ICT/SMART Group Evening Seminar - "Solder Mask Application Techniques Explained". (Full Review).
21st November 2001
Arundel continues to be a consistently popular venue for evening technical symposia, and the joint ICT-Smart Group event on 21st November gave an audience of about thirty people the opportunity to get up-to-date on solder mask application techniques. Expert presentations from Shaun Tibbals and Ian Gardner described the present state of the art, explained the advantages and disadvantages of methods in current use, and offered a privileged glimpse into the future at the emergence of techniques based on ink-jet printing.
Shaun Tibbals, Technical Support Manager of Electra Polymers, quoted statistics which indicated that the use of screen-imaged solder mask has now declined to less than 10% of world-wide consumption, and that liquid photoimageable now represented 90% of the market. Although pure epoxy chemistry had characterised the first generation of the LPI market, epoxy-acrylate systems had now become the industry standard, combining fast exposure with aqueous developing to give improved resolution and durability. Shaun's presentation focused on these materials, and made a remarkably comprehensive comparison of the effects of different application techniques on their processing characteristics and their ability to satisfy performance requirements.
Four main coating methods were in current use, flood-screen printing being the longest established, followed by curtain coating and electrostatic spray coating, and air-spray coating had more recently emerged as a popular technique.
Flood-screen printing was well characterised, and gave some opportunity to control thickness by mesh count and squeegee settings. It could be carried out as a manual, semi-automatic or fully automatic operation in single-sided or double-sided, horizontal or vertical mode. Screen printing used a high-viscosity ink and gave good coverage on track heights up to 100 microns, above which it might be necessary to double-coat in order to achieve acceptable thickness and avoid skips between tracks.
Curtain coating used an ink of substantially reduced viscosity compared with screen printing. Passing the work through a falling liquid curtain gave uniform coverage and very little wastage, but the technique was essentially single-sided and horizontal. On work with high tracks, the low ink viscosity could result in flowing off track edges and difficulty in maintaining adequate thickness.
Electrostatic spray coating relied on a low-viscosity ink being pumped to a bell rotating at speeds up to 50,000 rpm, the ink being atomised and the droplets receiving a negative charge which caused them to be attracted to the earthed circuit board. The method could be used in horizontal, vertical, single-sided or double-sided configurations. Electrostatic spraying could suffer from some distribution effects related to conductivity, large areas of copper tending to thieve solder mask from narrow tracks, and some areas of bare laminate might not coat. The equipment was expensive and there were few installations.
Air spraying was a relatively straightforward technique, and machines specifically designed for coating LPI had been used since the mid-1990's. Conventional spray guns were employed, generally with heated air to compensate for the cooling effects of atomisation and solvent evaporation. Machines could have single or multiple guns, and could operate horizontally or vertically, single or double-sided The set-up of multiple-gun systems was critical to avoid stripes in the coating, and the ink required to be specially formulated to atomise easily, or else mottling effects or poor track-edge coverage could be experienced. Under optimum conditions, tracks as high as 400 microns could be successfully coated, and 100 micron gaps successfully filled.
Shaun emphasised that the rheological characteristics of the ink were fundamental to successful coating, and that optimum characteristics were widely different for each coating process, although certain compromises had to accepted. For example, a high level of thixotropy was desirable for successful air spray coating, but could potentially have an adverse effect on the efficiency of atomisation; dilution of the ink could improve atomisation at the expense of track edge coverage and transfer efficiency.
The coating technique needed to be considered in close conjunction with the drying process. The objective of tack drying was removal of coating solvent whilst avoiding activation of the thermal curing chemistry. Double-sided coating processes were to be preferred as they allowed single-stage drying, without the side-to-side bias typically associated with curtain coating. Solvent content and thickness distribution of the different processes had significant effects on drying time. Curtain coating was by far the slowest, screen printing the fastest, and electrostatic and air spray somewhere in between. A further consideration was the extent to which different processes allowed ink to get into holes, which had significant consequences both in the drying process and subsequently in the developing process.
High resolution imaging was favoured by the spray coating techniques, since these tended to yield a thinner deposit on the substrate adjacent to conductors and hence require less UV energy to expose and be less susceptible to undercut by developer. A peculiarity of the developing process for LPI was that the primary objective was to remove ink from holes, and this could compromise optimum development of the surface pattern. The air spray process caused least problem. Control tests had demonstrated that, other factors being equal, an air sprayed coating required one quarter of the exposure energy and one third of the developing time to achieve clean 0.2mm holes and 50 micron solder dams, compared with the equivalent screen printed coating.
Reviewing the numerous performance demands placed on liquid photoimaging solder masks, Shaun demonstrated how key success factors could be evaluated against the attributes of different methods of application, and concluded that although an application method could not satisfy all the demands put on an LPI, careful selection of coating technology could open the process window, improve quality and reduce costs, and that air-spray application offered significant processing and end-user advantages.
The second paper of the evening introduced the concept of piezoelectric ink-jet printing of solder mask materials. Inkjet printing had become a familiar imaging technique, particularly as a consequence of the popularity of the ubiquitous desk-top machine, and had enjoyed spectacular success in the graphic arts industry. There had been much discussion and speculation in recent years as to the scope for applying ink-jet printing to printed circuit manufacturing technology.
Ian Gardner, Technical Director of Jetmask explained how his company had taken the initiative to become the ink-jet pioneers. Having already commercialised ink-jet as a means of direct imaging of legends, idents and serial numbers onto printed circuits, Jetmask and were at an advanced stage of development of the ink-jet technique for application of solder mask.
Compared with conventional methods of solder mask application and imaging, in By working direct from data, ink-jet eliminated the need to prepare photo-tools or stencils and enabled very fast set-up, with the flexibility to change-over instantly between designs and the freedom to optimise registration by adjusting the scale and orientation of the image in software.
Ian went on to explain the mechanism of the drop-on-demand ink-jet technique, and the action of a printing head, which worked by ejecting a droplet of ink each time a voltage pulse was applied to the cavity associated with an individual nozzle, the piezoelectric effect causing a mechanical contraction of the walls of the cavity. Typical printing heads being evaluated for solder mask application had 500 individual nozzles, arranged at a pitch of 180 per inch, each capable of delivering 4 billion droplets of ink during its working life. The limiting factor in generation of droplets from an individual nozzle was the finite time taken for the cavity to re-charge itself with ink after ejecting the last droplet, and for the print-head overall, the capacity of the raster image processor to produce the driver output from the image data.
Increased resolution required more drops per linear inch and multiple passes to interlace successive image columns. Cycle times could be reduced by utilising multiple print heads. Using a series of dot images of increasing density overlaid on a surface-mount footprint, Ian demonstrated that to achieve acceptable image definition, a resolution of at least 720 dpi was necessary.
Solder mask inks were being developed with specific fluid properties to suit ink-jet application, whilst retaining the material properties necessary to satisfy performance specifications. The spreading characteristics of the ink once the droplets had been deposited on the surface were particularly significant, and to maintain the balance between coalescence of adjacent droplets and sharp edge-definition of image features a method was required of arresting the flow very rapidly after deposition. UV-curing materials had been formulated which enabled the print-to-cure time to be precisely controlled within hundredths of a second, and Ian showed a series of photographs which illustrated how the parameters had been optimised. The solder masks in development were single-pack 100% polymerisable formulations which contained no volatile organic compounds.
Jetmask's development programme had characterised factors such as drive pulse waveshape and printhead height, which governed droplet velocity, and had studied the physical interactions between solder mask ink and substrate surface from such aspects as surface tension, viscosity, liquid density, solid surface energy, surface morphology and surface topography.
Reviewing the position of ink-jet printing in the printed circuit industry, Ian observed that it was a disruptive technology which was causing fabricators to reconsider their traditional perception of imaging processes for solder mask, and interest in ink-jet was growing rapidly. Strategic partnerships between ink-jet equipment manufacturers and solder mask suppliers were driving development such that the technology would soon become commercially available.
Pete Starkey
ICT Council
SMART Group Technical Committee