calce and calce pb-free research activities

1 University of MarylandCopyright © 2007 CALCE

Center for Advanced Life Cycle Engineeringhttp://www.calce.umd.edu

APEX 2007Reliability Summit

“Keeping industry reliability test protocols current with rapidlychanging markets”

CALCE and CALCE Pb-freeResearch Activities

ISO 9001:2000Certified, 1999

Formed 1987

Michael Osterman, [email protected]

Center for Advanced Life Cycle EngineeringUniversity of MarylandCollege Park, MD 20742

(301) 405-5323http://www.calce.umd.edu



What is CALCE?Center for Advanced LifeCycle Engineering(founded 1987) is dedicatedto providing a knowledgeand resource base tosupport the developmentand sustainment ofcompetitive electroniccomponents, products andsystems.

Areas of research

• Physics of Failure• Design of Reliability• Accelerated Qualification• Supply-chain Management• Obsolescence• Prognostics

~24 Faculty and Research Staff~20+ M.S. students~60+ Ph.D. students

CALCECALCEElectronic Productsand Systems Center

~$5M/Year

CALCECenter for Advanced

Life Cycle EngineeringRisk Mgmt in

Avionics Systems

• Manufacturing for sustainment(USAF ManTech Program)

• IEC and avionics workinggroup collaboration

LabServices

• Small jobs• Fee-for-service• Proprietary work• Use of CALCE Tools &

Methods• Turnkey capabilities• “Fire-fighting”

MEMSTechnology

• Combined RF MEMS andSi/Ge Hetrojunction BipolarTransistors (HBTs)

• MEMS chip-to-chip

bonding reliability

ResearchContracts

• Larger programs• Some past programs:• Power Electronics (Navy)• Embedded Passives (NIST)• Risk Management (USAF)• Life Assessment (NASA)• MEMS (NASA,NSWC)

• Risk assessment, mitigationand management of electronicproducts and systems

CALCEElectronic Products

and SystemsConsortium

EPSConsortium

• 40-45 companies• Pre-competitive research• Risk assessment,

management, andmitigation for electronics

http://www.calce.umd.edu

• Risk assessment, mitigationand management of electronicproducts and systems

CALCEElectronic Products

and SystemsConsortium

PHMConsortium

• Pre-competitive research• Research in fundamental

methodologies to developand implement prognosticsand health managementsystems.

Education• MS and PhD EPS

program• International visitors• Web seminars• Short courses for

industry• Publications

Long-termPb-free Reliability

Study• Aging• Intermetallic formation• PCB surface finish• ECM• Solder interconnects



CALCE Consortia Members

• Arbitron• Aavid Thermalloy

• ACEL, China

• Argon

• B & G , Sequel, CA

• BAE Systems

• Beihang University, China

• Boeing• DBD, Germany• Defense Logistics Agency• Dell Computer Corp.• EADS CCR, France• Emerson• EMC Corp.• Ericsson AB, Sweden

• ERS

• Faraday

• GCAS

• General Dynamics, AIS

• GE Corp.

• Goodrich Engine Control,

UK

• Grundfos, Denmark

• Hamilton Sundstrand

• Harris Corp.

• Hewlett-Packard Co.

• Honeywell

• Instit. Nokia de Tecnologia, Brazil

• Lutron

• Medtronic, Inc.

• Mercury Computer Systems

• Motorola

• NASA

• NAVAIR

• Northrop Grumman Corp.

• Naval Surface Warfare Center

• Nokia Research Center, Finland

• Philips Electronics, The Netherlands

• Qualmark

• QinetiQ Aquila, UK

• Reactive Nano Tech

• Research In Motion, Ltd., Canada

• ReliaSoft Corporation

• Rockwell Collins, Inc.

• Rolls Royce

• SAIT, Korea

• Samsung Memory, Korea

• Samsung Techwin Co., Ltd.,

Changwon-si, Korea

• Sandia National Labs

• Schlumberger Oil Field Services

• Seagate Technology Inc.

• Siemens AG, Germany

• Smiths Aerospace

• Sun Microsystems (StorageTek),

• Tessera

• TRW Automotive, UK

• Toshiba, Japan

• U.S. AMSAA

• U.S. Army Research Lab.

• U.S. Army Picatinny

• Whirlpool Corp.



CALCE Test Services and Failure AnalysisCALCE Test Services and Failure AnalysisLaboratoryLaboratory

ServicesServices• Benchmarking of part and contract manufacturers

• Review of mechanical and electrical design

• Review of failure-mode mechanisms andeffects-analysis (FMMEA)

• Recommendations for accelerated life testing

• Accelerated testing to determine product robustness

• Virtual qualification and reliability assessment

• Identification of critical-to-quality parameters

• Materials characterization

• Failure analysis and uncovering new failures

– Electrolytic capacitor failures due to aqueous electrolyte formulation

– Vertical filament formation (VFF)

– Cracking in base metal electrode (BME) ceramic capacitors in humidenvironments

Failure ofelectrolyticcapacitors

Failure ofPWB due

to VFF



CALCE EP&S Consortium Research ProgramCALCE EP&S Consortium Research Program

Research Program Theme(Identification and development oftechnologies, methodologies, andguidelines for assessing, mitigating, andmanaging the risks associated with thedesign, manufacture and fielding ofelectronic products and systems)

FailureIdentificationand ReliabilityModeling

Research RoadmapThrust Areas

FY07 Research Areas

Risk-informedTechnologyInsertionMethodologies

Environmentaland OperationalAssessment ofProducts

Pb-free Electronics

Failure Mechanisms

New Technologies, Materials andProcesses

Virtual Qualification

Part Management and Sustainment

Thermal Design and Management



CALCE EPSC Pb-free Research

Solder material testing (constitutive and durability properties)NCMS (4 solders)Sn/3.9Ag/CuSn/AgSn/CuOthers (Sn-In/Bi/Zn/Al/Sb/3.0Ag/?)

PWB/component finish & interface integrity (OSP, Imm Au/Ag/Sn, Au/Ni, HASL,SnPb) (Mixed technology issues eg Pb-contamination; Post-aging tests)

Overstress (ball shear, PWB flexure, shock)Cyclic durabilityNoble platings and creeping corrosionWhiskering

Connector fretting corrosionConductive and nonconductive adhesives

Soft particles (Au-plated polymers)No particles

Accelerated testing (SnAgCu, ?, multiple finishes, mixed tech, pre-aging)Thermal cycling/shockVibrationCombinationsMechanical shock/impactHumidity

Virtual qualification software/Model calibrationManufacturing/Rework QualityBusiness risk assessment, IP & liability issues

Virtu

al

qu

alifica

tion

,D

esign

trad

eoffs

Accelera

tedtestin

g,

Hea

lthm

on

itorin

g

---------96 97 98 99 00 01 02 03 04 05 06



CALCE Pb-free Solder Temperature CyclingReliability Testing

• Solders Completed– Indium SMQ 230 Sn95.4/Ag3.9/Cu0.7– Indium SMQ 230 Sn96.5/Ag3.5– Indium SMQ 92J Sn63/Pb37

• Solder Under Test– Aim SN100C Sn/Cu/Ni(.5) w/254 flux– Aim SAC 305 w/254 flux

– Indium SMQ92J Sn61.5/Pb 36.5/Ag2

Test details• 16 samples in each test condition• Resistance of each chip is monitored by a data

logger.• Temperature is recorded at the center of each

card.• Test continues until 100 % failure occurs.• Cross sectioning was performed on failed test

specimens to verify a solder interconnect failure.

Packages Under Test• 68-pin LCCC: 24mm 24mm• 84-pin LCCC: 30mm 30mm• PCB Board: 130 x 93 x 2.5 mm,FR4



Experimental Test Matrix

Completed757085158

Completed1510050-509#

85

100

125

75

125

75

100

Max.Temp.(C)

Completed1570157

Completed7510006

Completed75100255

Completed

Completed

Completed

Completed

Status

75

15

15

15

Dwell Timeat Max

temp* (min)

10004

100253

100-252

10001

Temp.range(C)

Min.Temp.(C)

Test

*Dwell at minimum temperature is set to be 15 minutes.# SnPb had only four cards



Test Results for Tmean = 50oC, T=100oC

x 10x

1.00

5.00

10.00

50.00

90.00

99.00

Probability - Weibull

Time, (t)

Unre

liabili

ty,F

(t)

8/24/2004 09:43CALCE Center

WeibullLccc68 Sn/Ag

W2 RRX - SRM MEDF=16 / S=0Lccc68 Sn/Ag/Cu

W2 RRX - SRM MEDF=16 / S=0Lccc68 Sn/Pb

W2 RRX - SRM MEDF=15 / S=0Lccc84 Sn/Ag

W2 RRX - SRM MEDF=15 / S=0Lccc84 Sn/Ag/Cu

W2 RRX - SRM MEDF=16 / S=0Lccc84 Sn/Pb

W2 RRX - SRM MEDF=16 / S=0

At lower cyclicmean and maxtemperatures thePb-free soldersoutperformed SnPbsolder. This may berelated to the lowercreep rate of thePb-free soldersunder test ascompared to theSnPb solder.

SnPb

Pb-free



Comparison of Time to Failure(68 IO Package)

For the Pb-free solders, increasing the average cyclic temperature showed adecrease in time to failure. As can be seen in the above chart, the behavior of theSnPb solder at the 100 and 125oC peak temperature shows non-monotonicallydecreasing behavior.

Peak Temperature (oC)(Dwell at Peak (min) )

Normalized



CALCE Strain Range Based Solder FatigueRapid Assessment Model

For eutectic solder,

• f Constant

•

1

1

2 2

cp

f

f

N

• Nf : mean number of cycles to failure

• p : inelastic strain range, g (package type,geometry, dimension, material property, load profile)

• K lead stiffness

• f model calibraton factor

• f = material constant, fatigue ductility coefficient

• c = material constant, fatigue ductility exponent, h(load profile)

• Tsj: mean cyclic temp. (°C)

• tD: dwell in minutes at the max. temp.

This model is preferred since it provides a physical relationship between thetemperature cycle load and failure and it has a quick solution time.

h

p

Ld

Ld (T) /2

Engelmaier, W., "Fatigue Life of Leadless Chip Carrier Solder Joints During Power Cycling,", Components,Hybrids, and Manufacturing Technology, IEEE Transactions on, Volume: 6 Issue: 3 , Sep 1983, pp. 232 -237

d

sjt

cTccc360

1ln210

Ah

TLfK dp

200

2

Leaded

Leadless

h

TfLdp

2



Fitted Strain Range Model Parameters

2.253.472.25f*[1]

0.8980.9660.980R^2

1.40E-027.83E-031.45E-02c2

-2.10E-03-1.74E-03-7.34E-04c1

-0.416-0.347-0.502co

SASACSnPbSolder

Parameters

[1] The fatigue ductility constant was derived onlyconsidering the neutral distance length of thepackage. This value must be adjusted withrespect to the cyclic temperature range, CTEmismatch between the package and the board, andthe effective solder joint height.

d

sjt

cTccc360

1ln210

f = Constant

c

f

fe

TLN

1

]1[22

1

Norm

aliz

edF

atig

ue

Lif

eA Strain Range Based Model for Life Assessmentof Pb-free SAC Solder Interconnects , M. Osterman,A. Dasgupta, B. Han, 56th Electronic Component andTechnology Conference, pp. 884 - 890, May 30-June2, 2006 Dwell Time (hrs)



2 mm thick board contained PBGA,TSOP, TQFP, CLCC packages. Thesimulation model is based on testingconducted under the JGPP/JCAA Pb-free Solder Test Program.

Test assemblies were subjected to a-55 to 125oC temperature cycle and a-20 to 80oC cycle condition

calcePWA Model

•JCAA/JG-PP No-Lead Solder Project:-55ºC to +125ºC Thermal Cycle Testing Final Report, David Hillman and Ross Wilcoxon, March 15, 2006

Comparison with Stain Range SimulationModel



Comparison of Simulation Results

Experiment Results

Sim

ula

tion

Results

TSOP (-20 to 80oC)

TQFP (-20 to 80oC)

Parts include (BGA,CLCC, TSOP, TQFP).Current model overestimates life of leadedparts. This result islikely due to the squareof T in estimating thestrain range for leadedparts.

•Data was obtained from

•JCAA/JG-PP No-Lead Solder Project:-55ºC to +125ºC Thermal Cycle Testing Final Report, David Hillman and Ross Wilcoxon, March 15, 2006

• Communication with T. Woodrow



Leadless Strain Range for Leaded Parts

Experiment Cycles to Failure

Sim

ula

tion

Cyc

les

toF

ailu

re

TSOP and TQFPparts modeledwith the leadlessstrain rangemodel.



Parts to bereworked

CALCE Thermal Aging and Rework Study- Test Board -

Board parameter• 8”x7”, single layer, one sided

(FR4) board• Board thickness: 62 mils• Parts to be reworked: 2 BGAs,

4 QFPs, 8 resistors (2512)• Glass transition temperature

(Tg): 130 ºC• Pb-free Board – ImSn Finish• SnPb Board – SnPb HASL

• Pb-free partsBGA (SAC305), QFP(Sn0.7Cu, Sn2Bi, Sn), 2512Resistors - (Sn)

• SnPb partsBGA (SnPb), QFP (SnPb),Resistor (SnPb)

Sn

Sn

SnCu

SnBi

Pb freePb-free Sn

Sn

SnBi

SnBiSnBi

SnCu

SnCu SnCu

Sn

Sn

Sn

Sn

Pb free Pb free

2512 resistors



SnPb Solder Joint Durability- BGA(Sn37Pb) – Sn37Pb Solder – HASL Pad

Finish -

1000.00 10000.001.00

5.00

10.00

50.00

90.00

99.00

Time (t) Hours

Unre

liab

ilit

y,F

(t)

WeibullSnPb- AgedW2 RRX - SRM MED

F=8 / S=0SnPb-Non agedW2 RRX - SRM MED

F=8 / S=0

=8.70, =1970, =0.95

=8.14, =2090, =0.97

Under a -40 to 125oC 1 hr cycle with 15 minute dwells, a decrease in life of5% was observed between aged (350hr/125oC) and non aged assemblies.



Pb-free Solder Joint Durability- BGA (Sn3.0Ag0.5Cu) – Sn3.0Ag0.5Cu Solder – ImSn Pad

Finish

100.00 10000.001000.001.00

5.00

10.00

50.00

90.00

99.00

Time (t) Hours

Un

reli

abil

ity

,F

(t)

WeibullPb-free-AgedW2 RRX - SRM MED

F=7 / S=1Pb-free-Non agedW2 RRX - SRM MED

F=6 / S=2

=1.59, =1630, =0.97

=2.27, =2230, =0.93

Under a -40 to 125oC 1 hr cycle with 15 minute dwells, a decrease in life of25% was observed between aged (350hr/125oC) and non aged assemblies.



Random Vibration Tests

A

C

B NM

D

E F

G H

I J

K L

O P

Q R

T U

V W

X Y

RNET1,2,3,4

RNET5,6,7,8

R1NET5,6,7,8

R1NET1,2,3,4

Step Stress Test Applied0.02 G2/Hz – 6 hrs0.05 G2/Hz – 6 hrs0.1 G2/Hz – 6 hrs0.2 G2/Hz – 18 hrs

Thermally aged (100 hr/125oC and350hr/125oC) conventional tin-leadassemblies (SnPb solder/HASL boardfinish) and lead-free (SAC305 solder/OSPboard finish) were subjected to a randomvibration stress step tests.



Time To Failure Results Under Random VibrationAssembly comparison after 125C/100hrs aging

100

120

140

160

180

200

220

240

260

280

400 600 800 1000 1200 1400 1600 1800 2000

Strain (μe)

Tim

e-t

o-f

ailu

re

SnPb

OSP

Assembly comparison after 125C/350hrs aging

100

120

140

160

180

200

220

240

260

280

400 600 800 1000 1200 1400 1600 1800 2000

Strain (μe)

Tim

e-t

o-f

ailu

re

SnPb

OSP

At low strain levels,below approximately 700e SAC out lasts SnPb.The slow of SAC isshallower than SnPb.Both SAC and SnPbshow a degradation dueto thermal aging. Earlierfailures were ball gridarray parts.



Test Setup: High Speed Flexure and Drop

Instrumented PWA with BGA

Test Specimen 4-Point Bend Fixture



Failure Analysis: High Speed FlexureBulk solder failure

Component

Sn-Pb Solder

Sn-Pb Solder

Board

FR4 board failure

Failure site moves from bulk solderto intermetallic or copper trace asthe PWA flexure rate increases.

Board

Component

Sn-PbSolder

Copper

Intermetallic failure

BoardCopper trace failure

Sn-PbSolder

Sn-PbSolder

Board

Component

Cu-SnIntermetallic

Crack

Sn-Pb Solder

Copper



1E5

1E4

1E3PW

AS

train

(με)

1E0 1E1 1E2 1E3

SAC

Sn-Pb

Strain rate = 1E5 με/sec

Cycles to Failure (N)

Solder failure Cu trace failure

1E5

1E4

1E3PW

AS

train

(με)

1E0 1E1 1E2 1E3

SAC

Sn-Pb

Strain rate = 1E5 με/sec

Cycles to Failure (N)

Solder failure Cu trace failure

Un-aged Aged

Dynamic Durability: Sn37Pb vs SAC305

• PBGA-256 interconnects (1 mm pitch balls on OSP pads)• Dynamic 4-point bend test• Effect of thermal aging (100 hrs at 125o C)

• SnPb outperforms SAC• Failure modes include solder failure and Cu-trace failure• Aging reduces drop durability



Electrochemical Migration Test Facilities

Control unit and data acquisition system

Temperature/humidity chamber



250 μm

Silver Migration on Lead-Free PCBDuring THB Testing

Sn-3.5Ag Solder onPolyimide Substrate withImmersion Sn Plating

EDS mapping revealed thatsilver migrated between thetwo electrodes. Sn-3.5Agwas the only source ofsilver in this sample.



Flex crack

Printed circuitboard Solder pad

Capacitor termination

Solder joint

ElectrodeCeramic dielectric

Illustration of Flex Crack in MultilayerCeramic Capacitor (MLCC)

250 m

Flex crack



Pb-Free Solder and MLCC Flex Cracking

Pb-free SolderMLCC parts weregenerally found tohave as good orbetter resistance toflex cracking ascompared to Sn37Pbsolder.

140 mm

Load span = 80 mm

Support span = 100 mm

85

mm

Strain on boardat 50% failure

(ε)

Strain on boardat 10% failure

(ε)

Strain on boardat 1% failure (ε)

Manuf.DielectricSizeSolder

18,6007,4003,300Vendor BX7R1812Sn3.0Ag0.5Cu

2,8001,9001,500Vendor BX7R1812Sn37Pb

4,8002,3001,700Vendor AX7R1812Sn3.0Ag0.5Cu

2,6001,7001,300Vendor AX7R1812Sn37Pb



http://www.calce.umd.edu/lead-free/tin-whiskers

CALCE Tin Whisker Study

In addition to conducting multiple research projects on lead free solder issues this past year, CALCEjoined with a number of companies to author an alert regarding the use of pure tin as a surface finish.

This alert was followed closely by a mitigation guide authored by CALCE with inputs fromcompanies participating in the Tin Whisker Alert Working Group.



Effect of Reflow Temperatures on WhiskerFormation

Conflicting results have been presented on the impact ofexposure to solder reflow temperatures on whisker formation.



Whisker Length and Growth AngleMeasurement

• Whisker length analysis (Best fit to lognormal distribution)

• Growth angle analysis– Angle growth demonstrates orientation preference. For this case, angle

distributed preferably in the range of 40 ~ 90

– Whisker growth angle appears to be independent of time

0

0.01

0.02

0.03

0.04

0.05

0 10 20 30 40 50 60 70 80 90

Length (µm)P

robab

ilit

y 8-month 13-month

18-month

13.0

50~60

11.4

40~50

6.7

30~40

18.1

60~70

7.9

20~30

15.420.54.32.8Percentage (%)

80~9070~8010~200~10Angle range ()



CALCE Tin Whisker Risk Assessment Software

A software package that calculates the probability of tin whiskerfailure for circuit card assemblies and products. Based on long-termtest data.



BGA(SnAgCu)

(Sn)

(Sn) (Sn)

(SnCu) (SnCu)

(SnCu)

(SnBi)

(SnBi) (SnBi)

QFP

FR4 board

CALCE Long-term Reliability of Mixed SolderStudy

Objectives:To provide participants with a critical assessment of the reliability of solderjoints formed with lead-free parts and Pb-based solder. This is particularly aconcern for companies that are attempting to maintain Pb-based solder.

Temperature Cycle Test Board

• Evaluations

– Intermetallics characterization underisothermal aging at 125ºC for 100,350 and 1000 hours

– Lead pull tests

– Thermal cycling (-40ºC to 125ºC)

– Vibration (step stress test)



Long-term Reliability of Pb-free ElectronicAssemblies in Contaminating Environments Study

Objective: To assess the long-term reliabilityof lead-free electronic assemblies subject tocontaminating and corroding environments.

CALCE Team: Sheng Zhan, Michael H. Azarian,Michael Osterman and Michael Pecht

Approach: Select board material and metalfinishes will be evaluated undertemperature/humidity/bias conditions withtest cells exposed to defined contaminationexposures as per Telecodia and surfaceinsulation resistance (SIR) will bemonitored.

Expected Benefits: Documented corrosionresistance measurements for a range of Pb-free materials and impact of contamination ofcorrosion based time to failure.

Electrochemicalmigration (ECM)failure site

Measured ECM Failure



FY07 CALCE Pb-Free ResearchC07-07 Solder Joint Reliability of Reworked/Repaired SMT AssembliesC07-01 Reliability of Pb-free and Reballed PBGAs in SnPb Assembly

ProcessC07-04 Solder Joint Reliability of Solder Dipped (SAC/SnPb) Leaded SMT

Packages in a SnPb Assembly ProcessC07-03 Effect of Long Dwell on Thermal Cycling Fatigue Damage for Pb-

Free Solders (Continuation of C06-03)C07-06 Effect of Temperature Cycle on the Durability Pb-free Interconnects

(Sn96.5Ag3.0Cu0.5 and SnCuNi) (continuation C06-06)C07-48 Characterization and Reliability Assessment of Lead-Free Solder

Alloys in High Temperature ApplicationsC07-02 Accelerated Qualification of SAC Assembly: Combined

Temperature Cycling & Vibration (Continuation of C06-02)C07-05 Tin Whisker Growth and Risk Assessment UpdateC07-08 Characterization of Tin Pest Formation in Pb-free Solder JointsC07-27 Characterization of PCB Laminate Materials Properties after Lead-

free Reflow CyclesC07-47 Acceleration factors in electrochemical migration



FY07 CALCE EPSC ResearchFailure MechanismsC07-10 Effect of solder fillet and composition factors on MLCC crackingC07-11 Reliability of Large Electrolytic CapacitorsC07-14 Electronic Component Failure Categorization under High G LoadingC07-34 Area Array Components Warpage Sensitivity

New Technologies and ProcessesC07-09 Hygroscopic Characterization of Polymer Materials beyond Glass

Transition TemperatureC07-12 Failure Mechanism & Reliability Assessment for System-in-Package

Technologies (Continuation of C06-12)C07-16 Detection of interconnect degradation using impedance analysisC07-17 Accelerated Testing of Flex AssembliesC07-18 Fundamental Understanding MEMs Structures Subjected to High Shock

LoadsC07-20 Qualification of a Stamped Metal Contact SocketC07-22 Measurement of Effective Chemical Shrinkage of Polymeric MaterialsC07-30 Characterization of Moisture-induced Degradation of Polymer InterfaceC07-31 Solder Assessment for SiC Device AttachmentC07-45 Power Cycling Durability of Advanced Power Modules



FY07 CALCE EPSC Research

Part Management and SustainmentC07-23 Life Cycle Supply Chain Data Mining and InterpretationC07-24 Implementation Cost of Lead-Free and Tin Whisker Mitigation

Performance StandardsC07-25 Life Cycle Cost of Component Management: Understanding the

Component Reuse Business CaseC07-32 PoF Guidelines for Qualifying FPGAsC07-41 Counterfeit Electronic PartsThermal ManagementC07-26 Enhanced Conventional Thermal Solutions for LED PackagesC07-29 Thermal Performance and Reliability of Thermal Interface MaterialsVirtual QualificationC07-15 Shock –calcePWA Shock Model Improvements (continuation of C05-15)C07-21 Modeling Mechanical Torsion of PWA in calcePWAC07-28 Probabilistic PoF Modeling: Effect of Number of I/O on Thermal Cycling

Durability PredictionsC07-46 Model-based Design Guidelines for Shock & Drop Loading (Continuation

of C06-46)

calce and calce pb-free research activities

Documents