Having looked at some polymer basics, we are moving on to applications, encouraging you to explore as many as possible of the ways in which polymers are crucial for the successful manufacture of electronic assemblies.
As part of Activity 1 in polymer types, you will have created a list of polymer materials used in electronics. Why not review that list now, and check it against the items in the following notes?
We have identified the following uses in electronic assemblies for polymeric materials:
In circuit assembly processes, polymer applications include:
And we still haven’t identified the largest user, which is the enclosure. Typically this will be an assembly of different parts, often made in different materials, although doing this makes recycling more difficult.
In the sections that follow, we are looking at some of the applications above, and giving an idea as to the types of materials used. Bear in mind of course, that the list is neither exhaustive nor definitive, because polymers are compounded for the application, rather than made to standard specifications.
Apart from the enclosure, the printed circuit board is physically the largest component in most electronic products. The majority of laminates used in professional circles are of epoxide resin, reinforced with glass fibre, although phenolic resins reinforced with paper are also used. More information on their make-up and manufacture will be given in Units 6 and 7. However, at this stage you should realise that the laminate on forming the heart of the board is not the only use of polymers within the PCB structure.
The photoresist is another polymer material used routinely during board manufacture, applied as a thin film by either wet or dry film techniques. The most commonly used photoresists are ‘negative-acting’, which means that they polymerise on exposure to ultraviolet light and hence become insoluble in a direct developer. Non-polymerised resist is removed by ‘developing’ to expose copper areas ready for electro-plating.
Another polymer used in the manufacture of PCBs is the solder resist or ‘solder mask’. This is a thin layer which coats the outside surface of the board while leaving the component pads clear and solderable. By resisting solder wetting, the solder mask reduces the number of solder shorts on an assembly, as well as protecting the tracking from chemical and mechanical damage, and increasing the humidity resistance and insulation resistance of the board.
Where surface mount and traditional through-hole components are combined, it is necessary to use an adhesive to attach SM devices to the underside of the mixed technology board. Leaded components, having been inserted through holes in the board, are held in place mechanically while the board passes over a molten solder bath; unsecured SM devices under the board would just float off into the solder wave!
Adhesive placed between solder pads also secures the device to the board, and has the spin-off benefit that it will also provide a degree of strain relief for thermal mismatch between components and PCB.
Although solid preforms of adhesive are available, liquid glues are generally used. These can be applied to the board by screen/stencil printing, pin transfer, or syringe dispensing. After the device is placed, the wet adhesive must have sufficient wet, or ‘green’, strength to hold the device in position until cured. The cured adhesive must then have sufficient strength to hold the device to the board during the solder wave process.
Once the PCB has been soldered effectively, the hold-down function of the adhesive is complete as the soldered joints will be much stronger than the adhesive bond. However, the adhesive will still be present and so must not adversely affect the performance of the circuit in any way. In particular, the adhesive must maintain sufficient insulation properties to avoid electrical shorts, and it must not create long-term corrosion problems.
Correct selection and use of the adhesive can reduce assembly defects, particularly with the latest high-speed dispensing equipment. Recent developments have enabled a faster process and higher yields. However, this has only been possible by understanding and controlling the process surrounding the adhesive.
Glues are commonly applied by three methods:
An adhesive which is suitable for one of the above methods may not be suitable for the others. Ideally a chip-bonding adhesive should satisfy the following requirements:
There are two main adhesive families in general use, epoxies and acrylics, although experiments have been carried out with both hot melt resin systems and hybrid urethanes.
Epoxies were the commonest of the first surface mount adhesives, developed from resin systems already in use in hybrid circuit assembly. Their positive and negative features are listed below:
Acrylic adhesives are however strong competitors:
Hybrid urethane adhesives are tough, non-corrosive and adaptable. They can be chemically engineered to be suitable for pin transfer, syringe dispensing or printing, and can be cured by heat or by relatively safe soft radiation. Highly thixotropic, these adhesives have excellent dot definition and the ability to give higher dots to improve contact on components with a greater stand-off height.
Contact-pressure adhesives, which are active long after application, could be applied to the board whenever convenient before component assembly, even by the PCB manufacturer.
Hot melt systems, in which an applicator heats and melts adhesive granules, are not yet established in the electronics industry, but have the potential to provide highly raised discrete dots with minimal slumping.
Work is being carried out on other thermoplastic adhesives, which are already being used successfully to attach substrates to packages, housings and main boards, even where there is a substantial CTE mismatch. Since there is no chemical reaction during or after bonding, these adhesives can be used in hermetically sealed units.
Thermoplastic performance is generally comparable to the thermosets, but the feature which sets them apart is that they soften on reheating, which allows easy rework. This feature is especially important with die-bonding on complex multi-chip modules (MCMs). Solvent-removable thermoplastics can also be used to bond bare die temporarily for test and burn-in, with the bond ‘unglued’ with a safe solvent which leaves no residue.
The ability to control dispensing determines whether or not a material is considered to be a satisfactory surface mount adhesive. The dot profile produced by a particular adhesive will be defined by its thixotropic recovery rate, viscosity at zero shear rate and surface tension. But it will also depend on non-adhesive parameters such as the wetting characteristic of the PCB surfaces.
Other factors that influence good dispensability and dot shape control are:
In practice, machine operation settings have to be established by process trials. Unfortunately, such settings are not universal for every adhesive grade because of different rheological characteristics. Simple substitution of one adhesive grade for another is not always possible without some tuning of the dispense machine settings.
Experience has shown that dispense nozzle cleanliness is perhaps the most common source of dispensing problems. Dot size inconsistency can often be traced back to partial clogging of the nozzle. Adhesive build-up can occur inside the nozzle and restrict flow. Partial curing of adhesives in the nozzle may occur if it is left for long periods in a warm environment. Changes to different adhesive grades can cause cross-contamination and nozzle blockage. For these reasons, it is essential to clean nozzles regularly and thoroughly.
Most polymers are good insulators, with resistivities in the range 1010 W.cm for PVC to 1018 W.cm for PTFE (Figure 1), and this insulating property is a major reason why these materials are widely used throughout the electrical and electronic industries in applications such as insulators, dielectrics, and resists.
However, adding metallic fillers to polymers both makes them conduct electricity and enhances their thermal conductivity.
The conductivity achieved depends on the type of metal, the particle size and shape of the filler, and the percentage loading of filler.
There are three forms in which polymers are used in a conductive role:
The first two technologies were devised primarily for board to board connection, the third for component attachment. Of these, the last is by far the most common, having attracted much attention as a potential solder substitute both because of increased environmental concern about lead contamination and in response to the drive toward materials which are more fatigue resistant than solder.
However, during the 1970s, research began on the development of a new class of polymer materials that exhibited intrinsic conductivity without the need to incorporate metal fillers, and these are beginning to be used for specialist applications. The four sections which follow deal with each in turn.
Polymers are used widely in demountable connectors. Usually these are assemblies of pressed metal contacts retained in a thermoplastic body, but alternatives proposed for making an array of conductive connections between boards include inserting metal columns or plugs into an insulating adhesive film in the required positions (Figure 2).
The advantages are guaranteed isolation between columns and controlled separation between the substrates to be bonded. However, there are a number of drawbacks to this general approach as it requires additional design, fixturing and processing which will increase the cost of assembly, and accurate alignment will be required between the different layers to be bonded.
One attempt to circumvent these drawbacks is the Cinch Cin::Apse™ connector. In its most simplistic form, this consists of an insulating substrate in which are positioned randomly wound cylinders of wire 0.5mm or 1mm in diameter. These ‘button contacts’ (Figure 3) are designed to protrude slightly, so that when two metallised surfaces are pressed against the substrate, the contacts will deflect and form the electrical connection between the two. Depending on the application, these metallised surfaces might be pads on printed circuit boards or flex circuits, semiconductor packages or connectors.
The insulating substrate, which is generally a moulded thermoplastic material:
The button contact is made of a single strand of randomly formed wire (below), resembling a cylindrical ‘Brillo pad’. The randomness:
The contact has very low resistance and inductance, so is useable at microwave frequencies, and its very low mass and lack of resonance makes it resistant to shock and vibration.
Although the concept can be applied to a range of wire and plating materials, a common choice is gold-plated molybdenum, to withstand the military temperature and corrosion environment.
Another approach to making a conductive connection across a film only at the appropriate places is to hold large conductive particles within the polymer film, so that during bonding they become trapped between protruding pads on the two substrates, thus forming the required connection (Figure 4). This approach requires accurate alignment between the surfaces to be bonded, and precise process control to avoid a range of short and open joint failure possibilities.
This concept of ‘anisotropic conductive adhesives’ (ACAs) has been around since the 1950s, when Barrows developed dielectric coatings which contained randomly dispersed particles to provide electrical connections only in the vertical (‘Z-axis’) direction. There were other patents, but no real products until the 1980s, when Z-axis bonding films were found to provide the answer to the problem of connecting to Liquid Crystal Displays, temperature-sensitive glass parts whose conductors are thin vacuum-deposited coatings.
These ACAs disperse conductive particles in a non-conductive film or paste, keeping the particle loading low enough so that the individual particles do not contact one another to produce unwanted electrical paths in the X and Y plane. When the ACA is sandwiched between opposing conductors, the non-conductive polymer is squeezed out, allowing a monolayer of conductive particles to bridge the gap.
The conductive particles are typically <25µm in diameter so non-coplanarity must be very tightly controlled to ensure good contact (Figure 5).
after Gilleo 1996
ACAs perform best when one adherend is compliant, allowing the bonding pads to be forced into coplanarity with the opposing surface, and for this reason conventional ACAs have found favour with users of flexible printed circuits.
The coplanarity problem can potentially be reduced by building compressible conductive particles. A preferred method is to metal-coat elastomeric spheres, which can deform under bonding pressure, reducing the effects of non-coplanarity and also helping maintain force on the pressure contacts forming the junctions.
Many types of polymer matrix have been used: thermoplastics process faster, require less pressure, and can be repaired; thermosets usually require a post-bake after bonding for maximum performance, but are available as pastes for dispensing. A key factor is the polymer’s coefficient of expansion relative to that of the particles: if the polymer expands more than the conductor, the pressure contact connection can open.
Ongoing developments in ACAs include:
Isotropic Conductive Adhesives (ICAs), which conduct equally in all directions, are mixtures of a polymer material and a metal conductor in the form of fine powder. The components are thoroughly mixed, so that metal is spread evenly throughout the mixture. The composition and metal loading are designed to ensure conductivity throughout the mixture, the conductive path being formed by contact between the particles of the metal powder (Figure 6).
Adding metal also significantly increases thermal conductivity, although the values for silver and copper-loaded resins are several orders of magnitude below those of the metals themselves.
The polymer used is normally a one-part epoxy, and the mixture supplied uncured. However, the principle can be applied both to other thermoset materials and to thermoplastics.
The properties of metal-filled polymers vary with the percentage of metal, and this is limited by the amount which can be incorporated before the compound becomes unworkably viscous. As with solder paste, this will depend on the particle shape and size distribution and not just on the metal loading.
ICAs have been developed as an alternative to solder paste and are intended to fit into a normal SM assembly process. They are printed onto the component pad sites on the board, components are placed into the soft adhesive, and then the assembly passes into an oven to cure the adhesive and fix the components to the board.
Adhesives have been shown to perform well for SM component assembly and are capable of use even at the finest of pitches. However, whilst nearly every conductive adhesive forms a stable junction with precious metal-coated components, most are incompatible with the base metal finishes used on SMDs and boards. Most electrical failures occur in less than a week during the standard 85% RH/85°C humidity test. Mechanical failures also occur during temperature cycling, but researchers at DuPont suggested that mechanical and electrical performance are substantially independent.
One material tested by Alpha Metals showed greatly increased stability under heat and humidity ageing, the proposed mechanism for this being oxide penetration by small conductive particles as the adhesive hardens and shrinks (Figure 7). Further work indicates that the elastic modulus of the material is critical, and there appears to be a ‘window’ of values giving both good junction stability and adequate thermal cycle performance.
Apart from the reliability concern, rework is a major drawback– the polymer which bonds the metal powder to the PCB pads and component will not melt at the same low temperature at which it was cured. Even at relatively high temperatures, close to the maximum temperature which packaged components can withstand, the polymer softens but does not melt. Although removing the components is possible, cleaning the component legs or pad sites is a time-consuming and imperfect process.
after Gilleo 1996
Explain what options are available for replacing solder in making connections between a rigid board and a fine-pitch QFP integrated circuit.
Compare your answer with this one.
Although electrical cables are extremely diverse, with specifications to suit particular applications, simplistically they are arrays of polymer coated wires or co-axial cables within a sheath which gives mechanical and environmental protection:
Polyethylene and PVC are the main materials used for insulation and sheathing, often compounded with a small amount of carbon black to give UV stability. However, PVC has a higher relative permittivity due to its polar structure, and for many applications polyethylene is the ideal polymer. It has excellent electrical properties, is relatively cheap and easy to process, and has a wide range of physical properties which can be optimised by selecting the correct molecular weight and degree of branching. Linear low density polyethylenes, and their copolymers with vinyl acetate are also used to enhance resistance to stress cracking.
The flammability of cable is important, especially in critical applications or hazardous environments. PVC has been widely used, because its halogen content gives it relatively low flammability, and halogenated flame retardant additives are incorporated in other polymers to reduce flammability. However, once halogenated materials ignite, the smoke given off is frequently dense and toxic, and the gases emitted, such as hydrogen chloride from burning PVC, are corrosive and can seriously damage electrical and electronic equipment near the site of the fire. These problems have led to a much greater emphasis on the use of non-halogenated materials for low flammability constructions.
For cables, one approach is to incorporate a high proportion of fire retardant filler such as alumina trihydrate. Between 160–260°C, this decomposes, absorbing heat and releasing up to 30% of its weight as water. The filler suppresses ignition and flame spread by absorbing heat and excluding oxygen.
A parallel development has been in fire retardant and high temperature resistant ‘high performance’ polymers, such as polyetherether ketone (PEEK). These combine thermal stability with excellent all-round physical properties, but need a high processing temperature and are rather expensive, so tend to be reserved for more specialised applications.
Conformal coating is the process of coating the assembled PCB and components in a thin layer of protective ‘varnish’, which ‘conforms’ to the profile of the assembly. Mainly used in the assembly of systems for harsh environments, such as automotive, aerospace and military applications, these coatings protect the boards from a variety of environmental problems such as:
Vacuum-applied solder masks provide enough protection to the PCB in moderately aggressive environments, but fail to protect components and solder pads.
Although coatings would seem to be a solution to many potential problems, they do add difficulties of their own, and a conformal coating is generally applied only if the application demands it. The first issue is cost, as an extra process is involved, but there are also yield and reliability issues:
The coating type must be chosen to meet the requirements of the application such as solvent and chemical resistance, ease of application and the possibility of repair. A wide variety of resins is used, including acrylics, epoxies, silicones, and urethanes:
|more sensitive to the action of moisture||generally good moisture performance|
|most suitable for applications requiring protection against mechanical stress||offer the best protection against occasional high humidity|
|significant differences between products supplied by different manufacturers||no significant
differences between products from different
|function only to 125°C||work up to 200°C|
|best products for hardness/adhesion||good hardness and adhesion||worst adhesion||good adhesion|
|no improvement in insulation characteristics||improved insulation characteristics|
|marked Q reduction after immersion test||no effect on high frequency parameters||some Q reduction after immersion test||no effect on high frequency parameters|
|correct thermal shock performance, but lifting and darkening in moisture resistance tests||
pass thermal shock and moisture resistance test, but slight iscoloration
|no failures on moisture test||excellent moisture resistance|
|ageing caused severe discoloration||some pores caused by adhesion defects in thermal shock test||some small blisters on ageing|
|non-flammable or self- extinguishing||mostly non-flammable||non-flammable|
|difficult to repair; cannot be removed with a soldering iron||poorly repairable||easily repaired with a soldering iron||easily repaired with a soldering iron|
|not removable||solvent removable|
Conformal coatings can exert a hydraulic force between component and substrate during cure or subsequent thermal cycling, and this stress may be increased as a result of differences in the CTEs of the coating, components and board. This has been known to cause glass diode breakage in through-hole constructions, and in surface mount assemblies can fracture solder joints.
Test results on sensors have confirmed that conformal coating can produce permanent physical stress and that the level of stress is related to the thickness of the coating and differences in material CTEs. Effects on devices are most marked at low temperatures, and some materials demonstrate a hysteresis effect after temperature cycling.
Circuit failure has also been reported to be caused by moisture penetration to uncoated areas under surface mount devices. However, trying to remove this failure cause by increasing the coating thickness, so that it virtually encapsulates the on-board components, introduces further problems:
There are four methods by which conformal coatings of conventional
polymers are applied to assemblies:
The last three processes are all used in volume production, and their repeatability depends on controlling the viscosity of the material used and selecting material and process to reduce the extent of runs and slump in the coating.
The ultimate in thinness and evenness of coating is produced by vapour deposition, a process where a vapour condenses onto the boards to form an adherent coating. The material most frequently cited is Parylene™, whose reactive monomer (paraxylylene) polymerises onto a cold substrate to form even, adherent pinhole-free layers over a range of thickness from 0.1µm to >100µm. Unlike liquid resin materials, the coating is the same thickness at the edges or over protrusions as it is on flat surfaces or in corners. Also, because the deposition takes place at ambient temperature, no stresses are induced. The process is, however, restricted in its commercial application as it uses specialised vacuum deposition equipment and takes a substantial time – the 25µm typically applied on a PCB takes five hours to deposit onto only 3.5m2 of assembly surface area.
All the processes in this section can be grouped under the terms 'embedding' or 'encapsulation', where resin completely encloses the part to be protected.
In casting, a polymer material is heated so that it is sufficiently fluid, then poured into a mould, and cured without pressure. Curing is carried out at room or elevated temperature depending on the resin used. For heavily filled materials, and castings with small clearances around embedded components, there is always a danger of entrapping air, leaving voids in the casting. To minimise this, it is common to preheat the mould and embedded components, to ‘outgas’ the casting resin, to fill in several stages, and to apply a vacuum to the filled mould while the resin is still fluid.
Advantages of casting are low mould cost, its ability to produce large parts with thick sections, the good surface finish on parts, and the fact that few finishing operations are required. The disadvantages are that the process is limited to simple shapes and is slow. Most casting polymers are thermosets, although thermoplastics such as nylon have been used.
Potting is very similar to casting, except that in the potting operation the mould remains a part of the final product. Ranges of standard moulds are available, typically as ‘empty nylon boxes’, and this approach is common for small quantity production of modules which need environmental protection.
Impregnation fills the interstices of a component (such as a coil or motor windings) with a low-viscosity resin system in order to consolidate the structure.
All the processes above are restricted to liquid resin systems which are ‘100% solids’, with no solvents present. These can be one-part or two-part materials, handled as described in the next sub-section.
In dip encapsulation, the part is completely covered by dipping in resin, a process which works well with components which are irregular in shape. A typical application is for the ‘tantalum bead’ axial capacitor. The dipping process is normally carried out with ‘100% solids’ liquid epoxy resin systems, but phenolic resin systems with a solvent carrier (such as DurezÔTM) have also been used. A significant limitation of the process is the length of time for which the resin bath is exposed to air.
Potting is a method of protecting assemblies by filling small spaces or entire surfaces with a material that will protect components from physical and environmental elements. Potting compounds can be made in a range of viscosities, depending on application. Filling small gaps requires less viscous material than filling large cases.
Potting involves adding a different material with yet another coefficient of thermal expansion, and can therefore add more stresses to the assembly. In addition, the potting process itself, which usually involves dispensing the polymer, cannot be relied upon to provide 100% fill or repeatable results, even when the casting precautions given above are taken. Therefore potting tends to be used only in the cases where it is absolutely essential for environmental reasons.
Your latest product includes a number of identical high-voltage circuit elements, mounted on a printed circuit board, and is intended for use in a country with consistent high temperature and high humidity. How might you use polymers to help the product survive its environment? And what would be the implications for its manufacture?
Hint! Don’t restrict yourself to thinking about just a single way of helping the product survive.
Just a reminder for you to look again at an assembly, and see if you can spot other applications for polymers. We don’t offer a prize for the longest list, but would certainly like to receive your comments and ideas!