Evaluation of risks due to thermal stress before physical failure appearance Michael Hertl – Jean-Claude Lecomte INSIDIX – 24 rue du Drac – 38180 SEYSSINS France Tél. : +33 (0)4 38 12 42 80 – Fax : +33 (0)4 38 12 03 22 E-mail :
[email protected]
EUFANET workshop 2006 – Wuppertal
1
Challenges
Tendencies ¾ Assemblies getting smaller and smaller, more and more integrated ¾ RoHS consequence : Solder temperature +34°C ¾ Customers: Increase junction temperature: 150 -> 175-200°C
Dilemma ¾ Increase of thermal stress, but decrease of space and time for elimination of heat ¾ More and more thermal stress is seen by the components
EUFANET workshop 2006 – Wuppertal
2
Materials and interfaces
An electronics component constitutes an assembly of multiple very different materials ¾ ¾ ¾ ¾
Variation of the coefficient of thermal expansion (CTE) of the different materials Interface behavior under varying temperature Fatigue …
Die pad
Die
Interconnection Bump
Substrate upon protection layer
Molding compound Via Substrate
Solder mask Electric via Thermal Via Technological design example
PCB Pad
Interconnection Bump EUFANET workshop 2006 – Wuppertal
3
Thermo-mechanic stress: A key problem
Thermal stress ¾
Constant temperature : T [°C]
¾
Dynamic situation :
¾
Thermal flux :
dT dt dT dz
[°C s-1] (x , y , z)
[°C z-1] (x , y )
Mechanical stress ¾
Screwing of PCB
¾ ¾
Clamping of components in sockets Mechanically fixed and tightened heat sinks
¾
…
EUFANET workshop 2006 – Wuppertal
4
Experimental problem
How to characterize the thermo-mechanical behavior : ¾
Of an electronic component, before and after assembly ?
¾
Of a PCB, before and after soldering of the components ?
¾
For different temperatures, and different types of temperature variation (e.g. during a reflow profile)
¾
In dynamic situations : ON / OFF
¾
Under mechanical or thermo-mechanical stress (e.g. component mounted on a heat sink)
¾
…
EUFANET workshop 2006 – Wuppertal
5
Classical development and analysis techniques
Thermo-mechanic simulation codes ¾
Easy access to parameter studies
¾ ¾
Time demanding Material and interface characteristics often unknown
¾
Failure appearance often related to “unknown” effects
Scanning acoustic microscopy (SAM) ¾
Powerful tool for delamination characterization
¾ ¾
Consequences of stress only detectable post-mortem No predictive power, no “warning” before failure
¾
Only operational at ambient temperature
EUFANET workshop 2006 – Wuppertal
6
Reliability evaluation by TDM
Delamination may be considered as a relaxation of stress •
through formation of cracks
Consequences of stress BEFORE delamination : •
Volume deformation in all three directions (x,y,z)
•
These volume deformations cause surface deformations • Out of plane deformation : Δz • In-plane deformation : Δx, Δy
THUS : Deformation is a failure risk indicator
EUFANET workshop 2006 – Wuppertal
7
Deformation under temperature variation
S2 S3
S1 z1 z2 z3 0
T = - 40°C
T = 25°C
Δl/l CTE
Strain
Compression
Δx1 - CTE 1 Δx2 - CTE 2
Δz1 Δz2
T = 260°C Temperature variation
Δz3
Δx3 - CTE 3
Δl/l measurement CTE mismatch evaluation
z
z1 + Δz1 z2 + Δz2 0 z3 + Δz3
Δz
Warpage (Δz) measurement delamination risk
EUFANET workshop 2006 – Wuppertal
8
New experimental approach TDM: Topography and Deformation Measurement under representative thermo-mechanical stress Camera Light source
Top heating Top Cooling
Top thermocouple
Sample holder Bottom Cooling
Bottom thermocouple Bottom heating
• 3D absolute topography and deformation analysis • Spatial resolution : ~ 2 µm • JEDEC type temperature profiles capability EUFANET workshop 2006 – Wuppertal
9
Application 1 : Brazing Al2O3 on Cu
Cu
Geometry variation during brazing process Al2O3
B
Al2O3: convex Cu : concave
B’ A
A
A’
B
B’
A’
EUFANET workshop 2006 – Wuppertal
10
Application 1 : Brazing Al2O3 on Cu
During cooling : The hole generated by the opposite bending of the two components (due to different CTE) is filled up with brazing alloy in liquid state. After several thermal cylces : Initiation de delamination detected by SAM, in the 4 corners of the assembly. SAM image EUFANET workshop 2006 – Wuppertal
11
Application 2 : Delamination during reflow Problem ¾
2 families of identical BGAs, produced at 2 different production sites
¾
One family ok, the other one fails during reflow
profils de température sur les differents composants
300
250
Tmax = 245°C temperature (°C)
200
150
comp8 comp1et8
100
comp4et5 comp2et3 comp2et3 comp4et5 50
comp1 comp6et7
0 0
60
120
180
240
300
360
420
480
540
600
660
720
780
840
900
960
tim e (s) EUFANET workshop 2006 – Wuppertal
12
Application 2 : Delamination during reflow BGA during reflow: Deformation analysis
t0 = initial
143 °C *
150°C
165 °C
173 °C
183 °C
200 °C
157 °C *
170 °C *
200 °C *
225 °C *
245 °C Z [µm]
t0 T max
t1 = final T ambiant
t1
EUFANET workshop 2006 – Wuppertal
13
Application 2 : Delamination during reflow Analysis for 2 components of each family comp1
comp2
comp3
comp4
T 23°C initial (t0)
Warning rate z different = 260µm
T max (245°C)
T 23°C final (t1)
EUFANET workshop 2006 – Wuppertal
14
Application 2 : Delamination during reflow Delamination checked by SAM Series 1 and 2: no delamination Series 2 Series 1
Comp. 1 1
5
Series 3 and 4: delamination Series 3 Series 4
Comp. 2 2
Comp. 3 3
Comp. 4 4
6
7
8
EUFANET workshop 2006 – Wuppertal
15
Application 3 : CTE mismatch analysis Iso-displacement fields Displacement field (vector x100)
X direction
Y direction ΔY~1pix.~16µm
ΔX~1pix.~16µm
¾ Axi-symmetric displacement field
Strain fields X direction
(ΔL/L)
Y direction
¾ Mean strain ~1700 ppm ¾ CTE ~ 14 × 10-6 /°C
EUFANET workshop 2006 – Wuppertal
16
Application 3 : CTE mismatch analysis Iso-displacement fields Displacement field (vector x100)
X direction
Y direction ΔY~0.8pix.~13µm
ΔX~0.7pix.~12µm
Strain fields X direction
(ΔL/L)
Y direction
¾ Mean strain ~ 1200 ppm ¾ CTE ~ 10 × 10-6 /°C
EUFANET workshop 2006 – Wuppertal
17
Application 3 : CTE mismatch analysis pixel
X direction
Iso-displacement fields Displacement field (vector x100)
pixel
Y direction ΔY~0.9pix.~14µm
ΔX~0.8pix.~13µm
¾ Axi-symmetric displacement field
X direction
Strain fields (ΔL/L)
Y direction
¾ Mean strain ~ 1350 ppm ¾ CTE ~ 11 × 10-6 /°C
EUFANET workshop 2006 – Wuppertal
18
Conclusions
Challenges Thermal management
TDM Contribution
PCB deformation during reflow
3D deformation measurement
BGA balls breaking
Realistic thermal stress
Delamination
µm range resolution Fully PC controlled
TDM Benefits Prevention and anticipation Failure understanding Reflow analysis Fatigue analysis EUFANET workshop 2006 – Wuppertal
19
Thank you for your attention ----------INSIDIX : Systems and Service Topography and Deformation Measurement: Insidix TDM Acoustic Microscopy: Sonix SAM X-Ray Imaging and Tomography: Fein Focus X-Ray X-Ray Micro-Fluorescence: Edax/Roentgenanalytik Eagle
EUFANET workshop 2006 – Wuppertal
20