copyleft patent license for computational quantum chemistry?

[Justin Wells]

Some people I know are setting up a couple of clusters to do computational 
quantum chemistry. They have a small grant now to buy a bunch of computers
and begin chugging away. Once that's done and they understand the ins and 
outs of it in their own labs, they're going to consider trying to
create some kind of distributed system, similar to distributed.net/SETI.

The goal in computational quantum chemistry is to work out possible reactions
through simulations and other types of analysis. When you find one, you
basically have something that could be patented: a new method for
synthesizing some compound from some materials. Not every method that 
comes out of the simulation is going to work in real life--but the idea 
is rather than running a huge (and expensive, and time consuming, and 
tedious) matrix of trials covering all the possible combinations,
you use a computer to cut down the experiments you actually have
to perform (experiments which often involve expensive heavy metals
as catalysts, materials that make platinum look extremely cheap.)

Normally what they would do when they computed some method is acquire a 
patent, like everyone else would, the ordinary way, through their U's
legal department.

With general public involvement, they would still want to acquire
a patent, but it would be an "open patent"--existing only as a
pre-emptive measure to prevent a multinational from patenting the
same method later, and ensuring that the original discovers receive
credit for it. Everyone would be allowed to follow the patented
method, without any fee. (The chemistry geeks still have to do hard
brain work to figure out what computations should be done, and
verify that the what the computer proposes actually works in a real
lab-- they would get credit for that, but with help from the public,
they wouldn't expect to restrict access to the work).

I described the GPL to one of the principals in the project, and they seemed
pretty interested in the copylefting concept. The public license for the 
patented method would then demand that any derived method be made public, 
and subject to the license of the original patent. 

Then if some multinational worked out a new way of synthesizing something
else, which made use of the work done with public cycles, they would 
have to release that discovery to the public under a similar patent. 
Get one or two key patents under a copyleft, and you could start a big
snowball rolling.

Does there exist something with a copyleft which would be suitable for 
patents in traditional, non-software fields of research?

Thoughts? Suggestions? Pointers?

It's not my project, and I don't really know a whole lot about the 
mathematics and other details--I'm looking for somewhere or someone or 
something to connect up with the people who are doing this, since they 
exist outside the free software world (they're chemistry geeks, not 
computer geeks).

There might be administrative hurdles to clear--possibly the U. would 
object to getting involved with copylefting. But who knows? Maybe a 
large enough public project wouldn't need much help from the U.

Justin

ps: If you are interested in the project, and have relevant background 
    in chemistry, there's a mailing list you could join, it's just 
    being set up now. Ask me for details.