[Pw_forum] band symmetry analysis for Pt (example04 in PP directory)

[Paolo Giannozzi]

On Sun, 2013-04-14 at 11:23 +0800, Miao Gao wrote:

> Considering the differences for degree of degeneracy and values of energy
> bands between reference and the repeated result, which one should I trust? 

There is no difference between the two sets of results, apart from 
the presence of two additional states. Notice that when these two
states are found, the upper manifold of states (fourfold degenerate)
is "cut into half" because the number of calculated states is fixed. 
Of course the code gets confused and doesn't find a suitable symmetry
label for the "half manifold", for a good reason: there isn't one.

> X point,           energy bands,     reference results, PRB 2, 883 (1970) 
> e(  1 -  2) =     10.44178  eV     2   --> G_6+  M_6+ | G_6+
> e(  3 -  4) =     10.87347  eV     2   --> G_7+  M_7+ | G_7+
> e(  5 -  6) =     17.37445  eV     2   --> G_7+  M_7+ | G_7+
> e(  7 -  8) =     17.67776  eV     2   --> G_6+  M_6+ | G_6+
> e(  9 - 10) =     18.65959  eV     2   --> G_7+  M_7+ | G_7+
> e( 11 - 12) =     19.10266  eV     2   --> G_6-  M_6- | G_6+ Inconsistent
> e( 13 - 14) =     26.26903  eV     2   --> G_6+  M_6+ |

Maybe the last G6+ state of PRB 2, 883 (1970) corresponds to states
13-14 of today's calculation? While it is quite possible that bands.x
incorrectly labels some irreps, it is exceedingly unlikely that it
confuses an odd-parity irrep with an even-parity one

> L point,           energy bands,     reference results, PRB 2, 883 (1970)
> e(  1 -  2) =     10.17424  eV     2   --> G_4+  L_6+ | L_6+
> e(  3 -  4) =     13.14245  eV     2   --> G_5+  L_4+ | L_4+
> e(  3 -  4) =     13.14245  eV     2   --> G_6+  L_5+ | L_5+
> e(  5 -  6) =     14.15869  eV     2   --> G_4+  L_6+ | L_6+
> e(  7 -  8) =     16.90324  eV     2   --> G_4-  L_6- | L_6+ Inconsistent
> e(  9 - 10) =     17.29982  eV     2   --> G_4+  L_6+ | L_6- Inconsistent

as above: most likely in the old calculation the ordering of the 
two doubly degenerate bands is reversed wrt today's calculations

P.
-- 
Paolo Giannozzi, Dept. Chemistry&Physics&Environment, 
Univ. Udine, via delle Scienze 208, 33100 Udine, Italy
Phone +39-0432-558216, fax +39-0432-558222