Evaluation by three-point-bend and ball-on-ring tests of thinning process on silicon die strength

Barnat, Samed; Frémont, Hélène; Gracia, A.; Cadalen, Eric

doi:10.1016/j.microrel.2012.06.093

Cited by 16 publications

(13 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Also, this SBS range with such die thicknesses was within the variations that were estimated from [9-11, 19, 35, 36]. The differences could be due to the discrepancy of the sample preparation, different silicon type and with the combination of 4PB and those ball ring tests setup [12,[15][16][17][20][21][22][23]. However, a slight improvement at a lower thickness value was observed; this may be attributed to the extra removal of silicon defects and damage layers during the thinning and surface treatment processes.…”

Section: Influence Of Silicon Thicknesses On Sbssupporting

confidence: 57%

“…Those bending stresses are mostly developed through a defined system curvature with a wafer bowing effect. Higher strength is expected with extra treatment on the silicon surface without major influence of edge defects or chipoff [12,[15][16][17][20][21][22][23].…”

Section: Influence Of Silicon Thicknesses On Sbsmentioning

confidence: 99%

“…Silicon strength is normally evaluated with various lab scale bend or loading tests. There are many newly redefined testing methodologies and analysis approaches for the material strength study which include 3PB [7][8][9][10][11], four point bending (4PB) [12][13][14], ball/ ring on ring [15][16][17] and others that offer different advantages. In general, the 3PB approach will result with an SBS range similar to the 4PB [12][13][14].…”

Section: Introductionmentioning

confidence: 99%

“…Recently, an improved 4PB application has been developed by combining with the indentation test; nevertheless, this does not simplify the application of the 4PB test [18,19]. Even though higher strength is estimated with the ball or ring related tests from the recent trend of literatures, it is suggested that those ball on ring (BoR), ring on ring (RoR) and ball breaker (BB) tests are only good in evaluating the silicon surface related strength, either without influence from the assembly process or with an intended focus on the potential treatment improvement on the silicon surface [12,[15][16][17][20][21][22][23][24]. The critical impact, such as edge flaw or chipping, is excluded due to the nature of the test setup.…”

Section: Introductionmentioning

confidence: 99%

“…The applied force at break (PB, in Newton), the support span (L, in millimetre), the width of the test specimen (b, in millimetre) and the thickness of the specimen (h, in millimetre) will be calculated to estimate the flexure stress ( fb) of silicon with a linear stress/ strain behaviour. The detail setup is documented in various literatures [7][8][9][10][11]17]. Jie-Hua et al stated that 3PB is a simple test but it is modified from a method used to measure the strength of brittle materials, such as ceramics [15].…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Experimental evaluation on the silicon mechanical performance of electronic packaging

Fong¹,

Koay²,

Azid³

2017

J. MECH. ENG. SCI.

View full text Add to dashboard Cite

Semiconductor packaging is trending towards a miniaturisation in size but an increase in functionality. Hence, the thickness of the silicon wafer has decreased dramatically with a concern on the possible degradation of the strength of the thinned wafer. In this paper, 3-point bend (3PB) on bare silicon is selected as the preferred silicon break strength (SBS) test methodology due to its setup effectiveness and subsequent application in the prediction study. This experimental testing study focused on evaluating the SBS from different thickness ranges. The study was then followed by evaluating the influence of a possible impact from flaw creation with laser marking on silicon surfaces. The results show that the SBS is consistent although with significant differences in the silicon thickness ranges. It is also revealed that the onset breaking load may not be a suitable metric and could be over sensitive on gauging the SBS comparison. The flaw creation from the engraving process revealed a significant drop of 75% SBS although with an approximate depth removal of <10% from the total thickness. The consistence failure mode is clearly visible and this left the silicon very vulnerable to catastrophic failure. This indicates that extra care is needed on ultra-thin silicon during the assembly process as fractures may happen even before any reliability stress test is conducted. Furthermore, the completion of the package level 3PB failure mode verification has helped to demonstrate that an SBS study can be conducted with a simplified bare silicon level testing. In short, the study with this simplified 3PB has successfully proven its usefulness in SBS estimation. The observed ultra-thin silicon SBS has degraded and strongly depended on the critical flaws especially from the surface defect and impact from the assembly handling.

show abstract