Page 1
TM
Freescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006.
Sn3.5Ag PbSn3.5Ag Pb--Free Alloy for BGA ApplicationFree Alloy for BGA Application
Customer PresentationCustomer Presentation
Min DingPSD/TSOJuly 20, 2009 Updated
Page 2
TMFreescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006. 1
Outline
► Background
► Component Level Characterization
► Board-level Solder Joint Reliability
► Summary
Page 3
TMFreescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006. 2
Introduction of Sn3.5AgWhat is Sn3.5Ag?
• Composition: 96.5wt%Sn3.5wt%Ag• Alternative Pb-free solder ball alloy intended for BGA applications. • No impact to customer board assembly process due to similar liquidus
melting temperature range as current solder ball alloys (Sn4.0Ag0.5Cu)
Market Trend on Sn3.5Ag• >10 years of research/application history • Sn3.5Ag is readily available from major solder alloy suppliers• Major package assembly houses are qualified with Sn3.5Ag for BGA
production
Page 4
TMFreescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006. 3
Component Level Characterization
► Evaluation Matrix and Test Vehicles
► Manufacturability
• Geometrical profile
► Mechanical Strength Test
• Ball Shear Results
• Cold Ball pull (CBP) test
Introduction
CBP test results
• Packing Drop Test
► Solderability
Page 5
TMFreescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006. 4
Evaluation Matrix• Solder Alloy: Sn3.5Ag and Sn4.0Ag0.5Cu
• Melting Point
• Sn3.5Ag: 221oC
• Sn4.0Ag0.5Cu: 217~221oC
• Ball Attach Reflow Profile
• Ramp-to-Peak profile
• 240oC peak temperature
• >230oC 20sec dwell
Page 6
TMFreescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006. 5
Test Vehicles
Supplier 30.60.529x291.0Thermally Enhanced PBGA
Supplier 1
Electrolytic Ni/Au
0.760.637.5x37.51.27TBGA
Supplier 10.660.4537.5x37.51.0TBGA
0.6
Ball Size(mm)
0.5
Pad Opening(mm)
Supplier 227x271.0PBGA
Substrate supplier
Pad FinishBody Size(mmxmm)
Pad Pitch(mm)Package Type
Page 7
TMFreescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006. 6
Manufacturability— Statistical Process Control of Ball Geometrical Profile
02468
1012
Coplanarity Ball Height Ball Diameter Coplanarity Ball Height Ball Diameter
After Assembly After Test
Cpk
SAC405
SnAg
1.67
• Sn3.5Ag and SAC405 have good process stability in ball geometry with Cpk > 1.67
• Statistical analysis shows insignificant difference between Sn3.5Ag and SAC405 in terms of co-planarity, ball height and ball diameter for all the package types tested
Page 8
TMFreescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006. 7
New Mechanical Strength Test— Ball Shear and Cold Ball Pull
Ball Shear Cold Ball Pull
• Good for testing solder ball bulk properties• Not sensitive enough for detecting
intermetallic bond strength & brittle failure• Exert mixed force (shear & tensile)• Side support from solder mask• Speed too low (<700um/s)
• Good for testing intermetallic bond strength & brittle failure
• Mainly tensile stress
• 10x faster loading speed than shear
Shear Blade Pull Jaw
Page 9
TMFreescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006. 8
Mechanical Strength Test— Ball Shear Test
Failure Modes in Ball Shear Test
• Test speed: 300um/sec
• Sample size: 40 balls/read point
Page 10
TMFreescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006. 9
Mechanical Strength Test— Ball Shear Strength
• Both Sn3.5Ag and SAC405 showed good ball shear strength with Cpk>1.67
• Sn3.5Ag has lower shear strength reflecting the lower yield strength and hardness. This could help to reduce the stress inside the solder joint during mechanical impact.
1.90
92.2
1525
1343
1856
Sn3.5Ag
1.27mm TBGA
2.41
87.4
1631
1398
1816
SAC405
2.46
64.9
1067
913
1164
Sn3.5Ag
1.0mm TBGA
2.24
86.7
1334
954.3
1286
SAC405
75.762.571.158.8Std.Dev. (g)
2.452.172.822.40Cpk
1256110713001121Mean (g)1078100612001018Min (g)1364126014511269Max (g)
SAC405Sn3.5AgSAC405Sn3.5Ag
1.0mm TEPBGA1.0mm PBGA
Page 11
TMFreescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006. 10
Mechanical Strength Test— Ball Shear Results
• Sn3.5Ag and SAC405 samples showed 100% ductile failure (Mode 1) in ball shear
• No significant difference in failure mode found between Sn3.5Ag and SAC405
1.27mm TBGA 1.0mm TBGA 1.0mm PBGA 1.0mm TEPBGA
0%0%0%
0%
1.27mm TBGA 1.0mm TBGA 1.0mm PBGA 1.0mm TEPBGA1.27mm TBGA 1.0mm TBGA 1.0mm PBGA 1.0mm TEPBGA
0%0%0%
0%
Percentage
SAC405 SAC405 SAC405 SAC405
1.27mm TBGA 1.0mm TBGA 1.0mm PBGA 1.0mm TEPBGA
0%0%0%
0%
1.27mm TBGA 1.0mm TBGA 1.0mm PBGA 1.0mm TEPBGA1.27mm TBGA 1.0mm TBGA 1.0mm PBGA 1.0mm TEPBGA
0%0%0%
0%
Percentage
SAC405 SAC405 SAC405 SAC405
Page 12
TMFreescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006. 11
New Mechanical Strength Test— Introduction to Cold Ball Pull Test
Failure Modes in CBP• CBP is a destructive test. Different failure modes are defined in Table 1
• Mode 2 (brittle failure) is the least desirable failure mode which can lead to missing ball incident
• Units with lower Mode 2 failure rate in CBP show lower missing ball yield loss in assembly and after shipping
• CBP is a highly accelerated test, the failure rate in CBP is much higher (>100x) than that in the field.
Table 1 Failure Modes in CBP
Page 13
TMFreescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006. 12
Stress Conditions
Reflow profile (Tpeak=260oC)1x, 2x, 3x, 6xMultiple Reflow
Baking at 150oC24, 48, 96,168, 504 and 1008 hours
High Temperature Storage
As-assembledT0
DescriptionStress Condition
Sample Size: For each reading point, totally 40 balls were pulled at 5mm/s. That is 8 balls per package and 5 packages per cell.
Page 14
TMFreescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006. 13
Mechanical Strength Test— Cold Ball Pull Test Results (TBGA)
• Sn3.5Ag reduces brittle failure rate significantly. No brittle failure is observed in CBP test with Sn3.5Ag spheres after assembly and test.
CBP Brittle Failure Rate - After Assembly & Test
78% 72%
0% 0%
90%96%
0% 0%0%
20%
40%
60%
80%
100%
After Assembly After Test After Assembly After Test
SAC405 SnAg
Brit
tle F
ailu
re R
ate
(Mod
e 2)
(%)
1.27mm 37.5x37.5 TBGA1.0mm 37.5x37.5 TBGA
CBP Brittle Failure Rate - Multiple Reflow
100% 100%
0% 0% 0% 0%0% 0% 0% 0%
100%100% 100%100%93% 93%
0%
20%
40%
60%
80%
100%
1x 2x 3x 6x 1x 2x 3x 6x
SAC405 SnAg
Brit
tle F
ailu
re R
ate
(Mod
e 2)
(%) 1.27mm 37.5x37.5 TBGA
1.0mm 37.5x37.5 TBGA
CBP Brittle Failure - High Temperature Storage (hours)
0% 0% 0% 0% 0%0% 0% 0% 0%10%
90% 85% 93% 88%
54%47%
100% 100% 100% 100% 100% 100%
63% 68%
0%
20%
40%
60%
80%
100%
24 48 96 168 504 1008 24 48 96 168 504 1008
SAC405 SnAg
Brit
tle F
ailu
re R
ate
(Mod
e 2)
(%
)
1.27mm 37.5x37.5 TBGA
1.0mm 37.5x37.5 TBGA
Page 15
TMFreescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006. 14
• PBGA with SAC405 shows <10% brittle failure in CBP test. This is consistent with subcon’s data
• After stressing at high temperature storage, brittle failures increase due to IMC growth
• Sn3.5Ag continues to show no brittle failure until after 504 hours HTS, and the rate is still much lower than SAC405
Mechanical Strength Test— Cold Ball Pull Test Results (PBGA)
CBP Brittle Failure Rate - After Assembly & Test
8%
0% 0% 0%1%
0% 0% 0%0%
5%
10%
After Assembly After Test After Assembly After Test
SAC405 SnAg
Brit
tle F
ailu
re R
ate
(Mod
e 2)
(%)
1.0mm 27x27 PBGA1.0mm 29x29 TEPBGA
CBP Brittle Failure - High Temperature Storage (hours)
0% 0% 0% 0% 0%13%
0% 0%
38%50%
0%0% 8%18%
8%18% 10%
75%
0%
28%
0% 0% 0% 0%0%
20%
40%
60%
80%
24 48 96 168 504 1008 24 48 96 168 504 1008
SAC405 SnAg
Brit
tle F
ailu
re R
ate
(Mod
e2)
(%) 1.0mm 27x27 PBGA
1.0mm 29x29 TEPBGA
Page 16
TMFreescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006. 15
Mechanical Strength Test— Cold Ball Pull Strength
• Sn3.5Ag shows significantly higher CBP strength than SAC405. That suggests higher mechanical robustness.
58
1872
1743
1994
Sn3.5Ag
1.0mm TBGA
75
1772
1564
1941
SAC405
1839817879Std.Dev. (g)2027210319382067Mean (g)1569187515421925Min (g)2319229122992248Max (g)
SAC405Sn3.5AgSAC405Sn3.5Ag
1.0mm TEPBGA1.0mm PBGA
Page 17
TMFreescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006. 16
Mechanical Strength Test— Packing Drop Test
• Purpose - Simulate the shock/knock during shipping
• Sample size - 60 good units / box (lot)• 2 bundles per box• 5 full trays + 1 empty cover tray (5+1)/bundle• 6 good units per tray, 4 at corners and 2 at center, dummy units for the other
pockets• Drop – 3 feet to hard surface from 10 different angles• Freescale’s current practice in production
• No missing ball should be found after 1 cycle of drops (drop from 10 angles)
2 bundles into one box
Drop from 10 different angles
Page 18
TMFreescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006. 17
Mechanical Strength Test— Tray Drop Test
• Purpose - Simulate the shock/knock in handling and shipping• Test configuration – 6 good units per lot
• 10 trays banded together• 6 good units in the 10th (bottom) tray and dummy units fill the rest of the pockets• Drop the bundles 10cm to the surface (hard material) in 5 angles shown by the red arrows• Not a production control, for engineering purpose only
Page 19
TMFreescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006. 18
Mechanical Strength Test— Packing Drop Test Results
• “Test to failure” approach - parts are dropped multiple cycles until missing ball is seen
• Test vehicle – 1.0mm, 37.5x37.5mm TBGA
1717Sn3.5AgTray Drop
22SAC405
Packing DropNo ball missing20Sn3.5Ag
33SAC405
No. of Cycle when 1st Ball
drop observed
No. of Cycles Tested
• Both Sn3.5Ag and SAC405 can meet Freescale’s current requirement – No missing ball after 1 cycle of test
• Sn3.5Ag showed considerable improvement over SAC405 in packing drop test
Page 20
TMFreescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006. 19
Solderability— Ceramic Plate Test
• No solderability failure in both alloys• Steamage: 8hrs. +/-15min + 24 hrs of baking at 125 deg.C• Reflow peak temp: 240C
1.0mm PBGA
1.0mm TEPBGA
1.27mm TBGA
1.0mm TBGA
SAC405
SAC405
SAC405
SAC405
Page 21
TMFreescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006. 20
Board Level Solder Joint Reliability
►Air Thermal Cycling (ATC) Evaluation• Test Matrix• Board Mounting Process• Results
Page 22
TMFreescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006. 21
Air Temperature Cycling— Test Matrix
Sn4.0Ag0.5Cu
Control New Alloy
Sn3.5AgSnPb2AgSolder Alloy
Air Temperature Cycling
Condition
-40~125oC
-40~125oC
27x270.50.61TEPBGA
27x270.50.61PBGA
Package size (mmxmm)
Pad Opening
(mm)
Ball Dia(mm)
Pad Pitch(mm)
PackageType
Page 23
TMFreescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006. 22
Air Temperature Cycling— Board Mounting Process
► Solder Paste• SnPbAg – SnPb Paste• SAC405 and SnAg – SAC387 Paste
► Reflow Profile• SnPbAg – SnPb board mounting profile • SAC405 and SnAg – Pb-free profile (peak temperature >230C)
Page 24
TMFreescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006. 23
Air Temperature Cycling (-40~125oC)– 1.0mm Pitch 388 PBGA (2L Substrate)
Cyc
les
to F
irst
Fai
l
• Pad pitch 1.0mm; Ball size 0.6mm; body size 27x27mm• All Pb-free alloys have similar mean-cycles-to-failure and
significantly larger than SnPbAg• Both SAC387 and SnAg showed good repeatability
Cyc
les
to 5
0% F
ailu
re
SnAgSAC387
SnPbAg
S10
1000
2000
30004000
5000
6000
7000
8000
9000
SnAgSAC387
SnPbAg
S10
1000
2000
3000
4000
5000
6000
7000
8000
100001000
99
90
8070605040
30
20
10
5
3
2
1
Cycles-to-Failure
Cum
ulat
ive
Failu
re (
%)
19.2554 5079.53 0.974 30 09.9916 8653.47 0.991 29 1
16.8491 8154.51 0.970 22 811.6398 8923.95 0.959 25 58.6615 9580.49 0.995 25 5
13.0520 8565.14 0.958 28 2
Shape Scale C orr F CTable of Statistics
SnPbAgSAC387SAC387_RptSnAgSnAg_RptSAC105
BGA Alloy
388 PBGA 1.0mm Pitch, 27x27mm Body Size - -40~125CWeibull
Page 25
TMFreescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006. 24
Liquid-to-Liquid Thermal Shock (-55~125oC)– 1.0mm Pitch 388 PBGA (2L Substrate)
• Pad pitch 1.0mm; Ball size 0.6mm; body size 27x27mm
100001000100
99
90
8070605040
30
20
10
5
3
2
1
Cycles-to-Failure
Cum
ulat
ive
Failu
re (
%)
4.30464 3518.53 0.907 28 03.82345 5958.10 0.980 23 72.26991 4018.90 0.980 29 15.57965 5852.72 0.994 23 72.18499 6245.22 0.972 22 8
Shape Scale C orr F CTable of Statistics
SnPbAgSAC387SAC387_RptSnAgSnAg_Rpt
BGA Alloy
PBGA 1.0mm Pitch, 27x27mm Body SizeLiquid-to-Liquid Thermal Shock -55~125C
Page 26
TMFreescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006. 25
Air Temperature Cycling (-40~125oC)– 1.0mm Pitch 516 TEPBGA
Cyc
les
to F
irst
Fai
l
• Pad pitch 1.0mm; Ball size 0.6mm; body size 27x27mm• Compared with SAC387, Sn3.5Ag similar 1st failure cycles, larger mean-cycles-to-failure• Both SAC387 and SnAg showed good repeatability
SnAgSAC387
SnPbAg
0
500
1000
1500
2000
2500
3000
3500
4000
SnAgSAC387
SnPbAg
0
1000
2000
3000
4000
5000
6000
7000
Cyc
les
to 5
0% F
ailu
re
100001000
99
90807060504030
20
10
5
3
2
1
Cycles-to-Failure
Cum
ulat
ive
Failu
re (
%)
15.8437 3302.58 0.945 31 111.3770 5034.43 0.963 32 015.6748 4563.81 0.972 32 08.0967 6238.57 0.990 32 09.2030 6599.86 0.979 31 1
Shape Scale C orr F CTable of Statistics
SnPbAgSAC387SAC387_RptSnAgSnAg_Rpt
BGA Alloy
TEPBGA 1.0mm Pitch, 27x27mm Body Size - -40~125CWeibull
Page 27
TMFreescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006. 26
• Sn3.5Ag showed similar life of cycles as SAC387
Cycles-to-Failure
Perc
ent
100000100001000100
99
90
807060504030
20
10
5
3
2
1
Table of Statistics
14 26.09270 4554.73 0.974 16 06.93027 5330.58 0.978
Shape
16 0
Scale Corr F C6.83572 2351.75 0.938
Variable
SnAg
SnPbAgSAC387
TEPBGA 1.0mm Pitch, 27x27mm Body Size - Varied AlloysLiquid-to-Liquid Thermal Shock (-55~125C)
Liquid-to-Liquid Thermal Shock (-55~125oC)– 1.0mm Pitch 516 TEPBGA
Page 28
TMFreescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006. 27
Mechanical Bend Test– IPC-9702 Monotonic Bend Test
Page 29
TMFreescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006. 28
Mechanical Bend Test– Example Bend Test Board Top
Page 30
TMFreescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006. 29
Mechanical Bend Test– 4 Point Bend Fixture
Page 31
TMFreescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006. 30
Mechanical Bend Test– Example Strain and Resistance Versus Time
Page 32
TMFreescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006. 31
Mechanical Bend Test– TBGA Weibull Plot for Break Strain
• Sn3.5Ag showed similar break strain as SAC405
Break Strain (microstrain)
Cum
ulat
ive
Failu
re (
%)
100001000
99
9080706050403020
10
5
32
1
17.8148 5850.06 0.920 3 011.7112 5904.42 0.969 5 022.1653 5439.52 0.897 5 0
Shape Scale Corr F CTable of Statistics
SnPbAgSAC387SnAg
Variable
1.27mm Pitch, 37.5x37.5mm TBGA Varied AlloyIPC - 9702 Monotonic Bend Test
Break Strain (microstrain)Cu
mul
ativ
e Fa
ilure
(%
)
100001000
99
90
8070605040
30
20
10
5
3
2
1
6.3882 5324.69 0.988 12 210.1465 4339.95 0.990 15 011.8291 4624.99 0.878 15 3
Shape Scale Corr F CTable of Statistics
SnPbAgSAC387SnAg
Variable
1.0mm Pitch, 37.5x37.5mm TBGA Varied Solder Ball AlloyIPC - 9702 Monotonic Bend Test
SAC405 SAC405
Page 33
TMFreescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006. 32
Board Assembly of Sn3.5Ag
Studied the performance of Sn3.5Ag and SAC405 at board level
• Manufacturability – robustness to reflow profile and solder paste
• Cu pad consumption
• Solder joint reliability with SnPb paste
Page 34
TMFreescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006. 33
Pure Pb-free PCB Board Mounting• Pb-free spheres: Sn3.5Ag and SAC405• Pb-free solder paste: SAC387• Pb-free reflow profile: 230~250oC peak temperature• Test vehicle: TBGA, TEPBGA• Response
• Solder sphere formation after board mounting
• Cu consumption after• 3x reflow at 240C• High temperature storage at 150oC up to 1008 hours
230,240,250
230, 240, 250
230, 240, 250
Reflow Peak Temperature
(oC)
SAC387
SAC387
SAC387
Solder Paste
37.5x37.51.27TBGA
37.5x37.51.0TBGA
27x271.0TEPBGA
Body Size(mmxmm)
Pad Pitch(mm)
Package Type
Page 35
TMFreescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006. 34
Pure Pb-free Board Mounting Process
Sn3.5Ag SAC405
230oC Peak Temperature, SAC387 solder paste, 1.0mm TEPBGA
• Sample size – 3 units per alloy per temperature• Both Sn3.5Ag and SAC405 collapsed and formed solder joint properly• No open failure was observed for both alloys in all the reflow profiles studied
Page 36
TMFreescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006. 35
Manufacturability— Board Mounting Reflow Process Evaluation at Customer
• Test Vehicle: 1.27mm TBGA with Sn3.5Ag
• Reflow profile: Pb-free board mounting with 225, 230 and 240C peak temperature
• Paste: SAC 305
Page 37
TMFreescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006. 36
Pure Pb-free Board Mounting Process— Cu consumption
Cu consumption = C - B
Page 38
TMFreescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006. 37
Pure Pb-free Board Mounting Process— Cu consumption (1.0mm Pitch TEPBGA)
• At T0 (after 1x board mounting reflow at 240C), Sn3.5Ag consumes about 3um more Cu than SAC405
• After 3x reflow at 240C reflow, Sn3.5Ag consumes 3um more Cu than SAC405
• Under high temperature storage, the Cu consumption increases with storage time for both alloys. The differences between SAC405 and Sn3.5Ag remains about the same
• The difference in Cu consumption is mainly due to the extra Cu dissolution into the Sn3.5Ag during reflow
0
2
4
6
8
10
T0 3X Reflow (240C)
Cu
Con
sum
ptio
n (u
m)
SAC405SnAg
0
2
4
6
8
10
12
T0 336hr 700hr 1008hr
Cu
Con
sum
ptio
n (u
m)
SAC405SnAg
Sample size: 14 measurements for each data point from 2 units
Page 39
TMFreescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006. 38
Backward Compatible PCB Board Mounting
•Pb-free spheres: Sn3.5Ag and SAC405•SnPb solder paste•Pb-free reflow profile: 221 and 225oC peak temperature•Test vehicle: PBGA, TEPBGA•Response
• Solder sphere formation after board mounting• Solder joint reliability
221, 225221, 225
Reflow Peak Temperature
(oC)
SnPbSnPb
Solder Paste
27x271.0PBGA27x271.0TEPBGA
Body Size(mmxmm)
Pad Pitch(mm)
Package Type
Page 40
TMFreescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006. 39
Backward Compatibility
0
20
40
60
80
100
Yiel
d (%
)
225C 221C
Sn3.5AgSAC405
0
20
40
60
80
100
Yiel
d (%
)
225C 221C
Sn3.5AgSAC405
1.0mm PBGA, SnPb Paste 1.0mm TEPBGA, SnPb Paste
• Yield loss was due to open solder joints after board mount
• SAC405 showed significant amount of open failure at 221C while Sn3.5Ag showed none
• SAC405 showed solder sphere to solder paste separation at 221C
Sample size: 13~15 units per temperature per alloy
Page 41
TMFreescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006. 40
Backward Compatible PCB Board Mounting— Solder Joint Reliability
Temperature Cycle-40~125oC
768
SAC405
SnAg
SAC405
SnAg
SAC405
SnAg
SAC405
SnAg
SAC405
SnAg
Alloy
14351x 221oC
TEPBGA
>20001x 225oC
3x 240oC27x27
Package size
(mmxmm)
1x 225oC >2000
1x 221oC
1.0
PBGA
1st Faiilure Cycles
Assembly Reflow Profile
Pad Pitch(mm)
PackageType
Sample size: 15 units per temperature per alloy
Page 42
TMFreescale Semiconductor Confidential and Proprietary Information. Freescale™ and the Freescale logo are trademarksof Freescale Semiconductor, Inc. All other product or service names are the property of their respective owners. © Freescale Semiconductor, Inc. 2006. 41
Summary
► Sn3.5Ag shows equally good manufacturability and board level solder joint reliability as SAC405
► No impact on FSL and customer manufacturing processes that have been set up for SAC405
► Sn3.5Ag has significantly better mechanical robustness at component level than SAC405 that could reduce the yield loss in package assembly and test significantly