Added question)

[ljk]

Hello, a few month ago, I questioned about qstem and got answers. But, I added another question, which might be missed from you.

Here is the link and Could you give me answers? 

https://groups.google.com/d/msg/qstem/y_eTQ9OmKSs/UhybynVfCwAJ

Thanks.

Sincerely

Lee

[Christoph Koch]

Dear Dr. Lee,

 

It seems that your sampling of the STEM image is not fine enough, so some atoms are hit by the electron beam, and some are not. In the configuration you call ‘qstem’ (top screen shot), the electron beam is right on top of the oxygen columns, but it does not hit any of the heavier atoms (the e-beam positions are the black crosses in the screen shots below).  

 

In what you call ‘qmb’ configuration, the unit cell is rotated a bit with respect to the scan, so things are a bit difficult to compare. Certainly, the oxygen columns are never hit is precisely, as in the case of the other configuration.

 

I admit, I do not know what

 

STO100Ex_DwEx_300K.qsc:

 

STO100qmb_DwEx.qsc:

 

it is important to take a look at the Debye-Waller factors in your different .cfg files. When you do not specify any Debye-Waller factors, qstem will assume default values which are inverse proportional to the mass. The STEM-signal is very sensitive to the DW-factor, since the majority of the signal on the HAADF-STEM detector is due to TDS.

 

Here are excerpts of your different files:

SrTiO3Dw.cfg:                                 DW_Sr = 0.62,      DW_Ti = 0.43,     DW_O = 0.73

SrTiO3Dw_qmb100.cfg:                DW_Sr = 0.62,      DW_Ti = 0.43,     DW_O = 0.73

SrTiO3Dwdf.cfg:                             DW_Sr = 0.16,      DW_Ti = 0.28,     DW_O = 0.78

STO100qmb_Dwdf100.cfg:          DW_Sr = 0.1658, DW_Ti = 0.2864, DW_O = 0.7879

 

Also, when I look at your images, I can also see the oxygen in the qmb case:

It really seems to depend on what values you use for the source size. Did you always use exactly the same number?

 

I suggest that you try your tests gain with equal orientation of the lattice and z-offsets of your slices, in order to compare apples with apples. I noticed that your slices are rather thick (the multislice simulation should converge in the limit of very thin slices), but more importantly, your qmb model is missing a layer of atoms at the top:

 

With best regards,

Christoph.

--
You received this message because you are subscribed to the Google Groups "QSTEM" group.
To unsubscribe from this group and stop receiving emails from it, send an email to qstem+un...@googlegroups.com.
To post to this group, send email to qs...@googlegroups.com.
Visit this group at https://groups.google.com/group/qstem.
For more options, visit https://groups.google.com/d/optout.