Doug
On the whole, I agree with you but would like to add a couple of kopecks
worth.
Ion chromatography, as well as ionic contamination, testing also depends
on tha ability to extract the ions out of the laminate. There is also a
risk of extracting the NaCl which is inherently a component of the epoxy
resin, whether you like it or not. There is no way you can differentiate
between the chloride sources.
Some of the HASL fluxes have replaced hydrochloride activators with
hydrobromide ones. These are apprantly more prone to be a cause of
conductive anodic filaments (Laura Turbini dixit). Why this should be I
have not understood. The bromine atom is alrger than the chlorine one,
so are the halide molecules and ions. I would therefore tend to assume
that absorption of the bromides into an epoxy resin matrix would be
less, hence nearer the surface, where there would be a better chance of
being able to leach them out again during the washing process. Bromine
also has a lower electronic bonding force to other atoms than chlorine,
so would dissociate more readily from valency-bonded (ionic) molecules,
as I understand it. I must admit that I am a little puzzled about the
differences.
I'm also doubtful about the role of the glass reinforcement and its
silane treatment (plus the glass>silane and silane>epoxy bonding
strengths). In a past life, I have taken some 7628 glass cloth and
measured the ionic contamination on it before resin impregnation. It was
small, but very finite (can't remember the actual figures but I think it
was just under 0.1 ug/cm2 eq. NaCl). I did not determine whether the
ionics came from the glass or the silane.
Finally, I don't know whether CSL (or anyone else) have ever determined
the actual absorption by microanalyses at different Z-axis depths from
the laminate surface and different X-Y-axes distances from a drilled
hole after HASL (theoretically, the latter should be zero if the copper
is watertight, but why does Laura claim that bromide HASL fluxes give
different results to chloride?)
I think we still have a lot to learn about the physical and chemical
mechanisms of laminates from cradle to grave. I forecast that, as we
move away from tetrabromobisphenol A as the flame retardant of FR-4 or
equivalent boards, we are going to see a lot more problems of this
nature arise.
Brian
Douglas Pauls wrote:
>
> *<sigh> I always try to avoid responding to these questions, but it draws
> me like a moth the the flame......
>
> refer to IVF Research Publication 96846 - Cleanliness and Reliability:
> Evaluation of Test Methods and the Impact on Contamination from Production
> Processes on the Reliability of Printed Circuit Board Assemblies - Dr
> Per-Erik Tegenhal 1996
>
> He states:
>
> "It was demonstrated that the accuracy of solvent extract conductivity
> measurements, i.e. the most commonly used method for cleanliness
> measurements, is very poor. In one case, it was found that less than one
> percent of ionic contaminants present on a hot solder levelled PCB of FR-4
> type was detected by a commercial tester. The main reason for the poor
> accuracy is that flux residues are absorbed into the epoxy resin during the
> soldering process. Therefore it is very difficult to clean FR-4 boards but
> also to analyze the cleanliness of such boards. The use of equivalence
> factors for various testers is not a feasible way to handle this problem."
>
> *I agree with Dr. Tegehall in that the repeatability and reproducibility of
> these automated ROSE or SEC testers is very poor. The IPC ionic
> conductivity task group, in conjunction with the EMPF proved this in the
> early 90s. IPC has the 200+ page report if desired. So does EMPF. I
> would disagree that the reason you don't see the flux is that it completely
> absorbs into the laminate. Some FR-4 laminates will absorb flux. Some
> solder masks will. Others don't. The better the chemical resistance of
> the laminate and mask, the more it will resist absorption of flux. The
> main reason that you don't get all the flux detected is a solubility, time
> and temperature phenomenon. ROSE and SEC are extraction tests. If you
> can't dissolve the contaminant, you can't detect it. Room temperature
> IPA/water will extract a certain amount. The ROSE testers that use heated
> solutions get a higher value because hot IPA/water and turbulent actions
> tend to increase the extraction into solution. It is not really difficult
> to clean an FR-4 board, we have been doing so for years. But it is
> difficult to get an accurate measure of cleanliness and even more difficult
> to determine how clean is clean enough.
>
> If the 1% of ionic contaminants are removed from the surface of the board
> then what is to stop the highly mobile chlorides from migrating out of the
> "spongy" FR-4 substrate after cleaning in the direction of the newly
> established concentration gradient and eventually reaching pre-clean
> contamination levels? What effect does this have on reliability?
>
> *If you DO have absorbed flux in the laminate and mask at appreciable
> levels, then you can get electrical leakage under humid conditions,
> electrolytic corrosion, conductive anodic filament growth, and
> electrochemical migration. You probably won't ever get to pre-clean
> conditions. The thermal processes of soldering have an effect on the
> materials, so you can never truly get back to your starting condition.
>
> Logically then, taking chlorides for example, the reliability of the PCB
> will NOT correlate with the concentration of the surface chloride levels,
> but with the relative humidity and "wetting" of the PCB surface. A "wet"
> surface would allow the chloride to migrate out of the substrate,
> compromise the SIR and this may correlate more highly to PCB failures.
>
> What levels of chlorides have been found in and around the solder pads of
> HASL/FR-4 type boards?
>
> *You need to keep in mind the difference between chloride residues on the
> board, and the results of the ROSE test. When the ROSE test gives you a
> result, it is in micrograms of sodium chloride equivalence per unit area.
> This is NOT the amount of chloride or sodium you have on your board. If
> you want to learn more about this, look in the recently published J-STD-001
> Handbook, where I give a full run down on why this is so. To answer your
> latter question about chlorides on HASL boards, CSL has done extensive work
> in that area using ion chromatography as the measurement tool. Much of
> that work was published in the monthly Circuits Assembly magazine column,
> Process Rx, between Sept 96 and Feb 01. You can also access that through
> the CSL web site at www.residues.com.
>
> In this situation. If the spot levels of chlorides in and around the solder
> pads of HASL/FR-4 boards is high (how high I do not know), what then is the
> relative contribution to failure rates by environmental chloride
> contamination? Environmental contamination may only contribute a fraction
> of the total ionic contamination found in an around the solder pads.
>
> *Chlorides are a nasty electroactive material and are one of the primary
> contributors to electronics failures. If high, you can see the
> electrochemical failures I listed above. Environmental chloride can be an
> issue, depending on whether you conformal coat the boards and how excessive
> the chloride can be. It all depends.
>
> Doug Pauls
> Rockwell Collins
>
> ---------------------------------------------------------------------------------
> Technet Mail List provided as a free service by IPC using LISTSERV 1.8d
> To unsubscribe, send a message to [log in to unmask] with following text in
> the BODY (NOT the subject field): SIGNOFF Technet
> To temporarily halt delivery of Technet send the following message: SET Technet NOMAIL
> Search previous postings at: www.ipc.org > On-Line Resources & Databases > E-mail Archives
> Please visit IPC web site (http://www.ipc.org/html/forum.htm) for additional
> information, or contact Keach Sasamori at [log in to unmask] or 847-509-9700 ext.5315
> ---------------------------------------------------------------------------------
---------------------------------------------------------------------------------
Technet Mail List provided as a free service by IPC using LISTSERV 1.8d
To unsubscribe, send a message to [log in to unmask] with following text in
the BODY (NOT the subject field): SIGNOFF Technet
To temporarily halt delivery of Technet send the following message: SET Technet NOMAIL
Search previous postings at: www.ipc.org > On-Line Resources & Databases > E-mail Archives
Please visit IPC web site (http://www.ipc.org/html/forum.htm) for additional
information, or contact Keach Sasamori at [log in to unmask] or 847-509-9700 ext.5315
---------------------------------------------------------------------------------
|
