Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Topics - bubble

Pages: [1] 2
Hi,  I intend to invesigate a system (i.e.  a 2D material on the metal substrate) using SGF. The version of QATK used here is the newest, and for this system the metal substrate is set to half-infinite according to SGF (in other words, I intend to place the 2D material  on one surface of the metal substrate, while the other "surface" of the metal substrate being set to has the periodic boundry condition).

Now I have several questions about SGF. 

1. SGF method can be used to this "2D material + metal substrate" system?

2.  SGF method can be used to optimize the structure of this system ?

3. SGF method can be used to calculate the electronic structure of this system? (e.g. energy band, DOS etal.)

4. The charge transfer between the 2D material and the metal substrate can be calculated using SGF?

Thank you for your reply!

General Questions and Answers / Boundary Condition at finite bias
« on: December 18, 2017, 15:13 »
For a device simulation, when it is under the finite bias, I found the  default boundary  condition for the C direction (Left, Right) are both Dirichlet condition. To my knowledge, this condition means that 'The Hartree potential is zero at the boundary'. However, because the device is under finite bias voltage (that is, the chemical potential, or electropotential is different). Then, it seems that the Dirichlet condition can not apply to left-electrode-center-region interface and right-electrode-center-region at the same time ?  Should I change the default Dirichlet condition to other condition such as  Neumann?

Thank you.

Now I have a question. If I want to adjust the Fermi level of a device. For example, if I set VL=VR=0 V, then I can obtain the transmission spectrum at zero bias. However, If I set VL=VR=0.5 V, can I obtain the transmission spectrum (at equilibrium) corresponds to the whole Fermi level is shifted by 0.5 eV?

Thank you.

General Questions and Answers / a question about PLDOS
« on: November 10, 2017, 08:42 »

In the above tutorial, the last but one figure gives the PLDOS of a 'forward-bias' PLDOS (Vb=0.3V). One can see that in the bias window (energy bewteen eL and eR), there are no states in the right electrode (corresponding the area in the energy range eL-eR is black if z is larger than. e.g. 60 A).   However, the tutorial discussed that ' here the electrons do not have to overcome any potential barrier, so that the current will be higher. '

I want to know what states are used to carry these electrons? Since in the bias window, for the right electrode, it seems that there donot exist any states!

Thank you!

General Questions and Answers / A strange phenomenon in ATK
« on: November 10, 2017, 04:22 »
     Recently, I used ATK 2017 to calculate the I-V curve in a molecular magnetic tunneling junction (MTJ) at its antiparallel (AP) spin configuration (i.e. the spins on two electrodes are opposite).  I found a very strange phenomenon. i.e. the I-V curve is not symmetrical at high negative and positive bias voltages (+-0.5 V). Please note that the geometric structure of MTJ is symmetrical (i.e., there is a mirror symmetry pendicular to z axis, only the spins on two electrodes are opposite ).
     Then, I carefully read some references, and found that this phenomenon is not unique.  For example, see DOI: 10.1039/c4ra09279a (Figure 3b), or DOI: 10.7498/aps.66.198503 (Figure 2b). Please note that although the second example is not written in English, it is not hard to understand the meaning of the figure. In the two examples, the geometric structures is symmetrical (this can also be verified from the LDOS, various eigenstates etc.) , but at high bias voltage, the spin current is asymmetrical. 

  At AP configuration, if the device is symmetrical (have mirror symmetry), one spin in one electrode should be identical to another spin in another electrode. According to symmetry, at +-Vb, the spin currents should be 'identical', except a sign indicating the direction. This should be indenpent on the details of the device. So, is there anything wrong in the calculation?

Thank you in advance.

General Questions and Answers / Is it a bug?
« on: October 24, 2017, 16:58 »
   recently, I investigated a 1D nanodevice with spin-polarized calcualation. The attachment is three files. You can see that for the down spin, the average T(E) and the k-resovled transmision coiffiences is of 10^(-6) order of magnitude (the first figure). Therefore, the IV show that Idown is nearly zero (the third figure). However, the calculated the eigenvalue of the spin-down channel is about 0.38 (see the left-down cornor of the first figure) and the corresponding states (the second figure) are delocalized.

I want to know whether it is a bug? OR, how to explain this strange phenominon?

Hi, I have three questions
1. In the log file, what dose DD mean?
2. Do the DM(U) and DM (D) represent the spin-up and -down components of density matrix?
3. For a given calculation, we use PBE-SGGA and SZP basis set, even though the basis set is relatively small, for my system the SZP contains, for example, 2s, 2p,3d orbitals. I wonder why in the loge file for a given atom,the  DM(U) and DM (D) are only two numbers rather than matrix?   

Thank you!

General Questions and Answers / Why Ne=48?
« on: September 22, 2017, 10:24 »

Recently, I found in the above tutorial, in the 'unit-of-work' section, the tutorial indicates that the 'contour energy point' Ne is usually 48, and this confused me.

To my knowledge, the  contour intergral is divided into equilibrim and non-equilibrim parts. For the former the default value is 30, as indicated in the 'New calculator' panel. However, for the latter, only the default 'Real axis point density', 0.001Ha (about 0.0272eV), is given in the 'New calculator' . What I want to know is:

1. Is the energy point in the  non-equilibrim part equal to (about) 18 (i.e. 48-30=18)?
2. How to fix the number 18 or 48? To my knowledge, in the real axis the integral length should be dependent on the chemmical potential difference (i.e. uL-uR)? If so,  whether the energy point used in the real axis should be equal to (uL-uR)/0.001Ha, rather than 18?

Thank you!

General Questions and Answers / Spectra current
« on: September 19, 2017, 14:14 »
In the IV-analyzer,  what is the expression for the 'spectral current' ?  is it equal to T(E)(fL-fR)? (Here, T(E) transmission spetrum of the given bias, f the fermi function of the L, R electrode), i.e. the integral of the spectra current is the corresponding current?

Thank you!

In the tutorial
For the "Check the electrode geometry and size" section, the "Note" indicates that 'The reason for this is that only the nearest-neighbor interactions of the electrode along C are kept in the device calculation (this is necessary in order to describe the electrodes as semi-infinite rather than fully periodic along C).'

Is this point only suitable for the SE method considered here ? Is is suitable for the DFT method? and how to understand it?

It means that the NN interaction along, for example, A or B direction is discarded? I think this will cause incorrect results for the actural devices.

 Thank you!

General Questions and Answers / The phonon spectrum
« on: September 13, 2017, 15:39 »
   Recently, I want to calculate the phonon sprectrum of 2D boron sheet. The experiment has shown that this system is stable. But I found negative frequencies at PBE/DZP level. I wonder whether the theoretical level that I considered is too low? Any suggestion?

Thank you!

I have a question. Now I want to get the degeneracy of a molecule. Using CH4 as an example (ATK 2017 version), I calculated its molecular energy spectrum. The input py file is in the attachment. But after calculation in the log file there exits a mistake, i.e., '  'float' object has no attribute '__getitem__'.

According to the manual, the return type of degeneracy() in keyword MolecularEnergySpectrum is list of float. Whether the notation degen[0]【i】 that I used in the script is wrong? If so, however, I found the notation of 'occupation()' (the return type of which is also list of float) can give correct results in the log file. In addition, if I change the notation degen[0]【i】 to degen, the result is also correct.

Thank you for your reply.

   Does the ATK2017 version provide different components of DOS? To my knowledage, the old version could only provide s p d...for DOS. In the new version, are the px, py, pz, dxy, dyz.... components avaible?

Thank you!

General Questions and Answers / transmission eigenstate
« on: July 28, 2017, 02:50 »
   What is the difference between the left and right contributions for the transmission eigenstate (TE) calculation? 
   For example, If the bias voltage is finite, I think that, for a given energy, the electron can be viewed as incident from the left and right electrodes at the same time, and the final current is the result of, e.g.  'right contribution minus left contribution'. And based on this consideration,  we can obtain the left and right contributions for the TE?
  Am I right?
Thank you !

General Questions and Answers / projection
« on: July 21, 2017, 08:07 »
   In ATK2017, is it possible to obtain the projected DOS more accurately? For example, to my knowledge, the old ATK only provides the d-states as a whole for projected DOS, in the new version the d-states projected DOS can be further divided into dxy, dyz, dzx...?

Pages: [1] 2