Previous Tech Talk articles^1,2 looked at the limitation of forming fine line circuits due to undesirable lateral etching (undercut), not just in the print and etch process but also in pattern plating, panel plate/tent and etch, and even in semi-additive processing. It appears that neither photoresist resolution nor plating capabilities but rather etching presents the ultimate limit for these traditional circuitizing techniques. This column looks at ways to extend the subtractive technology to achieve finer features. Some technical advances described here have found commercial acceptance whereas others don’t appear to have gained traction.

There are several problem areas associated with etching, some interrelated, some independent (see Figure 1). Benefits from process remedies may be seen in one or more of these problem areas. The first problem area, shown in Figure 1, is addressed here in Part A, and the latter three will be dealt with in Part B.

Addressing etch undercut first (see Figure 2), we have to acknowledge that wet etching is basically isotropic, and all techniques we apply to promote downward etching do not completely eliminate lateral etching.

Enlarge this picture Figure 2 Approaches to Reduce Lateral Etching

LP-Chemie’s impulse process³ creates spray pulses of etchant droplets with high vertical impingement force onto the copper surface that easily penetrate an etch inhibitor layer (banking agent) at the bottom of the etch channel but leave the inhibitor layer on the etch channel sidewalls largely undisturbed. This technique results in enhanced downward etching (i.e., a more favorable etch factor) and higher etch speed. The pulses are generated mechanically by the rapid rotation of a coaxial pipe sleeve with slots around the spray bar. This process is combined with Pill’s top side suction bar approach that addresses top-to-bottom etch non-uniformity due to puddling.⁴

It has been reported that controlling acid etchants at very low free acid normality improves the etch factor.⁵ These findings are in line with earlier work that showed that using NaCl instead of HCl as a source of chloride ions in acid etchers, to complex copper salts and keep them from precipitating, results in an improved etch factor.

Enlarge this picture Figure 3 Relationship Between Etch Factor, Etch Channel Width, and Resist Thickness for Three Etchants (Source: Oak Mitsui)

Different etch chemistries result in slightly different etch factors. Alkaline etching typically yields a less favorable etch factor than acid etchant. Some studies show a better etch factor with ferric chloride than with cupric chloride, but results are not always conclusive. The study of T. Yamamoto et al.6 (see Figure 3) shows the beneficial effect of wider etch channels and thinner resist but gives no clear distinction between ferric and cupric chloride.

The study in Reference 6 has shown that the etch factor deteriorates as the etch channel aspect ratio increases and as the absolute dimension of the etch channel width shrinks. This suggests that wherever design and image transfer method allow a change to thinner copper and/or thinner resist, one should move in that direction to improve selective downward etching. A special application of this approach can be found in panel plate/tent and etch. In the classic tent and etch process, a relatively thick dry-film resist is used to protect with a tent the through-hole copper from etchant attack. This thick resist adds height to the aspect ratio of the etch channels, which is known to result in a less desirable etch factor, thus limiting fine line etching. The use of a much thinner etch resist that coats the plated through-hole instead of protecting it with a thick tent will improve the etch factor. For example, a positive ED resist coating is much thinner than the tenting film, enabling finer lines on a panel plated board. Positive resist is preferred over negative working resist in this application because no radiation needs to reach the ED resist in the hole during exposure to form a coating resistant to the developer chemistry. Alternatively, the thin coating could be immersion tin, as in the Siemens process. The tin is selectively laser ablated from the copper surface where copper will be etched. Similarly, a process, developed by Atotech, uses a laser-ablatable non-photodefinable organic coating for this application.

References

Advertisement

1.    Dietz., K. H., “Fine Lines in High Yields (Part LXXXII): Fighting the Etch Factor and Etch Non-Uniformity,” CircuiTree, July 2002.
2.    Dietz., K. H., “Fine Lines in High Yields (Part CXXV): Fine Lines – Beyond the Limits of Semi-additive Processing?,” CircuiTree, February 2006.
3.    Feyerabend, V., “Kombinierte Verfahren im Aetzprozess,“ PLUS (Produktion von Leiterplatten und Systemen), October 2007.
4.    Feyerabend , V., “Vakuum-Aetz-Technologie, ” PLUS (Produktion von Leiterplatten und Systemen), August 2007.
5.    www.oxfordvue.com
6.    Yamamoto, T., Kataoka, T., & Andresakis, J., “Allowable Copper Thickness for Fine-Pitch Patterns Formed By a Subtractive Method,” CircuiTree, June 2000.