[QE-users] vc-relax produces weird result with PAW pseudopotentials

[mkondrin]

Hi, Pietro,

Many thanks for suggestions. Indeed my ecut=40 Ry was far below the 
proposed minimum cutoffs listed in the pseudopotential files (44 Ry for 
Zn and 55 for Se). Setting ecut to 60 allows me to get decent stress for 
the experimental structure (-6 kbar).

Sincerely yours,
M.V. Kondrin


On 24.09.2020 12:12, Pietro Delugas wrote:
>
> Hi
>
> the last one is probably the calculation done at the end of the 
> vc-relax calculation with reinitialized plane wave basis set.
>
> It might be that, because of the lower cutoff the relaxation with the 
> PAWs is more affected by the non reinitialized basis set.
>
> You could simply increase the cutoff in the relaxation using PAW , or 
> try to use the  ecfixed, qcutz and q2sigma variables, to reduce the 
> number of plane waves.
>
> hope it helps
>
> Pietro
>
> On 24/09/20 09:39, mkondrin wrote:
>> Dear developers and users,
>>
>> I have encountered strange results while trying to relax simple 
>> sphalerite structure of ZnSe using PAW pseudopotentials from official 
>> site. I am starting from experimental lattice constant at zero 
>> pressure but the calculated stress was large and negative. The 
>> subsequent calculation converges to zero pressure but the final step 
>> also produces the large negative stress.
>>
>> $grep stress bulk1.out
>>      Computing stress (Cartesian axis) and pressure
>>           total   stress  (Ry/bohr**3)                   (kbar) P= 
>> -368.58
>>      Computing stress (Cartesian axis) and pressure
>>           total   stress  (Ry/bohr**3)                   (kbar) P=  
>> 112.26
>>      Computing stress (Cartesian axis) and pressure
>>           total   stress  (Ry/bohr**3)                   (kbar) P=   
>> 23.80
>>      Computing stress (Cartesian axis) and pressure
>>           total   stress  (Ry/bohr**3)                   (kbar) P=   
>> -3.11
>>      Computing stress (Cartesian axis) and pressure
>>           total   stress  (Ry/bohr**3)                   (kbar) P=    
>> 3.52
>>      Computing stress (Cartesian axis) and pressure
>>           total   stress  (Ry/bohr**3)                   (kbar) P=   
>> -5.37
>>      Computing stress (Cartesian axis) and pressure
>>           total   stress  (Ry/bohr**3)                   (kbar) P=    
>> 2.61
>>      Computing stress (Cartesian axis) and pressure
>>           total   stress  (Ry/bohr**3)                   (kbar) P=   
>> -0.40
>>      Computing stress (Cartesian axis) and pressure
>>           total   stress  (Ry/bohr**3)                   (kbar) P= 
>> -268.92
>>
>> I have also tried ONCV pseudopotential (with larger cutoffs 70 Ry of 
>> course) and the result is quite different -- the calculated total 
>> stress is small (22 kbar) and it rapidly and monotonically converges 
>> to zero.
>>
>> I am using QE-6.4.
>>
>> Sincerely yours,
>> M. V. Kondrin
>>
>> _______________________________________________
>> Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)
>> users mailing listusers at lists.quantum-espresso.org
>> https://lists.quantum-espresso.org/mailman/listinfo/users
>
>
> _______________________________________________
> Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)
> users mailing list users at lists.quantum-espresso.org
> https://lists.quantum-espresso.org/mailman/listinfo/users