Al annealing -- update

Beinn Muir beinnmuir at
Fri Mar 9 12:53:34 PST 2007

Dear all,

Thanks for all the replies regarding my Al deposition question. They have
all been very useful. What I have learnt so far is that I need to carefully
consider what surface morphology is really important for my application.

I should elaborate on the purpose of using Al. I need the Al layer for
electrochemical anodisation and the formation of porous surfaces, therefore
using metals other than Al (or alloys) is not currently an option. The
anodisation process appears to initiate at the pore boundaries, and
therefore the pore density is driven by the grain size. If I want larger
pores (less dense pores) then I want larger grains. It is perhaps not an
important requirement that the surface is very flat, although it is
possible that even if the grain size is small that if the surface is flat,
or periodically textured, then the pore density will be lower -- something
which is done with bulk Al anodisation during an initial electropolishing
step. I have tried this electrochemical etching step with the thin films
and it does not appear to change the surface morphology.

A number of you have pointed out that annealing will most likely cause the
roughness to increase. It also seems that the grain size is proportional to
the layer thickness, so the next thing I will do is to look at >500nm thick
films. In addition slow cooling will promote large grain size. One option
which was suggested was to deposit thick films and then etch back. Does
anyone have information on suitable etchants for Al, and whether etching is
likely to reduce or increase roughness (ie will it maintain the triangular
or hillock structures, or remove them)?

I am still interested in trying some annealing experiments with the Al
films, but I am not sure which parameters will be the most important, and
where I should be in experimental space. Is it common to anneal metals
below the mp for long times (hours) or above the mp for short times
(minutes)? Should the cooling be done over >12 hours, or is 1-2 hours slow
enough? Your thoughts would be most welcome.

Thanks again for all your help.

Best regards,

More information about the labmembers mailing list