Bad coats from Edwards E306A

[cogswell]

Hello friends.  I have, as many people here seem to have, a venerable Edwards E306A carbon coater, using the 3mm rods.  This thing was 20 years old when we got it used 25 years ago and I have literally coated thousands and thousands of samples with it since then with nary a problem.   

These days I am hit or miss with bad carbon coats.  The signs are there: little paper triangle is more brownish coloured than black, samples are charging in the instrument.  I almost never have two bad coats in a row.  If I get a bad coat (checking in the SEM), I usually reload the E306A and do another run and that one turns out fine.   

The penning during coating - sometimes - rises up above the 10^-4 level.  Sometimes it does not rise much and stays well down the gauge range. Despite what I would think, I have not seen what I think is any correlation between "that gauge went up" and "that was a bad coat", as I feel I have had many good coats with the vacuum gauge rising.

Last week I did a radical step and tried changing the gasket on the big bell jar - admittedly for another gasket that was 25+ years old but at least it was a different one.  I did two coats on polished metal samples that worked fine in the SEM, but man it turned the metals black.  Both unrelated samples were polished metal in silicone and epoxy mount and I'll attach pictures to this if I can.  These carbon coated samples worked fine for imaging and EDS, but they were pretty thick, thick enough to block the ebsd response - which came back when I rubbed off the coat.   These coats came off in an instant with a bit of alcohol.  Overall I had two successful coats and one bad coat on the two samples, and both successful coats seemed to be this very-thick dark stuff.   

Dismayed by these blackened samples, I put the original gasket back on the bell jar, did a coat and it was just fine.  A single data point?  Sure. I continue to test. 

I'm presuming I have some sort of air leak?  I cannot find it for the life of me.  I squirt ethanol around all the gasket edges and the penning gauge and the fittings and nothing makes the penning twitch when it's pumped down.  If I had bad coats every time it might be easier to diagnose but I'm like 1 in 3, or 1 in 4 are bad on average in recent history.

Rod is nice and bright, with sparks, during coating.

I've cleaned out the water lines to the diffusion pump, didn't make a difference. 

Some pictures.  Here's a sample where the vertical surface which is still clean after coating:
You cannot view this attachment.
sample top which is dark after coating (the middle had pen marks over it)
You cannot view this attachment.
A view showing top and side at the same time after coating:
You cannot view this attachment.

[Probeman]

The Edwards 306A is a great coater. Did you use some silicone vacuum grease on (both sides of) the gasket?

Might be worth trying a sample of polished brass to detect the actual coating thickness:

https://smf.probesoftware.com/index.php?topic=1252.msg8988#msg8988

[cogswell]

Hi there;

I do occasionally use a polished piece of brass, using that exact post from the forum here.  I show that picture to all the students who doubt the power of science.  Ironically I did not have it in for these "very dark" coats, because I am lazy and don't want to clean it. I usually get in the orange-to-red range when I do use it. 

Of course, since I changed the gasket back I had a good coat every time I tried (2-3 times) and I got a "dark" coat by purposely overcoating the thing.  I will try out the grease though.  In the past I have smeared a tiny amount of the apezion L where the gasket meets the metal surface of the coater but not on the glass-to-gasket.  I have some not-as-expensive silicone grease I can use I think. 

Brass stub with sample (this was a coat that worked):
You cannot view this attachment.

Brass stub:
You cannot view this attachment.

Fun fact.  I got the piece of brass from our old cryo-sem setup which when we decommissioned it I repurposed these nice thick pieces of brass as holders for cross section samples, since they sit nicely on their sides.  Pain in the butt to get this initial polish on that side though.