VOC Considerations in Printed Board Fabrication Process Chemistry

On November 15 1990 President George Bush signed the Clean Air Act into law, legally restricting the output of selected air pollutants by cities and industry. The Act addresses multiple sources and types of air pollution, but for the purposes of this discussion, we will deal with the “VOC” [volatile organic compound] component of “ground level ozone”, resulting from the material choices and methods used in Printed Board fabrication. “Ground-level ozone (O3) is the major component of smog. Ozone is not emitted directly into the air, but is formed through complex chemical reactions between precursor emissions of volatile organic compounds (VOC) and nitrogen oxides (NOx) in the presence of sunlight.”

A simplified schematic of the reaction follows:

This presentation is the direct result of an occurrence at one of our customers that illustrated several widely misunderstood aspects of VOC regulations, measurement and calculation, and operational considerations. These misunderstandings potentially can result in suboptimal process choices and/or excessive costs. We felt that a brief review in open forum would benefit all of us in the industry.

Impact of Regulation and Response Alternatives on Operations

The Clean Air Act and the accompanying regulations provide for variations in both the degree and implementation schedule required depending on the degree of “attainment” (i.e. the degree of compliance with the targeted levels of ground level ozone), as illustrated below.

Areas that are likely to be classified as Extreme, Severe, or Serious as of late-1990 are presented below.

In addition, there is the usual provision allowing local governmental entities to apply more (but not less) stringent limits on emissions.

Most process areas of printed wiring board fabrication operations have potential contributors to VOC emission (various film, surface, and screen cleaners, fluxes, etc.). This illustration is limited to the photoresist stripping operation.

Photoresist stripping concentrates (proprietary variations on the “aqueous” stripping chemistries) vary from near 0#/gallon VOC (on a formulary basis) for ~100% inorganic [caustic] formulations to over 6#/gallon for 100% organic [amines, stabilizers, surfactants, etc.]. The actual reportable VOC content for such concentrates is most commonly measured using a method described in 40 CFR Part 60 Method 24, generally shortened in conversation to “Method 24“. This is the value typically reported on the MSDS for the concentrated material.

This method is essentially a weight loss determination at specified time and temperature, mathematically deducting the water content (most properly determined using a Karl-Fischer titration) and assuming that the remainder of the weight loss is due to VOC content. Several factors can result in a difference between the actual (formulary) VOC content and the reportable (by analysis) VOC content, including interactions between constituents, volume/surface area ratio of the glassware selected, etc.

The actual VOC emissions required to be reported must be determined by examining the air discharge permit at the location in question. The simplest (though neither the most technically correct nor in many cases, the wisest) method is to assume that ALL the VOCs contained in the incoming concentrate are emitted as reportable VOC’s. This is the worst-case scenario, and can result in reporting quantities many times higher than actual emissions.

There are at least two factors contributing to these possible discrepancies:

1. Partial-pressure effects on volatilization, and
2. Distribution of VOC constituents in non-volatile matrices/non-reportable waste streams.

Partial pressure calculations can be used to estimate the equilibrium concentration of a material split between the liquid and vapor phase. Perhaps more importantly (from a practical standpoint) is the distribution of the organic fractions of a process chemistry between the various input and output streams of a process.

The quantity of VOC emissions to be reported can be very different depending on how the permit is structured, and depending on how detailed an analysis the user is prepared to undertake. If no reasonably well-documented accounting for the various exit streams is documented, it will be
extremely difficult to justify any deviation from the simplest assumption, i.e. the gross-pounds-of-VOC-content imported to the operation equals the reportable VOC emission. At the very least, the operator must work closely with the local air-emissions regulatory body (in the US, the EPA or it’s local designee) to structure an acceptable means of accounting for actual or calculated losses.

Responding to a Process VOC Challenge: A Case Study

In the specific case that inspired this discussion, a high volume commercial operation was faced with a requirement to significantly reduce the reported quantity of VOC emissions from their photoresist stripping operation, after an audit by an environmental consulting firm. The production operation was happy with the operational performance of the chemistry in use, but a less functionally desirable alternative chemistry was available. A sample (of unknown pedigree) of the alternative chemistry was analyzed by the contractor, and was reported to be much lower in VOC content than the incumbent product. The printed boards in production at this facility were primarily SMOBC, using tin as the etch resist, and a signficant proportion started with two ounce or greater foil. In this operation, that meant multiple passes through the entire strip-etch-strip line for the heavier copper panels, to avoid line purges and the attendant production interruptions.

Process changes were not acceptable, for capital and training reasons. The immediate request from the operator was for “…a product that works just like the one we’re using, just lower the VOC content…” Reportedly, no cost increases would be tolerated. In descending order of priority were yield, speed, cost per gallon, copper finish, and finally, loading. Simplicity and consistency in operation were more important than chemical “efficiency” to this operation.

In the classic custom formulation approach, this would suggest substitution of inorganics for some portion of the organic sources of alkalinity, but it was soon determined that the tin deposit/registration issues interactions were insufficiently robust to tolerate multiple passes through a higher-causticity resist stripper without increases in etchouts. This was shown to be true despite incorporation of several alternative metal protection chemistry “packages“, within the “no cost increase” constraint.

Given the emphasis on speed and simplicity of control, and with some careful balancing of the various amine species for optimized results, it was possible to formulate an all-organic resist stripper meeting their VOC target, while maintaining speed, finish, and yield. The trade-off was loading capacity, which (while it might not have been detectable in the operation as normally configured) was unacceptable to us in terms of “value for money“. In reporting this to them, we reiterated the possibility of determining actual VOC emissions (vs. the simple “total VOC in = reported emissions“, worst case approach). On this hearing, and now realizing more clearly the trade-offs involved, a cooperative approach involving stack sampling by the environmental consulting firm, stabilized (baseline) operation by the operator, and chemical monitoring by the supplier was agreed upon.

This is not without a cost, as the stack monitoring and analysis, and calculation of fugitive emissions will be time (read “money“) consuming for all parties. In this case, the benefits of the possibility of continued operation with a well-understood, high-performance chemistry outweighed the costs associated with the analysis. In another situation (for example, an operation where the scrubber was inadequate, or a smaller operation), it might have required a chemistry or procedural change.

Future Considerations

The possibility of enforcement “speed-up” or “stretch-out” in various regions as attainment levels are met, compliance targets are changed, or the political climate varies always exists. It is reasonable to assume that geographic areas not currently emphasizing VOC emission targets will do so in the future.
As to future regulation, who can tell?

Conclusion

It is possible to meet stringent VOC content targets by customized formulation, optimizing for a specific location’s requirements. However, consideration of the “big picture,” and open, cooperative communication between all parties is needed to achieve the “best” balance between operational, environmental, and cost considerations.