How to Avoid Conductive Anodic Filaments (CAF) Ling Zou & Chris Hunt 22 January 20 1 Your Delegate Webinar Control Panel Open and close your panel Full screen view Raise hand for Q&A at the end Submit text questions during or at the end 2 1
NPL EI Free Technology Webinars Benefits of Using Nitrogen During Soldering and determining PPM Levels Thursday 14 February Attraction of Sn Tin Whiskers Wednesday M arch BGA Reliability The Effects of Solder Joint Voiding and Uneven StandOff Height Tuesday 30 April Reuse of Electronic Products Wednesday 1 M ay Through Hole Reliability for High Aspect via Holes Tuesday June Functionalisiation and Applications of Nanomaterials for Electronic Applications M onday 1 July Environmental Robustness: Performance Coatings & High Temperature Interconnects Tuesday 3 September www.npl.co.uk/ei IPC/NPL Cleaning & Contamination Testing Center Live stand 14 AV & Seminars 30ft 20ft SIR Testing Ionic Tester Microscope Contamination Flux Test Kits AOI for Flux & Coatings Ion Chromatography Aqueous Cleaner Ionic Tester Bobs Office Table http://www.ipcapexexpo.org/html/main/cleaning.htm 2
How to Avoide Conductive Anodic Filaments (CAF) Ling Zou & Chris Hunt 22 January 20 Three main drive for CAF concern in past ten years CAF failure mechanism was first reported in the ;s by Bell laboratory. Three main drives: The drive to increase circuit density with smaller printed wiring board geometries. The rapid increase of the use of electronics in harsh environments and for high reliability and safety critical application. Leadfree soldering process affect laminate stability and increase CAF material choices. 3
Conductive Anode Filament (CAF) CAF formation inside the PCB is an important failure mode for electronic circuit. It is a electrochemical process, and initially caused by anodic Cu corrosion. CAF is Cu corrosion products grow along the glass/resin interface from anode to cathode R Mechanisms Ionic species Moisture Bias Low ph Anode corrosion Electrochemical cell Mobile corrosion products Migration down pathway Susceptibility reflow pathway CAF ph increase Deposition of salt 4
Electrochemical process for CAF v Cu HO M H 2 HO Cathode Cu n ne Cu H 2e H 2 O 2 H 2 O 4e 4OH V M n H C H O Cu 2 2OH Cu (OH) 2 H M n Cu Corrosion of metal Anode Cu Cu n ne 2H 2 O O 2 H 4e Cu (OH) 2 Cu O H 2 O CuO: Black CuCl 2 : Yellow brown CuSO 4 H 2 O: Blue The Cu salt or Cu (OH) 2 can be deposited at high ph Conditions for CAF formation Electrical charge carriers must be present to form Electrochemical cell Ionic species inside PCB, H and OH from water Water must be present to dissolve the ionic material and sustain them in their mobile ionic state Moisture, humidity Acid environment around conductors is needed to initiate Cu corrosion at anode. Acid contaminations from glass fibre surface, fluxes from assembly process or/and acid residues from plating process Pathway is needed for Ions to move Delamination between glass fibre and resin due to reflow Bias acts as driving force for ion transport Circuits need to be powered up in service
Traditional CAF measurement Test PCBs at high temperature and humidity ( C/% RH) condition with bias (0V) for 00 hours. Monitor Insulation Resistance (IR) continually. Take long time for the evaluation. Time to failure Log SIR ( ) (ohm 21 hours 21 hours 21 hours 0 0 0 200 20 300
Traditional CAF measurements 00µm A C G 300µm 0µm Inline Inline Vertical Vertical E Inline Horizontal 400µm Inline Horizontal I 400µm AntiPad Inner Pads Ground Anode K 400µm AntiPad No Pads Ground Anode B 400µm Inline Vertical D 00µm Inline Vertical F 400µm Staggered H 400µm geometry J 400µm AntiPad Inner Pads Via Anode L 400µm AntiPad No Pads Via Anode No via InLine /staggered combs m 300 400 0 00 Via to via _ Fibre Weave InLine 14 Fibre Weave Staggered
Antipad combs Inner Layer Pads Inner Layer Cathode Via Comb I Comb K Anode Via Comb J Comb L 1 Parameters tested Factors affect CAF formation: Via wall gap Bias Pattern orientation Pattern arrangement Via polarity Presence of Innerlayer pads Surface finish Drill speed Reflow conditions PCB materials Board house 1
Effect of tested variables on CAF formationfailure time 1 Conclusions how to avoide CAF Via to via gap Fail faster with smaller gap Bias High bias significantly decrease the time to fail Laminate CAF resistance material reduce risk of CAF formation Identically specified laminates from different suppliers have different CAF performance Manufacture The in house process of laminate at board house resulting in different CAF performance Via and ground planes Anodic via fail faster than cathodic via Alignment to reinforcement Warp and weft direction Via staggered at 4 has greater CAF resistance Reflow condition High peak temperature in reflow is harmful to CAF performance of the PCB Board finish, drill speed, inner pads, Tg of laminate have a less significant effect on CAF performance 1 Wicking, & inter yarn voiding Images courtesy of Isola
Publication LeadFree and Other Process Effects on Conductive Anodic Filamentation Resistance of GlassReinforced Epoxy Laminate Ling Zou, Alan Brewin & Christopher Hunt, The ELFENT Book on Failure Mechanisms, Testing Method, and Quality Issues of LeadFree Solder Interconnects, 20 p221 Susceptibility of Glass Reinforced Epoxy Laminates to Conductive Anodic Filamentation, Alan Brewin, Ling Zou & Christopher Hunt, MATC(A)1, January 2004 1 Identifying CAF Typically CAF occurs in buried layers inside PCBs It is difficult to find and inspect, and hence it is difficult to study Work at NPL has investigated a method for controlling the growth of CAF, and then subsequently studying the growth. NPL has developed a sample structure where glass fibres are introduced, encapsulated, and then provide the correct electrochemical conditions to grow the CAF in relatively shot time. 20
Simulated Test Vehicle (STV) Simulated Test Vehicle (STV) was developed, which provides a controlled way to grow CAF. This enables us to investigate different variables separately. CAF formation: Different resin systems and glass fibres: STV & STV with drilled hole Reflow process: STV Desmear process: STV with drilling hole Glass bundle size: STV CAF can be easily seen using microscope with backlight 21 STV1 sample Polyimide substrate with 2 oz Cu Resin powder dissolved in acetone Resin cured at C for 0 minute 22
Test sample preparation Different variables 0.02ul 0.3mm mm Acid condition at anode (plating solution CuSO 4 H 2 SO 4 ) Ionic contamination inside PCB (NaCl coated fibres) Pathway between two conductors (Reflow process not necessary)
Test different resins and glass fibres STV Anode contaminated with or without plating solution Resin Glass fibres Suppler DICY Cured* Phenolic cured 0 finished 2 finished 2 finished 2 heat cleaned no finish A B 2 loom state no finish *DICY absent DICY cured resin 2 20 m Phenolic cured resin 2 heat clean fibre Not reflowed 0 30 0 0 0 Anode: no contamination Fibres: clean CAF formed on reflowed sample 1 2 3 4 Reflowed 0 30 0 0 0 1 2 3 4
Phenolic cured resin 0 finished fibre 1 2 3 4 0 30 0 0 0 0 30 0 0 0 1 2 3 4 Anode: 0% plating solution Fibres: clean No CAF formed on both reflowed and no reflowed samples CAF formation with different resins and glass fibres Anode contamination 0% plating solution None 0% plating solution Glass fibres Resin CAF formation Not reflow ed reflow ed 0 finished (A) 2 finished (A) x x 2 finished (B) DICY absent 2 heat cleaned DICY 2 loom state cured 2 heat cleaned x 2 loom state 2 heat cleaned x 2 loom state 0 finished (A) Phenolic 2 finished (A) cured x x 2 finished (B) 14
Different resins and glass fibres With anode contamination only: There is CAF formation for heat cleaned and loom state fibres with both resins (DICY cured and phenolic cured). There is no CAF formation for all finished fibres with both resins. Finished glass fibre DICY resin Anode Fibre Reflow 0% plating solution Yes 0% plating solution No 20% plating solution 3% NaCl Yes 20% plating solution No 0% plating solution Yes 0% plating solution No 20% plating solution 1% NaCl Yes 20% plating solution No Anode contamination Fibre coated in NaCl Reflow & no reflow 1
Conclusion for finished fibres Anode Fibre Reflow CAF formation 0% plating solution Yes Yes 0% plating solution No No 20% plating solution 3% NaCl Yes Yes 20% plating solution No No 0% plating solution Yes No 0% plating solution No No 20% plating solution 1% NaCl Yes No 20% plating solution No No Three factors must be met for CAF formation Low ph at anode (Plating solution contamination) Ionic contamination inside PCB (NaCl coated fibres) Pathway between two conductors (Reflow process) CAF formation different glass fibre bundle size Phenolic cured resin Anode: no contamination Fibres: Clean heat and loom state glass fibres Bundle size: small (~) & large (30~0) Reflowed 32 1
Phenolic cured resin Heat cleaned fibre loom state fibre Small bundle 0 30 0 0 0 Large bundle 0 30 0 0 0 1 2 3 4 1 2 3 4 33 Small bundle 1 2 3 4 0 30 0 0 0 Large bundle 14 1 1 0 30 0 0 0 CAF formation 34 1
STV with drilled hole Tap Cu Polyimide 0% 30% % STV is sitting on adhesive tape. Hole filled with 1µl different concentration plating solutions. Filling solution dried for 2 hours at room temperature before CAF testing. 3 CAF formation desmear process Nodesmear Desmeared 3 1
CAF formation without desmear Phenolic cured resin & 2 finished glass fibre (B) 0% plating solution 0 30 0 0 0 1 2 3 4 30% plating solution 0 30 0 0 0 % plating solution 14 1 1 0 30 0 0 0 3 More CAF formed with contamination increase CAF formation with desmear Phenolic cured resin & 2 finished glass fibre 0% plating solution 1 2 3 4 0 30 0 0 0 % plating solution 0 30 0 0 0 3 14 1 1 30% plating solution 0 30 0 0 0 Desmear process increase CAF formation, but not significant. 1
CAF formation on drilling sample 3 CAF formation 40 20
STV1 and STV1 with drilled hole A STV has been successfully used to evaluate the effect of different resin systems, different glass fibres, desmear process, reflow process and glass fibre bundle size on CAF failure. Heat cleaned and loom state fibres form CAF more easily than finished fibres. Loom state fibre has the highest propensity to form CAF compared with others. Phenolic cured resin promote more CAF than DICY cured resin, this is probably because the DICY was removed when the resin was dissolved in acetone in our sample preparation. Desmear process can increase CAF formation, but not significant. Reflow process increase CAF formation significantly. Large bundle size increase CAF formation, but not significant. Further developments of a STV are underway 41 How to Avoide Conductive Anodic Filaments (CAF) Ling Zou & Chris Hunt 22 January 20 42 21
Stand 14 Presentations Tuesday.00noon Cleaning and Contamination Failures Causes and Cures 2.00pm Cleaning Under Low Stand Off Components 3.00pm Testing Printed Board Assemblies For Cleanliness Wednesday.00noon 2.00pm 3.00pm Evaluating a Cleaning Process Performance Reliability of Flux Residues Under Low Stand Off Devices Cleaning Package On Package Assemblies Thursday.00am Conductive Anodic Filament CAF Failures and Defect Analysis.00 noon Testing Adhesion of Conformal Coatings and Failures 1.00pm Testing Components for Cleaning Process Compatibility Each presentation will be 0mins A copy of the presentation material can be obtained by visiting the stand http://www.ipcapexexpo.org/html/main/cleaning.htm NPL EI Free Technology Webinars Benefits of Using Nitrogen During Soldering and determining PPM Levels Thursday 14 February Attraction of Sn Tin Whiskers Wednesday M arch BGA Reliability The Effects of Solder Joint Voiding and Uneven StandOff Height Tuesday 30 April Reuse of Electronic Products Wednesday 1 M ay Through Hole Reliability for High Aspect via Holes Tuesday June Functionalisiation and Applications of Nanomaterials for Electronic Applications M onday 1 July Environmental Robustness: Performance Coatings & High Temperature Interconnects Tuesday 3 September www.npl.co.uk/ei 22