Reading time ( words)
To read Part 1 of Pete Starkey’s report from the recent gathering of the SMART Group at Longborough University, click here.
SMART Group Steering Committee member Ian Fox, from Rolls Royce Control Systems, discussed practical aspects of tin whisker mitigation from the viewpoint of a manufacturer of aero engine control electronics. A proper control plan was fundamental to lead-free component risk mitigation, and he described how to prepare one in accordance with IEC/TS 62647-1 and to carry out system assessment using the decision tree from IEC/TS 62647-2 Annex A. The Rolls Royce control plan fitted into an area between Levels 2B and 2C defined by GEIA-STD-0005-2. Their management plan for components was based on IEC/TS 62239-1 and referenced a database of approved lead-free components which listed all the components used, with an approved lead-free finish defined for each component. Any components not lead-free approved were verified by X-ray fluorescence testing on receipt. All product change notifications received were reviewed against the company’s requirements for lead-free component acceptability, and actions were taken to verify their continued acceptability. Alternatively, the component change could be rejected and remedial action taken. The PCN review procedure was extremely useful in indicating industry assembly and material trends.
Fox showed a range of component package styles, together with the acceptance tests and mitigation strategies for each type, including PCB design rules defining minimum component lead spacing. He also discussed the attributes of a range of conformal coatings: acrylic, polyurethane, silicone and parylene. Returning to the subject of component finishes, he commented on the debate as to whether the reflow process reduced or increased the propensity of tin to whisker, and concluded that if there was any doubt or there was a contractual obligation not to use tin, then refinishing of components was justified, using a controlled solder dip process that met the requirements of GEIA-STD-0006. “The only way to avoid whiskers is to avoid tin!”
Dr Mark Ashworth returned to deliver his second presentation on research at Loughborough, this time focused on post-plating mitigation methods. He commented that whatever might be achieved by way of optimisation of the tin electroplating process, whisker growth was so unpredictable that even the ‘best’ electroplated tin coatings could only be considered ‘whisker-resistant’ rather than ‘whisker-proof’, and additional precautions were required to further suppress the growth of whiskers. Three different techniques had been investigated.
The WHISKERMIT 2 research programme had set out to develop novel conformal coatings specifically designed to mitigate whisker growth by incorporating nano-fillers in the polymer formulation. Brass coupons electroplated with 2 microns of bright tin had been spray-coated with a proprietary acrylic formulation, modified by the addition of a nano-filler at 3%, 5% and 7% loading, and stored in an environmental chamber at 55°C and 85% humidity. The coupons had been periodically inspected for whisker growth using a stereo microscope. The modified coatings showed significant improvement in whisker mitigation compared with unmodified coatings, with the 5% nano-filler addition appearing to offer the optimum balance between mechanical strength and ductility.
Research published in 1994 suggested that tin whisker growth initiated at cracks in the surface oxide layer, so a second approach had been to investigate the effect of increasing the thickness of the oxide by electrochemical oxidation in borate buffer and potassium carbonate/bicarbonate solutions. An order of magnitude reduction in whisker growth had been demonstrated and the increased-thickness oxide layer had been observed to be still effective at mitigating whisker growth after three years.
The third approach reported by Dr Ashworth was atomic layer deposition, a proprietary thin film coating method using a self-limiting gas-phase chemical reaction to achieve thicknesses at the nanometre level. A range of pre-treatments and process conditions had been investigated. Coated and control samples had been inspected over a 12-month period using optical and scanning electron microscopy and significant reductions in whisker growth had been observed.
There had been several mentions of component refinishing during the seminar, in the context of the requirements of the GEIA-STD-0006 standard. Addressing the issues of continuation of supply and development of best practice, Mark Walmsley described how Micross Components had set out to develop and qualify an automated process for hot solder dipping electronic components that was compliant to the GEIA specification. The outcome was a seven-axis, multi-functional robotic machine with controlled pre-heat and the capability to manipulate and position its robotic arm to within 0.1 mm. The machine offered a choice of conventional wave, side wave or flat pot soldering, with control of depth, dwell, entry and exit speed, solder angle and exit angle, followed by controlled cool-down and in-line washing.
The qualification of process and equipment to the GEIA standard was undertaken in partnership with the University of Greenwich for mathematical modelling of package reliability, and National Physical Laboratory to investigate terminal finish and reliability. Greenwich investigated 10 different component types and concluded by non-parametric statistical testing that the refinishing process had no significant impact on the electrical performance of the components and that the null hypothesis, that un-refinished parts were the same as the refinished parts, could not be rejected at the 5% significance level. And confocal scanning acoustic microscopy detected no degradation after hot-solder-dip processing. NPL selected 25 devices from RoHS-compliant sources, representing QFP, BGA and through-hole package types, and these were characterised before and after re-termination by X-ray fluorescence spectroscopy, micro-sectioning and optical microscopy, solderability testing, BGA ball shear measurement, scanning electron microscopy and scanning acoustic microscopy. The re-terminated components showed solderability equal to or better than the original components. Ball shear results for BGA components were acceptable, scanning acoustic microscopy did not locate any differences between original and re-terminated components, and thermal cycle solder joint reliability was improved for re-terminated components compared with tin plated originals.
Micross had undertaken a two-year exercise with six industrial partners and two academic institutions to develop a European source for hot solder dipping. A number of technical papers had been published, and the overall conclusion was that hot solder dipping components was not a major risk to electronic components.
This SMART Group seminar was notably interactive. The presenters were happy to be interrupted with questions and many interesting points of discussion were raised. And the coffee and lunch breaks presented abundant networking opportunities. A few one-liners worthy of recording: “Whiskers are as trustworthy as politicians!” “People have been looking at whiskers for 70 years and we’re still looking!” “Anyone who tells you it’s whisker resistant is taking a bit of a chance!” An excellent day all round.
Pete Starkey is based in the UK, and joined I-Connect007 as technical editor in 2008. Starkey has more than 30 years’ experience in the PCB industry, with a background in process development, technical service and technical sales. To contact Starkey, click here.