Pb-Free BGAs in SnPb Assembly Process Project 08

Pb-Free BGAs in SnPb Assembly Process Project

Presented By:

Project Chair: Robert Kinyanjui, Ph.D. Sanmina-SCI Corporation

Project Co-Chair: Quyen Chu, Jabil

August 2008

1

Project MotivationProject Motivation

• The Effect of the European Union (EU) directive on

Reduction of Hazardous Substances (RoHS) has created

supply chain constraints on SnPb components availability especially for the RoHS exempted sector of the industry

• This has heightened the need for soldering Pb-free SnAgCu BGA solder ball terminations in a Sn-Pb soldering

process without yield loss or reliability impact

2

Project ScopeProject Scope

• For companies choosing to take the RoHS exemption and

continue to manufacture SnPb products beyond July 1,

2006, there will be a growing issue with the lack of availability of SnPb components

• Many companies may be compelled to use Pb-free BGAs in

a SnPb process, for which the process and reliability have

not yet been characterized

3

ObjectivesObjectives

• Phase 1

– Preliminary determination of lower boundary of process window for soldering Lead-free SnAgCu BGA in SnPb solder paste

– Establish process boundary condition for Phase 2 in-depth mixing level characterization and reliability understanding

• Phase 2

– Characterize homogeneity of Lead-free SnAgCu BGAs in a SnPb process with the following considerations:

• Package Size/Ball Volume

• Reflow Temperature

• Time Above Liquidus

• Solder Paste Volume

• Phase 3

– Characterize thermal stress integrity of various mixing level of LF SAC BGA in SnPb solder paste through accelerated temperature cycling (ATC)

4

Phase 1 Preliminary Investigation of Mixing Level

5

Overview of the Phase 1Overview of the Phase 1

• Phase 1 Objectives

– Preliminary determination of lower boundary of process window for soldering LF BGA in SnPb solder paste

– Establish process boundary condition for Phase 2 in-depth mixing level characterization and reliability understanding

• DOE Considerations

– Varying peak temperatures from: 205, 210, 215, 220, and 225C

– Stencil considerations: 6mil thick and with 10% reduction in stencil-to-pad opening

6

Mixed Alloy Solder Joints for a PBGA313 PackageMixed Alloy Solder Joints for a PBGA313 Package

• Observations

– Reflow temperature of 205C was determined to be the lower cliff that would provide marginal solder joint formation

– A peak reflow temperature of 210C and above should be primary focus for next phase development

Reflow @ 225oCReflow @ 205oC

Pb Elemental Map

7

Phase 2Mixed Solder Joint Characterization

8

ObjectiveObjective

• Characterize homogeneity of lead-free BGAs in a SnPb

process with the following considerations:

– Package size/ball volume

– Reflow Temperature

– Time Above Liquidus

– Solder paste volume

9

3Perimeter80.5132A-CTBGA132-0.5mm-8mm

3Array190.8288A-CABGA288 0.8mm-19mm

3Perimeter231.00324A-PBGA324-1.0mm-23mm

3Perimeter451.27600A-SBGA600-1.27mm-45mm

Quantity per

Board

Ball AlignmentSize(mm)

Pitch(mm)

I/OComponent Part Numbers

Note:

• SnPb components

of each type were

used for baseline

run. 1.0mm

PBGA324

0.8mm CABGA

0.5mm CTBGA132

1.27mm

SBGA600

Pb-free BGA ball alloy: SAC405

Components ConsiderationComponents Consideration

10

Populated Test VehiclePopulated Test Vehicle

• PCB Dimensions:

–6.800” x 4.075” x 0.093”

• Finish

–Electroless Nickel

–Immersion Gold (ENiG)

–Copper OSP

• Number of Layers

–8 Internal board Layers

• Thermal Properties

–Tg = 170C

–Td = 340C

Test vehicle populated with three of each of

the four types of components used

SBGA

600

PBGA324

CABGA288

CTBGA132

11

Assembly ConsiderationAssembly Consideration

• Assembly Setup

– Two thermocouples were placed at the center and corner joints of each component type on a sample test board

– Five profiles were generated

• Peak Temp: 210C, 215, and 235C

• Time Above Liquidus (TAL): 60, 90, 120sec

• Stencil

– 6 mil foil thick

– Two aperture openings (1:1 and 10% Reduction)

• For the 0.5mm pitch component, the same aperture opening was

used for better paste release

12

Summary of Assembly MatrixSummary of Assembly Matrix

13

Representative Reflow ProfileRepresentative Reflow Profile

Reflow Profile Conditions: Peak Temp. = 120oC; TAL = 60sec

14

Solder Joint Inspection after AssemblySolder Joint Inspection after Assembly

• A representative solder joint formation for the SBGA600 component reflowed at peak temperature of 210oC with 120 seconds above liquidus (Profile # 3)

• NOTE: At 120 TAL, the degree of mixing for the SBGA600 package between the SAC solder ball alloy and the SnPb solder paste appears to be partially mixed.

Non-uniformSolder jointmicrostructureMixed

Non

Mixed

15

Initial PCBA AnalysisInitial PCBA Analysis

• Analysis of as-reflowed BGA solder joints integrity through dye & pry and metallography

– Determine the solder joint structure

– Examine the degree of mixing between SnPb and SAC solder alloys

– Characterize the solder joint height/diameter

Dye & Pry

Metallography

Direction in reflow Oven

16

Representative Dye & Pry ResultsRepresentative Dye & Pry Results

1.0mm PBGA324

0.8mm CABGA288

PCB Side

Package Side

• No dye intrusion was observed no signs of open joints

17

Effect of Package Size (Pitch)Effect of Package Size (Pitch)

1.0mm

PBGA324

1.27mm

SBGA600

0.5mm

CTBGA132

0.8mm

CABGA288

Peak Temp: 210C and TAL: 60sec

The extent of alloy mixing is observed

to increase with increasing ball-to-

paste ratio for a peak temperature of

210oC and for all the TALs tested.

It was noted that approximately 40%

solder ball-to-paste volume ratio is

needed in order to achieve 100%

alloy mixing.

18

The degree of mixing increases linearly

with the time above liquidus of 60, 90,

and 120secs for the 1.27,1.00 and

0.8mm pitch packages.

The Effect of Time Above The Effect of Time Above LiquidousLiquidous

1.0mm PBGA324 1.27mm SBGA600

TAL:120secTAL: 60sec TAL:120secTAL: 60sec

0%

20%

40%

60%

80%

100%

40 60 80 100 120 140

Time Above Liquidus (sec)

Pe

rce

nt

Mix

ed

(%

)

1.2 mm

1.0 mm

0.8 mm

Linear (0.8 mm)

Linear (1.0 mm)

Linear (1.2 mm)

19

Little or no discernable difference was

observed in the degree of mixing for the

largest package (SBGA600) with pitch of

1.27mm.

However for the 1.0mm and 0.8mm pitch

BGAs, it was generally observed that

increasing the paste volume from a 90%

to 100% aperture opening increased the

degree of mixing.

90% Paste Volume 100% Paste Volume


90% Paste Volume 100% Paste Volume

0%

20%

40%

60%

80%

100%

120%

60 90 60 90 60 90

1.2 1.0 0.8

Package Pitch (mm), TAL (sec)

Pe

rce

nt

Mix

ed

(%

)

Aperture Opening 90%

Aperture Opening 100%

Effect of Paste VolumeEffect of Paste Volume

20

Tpeak = 215C,

TAL = 60sec

Increasing the peak temperature by 5oC

with TAL = 60 sec, leads to an increased

degree of mixing for both PBGA324 and

SBGA600.

The extent of mixing increases from 50% to

100% for the PBGA324 package


Tpeak = 210C,

TAL = 60sec

0%

20%

40%

60%

80%

100%

120%

1.2 1.0

Package Pitch (mm)

Perc

en

t M

ixed

(%

)

210C, 90%apt

210C, 100%Apt

215C, 90%Apt

Tpeak = 215C,

TAL = 60sec

Tpeak = 210C,

TAL = 60sec

Effect of Peak TemperatureEffect of Peak Temperature

21

Summary of ObservationsSummary of Observations

• The larger the package/ball volume, the lower the degree of mixing

observed.

– CTBGA132 (0.5mm Pitch) showed a 100% mixing at 210C and 60sec time

above liquidus

• Doubling the time above liquidus (from 60sec to 120sec), increased the

degree of mixing (~30% increase in mixing was observed).

• For the small packages (reduced SAC solder ball volume), increased

paste volume corresponds to increased degree of mixing.

– However, for the largest (SBGA600) package no significant change in mixing

was observed.

• Increasing the peak temperature from 210C to 215C led to a significant

increase in the degree of mixing (almost doubled) the extent of mixing.

22

Phase 3Mixed Solder Joint Reliability

23

ObjectiveObjective

• Characterize thermal stress integrity of various mixing

level of LF BGA in SnPb solder paste

– Accelerated thermal cycling (ATC) of -40C to 125C

24

459

309060215Sn-PbSn-Ag-Cu8

609060235Sn-Ag-CuSn-Ag-Cu7(Control)






639060210Sn-PbSn-Pb1(Control)

OSPENIG

Number of Boards for ATC Testing (Range: -40oC to

125°C)

Stencil Aperture Opening

(%)

Time Above Liquidus (TAL)

(seconds)

Peak Temp. (°C)Solder Paste Alloy

BGA Ball AlloyAssembly Process Flow

Assembly Test Matrix for ATC Test Assembly Test Matrix for ATC Test

25

Mixed/OSP/90% FixedTAL varies, 60, 90, 120

Characteristic life:

η60 =1055η90 =1267η120 =1331

• Observation:

– The longer the TAL, the better the solder joint reliability • Solder joint reliability is highest for TAL =120sec > 90 sec > 60 sec

Impact of Time Above Impact of Time Above LiquidusLiquidus on CTBGA132 SJRon CTBGA132 SJR

26

CTBGA132

The number of cycles to 63% solder joint failure for the CTBGA132 packages

increased with increasing TAL

Impact of TAL on Characteristic Life Impact of TAL on Characteristic Life -- CTBGA132 PackageCTBGA132 Package

27

Number of 1% Failures for Mixed CTBGA132 PackageNumber of 1% Failures for Mixed CTBGA132 Package

TAL = 60secs

TAL =120secs

TAL = 90secs

Reflow profile: 210oC, OSP, 90% solder paste

Cycle

s t

o F

ailu

re

• Observations:

– The 60 seconds TAL exhibits the least number of cycles at 1% cumulative failure.

– The number of cycles to failure increases by about twice when TAL is increased from 60 to 120 seconds

CTBGA132

28

Solder Joint Observed Failure ModesSolder Joint Observed Failure Modes

Optical & SEM images showing the

most commonly observed failure mode

for the CTBGA132 package

IMC Layer

The solder joints are completely open on the package side while the

crack is only partial on the board side

Representative Solder Joints for the CTBGA132 PackageRepresentative Solder Joints for the CTBGA132 Package

29

Mixed/OSP/90% Fixed

TAL varies, 60, 90, 120

Characteristic life:

η60 =473

η90 =446

η120 =619

η60 = 1007(SnPb)

η60 = 1281 (SAC)

Impact of Time Above Impact of Time Above LiquidusLiquidus on CABGA288 SJRon CABGA288 SJR

The solder joint reliability of the mixed joints is lower than that of “pure”

SnPb or SAC

30

Time Above Liquidus (secs)

Mixed Solder Joints

SnPb

Chara

cte

ristic L

ife

(Cycle

s t

o f

ailu

re @

63%

)

LF

PCB Finish: OSP

and Peak temp.

= 210C

The “pure” solder joints outperforms mixed solder joints

Impact of Time Above Impact of Time Above LiquidusLiquidus on CABGA288 SJRon CABGA288 SJR

31

Number of 1% Failures for the CABGA288 PackageNumber of 1% Failures for the CABGA288 Package

• NOTABLE OBSERVATIONS:

– At 1% failure, “pure” SnPb and “pure” SAC joints are better than mixed SnPb/SAC solder joints• “Pure” SnPb outperforms mixed joints by ~ 5 times

• “Pure” SAC outperforms mixed joints by ~ 3 times

Mixed

Pure

60 90 120 60 60

Cycle

s to F

ailu

re

SAC

SnPb

Test Condition: 210°C profile, OSP,

10% paste volume reduction

CABGA288Time Above Liquidus (secs)

32

Optical images showing the most commonly observed failure mode for

the CABGA288 package

The solder joints are completely open on the package side while the crack is only

partial on the board side

Representative Solder Joints for the CABGA288 PackageRepresentative Solder Joints for the CABGA288 Package


33

These fractures could not be confirmed by electrical testing during thermal cycling.

The most commonly observed joint fractures on the PBGA324 and SBGA600

packages

Representative Solder Joints for the PBGA324 and SBGA600 packages


34

018/189/99/918/189/918/1818/1818/189/918/18CTBGA132

018/1818/189/918/189/918/1818/1818/189/918/18CABGA288

0/1817/180/180/90/180/90/180/180/189/90/18PBGA324

0/184/180/180/90/180/90/180/180/189/91/18SBGA600

8765B5A4B4A321B1AAssembly Process Flow

OSPOSPOSPENIGOSPENIGOSPOSPOSPENIGOSPPCB Finish

Number of Failed Packages After 3559 ATC testing Cycles

• NOTABLE OBSERVATIONS:

– All the standard and mixed alloy CTBGA132 and CABGA288 packages failed at the end of the ATC testing

– All mixed alloy, SBGA600 and PBGA324 packages survived at the end of ATC testing

A summary of Number of Failed Packages at the end of 3559 Cycles

Decre

asin

g P

ackag

e S

ize

Electrical Testing Results at the End of ATC TestElectrical Testing Results at the End of ATC Test

35


• The CTBGA132 and the CABGA288 packages showed a significant number of failures across all of the assembly test conditions, while no failures were recorded for the PBGA324 and SBGA600 packages for the mixed alloy assemblies.

• The following conclusions are drawn from this work:

– For the smallest, CTBGA132 package, all three TAL conditions hadfull solder alloys mixing.

• The solder joint reliability of the fully mixed test assemblies for all TALs(= 60, 90, and 120 seconds) exceeds that of Sn-Pb and is at least equal to that of pure SAC.

– For the larger, CABGA288 package, increasing the TAL does not provide complete solder alloy mixing.

• The solder joint reliability of all TALs tested (TAL = 60, 90, and 120 seconds) is less than that of both pure Sn-Pb and pure SAC.

• The longest TAL condition displayed a better solder joint reliability among the three TAL conditions.

36


• For the two largest packages, PBGA324 and SBGA600, despite lack of complete solder alloy mixing, the solder joint reliability was better than that of either “pure” Sn-Pb or “pure” SAC solder joints.

• Full Sn-Pb and SAC solder alloys mixing is not a sufficient condition toguarantee good reliability.

– For small packages with low fatigue life requirements, full solder alloys mixing and homogeneous microstructure is required

• while for large packages with long fatigue/extended life requirement, full solder alloy mixing is not necessary for acceptable solder joint reliability.

• In general, the OSP-copper had better performance than the ENIG surface finish. However, the failure locations were almost exclusively at the package side of the solder joint and within the bulk solder.

– At this time no microstructural correlation has been identified linking surface finish and improved or reduced reliability.

37

Thank You

Questions?

38

Acknowledgement Acknowledgement

Documents

Pb-Free BGAs in SnPb Assembly Process Project 08