Materials for space electronics
How do you make a circuit board fit for service in space? Simon Frost reports from the Electronic Materials and Processes for Space workshop, in Germany.
When you’re taking electronics to the final frontier, rugged reliability is key. From radiation to extreme temperatures, everything in space has to be equipped for unearthly conditions – and while replacing a circuit board is easy enough on the ground, it’s a task that’s best avoided in orbit and beyond. Tin whiskers, solder damage, electromigration and contamination were some of the key challenges discussed by scientists attending the sixth Electronic Materials and Processes for Space (EMPS) workshop in April 2015 at the German Space Operations Centre, near Munich.
Alumina in disguise
Giles Humpston, Field Applications Manager at Suffolk-based Cambridge Nanotherm Ltd, opened proceedings with his company’s nanoceramic alumina coating process for aluminium printed circuit board (PCB) components, which enhances the material’s thermal and electrical properties.
Established methods for applying alumina coatings to a metal, such as anodising, plasma electrolytic oxidation and spray coating can already produce effective mechanical coatings. But their high surface roughness, low breakdown voltage and unpredictable deposition make them unsuitable for electronic applications.
Rather than simply coating, Nanotherm’s process converts the surface of aluminium itself into alumina, through electrochemical processing in an electrolytic cell filled with a benign alkaline solution, which can be safely disposed of down the drain. ‘It’s important to mention that this is a green technology,’ said Humpston – ‘one of the company’s founders demonstrated that it’s safe by putting a cup into the tank and drinking it, and he’s still alive!’
The resulting nanoceramic material maintains the robust mechanical properties of aluminium, but with thermal performance comparable with aluminium nitride – the boards survived thermal cycling from -40°C to 250°C without degradation, and PCBs with a thermal conductivity of 152W/mK can be manufactured by adhering copper foil to the top surface of the converted aluminium, Humpston said.
‘It’s alumina, but not as you know it’, he said. The composition of the nanoceramic layer is affected by the other elements present in the aluminium alloy, with chemical differences visible in the material’s colour – a purer grade produces an ivory finish, while alloys containing copper can be almost jet-black.
No space for error
‘By necessity, the space industry is risk averse’, began Dave Davitt, Materials and Process Engineering Manager at COM DEV International. Passive electronics are in continual demand for space applications because of their high reliability and low risk within satellite payloads.
But a difficulty for passive microwave devices is the transition of signal between layers from stripline circuits up to adjacent operational hardware. Conventional solutions have their drawbacks – adding plated vias to carry the radio frequency signal to the stripline surface increases the cost of processing, and issues with the integrity of the via hole can arise.
One solution COM DEV has developed is its Inhomogeneous Stripline Interface Transition (ISIT) technology, designed to remedy failures that plague conventional interfaces in high frequency applications. The coaxial connector is allied with the stripline circuit via a sliding contact interface installed in a cavity exposing the inner trace within the PCB. The cavities are partially filled with dielectric to maintain the stripline configuration at this interface. This design feature can also be used to optimise impedance matching in this region.
No area of technology exists in a vacuum (figuratively speaking), and presenters at EMPS crossed over from the space industry into medical, automotive and military spheres – one speaker even described to me his line of high-end fitness equipment.
Jan Vanfleteren, from Ghent University, Belgium, is one such scientist who came to EMPS having already applied his research to another sector. He brought with him a blue light therapy wrist wrap – a stretchable ‘2D’ circuit installed with blue LEDs on a polydimethylsiloxane surface that wraps around the wrist, designed to treat conditions such as carpal tunnel syndrome.
But ultra-thin, stretchable circuits are beneficial not only for medical applications, but any area where weight and space can be saved by removing the shape constraints of a rigid circuit board and where the circuit needs to be moveable – for example, components built in to spacesuits.
The elastic nature of the circuits Vanfleteren describes is achieved by the use of meander-shaped fine-line copper conductors embedded in a stretchable polymer. These joined circular segments offer the circuit an even stress distribution when the polymer is stretched, joining together the necessarily rigid components on a track that can change not only the direction of its path, but the distance between component islands. Circuits can remain stretchable or be thermoformed solidly to suit a required profile – Vanfleteren outlined to me light electronic components under development with the EU-funded TERASEL project that conform to the shape of a car’s dashboard.
Metallurgists will know tin whiskering as a troublesome phenomenon that occurs in electrical devices, especially between solder pads, causing short circuits. These little metal hairs are one of the reasons why lead is a desirable soldering ingredient – it slows the progress of tin whisker growth, and more than 3% tin eliminates the initiation of whiskers altogether. Whiskers are not a problem reserved for space electronics – they have caused failures in nuclear power stations and even pacemakers.
Researchers at the National Physical Laboratory (NPL), UK, have been measuring whisker growth and the effectiveness of mitigation techniques. One of the best methods is to coat tin-plated components, evidenced by one of NPL’s conclusions – tin whisker occurrence was dominated by eruptions at the edge of plates and right-angle bends, where coatings are liable to be less well deployed.
'It is my firm belief that it's not the choice of coating that matters but how well you apply it,' said Martin Wickham of the NPL. Whether coated with acrylic, polyurethane or silicone, sharp edges are the most common location for a whisker outbreak – it’s the integrity of the coating that really makes the difference. Wickham also noted that increasing the temperature of heat treatment and reflow delayed the occurrence of whiskers. He invited companies to support the ongoing studies, particularly in regards to the effect of vibration and forced air cooling on whiskers that emanate from tin-plated component leads.
Wiping the slate clean
Tin whiskers are not the only challenge to high reliability electronics created by the Restriction of Hazardous Substances Directive (RoHS) in 2006. The use of lead in solders was limited under RoHS and although high-reliability sectors such as
defence and space were excluded from this legislation, manufacturers now have difficulty obtaining ball grid arrays (BGAs) with sphere alloys that meet their needs.
The tin-silver-copper (SAC) alloys now commonly found on BGAs result in a less reliable solder joint for the demanding environments of space and defence.
Converting SAC BGAs to the reliable tin-lead (SnPb) finish for RoHS-excluded applications requires a complete flush of the SAC alloy before SnPb spheres can be attached. Solder wicking and vacuum desoldering allows residue of the original alloy to remain, resulting in unreliable interfaces and weak solder joints due to mismatches in metallurgy.
Don Tyler, Director of Corfin Industries, USA, explained how his company de-balls lead-free BGAs and replaces the array finish with tin-lead, prior to placing new tin-lead spheres.
Corfin’s 100% robotic system begins with dipping the array into a water-soluble acid flux. The excess flux is blown off and the BGA is preheated with hot air to activate and dehydrate the flux. The lead-free balls are then flushed off using a solder wave and hot air, and the remaining flux is removed in an ultra-filtered hot water flush before drying.
The new tin-lead balls are precisely positioned onto the flux-brushed array via laser-etched stencils and attached in a reflow oven. Tyler described a project that evaluated BGAs that had the entire process repeated three times to determine if occasional reworks were detrimental to the reliability of the devices. The company found no damage to the device nor any appreciable loss of nickel and copper by dissolution into the liquid solder.
Graham Naisbitt, Director of GEN3 Systems, UK, discussed ionic contamination, which can lead to process variability, an increase in defects, dendritic growth, tin whiskers and corrosion – all resulting in poor reliability at best, if not total failure.
He highlighted that existing cleanliness testing methods – resistivity of solvent extract (ROSE) and solvent extract conductivity (SEC) fall at the first hurdle. ‘They're measurements of the amount of soluble ionic contamination present on a circuit board, component or assembly. And the standard that set a pass/fail level for contractors in the mid-1970s unfortunately still survives to this day.’ He said.
For ROSE and SEC testing, isopropyl alcohol is mixed with de-ionised water, and the conductivity of the mixed solution is measured. The item to be tested is then added to the solution and agitated, and a second reading is taken to determine the μg per cm2 sodium chloride equivalence.
Naisbitt questioned the current industry standard – ‘The “acceptable” level is 1.56μg per cm2. That means that it's ok to leave up to that amount of salt on every cm2 of the assembly, where it may be localised underneath a ball grid array, for example.’ He said, adding a rhetorical, ‘Really?’
His recommendation for more thorough testing was to couple surface insulation resistance testing with GEN3’s automated, closed-loop method, process ionic contamination testing (PICT), which is under consideration by the International Electrochemical Commission and due for publication in 2016. The rapid, inexpensive add-on cycles test samples in a closed-loop de-ionised solution to determine its cleanliness more accurately.
Since 2010, the School of Engineering at University of Portsmouth, UK, has arranged at least one EMPS workshop every year. In previous years, the workshop, founded by Barrie Dunn FIMMM, has been held in France, Italy, Denmark, the Netherlands and, of course, Portsmouth, where it will return in April 2016. Dunn encourages materials science and metallurgy students to attend these workshops as the subject discussions will be of interest to them and it provides a good chance to network with engineers from different space agencies and spacecraft manufacturing companies.
For full details, visit emps.port.ac.uk, where presentations from this and all previous EMPS workshops can be downloaded.