[Wien] Is calculating the chemical shielding (orbital part) in paramagnetic system using x_nmr_lapw available ? x_nmr_lapw? lapwdm? DIPAN?

[林敏]

Dear Prof. Peter Blaha,

Yes, I am calculating the Li+/Na+ NMR shifts in cathode of battery materials (usually transition metal oxide).

Of course, most of Li+/Na+ NMR shifts is dominated by Fermi-contact shifts, which is related to :HFFxxx in the scf file. 

In some cases, the Fermi-contact shifts is small (several tens of ppm), so I don’t want to neglect the orbital shielding chemical shift casually even it’s range in several ppm. 

About the X_nmr, it do have -up/dn options. I test the program using LiCl: doing open shell SCF first, then running x_nmr -dn,  x_nmr -up respectively. 

I get the same results, as follows:

:NMRTOT001  ATOM:      Li1   1  NMR(total/ppm) Sigma-ISO =     45.40     Sigma_xx =     45.40   Sigma_yy =     45.40   Sigma_zz =     45.40
:NMRASY001  ATOM:      Li1   1  NMR(total/ppm) ANISO (delta-sigma) =     -0.00     ASYM (eta) = 0.000     SPAN =      0.00     SKEW = 1.000

:NMRTOT002  ATOM:      Cl1   2  NMR(total/ppm) Sigma-ISO =    460.40     Sigma_xx =    460.40   Sigma_yy =    460.40   Sigma_zz =    460.40
:NMRASY002  ATOM:      Cl1   2  NMR(total/ppm) ANISO (delta-sigma) =     -0.00     ASYM (eta) = 0.000     SPAN =      0.00     SKEW = 1.000

:NMRXIM     NMR(macroscopic-component/ppm, trace(xi)*8*pi/3) =     -5.33


When I using x_nmr -dn -up, also the same results of x_nmr -dn , or x_nmr -up. 

While the close shell calculations get following results: 

:NMRTOT001  ATOM:      Li1   1  NMR(total/ppm) Sigma-ISO =     90.24     Sigma_xx =     90.24   Sigma_yy =     90.24   Sigma_zz =     90.24
:NMRASY001  ATOM:      Li1   1  NMR(total/ppm) ANISO (delta-sigma) =      0.00     ASYM (eta) = 0.000     SPAN =      0.00     SKEW = 1.000

:NMRTOT002  ATOM:      Cl1   2  NMR(total/ppm) Sigma-ISO =    920.24     Sigma_xx =    920.24   Sigma_yy =    920.24   Sigma_zz =    920.24
:NMRASY002  ATOM:      Cl1   2  NMR(total/ppm) ANISO (delta-sigma) =     -0.00     ASYM (eta) = 0.000     SPAN =      0.00     SKEW = 1.000

:NMRXIM     NMR(macroscopic-component/ppm, trace(xi)*8*pi/3) =    -10.10

which is not same as the sum of -dn and -up. What is the problem behind these?

I can’t find further description of x_nmr on tutorial’s PPT and UG. Is there detail of -up and -dn options available?

The dipole part of hyperfine interactions is used for simulation of NMR spectrum. I also meet some problem and I am checking the mail list.

Best regards,

Min


———————————————
Min Lin
2018 Ph. D student
Physical Chemistry 
Chemistry Department Chemistry & College of Chemistry and Chemical Engineering
Xiamen University
China
e-mail: linmin at stu.xmu.edu.cn





> On Jan 28, 2020, at 4:38 PM, 林敏 <linmin at stu.xmu.edu.cn> wrote:
> 
> Dear Prof. Peter Blaha,
> 
> Thank you very much for your reply.
> 
> Very sorry for confusing you.
> 
> Here, the paramagnetic system, I mean the open shell system or spin polarize calculation, specifically is transition metal oxide, which usually are insulator/semiconductor.
> 
> From your reply, I still suppose that x_nmr doesn’t work for such system. The error message is no case.vsp file, while I have   case.vsp(up/dn).
> 
> About the dipole part, what I want to calculate is the hyperfine interaction between transition metal ions (TM) and observed nucleus (Li+/ Na+...), rather than TM nucleus themselves. The isotropic part is very easy done in normal SCF calculations. But the dipole part, can the magnetic field of TM be seen as external field for observed nucleus ? So I have to use DIPAN?
> Right?
> 
> Best regards,
> 
> Min
> ———————————————
> Min Lin
> 2018 Ph. D student
> Physical Chemistry 
> Chemistry Department Chemistry & College of Chemistry and Chemical Engineering
> Xiamen University
> China
> e-mail: linmin at stu.xmu.edu.cn
> 
> 
> 
> 
> 
>> On Jan 27, 2020, at 6:04 PM, 踢球者 <2236673859 at qq.com> wrote:
>> 
>> Dear Wien experts, 
>> 
>> Is calculating the chemical shielding (orbital part) in paramagnetic system using x_nmr_lapw available ?
>> 
>> I suppose not after reading the UG carefully.
>> 
>> I find that, there is a program in WIEN2k, lapwdm.
>> 
>> "RINDEX=3 LSINDEX=3: <X> is the orbital part of the hyperfine field at the nucleus (for a converged calculation at the very end)"
>> Is this for chemical shielding ?
>> 
>> By the way, in this program: RINDEX=3 LSINDEX=5: <X> is the spin dipolar part of the hyperfine field at the nucleus (for a converged calculation at the very end). Here, what is the difference between DIPAN program doing and "spin dipolar part of the hyperfine field" ?
>> 
>> About the DIPAN. Why I have to input spin moments of atoms?  Why not calculating using spin density directly?
>> 
>> Best wishes, 
>> 
>> Min
> 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20200131/a3a16a49/attachment.html>