AEC - Q006 - Rev -(2)

AEC - Q006 - Rev- June 8, 2015 Automotive Electronics CouncilComponent Technical Committee 4.1 Temperature Cycling (TC) This test highlights the differences in the coefficient of thermal expansion of package materials with Cu along with the increased hardness of Cu with respect to gold (Au). Perform per the requirements in AEC-Q100/Q101. 4.2 Biased Humidity (HAST/THB/H3TRB) This test can exacerbate corrosion along the Cu/bond pad intermetallic compound (IMC) interfaces. Perform per the requirements in AEC-Q100/Q101. 4.3 Power Temperature Cycle (PTC) / Intermittent Operation Life (IOL) This test can accelerate wearout by the combination of current/voltage and temperature. Perform per the requirements in AEC-Q100/Q101. 4.4 This test can accelerate IMC growth along the Cu/Aluminum (Al) interface to yield an open bond failure. It can also degrade the mechanical performance of the stitch bond. This is especially important for high temperature applications. Perform per the requirements in AEC-Q100/Q101. 4.5 5. 5.1 Board Level Stress Test Performance of this board-level temperature cycling test along with the test conditions, sample sizes and bill of materials to be used is to be agreed to between the user and supplier and justified by data. ANALYTICAL TESTS Delamination/CSAM Delamination of the mold compound over the Cu ball or stitch bond could lead to joint fatigue failure at either weld joint. The delamination criteria for various stages of qualification testing are shown in Table 1. High Temperature Storage Life (HTSL) / High Temperature Gate Bias (HTGB) / High Temperature Reverse Bias (HTRB) Page 3 of 10 AEC - Q006 - Rev- June 8, 2015 Automotive Electronics CouncilComponent Technical Committee Table 1: Delamination Criteria Read Point Mold Compound Delamination acceptance criteria No delamination at first (ball) or second (stitch/wedge) bonds unless otherwise (1)agreed between supplier and user. No delamination at first (ball) or second (stitch/wedge) bonds unless otherwise (1)agreed between supplier and user. No delamination at first (ball) bond. If any second (stitch/wedge) bond delamination found – no heel cracks or any other Cu-related fail mechanisms allowed. Evaluate the severity of any bond delamination found per Sections 5.2 and (2)5.3. Electrical T0 All components passing production test All components passing production test Post MSL PC Qualification Requirements 1X AEC grade X All components passing production test 2X AEC grade X (TC included if no BLR performed) All components passing (3)production test Minimum CSAM samples size: EITHER the same 11 components per lot through each readpoint (preferable) OR 22 random components per lot at each readpoint. Notes: (1) Agreement between the supplier and user would be achieved via the exchange of data that demonstrates that the form of delamination seen is not an issue for this part based on supporting data (field, monitor, in-process, etc.). (2) Method of evaluation to be determined by the user and supplier. (3) At 2x TC read point, passing production test means zero systematic Cu wire related issues. For example, if a failure was found to be related to solder ball or substrate, that is not considered a valid Cu wire failure. 5.2 Wire Bond Integrity The tests described below and where they are performed are a good gauge of the bond strength and weld formation of the ball and stitch bonds. They are done to demonstrate no reliability risk. The location of the hook for bond pull should be over the contact of interest (i.e., over the ball and over the stitch/wedge). ? Ball shear – ball bond area versus shear force (pre-packaged) ? Wedge bond wire pull (pre-packaged) ? Perform wire pull/ ball shear on first bond and wire pull for stitch bond (post packaged) ? Pad cratering test (pre-packaged) Wire pull / ball shear is performed after stress testing and decapsulation. A recommended process flow is described below: 1. Select components per the sample size specified in AEC-Q100/Q101 for wire pull and shear. Selecting worst-case components based on CSAM is desirable. 2. Carefully decapsulate these components so as to not damage or adversely affect the wire bonds but enough to be able to reliably conduct wire pulls and/or bond shears. Page 4 of 10 AEC - Q006 - Rev- June 8, 2015 Automotive Electronics CouncilComponent Technical Committee 3. The wire pull hook should be situated over the wedge bond for stitch bond pull and over the ball for ball bond pull. Stitch bond pull force results after stress testing may not be a reliable gauge of bond quality, as the act of pulling a wedge might not be repeatable and/or reproducible. 4. Compare these results with production data (i.e., before mold) to see if there is a level of degradation in the distribution of the data that could reasonably point to a potential reliability issue. If there are positively biased wires required in the test, ensure that they are included in this analysis, as they are thought to be more susceptible to corrosion. 5. In conjunction with pull/shear after decapsulation, a thorough inspection of the wedge bonds should take place to look for heel cracks or precursors for failure. For temperature cycling, pulls and shears at corner locations of the die/package are preferable. For moisture stressing, selecting random balls/stitches is acceptable (uniform moisture penetration) but ensure that both biased and unbiased pins are selected. Determination of which wires per device undergo ball shear, ball pull or stitch pull is left to the supplier to determine as long as the intent of inspecting all types of bonds is adequately addressed. 5.3 Cross-sectioning Inspection For initial supplier qualification of a new die/package (interaction) family/technology, components from the CSAM after TC stressing showing both no delamination and delamination over a bond(s) are to be used for cross-sectioning. The sample sizes, test conditions and acceptance criteria are specified in the overall process qualification flow shown in Table 2. Criteria of examination: ? Ball bond area o Amount and distribution of intermetallic - an alternative planar analysis method to evaluate ball bond IMC formation is also acceptable o Crack initiation/propagation o Corrosion after 1x ? Wedge bond area o Amount of contact o Wire angle to wedge o Crack initiation/propagation o Corrosion after 1x o Intermetallics formed in the bond area 6. 6.1 COMPONENT CHANGES Qualification Test Requirements for Cu Wire Changes The method of qualification of changes to already qualified and released components is outside the scope of this document. For those cases, AEC-Q100/Q101 is applicable and the qualification plan shall be based on an assessment of the tests needed for changes per Table 3 of AEC-Q100/Q101 and the relevant fail mechanisms, risk factors and best practices found in Appendix 1 of this document. If degradation models have been developed that can be technically justified via internal and external data to support the equivalent robustness of material and design changes to already-qualified Cu wire parts, this can be used with customer approval to then be allowed to perform AEC-Q100/Q101 testing. If there is limited knowledge or data on changes to already-qualified Cu wire parts, relevant tests should be performed per Q006 Table 2 conditions. Page 5 of 10 AEC - Q006 - Rev- June 8, 2015 Automotive Electronics CouncilComponent Technical Committee 7. QUALIFICATION REQUIREMENTS FOR Cu WIRE COMPONENTS The table below describes the individual steps required in a qualification flow for Cu wire components and the sample sizes required for each stress test. The qualification can be performed on a technology basis (define technology family for the purpose of Cu wire). Passage of the technology family allows subsequent components in the family to be qualified using the enhanced requirements below and allow them to be qualified. Subsequent qualification of parts in the same technology but not in the same product family would require performance to this table up to and including sequence #10 only. Table 2a: Integrated Circuit Qualification Test Requirements based on AEC-Q100 Sequence # HTSL PTC Stress Test Qualification Step Initial sampling CSAM @ T0 Preconditioning to MSLx CSAM after PC ATE Test Stress 1x CSAM post-1x stress ATE Test Wire pull Ball shear Cross-section Stress 2x CSAM post-2x stress ATE Test (1)(1)(1)(1) 1 2 3 4 5 6 7 8 9a 9b 10 11 12 13 Sample sizes as required Sample sizes as required 3x77 3x22 3x77 3x77 3x22 3x77 3x3 3x3 3x1 3x70 3x22 3x70 3x2 3x2 (2,3)(2)(2)(4)(4)3x77 3x22 3x77 3x77 3x22 3x77 3x3 3x3 3x1 3x70 3x22 3x70 3x2 3x2 3x1 (4)(4)(3)(4)(4)HAST/THB TC 1x45 1x22 1x45 1x45 --- 1x45 --- --- --- 1x45 --- 1x45 --- --- --- (3)--- --- 3x45 3x45 --- 3x45 --- --- 3x1 3x44 --- 3x44 --- --- 3x1 (3)14a Wire pull 14b Ball shear 15 Cross-section (2,4)(2,4)(2) 3x1 Notes: (1) Either 11 marked or 22 random parts per lot per Table 1 CSAM sample size criteria. (2) Performed only if board level reliability testing is NOT being performed. (3) Any failures beyond 1X must directly relate to the Cu wire bonding system for them to count as a legitimate failure requiring further evaluation (i.e., the projected lifetime of failure, effect of fail mode on product lifetime, corrective/preventive action). The method of approval is determined between the user and supplier. (4) Pull/shear as many as is possible per the number of wires per device to be qualified up to a maximum of 30 wires/balls from the total sample size specified. Page 6 of 10 AEC - Q006 - Rev- June 8, 2015 Automotive Electronics CouncilComponent Technical Committee Table 2b: Discrete Qualification Test Requirements based on AEC-Q101 HAST/H3TRB Sequence # Stress Test Qualification Step Initial sampling CSAM @ T0 Preconditioning to MSLx CSAM after PC ATE Test Stress 1x CSAM post-1x stress ATE Test Wire pull Ball shear Cross-section Stress 2x CSAM post-2x stress ATE Test (1)(1)(1)(1) 1 2 3 4 5 6 7 8 9a 9b 10 11 12 13 Sample sizes as required Sample sizes as required 3x77 3x22 3x77 3x77 3x22 3x77 3x3 3x3 3x1 3x70 3x22 3x70 3x2 3x2 (2,3)(2)(2)(4)(4)3x77 3x22 3x77 3x77 3x22 3x77 3x3 3x3 3x1 3x70 3x22 3x70 3x2 3x2 3x1 (4)(4)(3)(4)(4)3x77 3x22 3x77 3x77 --- 3x77 --- --- --- 3x77 --- 3x77 --- --- --- (3)3x77 3x77 --- 3x77 --- --- 3x1 3x76 --- 3x76 --- --- 3x1 (3)14a Wire pull 14b Ball shear 15 Cross-section (2,4)(2,4)(2) 3x1 Notes: (1) Either 11 marked or 22 random parts per lot per Table 1 CSAM sample size criteria. (2) Performed only if board level reliability testing is NOT being performed. (3) Any failures beyond 1X must directly relate to the Cu wire bonding system for them to count as a legitimate failure requiring further evaluation (i.e., the projected lifetime of failure, effect of fail mode on product lifetime, corrective/preventive action). The method of approval is determined between the user and supplier. (4) Pull/shear as many as is possible per the number of wires per device to be qualified up to a maximum of 30 wires/balls from the total sample size specified. Page 7 of 10 HTRB/HTGB --- --- IOL TC