[labnetwork] Black Layer Formation During Au E-Beam Evaporation

[Ryan Rivers]

Hi Chandan,

I've tried a few different pre-cleaning methods for the pellets (Rinsing
with Acetone/IPA/Water, sonication, etc), and none of them are nearly as
effective as just doing proper melt formation steps. Molten gold is well
above the temperature that breaks hydrocarbon-adjacent bonds to carbon
atoms (Including fluorine bonds on fluoropolymers), and in a vacuum chamber
those atoms aren't going to find each other to recombine before being
pumped by your turbo or cryopump (This is trace amounts of material, don't
worry about your cryo, it'll be fine). Carbon isn't terribly soluble in
gold, so it forms a surface interface between the carbon and the gold. That
surface has significant energy. When you heat gold to a fully liquified
state in ebeam there's a 200-300K differential temp between the top and
bottom of the melt, this induces convection currents in the melt, and those
currents will bring the gold up to the surface where it will float due to
surface energy, and eventually collect into nodules. The nodules become the
black film on top of the gold, which will not re-enter the melt due to
surface energy effects.

If there's enough contamination in the melt that the carbon forms a nodule
*before* getting to the surface, when it hits the vacuum interface the
Au-Carbon interface is suddenly replaced with an Au-carbon-vacuum
interface, which appears to result in a net energy release that ejects
material in the form of spitting.

This is not apparent in Ti, Al, Pt, Ni, Cr, etc. because those films form
metal carbides which precipitate to the solid/liquid interface of your melt
zone and then never move again (Ti/Al/Cr/Ni) OR they spit like crazy until
the carbide has been ejected from the material violently (Pt/Pd).

Which is all a long winded way of saying *burn out your gold melts in a
tungsten crucible at 5-10 A/s into an empty chamber for 2-3 minutes
whenever you add more than a couple pellets of material, then cool it off
and wipe it down. *If you don't do that, you're going to catch stray carbon
being ejected from the melt and that's going to leave black nodules in your
film. *If the melt is visibly spitting, it's contaminated, you're not done
burning it out.*

Also, buy your material from a proper vacuum materials company like
Plasmaterials, Lesker, VEM. If your precious metals material ever comes
packed in oil from the supplier they're not taking vacuum seriously, and
you shouldn't take them seriously as a supplier. (Wierd metals are
exceptions, we're talking standard materials here.)

-Ryan

On Fri, Sep 18, 2026 at 3:07 AM Chandan H B <chandanachar95 at gmail.com>
wrote:

> Thank you all for the feedback. I really appreciate the time and efforts
> made to address this issue.
>
> As most of you said, when the Au surface is wiped with a cleanroom
> compatible lint free wipe, it easily comes off and Au looks pretty clean. With
> that in mind, the primary suspect was carbon, but the question is, is it
> originating from the Au slug itself or as Grant said is it from the
> hydrocarbons from the chamber?
>
> In the case of hydrocarbons, I should see the same black particle/layer on
> any other material like Ti, Al, Pt, Ni, Cr etc., but I do not see any.
> Considering if that happens, how is it attracting the hydrocarbons at the
> very first place? Au can not be the first layer we coat as it needs an
> adhesive layer like Cr or Ti. Let me know more information if anyone has
> faced a similar kind.
>
> Having said that, I went ahead with EDS and XPS to check the carbon
> presence by giving out an Au pellet(99.999%). It did show 5~10 wt% of
> presence on the surface scan and 0~0.5wt% upon depth profiling. Considering
> these results, can source material undergo cleaning with IPA or DI water
> followed by Plasma Cleaning?
>
> This challenge is not new to many, so we look forward to hearing about how
> others have mitigated it.
>
> Thanks!
> Chandan H B
>
> On Fri, Sep 18, 2026 at 3:52 AM Ryan Rivers <rrivers at berkeley.edu> wrote:
>
>> Hi Chandan,
>>
>> We use a tungsten crucible liner to avoid the problem entirely. Mo is
>> also widely used as others have noted. I've had no issues with spitting
>> from W crucibles myself. That's usually a matter of source material
>> cleanliness and beam control. Spitting is almost always a function of your
>> source pellets having some amount of residual carbon on them.
>>
>> The other thing you can do with W or Mo crucibles is heat the gold to a
>> fully melted state with a liquid top (as opposed to a molten surface with a
>> breached liquid where the beam is hitting) and allow it to just sit for a
>> minute or two. If you're using a carbon spacer this is relatively trivial
>> to do without evaporating any appreciable amount of gold. When you cool
>> back down, all of the carbon in the melt will have floated up to the
>> surface and you can simply wipe the melt clean with a techwipe until you
>> get bright gold. You'll never see spitting out of the melt again until
>> someone contaminates it some other way.
>>
>> -Ryan
>>
>> On Thu, Sep 17, 2026 at 4:42 AM Chandan H B <chandanachar95 at gmail.com>
>> wrote:
>>
>>> Dear LabNetwork Community,
>>>
>>> We would like to seek inputs from cleanroom/fab facility managers and
>>> process owners regarding an issue observed during *Au deposition in our
>>> E-Beam Evaporators*.
>>>
>>> We have observed the formation of *black/dark layer on the Au charge
>>> after a few melt cycles*. The observation occurs during repeated
>>> melting of Au in a tungsten liner/crucible. We are currently investigating
>>> whether this could be related to *material surface contamination,
>>> crucible/liner condition, outgassing, chamber history, or the Au
>>> pre-cleaning/conditioning procedure*.
>>>
>>> We would appreciate your experience on the following:
>>>
>>>    - Have you observed *black/dark spots or a black layer forming on Au
>>>    during repeated E-beam melting/evaporation*?
>>>    - What *Au purity and material form* do you use?
>>>    - Do you perform any *cleaning/degreasing of the Au charge before
>>>    loading*? If yes, what procedure?
>>>    - Do you follow a specific *crucible/liner cleaning or replacement
>>>    procedure*?
>>>    - If you have identified a *root cause or corrective action* for a
>>>    similar observation, any details would be highly appreciated.
>>>
>>> Our objective is to compare practices across facilities and determine
>>> whether this is a *material-handling, source/liner, chamber-condition,
>>> or process-related phenomenon*.
>>> Any suggestions/inputs are highly appreciated.
>>> Thanks in advance!
>>>
>>> Regards,
>>> CHANDAN H B
>>> THIN FILMS ENGINEER
>>> _______________________________________________
>>> labnetwork mailing list
>>> labnetwork at mtl.mit.edu
>>> https://mtl.mit.edu/mailman/listinfo.cgi/labnetwork
>>>
>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mtl.mit.edu/pipermail/labnetwork/attachments/20260918/a346781f/attachment.html>