Posted by Steve Simon on April 18, 2014, 6:26 pm, in reply to "Itinerant Ferromagnetism"
Yeah, that is a tough question.
I think strictly speaking "itinerant ferromagnetism" is on the syllabus. But I think hubbard model is probably not (although it isn't clear -- sorry about this, I'm neither on the syllabus committee nor the exam committee, so I'm in the same position of trying to read their mind as you are).
My guess is that it would be surprising for the examiner to ask you to do a real (mean field) calculation with the hubbard model. However, they could ask you to write something about "what is itinerant ferromagnetism and why does it occur". You might then find it useful to refer to hubbard model in your answer -- but you could also answer such a question qualitatively -- explaining that electron-electron interaction drives itinerant ferromagnetism without using hubbard. It might be useful to be able to explain that if you have an on-site repulsive (hubbard!) interaction, this energy is dropped to zero if all the electron spins are aligned since Pauli would then prohibit two electrons from sitting on the same site, and this is the physics that makes spins want to align.
Similarly with antiferromagnetism. It might be useful to be able to argue why strong repulsive interaction with a half-filled band leads to antiferromagnetism --- and if asked you might want to invoke hubbard model. But I don't think it is likely that they would ask you to do an explicit calculation with the hubbard model.
I am guessing somewhat about what is examinable and what is not, but this is my guess.
35
Message Thread 
« Back to index