Al annealing

Baylor B Triplett baylortriplett at
Fri Mar 9 10:13:33 PST 2007

    Al-1%Cu is a reasonable option to reduce the crystallization 
described below unless you have extreme needs but you have to be careful 
to eliminate ions and water to prevent corrosion...shelf life of the 
devices may be limited without passivation of the Al alloy. Another 
alternative is Al-2%Si with much less susceptibility to corrosion but 
larger crystallite sizes. Below is a little of what I remember about 
this subject.
   [ "Surface roughness" with Al or Al alloys is usually simply grain 
growth resulting from crystallization. Although quite dependent on film 
thickness (the thicker the film, the larger the crystallite sizes), it 
is also dependent on cooling rate...rapid cooling will minimize the 
grain size formed. Aluminum is particularly bad since it has a low 
melting point.
For this reason, many in semiconductor technologies use Al-2%Si alloys 
or Al-0.5%, Al-2%Cu, and even Al-4%Cu to suppress grain growth and 
improve electromigration.
However, the Achilles heel of high Cu is the strong electrochemical cell 
formed between Al and the Al2Cu mixed phase which suppresses the size of 
the crystallites. About 10 years ago (before the advent of straight Cu 
metallization to improve electromigration), industry.] Depending on 
needs more refractory metals such as Ti can also provide
solutions and a dual layer of Ti plus Al alloy often works well at the 
expense of complexity.
                  Baylor Triplett, Consulting Professor
David W. Parent wrote:
> Dear Beinn,
> I have had a hard time fining data on this issue.  I have found though 
> brute force experimentation that the longer the pull time on any Al 
> film on SIO2, the greater the surface roughness, and the higher the 
> resistivity!  To reduce surface roughness and decrease RS I pull the 
> wafers out of our furnace (AT SJSU not SNF) as quickly as possible.
> (I mean I yank them out.)
> -DAve
> At 12:05 PM 3/8/2007, Beinn Muir wrote:
>> Dear Labmembers,
>> I am interested in increasing the grain size and reducing the 
>> roughness of
>> thermally evaporated films of Al (200-1000nm thick) deposited on Si with
>> native oxide. Do any of you have experience with annealing Al films, 
>> or can
>> reccommend any useful literature discussing the conditions and resulting
>> morphology changes? Another possible route I could take would be to 
>> etch /
>> polish the Al film to reduce roughness, but I have had limited 
>> success with
>> this.
>> I would be grateful for any information.
>> Best regards,
>> Beinn...

More information about the labmembers mailing list