I am not a metalurgist but I am fairly certain that there are things that you can do in the area of wire drawing process control, lubricants, and die angles, etc. that would be a whole lot easier then altering the chemical composition/make up of the trace elements in the wire. I know from personal experience that the smalles amount of silver added to copper makes a big change in annealability. Right now you have a single problem - wire breaks. If you change the chemical makeup you may introduce untold number of variables (problems) into the equation (process).
First I would check around and find out what is typical pounds/wirebreak. You might not have a problem. Secondly I would check with the die people to see what die profiles they suggest. I would also talk to the lubricant people and ask what they suggest. I would also take a good look at the breaks. What do they look like? The appearance of the breaks can tell you a lot.
I know this sounds elementary but I would also check die elongation and machine capstan speeds to make certain that my dies match my machine.
All of these things are easy to do compared to the playing with the trace metals! If you do all of these things and still have problems, come back let us know what you did, the results, and perhaps someone out there will have other or better suggestions.
I know: 4047 is aluminum:
Si CU Mg Fe Zn Mn AL
11.0 - 13.0% .30% .10% .8% .20% .15% Remainder
RB/Spectre