Gold is indeed very soluble in liquid phase tin-lead solders at normal 
pot tempertures. Virtually all the gold exposed to a wave for >2 sec will 
dissolve, and is widely considered a contaminant.

A presentation at the 8/95 SMI show by Dr. Jennie S. Hwang indicated the 
dissolution rate of gold into solder is 111 microinches/sec at 232 
degrees C. Given that most plated finishes are below that thickness, and 
that the operating temp of your wave solder pot is normal, you may expect 
to find dissolved gold in the pot in short order.

Since gold-tin intermetallics are considered highly detrimental to solder 
joint fatigue life, this may be a concern.

The other side of the argument may be that the fast dissolution is 
actually good.

If you plan to use gold as a solderability protective (or just do so 
without planning) for the PWB, perhaps you can use a very thin gold 
coating and tolerate the dissolution, as the pot may have say 200 kg of 
solder alloy and the dissolution can continue some time until the pot 
reaches a detrimental solution level of gold.

Manko recommends that the amount of copper contaminant plus _twice_ the 
gold should not exceed 0.35 % (wt/wt); he also states that gold as low as 
0.02% has been reported to cause wetting problems. 

Immersion coatings of as little as 5-10 microinches are possible (but at 
that thickness, I wonder if gold may not provide the protection against 
base metal passivation that is apparently desired). So I recommend that 
you calculate surface areas and production rates, see how many you can 
run before you have your pot at a level which may affect the metallic 
properties of all joints, or even the wetting  you'll see. I would 
definitely accelerate a regular pot testing program. 

See _Solders and Soldering_, 3rd ed. by Howard Manko for a good 
discussion of intermetallics, and lots of other related info.

I have heard some recent proposals of the use of Ni/Au as a solderability 
preservative, but there would seem to be some fundamental metallurgical 
concerns which make it sound less than fully safe for me. This said, I 
know that some bizarre sounding approaches have been shown to be workable 
in the past. I know some good companies who required gold plated finish 
on PWB's in the old days (meaning more than a few years in this 
business).

Seems strange that we now have some people saying that this plated finish 
is desireable, when for years I seem to remember people twisting my arm, 
raising prices, and using other techniques to get me to accept no gold or 
various gold replacements.

Perhaps you could simply dump and replace your pot at regular intervals 
and tolerate a level of say 0.02%. If so, assuming a 200 kg pot, you can 
have about 0.0002 x 200,000 g or 40 g of gold. At a thickness of 10 
microinches (0.25 um), that  represents 40 g = (0.000025 cm) (area, cm2) 
(19.3 g/cm3); or the area in cm2 you can dissolve before reaching 0.02% 
wt/wt is:

 40/.000025/19.3 = 82,900 or 13,000 sq inches.

That may be a lot of units you can wave solder, depending on the total 
surface area of gold you see on each unit. Since you have a dual wave 
machine, maybe the second pot will see relatively little gold, and the 
joints are actually being formed by the second (less contaminated) pot.

Keep in mind that 40 g of gold is more than one troy ounce, which may 
warrant recovery.

You will need a better metallurgist than me to predict the possiblity of 
harmful intermetallic formation using solder having low level gold 
contaminants, or what may happen with uneven dissolution and resultant 
higher concentrations at joint sites.

AND, the above is pretty straightforward, but what about SMT joints where 
ALL that gold will dissolve and STAY in the joint with normal printed 
paste/relow joining soldering techniques...

Lastly, keep in mind the Cowboys blasted the Giants 35-0 last night, this 
may correlate with higher than normal residual ethanol levels in my 
bloodstream and affect my calculations for a day or so. Check those 
numbers.

I'd like to hear what your experience is with this finish in regard to 
the actual solderability and joint formation.

Jerry Cupples
Interphase Corporation
Dallas, TX
214-919-9150
[log in to unmask]