Abstract

Advancements and evolutions in printed circuit board manufacturing, design, and electronics assembly have driven new research on high temperature organic solderability preservative (HT OSP) surface finishes. More specifically, developments in OSP chemical processes are aimed at producing adurable finish which ensures that a board surface retains solderability through more challenging and harsh Pb-free assembly conditions. From this, it is clear that advancements in OSP processing and coating performance require a solid understanding of the mechanisms associated with coating formation and thermally driven degradation. This work examines and describes OSP structure and composition and how it is affected by heat treatments. Additionally, mechanisms of degradation of OSP are proposed along with possible strategies to remedy it.

Introduction

Organic solderability preservative (OSP) is one of the most common final finishes in the PCB industry today. In addition to its ever attractive characteristics of low cost and a relative ease of application,perceived or real issues with other final finishes contribute to OSP’s appeal. For some fabricators OSP is the preferred final finish.

However, Pb-free processing is a challenge for OSP. Not only assemblers have been forced to optimize their soldering processes but also OSP vendors have reformulated their products in order to meet the challenge as well. In this paper, we attempt to examine the attributes and makings of a good high temperature (HT) OSP process and coating.

There are several of these attributes that one may wish to consider and several technical ideas or claims have been made. Under discussion are: type of azole, which is the primary chemical constituent of OSP, its volatility, its decomposition temperature, existence of an organometallic polymer between the azoleand a metal other than copper. It may be hard to see how these correlate with better solderability. On the other hand, we have found out that some other properties and phenomena occurring in the OSP coating maybe of greater importance.

OSP Coating Depth Profile – Before Heat Treatments

Enlarge this picture Figure 1. Profiles of C,O and Cu Concentrations Within As Coated OSP Layer Obtained by Dynamic SIMS.

It is of fundamental importance to understand how OSP coating is constructed. We carried out depth profiling using Dynamic SIMS measurements of as obtained OSP coated surface looking at concentrations of carbon, oxygen and copper. Oxygen was of interest because it is reasonable to assume that the Cu^I/Cu^IIlayer contained oxide resulting from rinsing and other pretreatments in the process prior to the actual OSP coating. It is not clear whether it is completely transformed into a Cu^I/Cu^II-azole layer or some oxide is retained at the interface.

The graph in Figure 1 shows that:

1. 1) There is a noticeable O rich layer at the Cu-OSP interface; this could be from Cu^I/Cu^II oxide or from an O richer additive deposited at that interface,
2. 2) The Cu concentration is increasing with depth and therefore is not the same throughout. This isimportant to note if one recalls schematic representations of OSP structure as a network of Cu andazole molecules in 1:1 ratio ( which assumes same Cu concentration throughout the coating).

The OSP thickness from this measurement is (taken as 1/10 of the maximum C signal) 0.31 microns (310nm).

OSP Coating Depth Profile – After Heat Treatments

Enlarge this picture Figure 2. Profiles of C,O and Cu Concentrations Within OSP Layer Reflowed Three Times Obtainedby Dynamic SIMS

We then repeated the same measurements using substrates that were reflowed three times in airusing a Pb-free reflow profile.

The graph in figure 2 shows that:

1. 1) The oxygen peak at the interface of metallic Cu and the organic layer (~0.22 μm) is bigger; anestimated increase in O concentration is 2 times,
2. 2) The values of Cu concentrations within the organic layer are higher and almost steady, rising onlyvery slowly with depth,
3. 3) The layer appears better defined and more compact; the OSP thickness after reflows (taken as 1/10of the maximum C signal) is 0.25 microns (250 nm).

Heat Treatments of OSP – What Happens?

Let us examine processes and phenomena that can occur when OSP coated Cu surface are heated during assembly.

1. 1) Azole Volatilization. As previously reported, we see only a small % loss of azole (less than 10%) during reflows. With appropriate safety margin for thickness in OSP application, this is not a problem.
2. 2) OSP (Azole) Decomposition. Again, as previously reported, we have not seen decomposition at temperatures below 260 ºC. In fact, a TG curve showed a clean transport into a vapor phase without exothermic events. Others reported results of thermal measurements suggesting a correlation between decomposition temperature and quality of an OSP coating. This correlation is unclear, especially if reported temperatures are well above 260 ºC, therefore not encountered in assembly and therefore irrelevant.
3. 3) Reaction with Oxygen. It is clear that oxygen has a negative impact. Samples processed under nitrogen solder better than those processed in air. It is not entirely clear though what is the responsible mechanism: oxidation of azole, oxidation of metallic Cu under the organic layer or other reactions with OSP components including ionic Cu scattered in it. We have seen the evidence of some degradation of our azole heated in air (color change) but analytical techniques such as NMR and UV/vis hardly indicated any change. Oxidation of metallic Cu seemed a more detrimental process especially in light of reports that OSP coatings are permeable to some extent to oxygen and our previous report which appeared to confirm this.
4. 4) “Hardening” of OSP. It may be impossible to isolate physical, heat induced changes within OSP, from the changes resulting from oxygen impact. We attempted no-flux soldering of panels with through holes which were heated under nitrogen. No hole fill was observed indicating that heat alone (and not oxidation) was detrimental. We clearly see that OSP coatings reflowed in air are not strippable with organic solvents, and this may be caused by migrated ionic Cu, which “hardens”OSP. One can, however, speculate that azole molecules reorient themselves during heat excursions increasing intermolecular interactions between azole aromatic rings rendering the organic layer less soluble. Interestingly, if the organic layer is less soluble in organic solvents, it can be expected that fluxes used in soldering are less effective. Looking from a different angle, the role of fluxes in soldering becomes more important.

We believe that the first two processes, azole volatilization and decomposition, are insignificant. The latter two processes, oxygen permeation and oxidation of Cu metal, however, are important. The difficulty in soldering OSP coated boards may be the result of oxide growth under the OSP, which is not readily accessible by flux that should normally remove the oxidation. Moreover, thermal excursions cause hardening of the organic layer, which is then more difficult to penetrate by flux.

Enlarge this picture The scheme above illustrates the phenomena that occur within the OSP coating as revealed by the measurements

Summary - Strategies for Improving Solderability

Our results and observations suggest that solderability of OSP coated boards after heat excursions may be more difficult due to organic layer hardening and copper migration. These two phenomena suggest more specific strategies for chemical formulators:

1. 1) Azole Selection. Within the families of imidazole and benzimidazole derivatives some compounds may be better. Of course, one needs to consider properties that make a compound suitable for formulation of an OSP bath, such as solubility, coverage, process window etc. In addition to these, however, some compounds may be less prone to hardening during heat treatments. For example,molecules of lower symmetry with geometrically obstructing substituents may undergo less compacting upon heat treatments.
2. 2) Additives. These may play particularly useful roles. They can a) limit oxygen permeation through the organic layer, b) limit migration of Cu ions and c) limit compacting of OSP. We can thus think of additives as playing roles of oxygen scavengers and fillers. Alternatively, we can select additives based on their known or perceived direct impact on solderability/wetting.
3. 3) Synergize with Flux. If flux is expected to dissolve OSP during soldering it may be useful to formulate OSP to work better with fluxes and optimize fluxes for OSP. It has been suggested that fluxing is very important for OSP and that some flux formulas may be more optimal for OSP.

Acknowledgements: The authors wish to thank Evans Analytical Group LLC in East Windsor, NJ for conducting the Dynamic SIMS measurements.