A quick question to add my bit into this conversation: where does the
definition of semimetal as "insulating (semiconducting) in one direction
and conducting in another one" come from? I have never heard of this; it
may the lack of my knowledge but I hold that it's the overlaping valence
and
You did not read up to the end (Graphite with a strong directional dependence
of the conductivity was an example, maybe it was not lucky to place it in front)
a little later you find:
"In some cases you have to check whether there might be an overlapp of the
valence and and conduction bands at
Thank you for your response, Prof. Gerhard. In my SOC calculation for
MnFe2O4 with GGA, the band structure is quite different than the DOS. Here
I have attached both the figures. I would like to know whether it is
possible to get this kind of difference in SOC calculations or not?
On Wed, 30 Jan
The Wikipedia article on metalloids says it's the overlaping valence and
conduction bands that define semimetals in physics.
Following textbooks in chemistry could lead to confusion as many chemists still
use the term for elements that cannot be classified definitively as metals or
non-metals
Just to be complete; here is the missing definition:
J. W. ALLEN; Nature 187, pages 403–405 (1960)
Gallium Arsenide as a Semi-insulator
but unfortunately there seems to be no Wikipedia page for semiinsulators
available.
In the calculations I find only TWO situations
1) bands are
With spin orbit interaction you can not distinguish the band structure for spin
up and down
however, you still vcan calculate the density of states for the two spin
directions
(was discussed several times in the forum)
Your band structure with spin orbit structure might therfore be similar
to a
I want to calculate the energy for a Ni vacancy at a Ni/NiO interface,
comparing the energy for the vacancy in the Ni layer at the interface to
the vacancy in the NiO layer at the interface. The interface structure is a
bit of a pain, but under control.
However, the killer is that this requires
Unfortunately, "ab initio" RPA does not work for correlated electrons
(statement from G. Kresse).
It is clear that such a problem (an interface between a localized
(highly correlated) and an itinerant (delocalized) system is very
difficult to treat with all current methods. In any case, with
8 matches
Mail list logo
