HybridSOGGA.HSE06 for device calculation showing error at central region

[sumitn60]

Hi,

I was trying to test whether the noncolinear SOC along with hybrid functional is implemented in version 2021.06-SP2 using HybridSOGGA.HSE06 in the place of SOLDA.PZ (which ran perfectly). The calculation stopped while calculating the density matrices of the central region showing the error:

atkpython: /slowfs/qatkdev2/bamboo/de02vlbamboo10/bamboo-agent-home/xml-data/build-dir/QL-Q2-BLD/c++/src/physics/blocksparseatompairmatrix.cpp:361: BlockSparseAtomPairMatrix& BlockSparseAtomPairMatrix::operator+=(const BlockSparseAtomPairMatrix&): Assertion `spin_type_ == source.spinType()' failed.

My question would be if it's not implemented why it's only showing problems while doing the calculation of the central region why not the electrodes (attached the log file for your reference)?

P.S: Previously I have asked about the availability of the noncolinear SOC with hybrid functional for device calculation within the NEGF framework is available or not and got the reply from Ulrik that it would be implemented in the newer release. Could you please confirm if the error I'm getting is due to the non-implementation?

Waiting for your quick response!

Thank you,

Sumit

[Anders Blom]

HSE SOC for NEGF was implemented in 2022.03

[sumitn60]

Dear Anders,

Thank you for your reply. I have tried to do the calculation in the 2022 version and the calculation ran without an error but the off-diagonal parts in the transmission matrix i.e. T(real-up-down) and T(imag-up-down) are zero. And weirdly it's printing the same values for T(up) and T(down). But the same system with LDA shows polarization in transmission: https://pubs.acs.org/doi/full/10.1021/acs.jpclett.2c03747, which reverses while we change the handedness of the helix. Could you please help me find out the problem?

Thank you,

Sumit

[Anders Blom]

That's a deeper scientific question that would require careful analysis to answer, provided the software does what it's supposed to. I think it would be interesting to test this system internally though, to verify there are no bugs. Do you mind sending me the input scripts for both the LDA and HSE cases by email?

Note though, that I am not sure there is any particular point in using HSE for this system, it's not making the Au electronic structure better (in fact, maybe worse). It is however still very cool that you can run HSE-NEGF-SOC with full transmission spectrum in 6 hours, don't you agree!

[sumitn60]

Yes, I agree that the calculations are fast, but what I'm concerned about is the results it is providing.

I have sent the input files through email. Please have a look.

Thank you,

Sumit