Post on 08-Sep-2020
Reliability Evaluation of Under Bump Metallugy (UBM)
in Flip-Chip Interconnects
Yifan GuoSkyworks Solution, Inc.4311 Jamboree Road
Newport Beach, CA 92660
Interconnect Failure Modes in Different Solder Systems
Sn-Pb Solder Sn-Ag Solder
UBM
Solder
INTRODUCTION
Interconnect Failure ModesInterconnect: base metal, UBM, intermetalics and all the interfaces
For a robust and reliable interconnect system, the crack should propagate through the solder, because solder usually has better fatigue resistance than the interconnect.
Crack initiation from flux contamination or local defect such as underfill voids
FM1, FM2
FM3, FM4
FM5
Crack PathsDie
UBM
Intermetalics
Base Metal
Solder
Underfill
Substrate
Passivation
FM1, FM2:- failure in the base metal and UBM
FM3, FM4:- failure in the UBM-IMCand IMC-solder
FM5:- failure in solder
INTRODUCTION
Stress
Prob
abili
ty
UBM StrengthDistribution
UBM StressDistribution
UBM Reliability
Failure Rate
Prob
abili
ty
Stress Free Temperature
150 C - 50 C
UBM StrengthDistribution
Stress
Minimum UBM
Strength
UBM StressDistribution
Process Variations
UBM Reliability
UBM Strength and Stress
UBM SolderDie
Die Pull Thermal Cycle
Force at Break Applied Force fromThermal Mismatch
UBM Strength UBM Stress
Substrate
UBM SolderDie
Substrate
Underfilll
>?
Underfilll
Die Pull
UBM Strength Test
• Tensile stress is applied to the UNB through solder• The stress level in the UBM is the same as in the solder
UBM Solder
Substrate
Die
UBM Solder
Die Pull
Substrate
Die
Minimum UBM Strength
9 7 % Pb / 3 % Sn pr oper t y @RT
6 3 % Sn / 3 7 % Pb pr ope r t y @RT
S t r a i n ( %)
S t r e s s ( MP a )
9 7 % Pb / 3 % S n solder stress level at taffy pull at room temperature
6 3 % Sn / 3 7 % Pb pr ope r t y @1 5 0 C
Eutectic solder stress level at taffy pull at room temperature
0
1 0
2 0
3 0
4 0
5 0
0 1 2 3 4 5 6 7 8 9 1 0
If solder fails when UBM is still intact: UBM Strength > Solder Ultimate Strength— The Minimum UBM Strength is established
Die Pull
UBM Solder
Substrate
Die
Die Pull Equipment & Process
1 2 3 4
• Self centering fixtures
• Cross-head speed of 5.1 mm/minute +/- 0.1
• Failure Modes: Solder failure or UBM failure
100 µm
Fail in Solder Solder Fracture Surface
Failure in Solder Material during Die Pull
( Courtesy of A. De Silva, Motorola Inc. )
( SnCu bumps on electroplated Ni )
30 µm
Failure in Solder Material during Die Pull
• ‘Taffy’ Pull in Solder• Average failure force (Newton) :15.42, SD: 1.36• UBM Fail: 0/168 (6 die x 28 bumps/die)
(Jin-Wook Jang)
Ni UBM
IMC
EutecticSnPb
Failure in UBM during Die Pull
( Courtesy of A. De Silva, Motorola Inc. )
Prob
abili
ty
Stress Free Temperature
150 C - 50 C
UBM StrengthDistribution
Stress
Test 495 bumps with zero failure in UBM (100% failure in solder)--- 99% confidence on 99% reliability
Minimum UBM
Strength
UBM StressDistribution
Process Variations
UBM Reliability
UBM Stress ModelThermal Cycle
-60
-40
-20
0
20
40
60
-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1
-70-5002050100150UBMstress
T em p. ∆T fromStress-free
T em p.
U nderfillC T E
Strain(% )
SolderC T E
Strain(% )
SolderM echanicalStrain(% )
U B MStress(M pa)
-50 -70 -0.49 -0.14 -0.35 -40
20 0 0 0 0 0
100 80 0.56 0.16 0.40 20
thermaltotalmechanical εεε −=
UBM SolderDie
Underfilll Underfilll
Substrate
( Stress-free temperature = 20 C )
Eutectic Sn/Pb Solder System
Stress (Mpa)
Strain (%)
Solder Strain:
• Stress-free temp. = 20 C• Solder CTE = 20 ppm/C• Underfill CTE = 70 ppm/C• Mechanical strain = 50 ppm/C
0
30
-100 -50 0 50 100 150
37Pb/63Sn 97Pb/3Sn
Temperature (oC)
Interconnect Stress (MPa)Strength of 37Pb/63Sn @ RT
--- Min. UMB Strength if 100% solder failure in die pull test
Interconnect Reliability in Flip-Chip Packages
An interconnect system is reliable if 100% solder failure is achieved in die pull test at room temperature.
40
-50
-40
-30
-20
-10
20
10
50
Strength of 97Pb/3Sn @ RT
--- Min. UMB Strength if 100% solder failure in die pull test
UBM Reliability in Lead Free Solder Systems:
SnAgSnCu
SnAgCu
Opening and cover of Teflon-coated moldfor reflow processing of solder
Solder tape from Indium Corporation used for specimens
4mm8mm
28mm
Solder Sample Geometry
Sample Preparation forLead-Free Solder Material Property Tests
Material Properties for Three Lead-Free Solders
SnAgCu
0
10
20
3040
50
60
70
0 1 2 3 4 5 6
Strain [%]
Stre
ngth
[MPa
]
SnAg
0
10
20
30
40
50
60
70
0 1 2 3 4 5 6
Strain [%]
Stre
ngth
[MPa
]
SnCu
0
10
20
30
40
50
60
70
0 1 2 3 4 5 6
Strain [%]
Stre
ngth
[MPa
] -40C -25C 25C 100C 180C
Temperature-dependent stress-strainproperties of Sn-Ag, Sn-Ag-Cu, and Sn-Cu
0
10
20
30
40
50
60
0.0 0.5 1.0 1.5 2.0
(Sn/Ag/Cu)
180 C
25 C
-25 C
-40 C
100 CUBM Stress
Stress (Mpa)
Strain (%)
UBM Stress
0
10
20
30
40
50
60
(Sn/Ag)
180 C
25 C
-25 C
-40 C
100 C
0.0 0.5 1.0 1.5 2.0
Stress (Mpa)
Strain (%)
0
10
20
30
40
50
60
0 0.5 1 1.5 2
180 C
25 C
-25 C
-40 C
100 C
(Sn/Cu)
Stress (Mpa)
Strain (%)
UBM Stress
UBM Stresses in Three Lead-Free Solder Systemsin Thermal Cycles
( Stress free temperature = -50 oC )
-10
0
10
20
30
40
-100 -50 0 50 100 150 200
SnAgSnAgCu SnCu
Temperature (C)
Stress (Mpa)
Ultimate Strength of SnCu
Ultimate Strength of SnAg
Ultimate Strength of SnAgCu
UBM Stresses in Three Lead-Free Solder Systems( Stress free temperature = -50 oC )
-10
0
10
20
30
40
-100 -50 0 50 100 150 200
SnAgSnAgCu SnCu
Temperature (C)
Interconnect Stress (Mpa)
Strength of SnAg @ RT (Mixed failure mode)
Strength of SnAgCu @ RT (Mixed failure mode)
INERCONNECT RELIABILITY— in Three Lead-Free Solder Systems
Strength of SnCu @ RT= Min. UBM Strength
100% solder failure was only achieved in the SnCu systemIn the die pull test at room temperature
FEM Model on UBM StressValidation of the Simplified Stress Modle
Die
Substrate
Underfill
Solder bumpUBM
FEM Validation on the Stress Model
-60
-50
-40
-30
-20
-10
0
10
20
30
-1 -0.5 0.5 1 1.5
FEM Model with gap
FEM Modelwith adhesion
Theoretical Model
UBM Stress (MPa)
Solder Strain (%)
( FEM Using FR-4 Substrate Property )
FEM Validation on the Stress Model( FEM Using Ceramic Substrate Property )
-60
-50
-40
-30
-20
-10
10
20
30
-1 -0.50
0.5 1 1.5
UBM Stress (MPa)
Solder Strain (%)
FEM Model with gap
FEM Modelwith adhesion
Theoretical Model
CONCLUSIONS ( I )
• A methodology was established in evaluating UBMreliability when different solder systems were used
• UBM Strength (minimum strength) was determined from die pull test
• UBM Stress (stress distribution) was determined form stress model
• By comparing the minimum Strength and theStress distribution, UBM reliability was determined
• Solder material has strong impact on the interconnect reliability.
• An interconnect system is reliable if 100% solder failure is achieved in die pull test at room temperature.
• This methodology can be used in reliability evaluations and predictions of different bumping technologies with different solder systems.
CONCLUSIONS ( II )
Acknowledgement
• Betty Yueng• Shun-Meen Kuo• Jang-Kai Lim• Lei Mecardo• Ananda De Silva