implementation of a collaborative industrial consortia ... coyle, nokia bell labs, murray hill, nj,...
TRANSCRIPT
Implementation of a Collaborative Industrial
Consortia Program to Characterize the
Thermal Fatigue Reliability of Third-
Generation Pb-Free Solder Alloys*
Richard J. Coyle
Nokia Bell Labs
Murray Hill, NJ, USA
(908-582-8062)
SMTA Capital Expo
August 30, 2016
* Content modified from work to be presented at SMTAI 2016
.
Authors and Affiliations
Richard Coyle, Nokia Bell Labs, Murray Hill, NJ, USA Project Chair
Richard Parker, iNEMI, Tipton, IN, USA Project Co-Chair [email protected]
Keith Howell and Keith Sweatman, Nihon Superior Co., Ltd., Osaka, Japan
Dave Hillman, Rockwell Collins, Cedar Rapids, IA, USA
Joe Smetana, Nokia, Plano, TX USA
Glen Thomas and Hongwen Zhang, Indium Corp, Utica, NY, USA
Stuart Longgood and Andre Kleyner, Delphi, Kokomo, IN, USA
Michael Osterman, UMD CALCE, College Park, MD, USA
Eric Lundeen, i3, Endicott, NY, USA
Polina Snugovsky, Celestica, Toronto, ON, Canada
Julie Silk, Keysight Technologies, Santa Rosa, CA, USA
Rafael Padilla and Tomoyasu Yoshikawa, Senju Metal Industry Co., Tokyo, Japan
Mitch Holtzer, Alpha Assembly Solutions, South Plainfield, NJ, USA
Jasbir Bath, Bath & Associates Consultancy, Fremont, CA, USA
Jerome Noiray and Frederic Duondel, Sagem, Paris, France
Raiyo Aspandiar, Intel, Hillsboro, OR, USA
Jim Wilcox, Universal Instruments, Conklin, NY, USA
.,
Outline
Background
- Attachment reliability and the thermal fatigue failure
mode in BGA-type packages
- Driving force for the project
- Basis for the collaborative consortium approach
- Overview of the iNEMI Alloy Alternatives Project as the
model for current project
- Member companies and project leadership
3rd Generation Pb free Alloy Project: Problem statement,
overview, and project scope
Experimental test matrix
Status and schedule
Attachment Reliability and Thermal Fatigue
Fatigue of solder attachments is the major wear-out failure mode for surface mount components in electronic assemblies. We don’t want products to fail in use over their design lifetime.
Exposure to temperature changes can cause fatigue failure of solder joints.
The severity of a use environment is determined by the difference in its temperature extremes (ΔT) and the peak temperature.
The coefficient of thermal expansion (CTE) mismatch between a component and a printed circuit board drives solder joint fatigue.
The components most vulnerable to solder fatigue are low CTE ceramic packages and QFN and BGA plastic packages with non-compliant attachments.
CTE Mismatch Drives Solder Joint Failure
Substrate
CTE = 13 - 15 ppm
Solder Ball
CTE = 25 ppm
Circuit Board
CTE = 14 - 21ppm
Mold Compound
CTE = 11 - 18 ppm
Silicon
CTE = 2.5 ppm
CTE values of the different constituents of an overmolded plastic
BGA (PBGA) assembled to a printed circuit (PCB)board
Metallographic cross
section of a thermal fatigue
failure in a Pb-free BGA
solder joint
Thermal Fatigue of Solders
Solder fatigue is low cycle fatigue involving plastic deformation and fewer cycles
to failure compared to the more familiar case of high cycle fatigue where stress is
low and deformation is primarily elastic.
Solder fatigue is cyclic deformation by a creep mechanism.
As application environments become more aggressive, solders are required to
operate at greater homologous temperatures.
Fatigue limit for 1045 steel
~106 cycles
Summary of BGA Package Reliability Drivers
Property Direction Attachment Reliability
Influence
Package Pad Size +++
Substrate Thickness
+++
Package Stand-off +
Die Size ++
Package Body Size Package CTE
Moiré fringes for
full array package
Die
location
Fringes pattern
indicates lower
CTE under Si die
Driving Force for the Pb-Free Alloy Project
Significant innovations in Pb-free solder alloy compositions are driven by volume manufacturing and field experiences.
The industry continues to see an increase in the number of Pb-free solder alloys beyond the near-eutectic Sn-Ag-Cu (SAC) alloys that replaced SnPb eutectic.
New Pb-free alloys are formulated to address shortcomings of near-eutectic SAC: poor mechanical shock performance, alloy cost, copper dissolution of plated through holes, and poor mechanical behavior of joints during board bending.
There are several technical and logistical risks. This project focuses on thermal fatigue resistance. High reliability end users that use SAC305, are aware of additional alloy offerings and are addressing the thermal fatigue reliability of those alloys.
The iNEMI consortium initiated the Alternative Alloy Characterization Project in 2008 to “provide thermal cycle reliability data on a variety of commercially and scientifically important alloys.”
If anything, this situation is more complex than it was at the onset of the original iNEMI Alloy Alternatives Project in 2009.
Background: Basis for Consortia Collaboration
The iNEMI Alternative Alloy Characterization Project is a R&D project based on
alloy science and application requirements. It has an open time frame because
it is tied to evolving alloy development and previous test results.
Phases 1-3 (2009-2015) compared performance of lower cost, lower Ag
content SAC (SnAgCu) alloys to high Ag alloys such as SAC305 and SAC405.
Phase 4 planning began In 2014 to evaluate so-called 3rd generation Pb-free
alloys. Phase 4 was launched officially in 2015 following SMTAI 2015.
In 2015, the High Density Package User Group (HDPUG) initiated a program to
evaluate performance of high reliability solders. Because of similar goals and
common core team membership, iNEMI and HDPUG agreed to collaborate
on a program to evaluate thermal fatigue performance of 3rd generation
Pb-free alloys.
The iNEMI team also includes CALCE and Universal AREA participation.
9
+ + +
Overview of the iNEMI Alloy Project (2009-2015):
Thermal Fatigue Reliability of Multiple Pb-free Alloys
Develop an area array test vehicle Validate the impact of Ag concentration in the range of 0 to 4% Evaluate the impact of commercially common dopants (microalloy
additions), such as Ni, on thermal fatigue performance. Evaluate the impact of dwell time Evaluate the impact of aging (isothermal preconditioning) on reliability. Provide basic thermal fatigue data for several of the emerging alternate
alloys on the market today, benchmarking them against SAC305. Work with the IPC Solder Products Value Council to establish standard
test methods for alloy properties and reliability evaluations.
We will leverage the test vehicle and the experience developed with thermal cycling, data analysis, and microstructural characterization from the original program.
iNEMI Alloy Test Vehicle
Components purchased as land grid arrays
(LGA) with subsequent ball attachment with
alloy spheres to create ball grid arrays (BGA).
No. BGA Ball Alloy
Trade Name or
Designation
Solder
Paste Comments
1 Sn-37Pb Eutectic Sn-Pb Sn-37Pb Control
2 Sn-0.7Cu+0.05Ni+Ge SN100C SN100C 0% Ag joint
3 Sn-0.7Cu+0.05Ni+Ge SN100C SAC305 Impact of [Ag]
4 Sn-0.3Ag-0.7Cu SAC0307 SAC305 Impact of [Ag]
5 Sn-1.0Ag-0.5Cu SAC105 SAC305 Impact of [Ag]
6 Sn-2.0Ag-0.5Cu SAC205 SAC305 Impact of [Ag]
7 Sn-3.0Ag-0.5Cu SAC305 SAC305 Impact of [Ag]
8 Sn-4.0Ag-0.5cu SAC405 SAC305 Impact of [Ag]
9 Sn-1.0Ag-0.5Cu+0.05Ni SAC105+Ni SAC305
Impact of
dopant
10 Sn-2.0Ag-0.5Cu+0.05Ni SAC205+Ni SAC305
Impact of
dopant
11 Sn-1.0Ag-0.5Cu+0.03Mn SAC105+Mn+Ce SAC305
Impact of
dopant
12 Sn-0.3Ag-0.7Cu + Bi SACX0307 SAC305
Doped
commercial
alloy
13 Sn-1.0Ag-0.5Cu SAC105 aged SAC305 Effect of aging
14 Sn-3.0Ag-0.5Cu SAC305 aged SAC305 Effect of aging
15 Sn-1.0Ag-0.7Cu SAC107 SAC305 Impact of [Cu]
16 Sn-1.7Ag-0.7Cu-0.4Sb SACi SAC305
Doped
commercial
alloy
Profile No. Company
Cycle (Min/Max/Dwell) Comment
1 ALU 0/100/10
Core DOE
2 IST 25/125/10
3 Henkel -40/100/10
4 Nihon -15/125/10
5 ALU 0/100/60
6 HP 25/125/60
7 HP -40/100/60
8 CALCE -15/125/60
9 CALCE -40/100/120 Long Dwell
10 Delphi -40/125/10Common; Auto
ATC-DOE Factors Level-1 Level-2
Max. Temperature, oC 100 125
Dwell-Time, minutes 10 60
ΔT (maxT – minT) , oC 100 140
12 Pb-free alloys (plus SnPb),10 temperature profiles, 6 test locations
iNEMI Alloy Phases 1-3: Experimental Scope
iNEMI Alloy Project: Examples of Thermal Cycling Data
iNEMI Alloy Project: Microstructure and Failure Analysis
16 conference papers and
two journal papers
JOM Vol. 67, no. 10, 2015
The original consortium
project documented the
findings using an “open
source” approach and this
approach will continue in
the new project.
iNEMI Alloy Project: Documentation and Bibliography
Pb-Free Alloy Characterization
Team Members and Contributors
16
25 companies representing:
Solder alloy suppliers, device suppliers, EMS providers, OEMs, industrial
consortia, test and measurement, PCB fabrication
12 companies are members of both iNEMI and HDPUG
Collaborative Project Leadership Team
Richard Coyle, Nokia Bell Labs, Chair
Rich Parker, iNEMI, Co-chair
David Godlewski, iNEMI Project Coordinator
Larry Marcanti, HDPUG Project Facilitator
3rd Generation
Pb-free Solder
Alloy Evaluation
Some Questions and Answers
What do we mean by “3rd generation” Pb-free solder alloys?
1st generation commercial alloys: near-eutectic SAC (SnAgCu)
alloys. SAC305 emerged as the alloy of choice circa 2006.
2nd generation alloys: low Ag and no Ag alloys to address
shortcomings of near-eutectic SAC, such as the poor mechanical
shock performance, alloy cost, copper dissolution of plated through
holes, and solder joint or laminate failure during board bending.
3rd generation alloys: both high and lower Ag alloys with solid solution
strengthening and microalloy additions to address specific
shortcomings of current offerings.
Is SAC305 so inadequate that we need more new alloys?
The performance of SAC305 is acceptable in many applications.
However, certain high reliability automotive, military and defense,
avionics, and telecom applications may require better thermal fatigue
resistance, better drop/shock or vibration resistance, lower processing
(melting) temperature, and lower cost.
Commercial Motivation for High Reliability Solders
Automotive electronic assemblies must perform in environments
characterized by increasing temperatures, thermal and power cycling,
vibration, and thermal and mechanical shock.
Automotive modules and components are qualified using -40/125 °C
testing. This is not considered a significant acceleration factor.
Commercial Motivation for High Reliability Solders
Electronics control a large number of critical and non-critical automotive functions.
Solder joint reliability is a critical requirement due to the aggressive use environment.
More than 10 years ago, a working group including Siemens, Bosch,
Henkel/Heraeus, Alpha Metals, Infineon, the Fraunhofer Institute and
others was formed to develop a solder that could meet severe
automotive requirements.
The outcome was the Innolot (90iSC) alloy, a near-eutectic SAC
composition modified with significant additions of Bi, Sb, and a
microalloy addition of Ni.
Since then, many other high Ag, high reliability alloys have been
introduced. Additionally, alloys modified with Bi and Sb have been
introduced with Ag content lower than SAC305 to address needs for
better drop/shock resistance, lower processing (melting) temperature,
and lower cost.
High Reliability Pb-free Solder Alloy Development
The primary strengthening mechanism in SAC solders is the formation of
networks of Ag3Sn precipitates at the Sn dendrite boundaries.
During thermal cycling or high temperature exposure, the Ag3Sn
precipitates coarsen and become less effective in inhibiting dislocation
movement and slowing damage accumulation. Precipitate coarsening is
driven by a combination of temperature and strain (ΔT).
Precipitate Strengthening of SAC with Ag
SAC305 as solidified SAC305 after thermal cycling
Ag3Sn
Solid Solution Hardening with Bi, Sb, or In
Dislocation movement or deformation is inhibited by distortion in
the β-Sn lattice caused by solute atoms such as Bi, Sb, or In.
Alloying Elements Can Effect the Melting Temperature and Melting Range
Bi and In can lower the liquidus temperature and Sb may raise the liquidus temperature.
Effect of Microalloy or Trace Additions
Ni – forms intermetallic precipitates that can strengthen the solder and also may strengthen by inducing Sn grain refinement.
Ge – improves wetting
Ce, La, Er, Nd – it has been suggested that microalloy or trace amounts of various rare earth elements can improve the wetting and enhance the mechanical properties of SAC solders. Some of these elements may promote Sn whisker growth.
The project will continue to focus on using thermal cycling to evaluate
the thermal fatigue reliability of various solder alloys.
Third generation Pb-free solder alloys include two prominent
development paths:
- Higher reliability alloys more suited for the industries like
automotive, military/defense, avionics, and telecom.
- Alloys with Ag content lower than SAC305 to address needs for
better drop/shock resistance, lower processing (melting)
temperature, and lower cost.
The current emphasis is to compare thermal fatigue reliability data
generated with the aggressive thermal cycling profiles used for
automotive and military/defense applications.
3rd Generation Pb-free Alloy Thermal Fatigue Evaluation
3rd Generation Pb-free Alloy Matrix
iNEMI Pb-free Alloy Daisy Chained
Thermal Fatigue Test Vehicle
Daisy chained BGAs and PCB to
enable in situ resistance
monitoring during thermal cycling.
Pb-free Daisy Chain Component Fabrication
The components are
purchased as land grid
arrays with subsequent ball
attachment performed for
the various alloys.
iNEMI Pb-free Alloy Test Vehicle Attributes
3rd Generation Pb-free Alloy Thermal Cycling
In situ resistance monitoring during thermal cycling in accordance with
the IPC-9701 Attachment Reliability Guideline:
0/100 °C (TC1) preferred for telecom applications
-40/125 °C (TC3) used commonly for automotive requirements
-55/125 °C (TC4) for military/defense requirements
-40/150 °C for very aggressive automotive requirements
3rd Generation Pb-free Alloy Thermal Cycling Matrix
15 alloys + SAC305 and SAC105 controls
Component sample size of 32 for each alloy test cell
Two replicate boards per test cell
OSP final finish with OSP and ENIG in -55/125 °C ATC profile
Baseline assemblies for microstructural characterization
4848 components and 167 assembled boards
Status
The solder ball attachment has been completed on all of the
components required to populate the thermal cycling and baseline test
matrix.
The stencil printing process and surface mount reflow profile were
developed using SAC305 components and solder paste during a pilot
build conducted at Rockwell Collins. Multiple stencils have been
ordered to facilitate the assembly of the complete test matrix.
All but one of the solder pastes have been delivered to Rockwell
Collins and are in refrigerated storage.
The surface mount assembly of the test boards will begin in early
September 2016.
Thermal cycling will be performed at four participant sites:
- CALCE (-40/125 °C)
- i3 (-40/150 °C)
- Nokia Bell Labs (0/100 °C)
- Rockwell Collins (-55/125 °C)
Acknowledgments
The authors wish to thank the following:
Pete Read from Nokia Bell Labs, Dennis Decker, Product Line
Manager, Micross Components, and the staff at Micross Components
for coordinating the ball attachment.
Michiharu Hayashi of CMK America and Jim Fuller of Sanmina for PCB
test vehicle support.
Neil Hubble and Ryan Curry of Akrometrix for warpage measurements.
Jörg Trodler of Heraeus, Shantanu Joshi of Koki, Mike Gesick of
Inventec, Brook Sandy-Smith of Indium, and Torey Tomaso of Alpha
Assembly Solutions for facilitating the acquisition of the various solder
sphere samples and solder pastes.
Grace O’Malley of iNEMI and Marshall Andrews of HDPUG for their
help in establishing the collaborative agreement and Dave Godlewski,
Larry Marcanti, and Robert Smith for their continued support in
coordinating the work between the cooperating consortia.
Richard J. [email protected]
(908-582-8062)
Richard Coyle is a Consulting Member of the Technical Staff in the Bell
Labs Reliability Engineering organization of Nokia in Murray Hill, NJ.
He is responsible for reliability and quality of electronic assemblies and
manages the work in the Bell Labs Interconnection Failure Analysis
Lab. His work often has focused on the reliability and metallurgical
issues affecting the conversion to Pb-free manufacturing. He received
his Ph.D. in Metallurgical Engineering and Materials Science from the
University of Notre Dame. He is a member of TMS, ASM International,
the CPMT of IEEE, AWS, and is on the Board of Directors of SMTA.
Thank You!