TECHNET Archives

June 2013

TechNet@IPC.ORG

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
Laura J Turbini <[log in to unmask]>
Reply To:
TechNet E-Mail Forum <[log in to unmask]>, Laura J Turbini <[log in to unmask]>
Date:
Sun, 23 Jun 2013 17:03:06 -0400
Content-Type:
text/plain
Parts/Attachments:
text/plain (183 lines)
Hi Brian,
I personally believe that the glycols, and the bromideions diffuse into the epoxy during the soldering process.  When the epoxy goes above its Tg it opens up its polymeric structure and allows the contaminants to enter.  There is a good description of the epoxy backbone in a web site called Macrogallaria. http://www.pslc.ws/macrog/level2.htm  You will note that there are a lot of places in the epoxy resin for hydrogen bonding to occur.  In the late 90's we were analyzing a field failure due to CAF.  The growth was around the 5th layer of a 10 layer board as I recall.  When my student polished down to that layer, he was able to extract the ionic residues and found bromide, but no chloride.  This board was processed with the high bromide HASL fluid.  We have also published some work  using different polyglycols in the flux, and we noted that the Cu and Cl ions in the matrix had a different morphology depending on the polyglycol used.

I agree with you that the traditional test method described as Surface  Insulation Resistance is in fact an electrochemical migration test.  According to the IPC, Electrochemical migration (ECM) is defined as the growth of conductive metal filaments across a printed circuit  board (PCB) in the presence of an electrolytic solution and a DC voltage bias.  The low voltage test could also be described as ECM but it does more accurately show the insulation resistance at the given temperature and humidity conditions because dendritic growth would be rather slow under those conditions.  

I hope you are enjoying the summer.
Regards,
Laura



-----Original Message-----
From: TechNet [mailto:[log in to unmask]] On Behalf Of Brian Ellis
Sent: June-21-13 7:28 AM
To: [log in to unmask]
Subject: [TN]

Laura,

Is is a long time since we had any conversation together or even crossed swords! But it's good to hear from you.

I used the term 'chemi-physio-adsorption', which I coined for the occasion, to describe what I believe is the combination of why glycols tend to stick to epoxies. If the surface were glass smooth, I believe the only mechanism would be hydrogen bond adsorption. The broken surface of etched epoxy probably allows for considerable absorption, which of course is purely physical. The crunch lies in the fact that some of the adsorption occurs in the hollows, where removal is very much more difficult. Of course, acetonitrile is a very effective general-purpose solvent which can remove both hydrophilic and hydrophobic organics. 
Provided that the bond strength between the acetonitrile and a contaminant is stronger than the bond strength between the contaminant on the substrate, then the contaminant will be dissolved in the solvent. 
In the case of glycols, both OH and H bonds can form simultaneously with the solvent, so that it is not surprising that it can remove them, at least partially.

Of course, being retired, I am totally out of touch with the latest developments. However, I'm a little surprised at your statement that bromide ions diffused into FR-4 as an affinity to the flame retardant bromine compounds. These are in no way ionic and, not only are the bromine atoms covalently bonded to the carbon, they are part of the cross-linking in the polymerisation of the resin. I would seriously suggest that the mechanism of migration of bromide ions is more likely to be due to an ion exchange mechanism with the residual sodium chloride molecules in the epoxy resin. This would also explain why chloride ions do not have any effect.

I know you are the guru on the electrochemical migration and its effects on surface insulation resistance, but this is only one mechanism. In my opinion (not humble, of course), if you refer to the various publications on the subject I have made since about 1986, you will see that I make a very distinct differentiation between surface insulation resistance and electrochemical migration resistance; although many confuse the two, they are horses of different colours. The test that I was proposing in my earlier message was true surface insulation resistance and, for this reason, I stated that the test should be conducted without any bias voltage and with the measurements made at 5 V or less for as short a time as possible. This is to prevent any electrochemical migration from altering the results or, any dissociation of the sodium chloride molecules in the epoxy resin from doing the same. 
If you apply a bias voltage, there is migration of the sodium and chloride ions within the epoxy structure and this can seriously change the apparent surface insulation resistance independently from that due to the presence of a contaminant such as any form of surfactant.

As you are probably aware, I pioneered the notion of such low voltage, unbiased, SIR tests when I developed the Insulohmeter IRMA. Much of the research into the effects due to contamination were done by myself, while the effects due to the structure of the epoxy was studied by a graduate student at the Swiss Federal Institute of Technology whose Master's degree dissertation was on this subject. Unfortunately, I cannot remember his name or details but I do remember that he worked under Prof Kausch who had the chair of polymers, at that time in the 1980s. As the guy had become so knowledgeable about the electrical characteristics of epoxy resins, I suggested to him that there must surely be many openings in the industry for a person with this experience; unfortunately, he chose to be attracted by American "big oil", rather than specialise in a very narrow field. I've never heard of him since! Incidentally, Kausch told me, after the adjudication of his dissertation (I was on the panel) that it was the best master's thesis he had ever had the pleasure of reading! He bought me a dinner in recompense of having lent the Department the Insulohmeter for three trimesters!

Of course that brings me to a remembrance of the dinner we had in Washington DC, together with Barbara K. I recall that we had some very interesting discussions in that Italian restaurant!

Life goes on in sunny Cyprus, unfortunately with the physically degenerative effects of old age, about two weeks short of my 81st birthday! How is it with you in your colder climes?

Best regards

Brian

On 21.06.2013 00:32, Laura J Turbini wrote:
> Hi Brian,
> You always add a little spice to the conversation including references which only us "old timers" know.  You are correct that Zado's worked focused on PEG and polypropylene glycol.  Today, there are block copolymers such as polyethylene propylene glycol (PEPG), and others that are used in HASL fluids.  Jack Brous showed in 1981-82 that the PEG absorbed into the epoxy (it was not a chemi-physio-adsorption) and he was able to extract it from the boards using acetonitrile.  When he evaporated the acetonitrile solution and took an FTIR spectrum of the residues he found PEG.
>
> More recently, my former student, Dr.  Antonio Caputo published a paper which included extraction of PEG and PEPG from water soluble fluxed  FR-4 test coupons. Ref.  A. Caputo, L.J. Turbini, D.D. Perovic, (2009), “Conductive Anodic Filament (CAF) Formation Part I:  The Influence of Water Soluble Flux on its Formation”, Journal of Electronic Materials, Vol. 39, 85-91 (2010).
>
> In another paper he also showed that if the HASL fluid contained a high bromide content (~15%), the bromide ions also diffused into the FR-4 (because for  brominated epoxy - like dissolves like).  Chloride from the flux did not diffuse into the epoxy. ref A. Caputo, L.J. Turbini and D.D. Perovic, “Characterization and Electrochemical Mechanism of Bromide-Containing Conductive Anodic Filament (CAF) Failure,” Journal of Electronic Materials, Vol. 40, No. 9, 2011.
>
> You commented below - The only valid way of determining the presence of hydrophilic surface phenomena is by non-biased, low voltage 50/90 or 85/85 SIR qualification tests.  What do you mean by non-biased, low voltage?  Aren't the two terms contradictory.
>
> The rate of electrochemical migration (dendrite or CAF growth) is affected by the contamination present, but also by voltage, temperature and humidity.  Using low voltage testing would require a longer time for the dendrite to form.  There is a rule of thumb that says that a chemical reaction doubles for each 10oC rise in temperature.  Thus, the use of a lower temperature would also require a longer time for dendrites to form.
> Regarding humidity, FR-4 boards will easily have enough layers of water molecules at 70% RH or higher, to allow the electrochemical migration to occur.  So whether it is 85% or 90%, the difference in the rate of dendrite formation will be small.
>
> Regards,
> Laura
>
>
>
> -----Original Message-----
> From: TechNet [mailto:[log in to unmask]] On Behalf Of Brian Ellis
> Sent: June-20-13 11:59 AM
> To: [log in to unmask]
> Subject: [TN]
>
> I'm afraid that some of what you say may be misleading. Frank Zado's paper, at the Anaheim and Philadelphia Nepcon conferences in 1979, explored mainly Carbowax (polyethylene glycol. PEG) of specific ranges of MW. Although he did some tests with polypropylene and higher glycols, these proved to be of much reduced effect. This was also specific to wave soldering. Also the effect was not due to an epoxy-OH bond; it was a hydrogen bond, exacerbated by the structural surface of the epoxy, left by the copper treatment. It could be described as a chemi-physico-adsorption. However, PEG fell largely into disuse in the 1980s, except for some tin-lead reflow and HASL processes in the FAB side. Of course, it was your famous OH group that potentially created any hydrophilic characteristics at the other end of the molecule!
> More particularly, as I have propounded many times since 1969 (Inter-Nepcon), in my book and other publications, in lectures and in my swansong paper in Circuit World, the water-break test is absolutely meaningless, with easily produced false negatives and false positives.
> IMO, anyone who uses it as determinant of any specific reliability 
> conditions needs his head examining. The only valid way of determining 
> the presence of hydrophilic surface phenomena is by non-biased, low 
> voltage 50/90 or 85/85 SIR qualification tests. The oracle hath 
> spoken! :)
>
> Brian
>
> On 20.06.2013 17:21, greg wrote:
>> It is true that WS flux should be removed in the cleaning process.
>>
>> However, many glycols actually bond to FR-4 epoxy through their -OH 
>> groups. Hence the surface after soldering and cleaning is 
>> hydrophillic. (Frank Zado showed this back in the early 80s.)
>>
>> An easy test is take a board that is clean but not WS soldered and drop DI water on it. It should bead up.
>>
>> If after WS soldering and cleaning a drop of DI spreads you have glycols bonded to the epoxy.
>>
>> Adding a no-clean (with dibasic acids) to a hydrophobic mix may be an iffy proposition.
>>
>> Better to use a Bellcore compliant flux for your final soldering.
>>
>>
>>>    -------Original Message-------
>>>    From: Steven Kelly <[log in to unmask]>
>>>    To: [log in to unmask]
>>>    Subject: [TN]
>>>    Sent: 20 Jun '13 09:02
>>>
>>>    Thanks Dave - customer has not done any testing and for years we 
>>> have only used no-clean for both operations so now I have some parts 
>>> to be done one way and some another for Class 3 medical. Not good in 
>>> my 2 cent opinion. Regards Steve Kelly
>>>
>>>    From: [log in to unmask] [mailto:[log in to unmask]]
>>>    Sent: June-20-13 9:15 AM
>>>    To: TechNet E-Mail Forum; Steven Kelly
>>>    Cc: TechNet
>>>    Subject: Re: [TN] Mixing solders,
>>>
>>>    Hi Steve - I'll have to pay Doug and use his tag line - "It depends"! Anytime you mix two different flux systems, especially a water soluble and a low residue (aka no clean), there may be an issue of incompatibility that could result in a really hard lacquer (best case) or a really cool corrosion cell (worst case).  My recommendation would be to advise the customer that the mixing of the two flux systems would not be advised unless some testing can be conducted to ensure no detrimental reactions would occur. A second option would be to check with the flux supplier to see if they have any compatibility data. If the fluxes come from two different suppliers, don't waste your time asking that question as they won't have the answer. Good Luck.
>>>
>>>    Dave Hillman
>>>    Rockwell Collins
>>>    [log in to unmask]<mailto:[log in to unmask]>
>>>
>>>
>>>
>>>    From:        Steven Kelly <[log in to unmask]<mailto:[log in to unmask]>>
>>>    To:        <[log in to unmask]<mailto:[log in to unmask]>>
>>>    Date:        06/20/2013 08:02 AM
>>>    Subject:        [TN] Mixing solders,
>>>    Sent by:        TechNet <[log in to unmask]<mailto:[log in to unmask]>>
>>>    ________________________________
>>>
>>>
>>>
>>>    Hi All,
>>>    I have been looking in the archives but can't seem to find what I 
>>> want an answer to . I have a customer who wants us to use water 
>>> soluble RoHS for the SMT process but wants no-clean RoHS for the 
>>> touch-up. Is this recommended? Regards Steve Kelly
>>>
>>>
>>>    If the recipient to whom this e-mail is sent has an NDA with PFC Flexible Circuits Limited this e-mail is considered confidential and is subject to any NDA agreements between the respective parties.
>>>
>>>    See PFC on "How It's Made`` coming soon on the Discovery Channel!
>>>
>>>
>>>    ______________________________________________________________________
>>>    This email has been scanned by the Symantec Email Security.cloud service.
>>>    For more information please contact helpdesk at x2960 or [log in to unmask]<mailto:[log in to unmask]>
>>>    
>>> ____________________________________________________________________
>>> __
>>>
>>>
>>>    ______________________________________________________________________
>>>    This email has been scanned by the Symantec Email Security.cloud service.
>>>    For more information please contact helpdesk at x2960 or [log in to unmask]
>>>    
>>> ____________________________________________________________________
>>> __
>>>
>>
>> _____________________________________________________________________
>> _ This email has been scanned by the Symantec Email Security.cloud 
>> service.
>> For more information please contact helpdesk at x2960 or 
>> [log in to unmask] 
>> _____________________________________________________________________
>> _
>>
>
> ______________________________________________________________________
> This email has been scanned by the Symantec Email Security.cloud service.
> For more information please contact helpdesk at x2960 or 
> [log in to unmask] 
> ______________________________________________________________________
>
> ______________________________________________________________________
> This email has been scanned by the Symantec Email Security.cloud service.
> For more information please contact helpdesk at x2960 or 
> [log in to unmask] 
> ______________________________________________________________________
>

______________________________________________________________________
This email has been scanned by the Symantec Email Security.cloud service.
For more information please contact helpdesk at x2960 or [log in to unmask] ______________________________________________________________________

______________________________________________________________________
This email has been scanned by the Symantec Email Security.cloud service.
For more information please contact helpdesk at x2960 or [log in to unmask] 
______________________________________________________________________

ATOM RSS1 RSS2