report
SZZA031 - December 2001
A Nickel-Palladium-Gold Integrated Circuit Lead Finish
and Its Potential for Solder-Joint Embrittlement
Donald Abbott, Douglas Romm, Bernhard Lange
Standard Linear & Logic
ABSTRACT
This gold (Au) embrittlement study evaluates TI’s original four-layer nickel-palladium (NiPd)
lead finish that was introduced in 1989 and TI’s nickel-palladium-gold (NiPdAu) lead finish
that replaced four-layer NiPd in 2001. Samples were prepared with three different Au
thicknesses for the NiPdAu components. The printed wiring board (PWB) finishes used were
organic solderability preservative (OSP) and electroless nickel-gold (NiAu). The latter were
made with two different Au thicknesses. The goal was to understand the effect of Au from the
component finish and the PWB pad finish on Au embrittlement in the solder joint. A matrix
of samples was built from the components and boards described above and exposed to 1000
temperature cycles. Lead pull testing and metallurgical analysis of the solder joints were
performed to determine if the variations in Au content, either from the lead or the board or
both, led to Au embrittlement of the solder joints.
The theoretical calculations (Appendix A) showed that 3 weight % Au would not be exceeded
in the solder-joint test. Based on the Au embrittlement literature, this predicts no solder-joint
embrittlement would occur. The lead pull data showed no evidence of catastrophic drop in
solder-joint strength that would be expected with Au embrittlement. The cross sections show
no SnAu intermetallics in the bulk of the solder for the solder joints made with normal
Au-thickness-range components and boards. Thin layers of SnAu intermetallics are seen at
the solder/component and solder/PWB interfaces. This is expected because intermetallics
form between Sn from the solder and Au from the PWB and component finishes. Only the
solder joints made with artificially high Au-thickness components and PWBs have the
acicular SnAu intermetallics that have been shown to cause Au embrittlement in some
instances in the literature. The NiPdAu leadframe finished components investigated for this
applicaton report showed no measurable Au embrittlement of the solder joint.
Contents
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Components and PWBs Tested . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Test Matrix for Au Embrittlement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Reflow Profile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Board-Mount Equipment and Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Theoretical Au Content Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Predicted Au Content of Solder Joints Using Thickness Data for ICs and PWBs . . . . . . . . . . . . . . .
3
4
4
5
5
5
5
1
SZZA031
Performance Measures and Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Visual Appearance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Lead Pull Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Lead Pull Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Observations From Lead Pull Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Statistical Analysis of Lead Pull Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Analysis of Variance (ANOVA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Box-and-Whisker Plots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Summary of Statistical Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Cross Sections of Solder Joints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Sample 6: NiPdAu Standard Finish, PWB Standard NiAu Finish . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Sample 5: NiPd Lead Finish, PWB Standard NiAu Finish After Thermal Cycling . . . . . . . . . . . . . . 12
Sample 7: NiPdAu 5× Au, PWB Standard NiAu Finish After Thermal Cycling . . . . . . . . . . . . . . . . . 14
Sample 2: NiPdAu Standard Finish, PWB OSP Finish After Thermal Cycling . . . . . . . . . . . . . . . . . 14
Sample 10: NiPdAu Standard Finish, 5× NiAu PWB With No Thermal Cycling . . . . . . . . . . . . . . . . 15
Sample 12: NiPdAu 100× Au, PWB 5× Au Thickness After Thermal Cycling . . . . . . . . . . . . . . . . . 16
Summary of Cross-Section Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Summary and Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Appendix A Calculations for Au Embrittlement Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
List of Figures
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
2
Structure for TI Four-Layer NiPd Finish . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Structure for NiPdAu Finish . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Reflow Profile for SnPbAg Solder Alloy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Cross Section of Solder Joint . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Lead Pull Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Lead Pull Results by Factor-Level Setting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Box-and-Whisker Plot for OSP Coating on PWB (No Au) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Box-and-Whisker Plot for Standard NiAu PWB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Box-and-Whisker Plot for PWB With 5× Au (0.4 µ to 0.65 µ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Sample 6 After Thermal Cycling (The Dark Gray Matrix Is the Tin-Rich Phase,
While the Bright Phase Is the Lead-Rich Phase) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Board Side of Sample 6 at 4000× After Thermal Cycling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Lead Side of Sample 6 at 4000× After Thermal Cycling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Spectra of Cu, Ni, Au, and Sn Taken Along Line 1–2 in Figure 12 . . . . . . . . . . . . . . . . . . . . . . . . . . 12
1000× Image of Sample 5 After Thermal Cycling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Board Side of Sample 5 at 4000× After Thermal Cycling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
4000× Image on the Lead Side of Sample 5 After Thermal Cycling . . . . . . . . . . . . . . . . . . . . . . . . . 13
Line Scan Along Line 1–2 in Figure 16 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Sample 7 at 1000× After Thermal Cycling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
1000× Image of Sample 2 After Thermal Cycling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
A Nickel-Palladium-Gold Integrated Circuit Lead Finish and Its Potential for Solder-Joint Embrittlement
SZZA031
20
21
22
23
24
25
26
27
1000× Image of Sample 10, No Thermal Cycling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Sample 10 at 4000× Board Side, No Thermal Cycling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Sample 10 Lead Side at 4000×, No Thermal Cycling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
Line Scan Along Line 1–2 in Figure 22 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
Sample 12 at 1000× After Thermal Cycling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
Sample 12 Pad Side at 4000× After Thermal Cycling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
Sample 12 Lead Side at 4000× After Thermal Cycling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Line Scan Along Line 1–2 in Figure 26 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
List of Tables
1
2
3
A–1
A–2
A–3
A–4
A–5
Group Numbers for Each Variation Tested . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Calculated Weight % Au in the Solder Joints for the Test Matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
ANOVA Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
XRD/EDX Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
XRF/EDX Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Calculations for Boards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Calculations for Components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Calculated Percentage of Au in the Solder Joints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Introduction
In 2001, TI converted from a nickel-palladium (NiPd) lead finish for components to a
nickel-palladium-gold (NiPdAu) lead finish. The NiPd finish for IC leads was first introduced in
the late 1980s.[1,2] By October 2001, more than 40-billion NiPd-finished IC packages were in
the field. The four-layer NiPd structure is shown in Figure 1. The structure of the NiPdAu lead
finish is shown in Figure 2. The introduction of Au on the top surface of the NiPdAu leadframe
finish suggested an Au embrittlement study would be useful.
Palladium (0.075 m min)
Gold (30 Å to 150 Å)
Palladium (0.02 m min)
Nickel (1.0 m min)
Nickel (0.5 m min)
Palladium/Nickel Strike
Nickel Strike
Copper Base Metal
Copper Base Metal
Figure 1. Structure for TI
Four-Layer NiPd Finish
Figure 2. Structure for
NiPdAu Finish
A very good description of Au embrittlement is provided by Glazer et al.: “Because Au has
essentially no solubility in either Sn or Pb, when the solder joint solidifies, Au is dispersed
through the joint in the form of brittle, elongated AuSn4 or Au(SnPb)4 intermetallic
precipitates.” [3] (see Figures 20 and 24)
A Nickel-Palladium-Gold Integrated Circuit Lead Finish and Its Potential for Solder-Joint Embrittlement
3
SZZA031
In the literature, the amount of Au that causes embrittlement of a solder joint has been
suggested to cover a range of 4 weight % to 6 weight % Au in the joint.[4,5,6,7] Glazer et al.
recommend that Au in the joint not exceed 3 weight %, based on the results of their study,
“Effect of Au on the Reliability of Fine Pitch Surface Mount Solder Joints”.[3] This careful study
considered only Au from the PWB as a contributor because the leads were SnPb finished.
Recently, Zribi et al. have shown a failure mechanism due to a ternary Au solder alloy that
occurs with Au concentrations as low as 0.1 weight %.[8] This work was done with ball grid
arrays soldered to NiAu-plated PWBs. Zribi’s work raises a question about the use of
NiAu-finished PWBs in general. However, it is not clear how the weight % Au is calculated in
Zribi’s study.
Au in the solder joint can come from the component lead and the PWB pad finish. For this study,
the thickness of Au on the component leads and on the PWB pads was varied from 0 to an
extreme for different samples.
Components and PWBs Tested
For the PWB pad finish, 5-m to 7-m Ni and 0.09-m to 0.1-m Au were targeted for the standard NiAu
PWB samples. The thick-Au samples had the same nickel thickness, but with an Au-thickness
target of 0.4 m to 0.65 m (this is denoted as 5× NiAu PWB in this application report). The third
(control) variation on PWB pad coating was an organic solderability preservative (OSP) applied
directly to the copper (Cu) pad.
The IC component style used was a 20-pin, dual-inline, surface-mount, small-outline package
(SOP), with lead pitch of 1.27 mm. TI package designator for the test component is NS. The
components were plated with NiPdAu with the Ni and Pd being held constant and the Au
targeted to 30 Å, 150 Å, and 3000 Å. These Au thicknesses are the minimum, 5× the minumum,
and 100× the minimum thicknesses indicated in the TI NiPdAu specification for Au. Throughout
this application report, we refer to the various component sample finishes as Std NiPdAu, 5×
NiPdAu, and 100× NiPdAu. TI’s four-layer NiPd was used as the control, contributing no Au to
the solder joint.
The Au actual thickness measurements for the components and PWBs used in the study are
shown in Appendix A, part B.
Test Matrix for Au Embrittlement
Relative contribution of Au from components and PWB pads for each sample is shown in
Table 1. The solder paste was 62Sn/36Pb/2Ag with a melting point of 179°C. The peak reflow
target temperature was 225°C.
Table 1. Group Numbers for Each Variation Tested
Au CONTRIBUTION
FROM COMPONENT
4
Au CONTRIBUTION FROM PWB PAD
OSP
Std NiAu PWB
5× Std NiAu PWB
NiPd (no Au)
1
5
9
NiPdAu (Std finish)
2
6
10
NiPdAu (5× Std finish)
3
7
11
NiPdAu (100× Std finish)
4
8
12
A Nickel-Palladium-Gold Integrated Circuit Lead Finish and Its Potential for Solder-Joint Embrittlement
SZZA031
Reflow Profile
The reflow profile was based on inputs from the solder-paste vendor (see Figure 3). This profile
reaches a preheat temperature of 120°C to 160°C for approximately 60 seconds before rising to
a peak temperature of 225°C to 228°C.
Temperature – °C
200
100
0
0
2
4
Time – Minutes
6
Figure 3. Reflow Profile for SnPbAg Solder Alloy
Board-Mount Equipment and Procedure
The solder paste stencil was laser-cut and polished 150-mm stainless steel. Printing of the solder
paste was performed manually. Prior to component placement, optical inspection of the printed
solder paste was performed to insure adequate paste height and complete printing. An optical
alignment tool for manual component placement was used to position the leads to the
solder-paste prints. For the reflow soldering process, a Rehm full-convection reflow oven with N2
purge was used. Remaining O2 was 500 ppm to 1000 ppm.
Theoretical Au Content Calculations
Appendix A, part A, shows that the Au thickness to achieve 3 weight % Au content for the solder
joints made using the OSP boards and ICs under test would be 0.757 m or 7570 Å–roughly 250
times the minimum thickness specified for Au on the component lead. This is ~50× the
maximum thickness specified for the Au on the lead. The weight % Au calculated assumes that
the layer of Au on the component dissolves entirely into the solder joint. The calculation
assumes no Au comes from the PWB pad. Assumptions about the surface area of the lead and
the characteristics of the solder paste also are in Appendix A, part A.
Predicted Au Content of Solder Joints Using Thickness Data for ICs and PWBs
Appendix A, part B, shows the measured Au thickness data for the PWBs and components in
this study. This data was taken using energy-dispersive X-ray analysis (EDX) for the ICs and
X-ray fluorescence (XRF) for the boards.
A Nickel-Palladium-Gold Integrated Circuit Lead Finish and Its Potential for Solder-Joint Embrittlement
5
SZZA031
Appendix A, part C, shows the calculations for Au content in the solder joints for the
embrittlement matrix in this study. The key conclusions about weight % Au in the joints are
shown in Table 2. The simplistic approach taken for these calculations was that the Au layer
dissolved completely and uniformly into the solder joint. The Au thickness was taken as the
average. The other assumptions are noted in Appendix A.
Table 2. Calculated Weight % Au in the Solder Joints for the Test Matrix
Au CONTRIBUTION
FROM COMPONENT
WEIGHT % Au IN THE JOINT (CALCULATED)
OSP
Std NiAu PWB
5× NiAu PWB
0
0.57
1.55
NiPdAu (Std finish)
0.01
0.59
1.57
NiPdAu (5× Std finish)
0.07
0.61
1.63
NiPdAu (100× Std finish)
1.13
1.70
2.66
NiPd (no Au)
The calculated contribution of the component lead finish to the percentage Au in the solder joint
in the standard and 5× NiPdAu thickness range is 0.01 weight % Au to 0.07 weight % Au. The
Au from the PWB, even at the standard level of Au on the board, overwhelms the contribution of
Au on the lead to the weight % Au in the joint. The Au flash on component leads poses a
negligible risk of contributing to Au embrittlement of the solder joints made with such
components.
Performance Measures and Results
Visual Appearance
The solder wetting and fillet height of the solder joint is shown in Figure 4. This substantiates the
assumption made in Appendix A, part A.2.d, that the fillet height of the solder joint is to the top of
the lead. The fillet height was used in calculating the surface area of the lead that was soldered.
Figure 4. Cross Section of Solder Joint
6
A Nickel-Palladium-Gold Integrated Circuit Lead Finish and Its Potential for Solder-Joint Embrittlement
SZZA031
Lead Pull Test
Lead pull testing determined the force needed to pull an individual IC lead from the PWB land
pattern after soldering. First, to allow access to an individual lead on the PWB, the leads were
cut near the package body. Next, with the leads separated from the package body, the PWB was
fastened in a test fixture. Finally, the lead was pulled perpendicular to the PWB surface until it
separated from the PWB. The rate of movement of the test device was 0.4 mm/sec vertically to
the board surface. The force to pull the lead from the PWB was measured and recorded. Lead
pull data was taken before and after temperature cycling.
Lead Pull Data
The temperature cycle excursion was –40°C to 125°C in 10-minute cycles. This is a
thermal-shock test with the boards being moved from a –40°C chamber to a 125°C chamber.
There was no ramp between the temperature extremes.
The minimum lead pull value specified by the SEMI standard for non-temperature-cycled
samples (with the lead cross-sectional area of the units tested) is 10 N.[9] All lead pull values for
temperature-cycled units are above the minimum requirement. The SEMI standard indicates that
the average lead pull force of the temperature-cycled units shall be greater than half of the
average lead pull force of the non-cycled units.[9] The lead pull values shown in Figure 5 for
post-temperature-cycled units also meet this SEMI-standard requirement.
In Figure 5, the average lead pull value for the group prior to temperature cycle exposure is
indicated with an X. A diamond shape indicates the average lead pull value for the group after
exposure to 1000 cycles of the temperature-cycle excursion stated previously. A square and
triangle denote the minimum and maximum lead pull values, respectively, for each group.
35
25
Average
20
Minimum
15
Maximum
10
PreTC
5
NiAu Std PWB
OSP PWB
NiAu (Thick Au) PWB
NiPdAu 3000 A
NiPdAu 150 A
NiPdAu Std
NiPd Std
NiPdAu 3000 A
NiPdAu 150 A
NiPdAu Std
NiPd Std
NiPdAu 3000 A
NiPdAu 150 A
NiPdAu Std
0
NiPd Std
Pull Force – N
30
Figure 5. Lead Pull Results
A Nickel-Palladium-Gold Integrated Circuit Lead Finish and Its Potential for Solder-Joint Embrittlement
7
SZZA031
Observations From Lead Pull Results
For the combinations of OSP PWB pad finish and NiPdAu, component finishes with standard Au
thickness and 5× Au show slightly higher lead pull values compared to NiPd component finish
(no Au). Lead pull values for NiPdAu finish with 100× Au are slightly less compared to the other
three samples. For the standard NiAu pad finish, lead pull values with NiPd, standard NiPdAu,
and 5× Au are all essentially equivalent. The NiPdAu component finish with 100× Au is only
slightly lower than the other three. These results are about 25% lower than the OSP results. For
the 5× Std NiAu PWB finish, the four lead pull values (post-temperature cycle) are essentially
the same as those of the standard NiAu-pad-finish board.
The most significant drop between pre-temperature-cycled and post-temperature-cycled lead
pull averages is noted for all four groups in the 5× NiAu PWB finish. The eight samples mounted
on NiAu PWB finish (both standard PWB Au thickness and 5× PWB Au thickness) appear to be,
as a group, lower than the four samples mounted on an OSP pad finish. The difference is about
25%.
Statistical Analysis of Lead Pull Data
Analysis of Variance (ANOVA)
To analyze the lead pull results from the various groups, statistical techniques were employed.
First, a multifactor ANOVA was used for the response of the lead pull (see Table 3). The ANOVA
results in Table 3 decompose the variability of the response (lead pull) into contributions due to
the inputs. In this experiment, the inputs analyzed were Au thickness on the component lead
(Lead), Au thickness on the PWB pad (PWB), and the interaction between the two (A×B).
Table 3. ANOVA Results
SUM OF
SQUARES
MEAN
SQUARES
A: Lead
593.0
197.667
8.4 %
B: PWB
4180.29
2090.15
59.3 %
A×B
380.02
63.3367
5.4 %
Residual
1892.81
4.04447
26.9 %
Total
7046.13
SOURCE
CONTRIBUTION
Main Effects
Interactions
100 %
The ANOVA results indicate that the Au on the PWB had the strongest contribution (59.3%) to
the variance seen in lead pull results between all test groups. The contribution of the Au on the
component lead accounted for 8.4% of the total variance, which is less than the contribution of
the residual error (26.9%). The interaction of the Au content in the component lead, with Au
content in PWB finish, accounted for 5.4% of the variance, which is also less than the
contribution of the residual error. ANOVA results indicate that the Au content of the PWB pad
had the strongest effect on lead pull performance, with the Au content on the component lead
being overwhelmed by the residual error of the experiment.
Graphical data analysis was used to understand the variation between the test groups. Figure 6
shows the lead pull results broken down into each factor-level setting. The graph shows Au
thickness on the component lead across the X-axis, lead pull values on the Y-axis, and Au
thickness on the PWB pad by line color (or shading).
8
A Nickel-Palladium-Gold Integrated Circuit Lead Finish and Its Potential for Solder-Joint Embrittlement
SZZA031
25
Pull Force – N
20
15
OSP
NiAu (Std)
10
NiAu (Thick Au)
5
0
30
0
150
3000
Au Thickness on Component Lead – Å
Figure 6. Lead Pull Results by Factor-Level Setting
Figure 6 indicates that the addition of NiAu to the PWB significantly lowers the lead pull strength
results when compared to OSP-coated boards. However, the pull strength on standard NiAu
PWBs or 5× Std PWBs were similar to one another. Pull tests of component samples with
100× Au were lower than those components with lesser Au thicknesses. For the line indicating
zero Au thickness on the PWB (OSP line) the data points for the Au thickness on the component
lead of 0, 0.30 Å, and 5× Au are similar, while the data point for 100× Au is different. For the
lines indicating average values of PWB Au thickness of either Std or 5× Std Au (bottom two
lines), the results are similar within each group, as well as being similar to one another. Even
with the difference in results between components mounted on OSP PWBs (no Au) and
components mounted on NiAu PWBs (various levels of Au thickness), all data points are above
the minimum required by the SEMI standard for noncycled and temperature-cycled
components.[9]
Box-and-Whisker Plots
Box-and-whisker plots were used to analyze the scatter of the data for each group. In this
data-presentation method, the box covers the middle half of the data. The middle line within
each box is the middle data point (median). The whiskers extend to the extreme data points of
each group. Outliers are indicated as small boxes beyond the whisker points. For our data set,
the shift in the pull-strength average is significantly different where the shift is large compared to
the variation in the data (distance between the whisker tips). If the mean of one box falls outside
the range of another box, a significant difference is present.
Figure 7 shows a box-and-whisker plot for the data set, with OSP coating on the PWB (no Au).
Au on the component lead for each group is indicated on the horizontal axis and the lead pull
values are indicated on the vertical axis.
A Nickel-Palladium-Gold Integrated Circuit Lead Finish and Its Potential for Solder-Joint Embrittlement
9
SZZA031
32
Pull Force – N
28
24
20
16
12
0
30
5×
100 ×
Au Thickness on Component Lead – Å
Figure 7. Box-and-Whisker Plot for OSP Coating on PWB (No Au)
The box-and-whisker plot for PWB Au thickness of zero shows that where the component-lead
Au thickness is 0, 30 Å, or 5× Au the data is similar. This is indicated by the overlap of the boxes
representing the middle half of the data. For the data set with component-lead Au thickness of
100×, the box is much lower and does not show much overlap, thus, a distinguishable difference
exists. Overall, there is a lot of scatter in the data, indicating other factors that have not been
accounted for in the experiment.
Figure 8 shows a box-and-whisker plot for the data set with standard Au (0.09 m to 0.1 m) on the
PWB. In Figure 8, results are similar between data sets with 0 and 30 Å of Au on the component
leads. Results are also similar for data sets with 0 and 5× Au on the component leads. The data
set with component-lead Au thickness of 100× is lower than data for the other data sets.
22
Pull Force – N
20
18
16
14
12
0
30
5×
100 ×
Au Thickness on Lead – Å
Figure 8. Box-and-Whisker Plot for Standard NiAu PWB
Figure 9 shows a box-and-whisker plot for the data set with 5× Au (0.4 m to 0.65 m) on the PWB.
In this figure results are similar, except where the Au thickness on the component lead is 5×.
10
A Nickel-Palladium-Gold Integrated Circuit Lead Finish and Its Potential for Solder-Joint Embrittlement
SZZA031
22
×
Pull Force – N
20
18
16
14
12
0
30
5×
100 ×
Au Thickness on Leads – Å
Figure 9. Box-and-Whisker Plot for PWB With 5× Au (0.4 µ to 0.65 µ)
Summary of Statistical Analysis
Statistical analysis indicates that addition of Au to the PWB dramatically lowers the pull strength
values. This result is reflected in the multifactor ANOVA, which showed that the Au on the PWB
had the highest percentage contribution to lead pull variance. While the presence of Au on the
PWB lowered the pull strength, the amount of Au (above zero) did not make a difference. A
lowering of the pull strength is noted when the component lead has 3000 Å of Au. There is not
much difference in the results when the Au on the component lead is 0 to 150 Å. Even in the
worst-case conditions of thick Au on the PWB and thick Au on the component lead, there was no
catastrophic drop in lead pull values after temperature cycling. All values are above the
minimum industry standard requirement for the lead pull test method.
Cross Sections of Solder Joints
Cross sections were done on the joints made from the sets of components (samples) discussed
in the following paragraphs. All of the sections were done on joints that had been through 1000
thermal cycles, except as noted for sample 10.
Sample 6: NiPdAu Standard Finish, PWB Standard NiAu Finish
The 1000× image shows a fairly uniform appearing solder joint with a Sn-rich phase in which
Pb-rich phases are embedded (see Figure 10). On the board side at 4000×, the Ni/Sn
intermetallic at the Ni/solder interface is visible (see Figure 11). Some small, Sn/Au/Ag,
intermetallic-phase areas also are present. The bulk Sn-rich phase does show some Au in the
EDX spectra, but not a distinct Au-rich phase. The Pb-rich phase shows no Au. The 4000×
image of the lead side shows the relatively thick Ni/Sn intermetallic and some Au/Sn
intermetallics scattered along the Ni interface (see Figure 12). In Figure 13, the percentage of
Cu, Ni, Au, and Sn is shown along the line marked 1–2 in Figure 12. Note the change from pure
Cu through an Ni-rich region to the SnAu intermetallic region to bulk Sn.
A Nickel-Palladium-Gold Integrated Circuit Lead Finish and Its Potential for Solder-Joint Embrittlement
11
SZZA031
Figure 10. Sample 6 After Thermal Cycling
(The Dark Gray Matrix Is the Tin-Rich Phase,
While the Bright Phase Is the
Lead-Rich Phase)
Figure 12. Lead Side of Sample 6 at 4000×
After Thermal Cycling
Figure 11. Board Side of Sample 6 at 4000×
After Thermal Cycling
Figure 13. Spectra of Cu, Ni, Au, and Sn
Taken Along Line 1–2 in Figure 12
(See Comments in Text)
Sample 5: NiPd Lead Finish, PWB Standard NiAu Finish After Thermal Cycling
The 1000× image of this section shows a predominantly Sn-rich area in which Pb-rich phases
and some rod shaped Sn/Ag phases are embedded (see Figure 14). These phases have been
seen before with SnPbAg solder paste. The 4000× image on the board side also shows a Sn/Ni
intermetallic phase on the electroless Ni plating and some scattered Sn/Au/Ni intermetallics of a
different form than the Sn/Au/Pd intermetallics on the lead side (see Figure 15).
12
A Nickel-Palladium-Gold Integrated Circuit Lead Finish and Its Potential for Solder-Joint Embrittlement
SZZA031
On the lead-side phase at 4000×, there is an Ni/Sn intermetallic layer at the Ni surface (see
Figure 16). Over this are some rods or plates of Sn/Au/Pd. The presence of Au at the lead
interface indicates that Au has migrated from the board side to the lead side because there is no
Au on the NiPd lead surface. In Figure 17 the scan along the line 1–2 indicated in Figure 16,
shows a distinct Ni region between the Cu and Sn regions. It also shows a lower Sn percentage
attributable to an SnAg intermetallic, which is the rod-shaped region in Figure 16 (the Ag is not
shown in the scan).
Figure 14. 1000× Image of Sample 5
After Thermal Cycling
Figure 15. Board Side of Sample 5 at 4000×
After Thermal Cycling
Figure 16. 4000× Image on the Lead Side
of Sample 5 After Thermal Cycling
Figure 17. Line Scan Along Line 1–2
in Figure 16 (See Comments in Text)
A Nickel-Palladium-Gold Integrated Circuit Lead Finish and Its Potential for Solder-Joint Embrittlement
13
SZZA031
Sample 7: NiPdAu 5× Au, PWB Standard NiAu Finish After Thermal Cycling
Sample 7 should be similar to samples 5 and 6 because the Au thickness on the board surface
overwhelms the Au effect from the lead (see Figure 18). In fact, there appears to be a slightly
more uniform layer of Sn/Au intermetallic at the board side. This may be caused by the roughly
7× thicker Au on the lead side migrating to the board side. It also could be caused by local
variation in the Au thickness on the PWB.
Figure 18. Sample 7 at 1000× After Thermal Cycling
Sample 2: NiPdAu Standard Finish, PWB OSP Finish After Thermal Cycling
The low-magnification image shows a solder joint with features similar to samples 5 and 6 (see
Figure 19). SnAg rod-shaped intermetallics are present. On the lead side, there is the expected
Sn/Ni intermetallic at the Ni interface and, on the board side, there is the Sn/Cu intermetallic,
which is thicker than the Sn/Ni intermetallic layer. No Au is seen in any of the images for this
system, which supports the idea that the Au from the board finish overwhelms any effect of the
Au from the lead finish.
14
A Nickel-Palladium-Gold Integrated Circuit Lead Finish and Its Potential for Solder-Joint Embrittlement
SZZA031
Figure 19. 1000× Image of Sample 2 After Thermal Cycling
Sample 10: NiPdAu Standard Finish, 5× NiAu PWB With No Thermal Cycling
This sample was not thermal cycled. Figure 20 shows large numbers of Sn/Au acicular
intermetallics dispersed in the bulk of the solder joint. The Sn/Ni intermetallics at both the board
(see Figure 21) and lead (see Figure 22) surfaces are thinner than for the thermally cycled
samples, which is expected. On the lead side there are Sn/Au regions that are detached from
the Sn/Ni and Sn/Au intermetallics on the surface. The Sn/Au intermetallic layer appears to
overlie the Sn/Ni layer that is on the Ni layer. The line scan, clearly shows that the SnAu
intermetallics at that location form because of the high Au content of the PWB and these
intermetallics are showing up at the lead side of the joint (see Figure 23). This shows the
mobility of Au in molten solder.
Figure 20. 1000× Image of Sample 10,
No Thermal Cycling
Figure 21. Sample 10 at 4000× Board Side,
No Thermal Cycling
A Nickel-Palladium-Gold Integrated Circuit Lead Finish and Its Potential for Solder-Joint Embrittlement
15
SZZA031
Figure 22. Sample 10 Lead Side at 4000×,
No Thermal Cycling
Figure 23. Line Scan Along Line 1–2
in Figure 22 (See Comments in Text)
Sample 12: NiPdAu 100× Au, PWB 5× Au Thickness After Thermal Cycling
With thick Au on both the board and the lead, Au/Sn intermetallics are expected in the bulk of
the solder. This is shown in the 1000× image in Figure 24. On the pad side, there is a thin Sn/Ni
intermetallic layer and thick Au/Sn intermetallic layers (see Figure 25). Similarly, on the lead
side, there is a thin Sn/Ni intermetallic layer over which is a thicker Au/Sn intermetallic layer (see
Figure 26). In Figure 26, there is also one of the acicular Au/Sn intermetallics. The line scan
shows that with the highest level of Au in the solder joint of any sample in this study, there are
two large SnAu intermetallics present along this scan line (see Figure 27).
Figure 24. Sample 12 at 1000×
After Thermal Cycling
16
Figure 25. Sample 12 Pad Side at 4000×
After Thermal Cycling
A Nickel-Palladium-Gold Integrated Circuit Lead Finish and Its Potential for Solder-Joint Embrittlement
SZZA031
Figure 26. Sample 12 Lead Side at 4000×
After Thermal Cycling
Figure 27. Line Scan Along Line 1–2
in Figure 26 (See Comments in Text)
Summary of Cross-Section Analysis
Both samples made with the PWBs at 5× Au showed large numbers of free-floating, acicular
AuSn in the bulk solder joint (Figures 20 through 27). None of the other samples examined
showed this phenomena. For the samples made with the standard PWB (Au at 0.2 m), there
were Sn/Au intermetallics at the solder interfaces, especially on the board side. There were a
few free-floating, nonacicular intermetallics in the 0.2-m PWB samples.
For the OSP board samples, no Au intermetallics were found for Au at any thickness from 0 to
100× Std NiPdAu. Any present may have been below the limit of detection. As expected, large
amounts of Sn/Cu intermetallics were on the OSP board.
The cross-section results support the notion that Au embrittlement of solder joints will not occur
because the Au coating on the surface of the component lead is very thin.
Summary and Conclusions
•
The theoretical Au content of the solder joints that would result from using the components
and PWBs in this study is less than the 3 weight % level cited by Glazer as the maximum Au
content for fine-pitch, surface-mount devices.
•
The theoretical calculations of Au content demonstrate that the contribution of the Au from
the lead is dwarfed by the contribution from the PWB.
•
The lead pull data shows that Au embrittlement did not occur, even after 1000 thermal
cycles.
•
The statistical analysis of the lead pull data indicates that in this study, Au on the PWB is the
prime contributor to lowering lead pull force (although results are still acceptable by industry
standards).
A Nickel-Palladium-Gold Integrated Circuit Lead Finish and Its Potential for Solder-Joint Embrittlement
17
SZZA031
•
•
The metallographic data shows that:
–
The solder joint is made to the Ni surface of the component lead.
–
There is no Cu migration through the Ni barrier layer of the lead.
–
In a system with no Au on the PWB and with a standard Au thickness on the lead, there
is no Au detectable in the bulk of the solder joint.
–
The Au from the PWB can migrate across the solder joint and appear at the lead/solder
interface in the case of the NiPd-finished lead.
–
At very high Au thicknesses on PWB and leads but give Au concentrations of less than 3
weight %, acicular SnAu intermetallics do form. These do not appear to be sufficient to
affect pull strength.
The risk of Au embrittlement caused by TI’s NiPdAu component lead finish is essentially nil.
Acknowledgments
The authors wish to recognize the following individuals for their professional assistance:
–
Kay Haulick and Martin Pauli for their board mounts, visual documentation, and lead pull
testing
–
Dr. Al Hopkins of the Texas Instruments Attleboro S&C Group for the cross sections and
SEM/EDX work
–
Bill Russell of Raytheon Technologies for support with statistical analysis of lead pull data
The authors wish to recognize Multicore for supplying the SnPbAg solder paste used in this
study.
References
1. D. C. Abbott, R. M. Brook, N. McLellan, and J. S. Wiley, IEEE Trans. CHMT, 14:567 (1991).
2. A. Murata and D. C. Abbott, Technical Proceedings, Semicon Japan, p. 415 (1990).
3. J. Glazer, P. A. Kramer, and J. W. Morris, “Effect of Au on the Reliability of Fine Pitch Surface
Mount Solder Joints”, Proceedings of the Technical Program, Surface Mount International
Conference & Exhibition, August 1991, pp. 629–639.
4. E. E. de Kluizenaar, Soldering and Surface Mount Technology, No. 4, February 1990. pp. 27–38.
5. M. F. Bester, Proceedings of Internepcon, 1968, pp. 211–231.
6. F. G. Foster, ASTM STP 319, 1963, p. 13.
7. R. Druckett and M. L. Ackroyd, Electropl. Met. Fin., Vol. 29, 1976, pp. 13–20.
8. A. Zribi et al., IEEE Transactions on Components and Packaging Technologies, Vol. 23, No. 2,
June 2000.
9. IPC-A-610C, Acceptability of Electronic Assemblies, January 2000.
18
A Nickel-Palladium-Gold Integrated Circuit Lead Finish and Its Potential for Solder-Joint Embrittlement
SZZA031
Appendix A
Calculations for Au Embrittlement Study
Part A. Theoretical
1. Sn mass per pad
a. Pad size is 0.7 mm × 1.4 mm = 0.98 mm2.
b. Stencil thickness is 150 m or 150 × 10–3 mm.
c. For a reflowed board, the solder thickness is 80 m or 80 × 10–3 mm.
d. Solder volume is 0.98 mm2 or 80 × 10–3 mm = 0.0784 mm3.
e. SnPbAg solder has a density of ∼8.4 g/cm3.
f. 0.0784 mm3 × 8.4 g/cm3 × 1 cm3/1000 mm3 = 6.59 × 10–4 g of solder per pad
2. Area of lead
a. Foot length is 1.05 mm maximum; multiply by 1.5 to allow for heel filet = 1.575 mm.
b. Foot width is 0.51 mm.
c. Foot thickness is 0.15 mm.
d. Surface area
Foot area is 1.575 × 0.51 = 0.8033 mm2.
Edge area is (assuming fillet goes to top) [(1.575 × 2) + 0.51] × 0.15 = 0.549 mm2.
e. Total soldered surface area per lead is 1.3523 mm2 per lead.
3. Calculated thickness of Au for >3 weight % Au in joint
a. 6.59 × 10–4 g Sn per pad × 0.03 = 1.3523 mm2 per lead × 19.32 g/cm3 (Au density)
× 1 cm3/1000 mm3 × Au thickness in mm
b. Solve for Au thickness in mm = 7.57 × 10–4 mm or 0.757 m or 7570 Å
c. Nominal thickness in production does not exceed 50 Å. To get to the 3 weight % Au level,
Au thickness would exceed the minimum by >150×, if there were no Au on the PWB pad.
Part B. Thickness Data on PWBs and Components Used in Study
The XRD/EDX data are for the PWBs and the components evaluated in this study. Ten leads per
component were evaluated, except as noted.
•
PWB pads (see Table A–1)
Table A–1. XRD/EDX Data
THICKNESS OF Au (m)
PWB PADS
(Au m)
AVERAGE
MAX
MIN
SD
5× NiAu
0.55
0.67
0.48
0.04
Std NiAu
0.20
0.28
0.09
0.05
A Nickel-Palladium-Gold Integrated Circuit Lead Finish and Its Potential for Solder-Joint Embrittlement
19
SZZA031
•
Devices (see Table A–2)
Table A–2. XRF/EDX Data
COMPONENTS
THICKNESS OF Au (Å)
AVERAGE
MAX
MIN
SD
3000 Å
2893
4325
2225
722
150 Å
×1835†
180
250
125
42
33
37
30
3
† Sub-50-Å data obtained by EDX
NOTE 1: 1 m = 10,000 Å = 10 × 10–3 mm
Part C. Calculations of Au Concentration in Parts Used in the Study
This part gives calculations of the percentage of Au in the joints we made, accounting for the Au
on the PWB pads. The averages of measured Au values are used in the calculations.
4. Boards (see Table A–3)
a. Pad size is 0.7 mm × 1.4 mm = 0.98 mm2.
b. Au thickness on board 38a is 0.55 m or 5.5 × 10–4 mm.
c. Au thickness on board 39a is 0.20 m or 2.0 × 10–4 mm.
d. Mass of Au on 38a is
0.98 mm2 × 5.5 × 10–4 mm × 1 cm3/1000 mm3 × 19.32 g/cm3 = 1.04 × 10–5 g Au/pad.
e. Mass of Au on 39a is
0.98 mm2 × 2.0 × 10–4 mm × 1 cm3/1000 mm3 × 19.32 g/cm3 = 3.79 × 10–6 g Au/pad.
5. Components (see Table A–4)
a. Total soldered surface area per lead is 1.3523 mm2.
b. For the nominal 3000-Å leads:
1.3523 mm2 × 2.893 × 10–4 mm × 1 cm3/1000 mm3 × 19.32 = 7.56 × 10–6 g Au/lead
c. For the nominal 150-Å leads:
1.3523 mm2 × 1.80 × 10–5 mm × 1 cm3/1000 mm3 × 19.32 = 4.70 ×10–7 g Au/lead
d. For nominal 30-Å leads:
1.3523 mm2 × 3.3 × 10–6 mm × 1 cm3/1000 mm3 × 19.32 = 8.62 × 10–8 g Au/lead
6. Percentage of Au in solder joints (general formula) (see Table A–5)
a. Percentage of Au = [(mass Au on lead + mass Au on pad)/(mass of Sn + mass Au on
lead + mass Au on pad)] × 100
b. Mass of solder joint = 6.59 × 10–4 g
Table A–3. Calculations for Boards
FOR
BOARDS
MASS OF Au
(g)
38a
1.04 × 10–5
3.79 × 10–6
39a
20
THICKNESS USED
IN CALCULATION
(m )
0.55
0.20
A Nickel-Palladium-Gold Integrated Circuit Lead Finish and Its Potential for Solder-Joint Embrittlement
SZZA031
Table A–4. Calculations for Components
NOMINAL
LEAD THICKNESS
(Å)
MASS OF Au
(g)
THICKNESS USED
IN CALCULATION
(Å)
7.56 × 10–5
4.70 × 10–7
2893
150
30
8.62 × 10–8
33
3000
180
Table A–5. Calculated Percentage of Au in the Solder Joints
Au CONTRIBUTION
A
FROM COMPONENT
NiPd (no Au)
CALCULATED WEIGHT % Au IN THE JOINT
OSP
Std NiAu PWB
(0.28 m)
5× NiAu PWB
(0.67 m)
0
0.57
1.55
NiPdAu (Std finish 30-Å Au)
0.01
0.59
1.57
NiPdAu (150-Å Au)
0.07
0.61
1.63
NiPdAu (3000-Å Au)
1.13
1.70
2.66
NOTES: 2. The contribution of the lead finish to the percentage Au in the solder joint in
the 50-Å to 150-Å thickness range is 0.01 weight % and 0.07 weight %. The
Au from the PWB, even at the level of 0.2-m Au on the board overwhelms the
contribution of the Au on the lead.
3. The percentage of Au in a joint can be estimated from EDX data or, possibly,
from a wet chemical analysis for the percentage of Au in the joint.
A Nickel-Palladium-Gold Integrated Circuit Lead Finish and Its Potential for Solder-Joint Embrittlement
21
IMPORTANT NOTICE
Texas Instruments Incorporated and its subsidiaries (TI) reserve the right to make corrections, modifications,
enhancements, improvements, and other changes to its products and services at any time and to discontinue
any product or service without notice. Customers should obtain the latest relevant information before placing
orders and should verify that such information is current and complete. All products are sold subject to TI’s terms
and conditions of sale supplied at the time of order acknowledgment.
TI warrants performance of its hardware products to the specifications applicable at the time of sale in
accordance with TI’s standard warranty. Testing and other quality control techniques are used to the extent TI
deems necessary to support this warranty. Except where mandated by government requirements, testing of all
parameters of each product is not necessarily performed.
TI assumes no liability for applications assistance or customer product design. Customers are responsible for
their products and applications using TI components. To minimize the risks associated with customer products
and applications, customers should provide adequate design and operating safeguards.
TI does not warrant or represent that any license, either express or implied, is granted under any TI patent right,
copyright, mask work right, or other TI intellectual property right relating to any combination, machine, or process
in which TI products or services are used. Information published by TI regarding third–party products or services
does not constitute a license from TI to use such products or services or a warranty or endorsement thereof.
Use of such information may require a license from a third party under the patents or other intellectual property
of the third party, or a license from TI under the patents or other intellectual property of TI.
Reproduction of information in TI data books or data sheets is permissible only if reproduction is without
alteration and is accompanied by all associated warranties, conditions, limitations, and notices. Reproduction
of this information with alteration is an unfair and deceptive business practice. TI is not responsible or liable for
such altered documentation.
Resale of TI products or services with statements different from or beyond the parameters stated by TI for that
product or service voids all express and any implied warranties for the associated TI product or service and
is an unfair and deceptive business practice. TI is not responsible or liable for any such statements.
Mailing Address:
Texas Instruments
Post Office Box 655303
Dallas, Texas 75265
Copyright  2001, Texas Instruments Incorporated
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertising