We're still using Bell Bright on the inside of our bell jars, so glad to hear from Anette that there's a replacement for that product.
On our modified Edwards carbon coater (306A), I generally attempt to catch the reflection of the arc in a polished brass disk as the samples and the brass disk rotate. Using that reflection I can generally detect the color changes even when the bell jar is pretty dark. But still about once every three to six months or so we need to clean the bell jar to avoid "flakes".
Some things I keep in mind when coating: I watch the backing pump pressure on the diffusion pump and when that starts to creep up I reduce the current to the electrodes until it recovers (usually in 30-60 sec). I also find that the carbon electrodes need to generate some small sparks (1 to 3 cm long) to get a good conductive coating. If the current is kept very low, I find that the resulting carbon coat is actually quite non-conductive. This may be unique to this Edwards coater. We also try to get the color on a polished brass disk just as it changes from red to blue, so a violet/purple color is what we shoot for. This gives a very reproducible 20 nm coating thickness, though 10 or 15 nm is probably enough thickness for conductivity, the coating reproducibility of a red-blue transition on brass is more important I think.
john
We are a community of analysts, that cares about EPMA