Gold Embrittlement in solder joints

[status1]

Hello,
The basic question was answered in another closed thread but I just wanted to find out a little more detail about it

How exactly and why is solder used to remove the gold plating ?
What I mean is that since gold melts at much higher temperature than solder than why would the solder remove the gold ?
Is there something else going on like a chemical reaction ?
Maybe the correct word is leaching ?
But I still don't quite understand why would gold leach out in the solder. Does this work for other metals or just gold ?

I am taking the IPC 610 training and I asked the trainer about it but he did not know exactly how that happens
I know it sounds like a stupid question but I am just curious

 

[IRstuff]

Close, gold forms eutectics with many other materials:

http://www.engineeringtoolbox.com/gold-alloys-melting-points-d_1434.html

TTFN

FAQ731-376

 

[status1]

Hello,
Thanks for the reply but that link says no document found.

Is there another link that has the information ?

 

[IRstuff]

> It works for me
> There's a thing called a "search engine"

TTFN

FAQ731-376

 

[eng1ne]

http://209.85.229.132/search?q=cache:ANMnKCVb9AwJ:

http://www.engineeringtoolbox.com/g...tic&cd=1&hl=en&ct=clnk&gl=uk&client=firefox-a

This is the cached version found on Google - essentially what IRStuff was pointing to...

http://209.85.229.132/search?q=cache:NUivav3Pv4MJ:

http://www.semlab.com/GoldEmbrittle...der&cd=1&hl=en&ct=clnk&gl=uk&client=firefox-a

 

[status1]

Hello,
Ok It's working today
From what I understand when gold is mixed with other metals it forms an alloy that has a lower melting point than pure gold

Should I assume that the gold plating on pc boards is not pure gold ?

How do I know what gold alloy is used on pc boards ?

 

[MacGyverS2000]

No, gold plating is pretty darn pure (like 99.999% or better). The gold is plated directly onto the fresh/clean copper traces. There is a slight mixing of gold into the solder during the soldering process, but a proper process will minimize this to irrelevant levels. And metal alloys do not necessarily have a lower melting point than the individual metals.

Dan - Owner

http://www.Hi-TecDesigns.com

 

[status1]

Sorry I am confused again
So the pure gold is plated over the copper so far it's clear

"There is a slight mixing of gold into the solder during the soldering process"

That's what I am curious about
If the pure gold melting temperature is much higher than the melted solder how does the solder mix with the gold ?
If solder can do that why not mix with the copper trace too ?

"And metal alloys do not necessarily have a lower melting point than the individual metals"

Ok. So what was the point of the link about the eutectic alloys of gold then ?

 

[IRstuff]

I don't understand what you are not understanding. Some metals, like gold, form eutectics, which have lower alloy melting temperatures; others do not.

TTFN

FAQ731-376

 

[status1]

I am just going by macgyvers2000's post that seems to be contradictory to your post

He is saying that the copper traces are plated with pure gold and follows that up with a statement that implies that gold alloys do not necessarily melt at lower temperatures Of course
he did not mention specifically gold alloys

I understand that some gold alloys have lower or higher melting point than pure gold

I am just trying to figure out specifically what is the process by which gold is removed from copper traces

Perhaps I need to understand what eutectic means
You said "gold forms eutectics with many other materials"
Could you expand on that a little bit more and be a little more specific as it relates to plated gold on copper traces ?

Which material to be more specific ?
According to the statement at the link you provided
"The melting point of an alloy of two (binary) or more metals depends on the relative proportions of the ingredients"
Assuming that the gold plating is pure how can it have other materials in it to have the right proportion ?

 

[eng1ne]

The PCBs are plated with gold; i.e. a layer on top of the copper.

Pay less attention to the copper.

Under certain conditions, e.g. with respect to temperature and gold concentration; when tin-containing solder is applied, gold and tin will form intermetallic structures within the joint. These structures weaken the joint because they are formed differently to the solder. Note though, that this happens only under certain conditions. If you push tin and gold together with your bare hands at room temperature, it will not work.

Bear in mind that the metals do not already have to be 'alloyed' before an instance like this is possible.

I think you have enough information to explore embrittlement further.

The title of your post searched on Google brought extensive resources for you to study; e.g.

http://www.semlab.com/GoldEmbrittlementofSolderJoints.pdf

plus MANY more!

 

[status1]

That was interesting to read but it all seem to be at a microscopic level

What do you think of this statement that I read somewhere ?
"Gold readily dissolves in molten solder"
Is that true ?

Assuming I have a bar of gold (I don't)and I place it in a solder pot at let's say 750F
Would the gold bar dissolve completely ,just a little bit on the surface or not at all ?

 

[MacGyverS2000]

Gold melts at nearly 2000F, so no, it's not going to completely melt at 750F... but it will mix slightly at the surface, and if the pot is stirred over a long period of time, the gold will continue to mix. From that standpoint, the statement that gold readily dissolves in solder is misleading in its simplicity... but nevertheless, it's true from a purity standpoint. The "microscopic" level is all that's necessary to completely taint an entire pot and make it worthless (at least as far as circuit reliability is concerned).

Dan - Owner

http://www.Hi-TecDesigns.com

 

[eng1ne]

"That was interesting to read but it all seem to be at a microscopic level"

Gold embrittlement on a 10 inch wide solder joint (if they exist) would be the same gold embrittlement that exists in your own solder joint. If you want to disregard the microscopic level, then what you might come across on a PCB would simply be a poor solder joint with an unknown diagnosis.

 

[status1]

Thanks for the explanation
I am starting to understand the process a little bit better now
So just to make sure I understand
If gold is heated by itself to 750F nothing happens but when solder is added then some of the gold is mixed with the solder at the microscopic level?

This gold that is mixed with the solder is it melted gold, gold molecules or gold that is flaked off the pc bd ?

I am just trying to grasp the magnitude

 

[eng1ne]

Status, take a look at this document that I posted earlier:

http://www.sem-lab.com/GoldEmbrittlementofSolderJoints.pdf

Read the article, look at the pictures and then you should grasp the magnitude and process by which it occurs. If you have any questions afterwards, please come back.

"If gold is heated by itself to 750F nothing happens but when solder is added then some of the gold is mixed with the solder at the microscopic level?" In a proverbial nut shell, yes.

 

[status1]

I read that article before but it's hard to determine how much gold is there It's also hard to tell which shade of gray is the gold

If that was a copper trace that was plated with gold does all of the gold come of or gets mixed with the solder in one application ?
If only a small microscopic amount is mixed with the solder I would imagine it would take a long time to remove all of it

Is the procedure in IPC 610 to add solder and the remove it and doing that twice enough to remove all the gold ?

 

[eng1ne]

Status, extracting the gold on such a scale would take more than a long time and is not practical. One would be better off creating a new joint.

I am not familiar with IPC 610, and given you are on a course for such; would it not be wiser to ask your trainer?

 

[IRstuff]

If the objective is to recover gold, then there are chemical processes that are more effective and selective.

TTFN

FAQ731-376

 

[status1]

The objective is not to extract gold but to remove the gold from the plated surface where something is going to be soldered to
This would be from at least 95% of surface to be soldered to

I did ask the trainer but he did not have an answer at the time and he said that he will get back to me and I thought maybe I can research it myself in the meantime

I found it hard to believe at the time that just by tinning a gold plated surface the gold would be gone so easily without actually melting the gold

Someone said in a previous post with the same title that tinning the gold plated surface and removing it by wicking it off with solder wick is acceptable while the instructor said this must be done twice maybe as an extra step in case not all of the gold was removed in the first step

I don't have the manual with me but I don't think IPC 610 specifies what method to use to remove the gold
I read somewhere else that a double dip (two solder pots) immersion or the use of a flowing solder wave is the way to go

I guess the only way to find out for sure is to use a method and somehow analyze the solder joint under a powerful microscope to determine how much if any gold is left in the solder joint

I want to thank everyone for their comments on this subject