Messages

31

Installation and License Questions / parall in Windows XP system

« on: March 6, 2010, 11:03 »

Dear all,
    I have the question that, can ATK do the calculation in parall on Windows XP system? If yes, how to deploy it?

thank you！

32

General Questions and Answers / Can the ATK2009.12 handle with the gate volatages for other system

« on: February 3, 2010, 12:35 »

Dear everyone,

Can the ATK2009.12 handle with the gate volatages for other system except the example about the Graphene Junction? it seems that it can not be applied  in this version, right?

thank you.

33

General Questions and Answers / Can the ATK2009.12 handle with the spin-dependent system

« on: February 3, 2010, 12:16 »

Dear everyone,
     I have the question that, how to set up parameters to investigate the the spin transport properties of one magnetic molecule in the version of ATK2009.12,  like that in the version of 2008.10, there are some paremeters about Initial spin of the electrodes and central molecule.
How to get the spinup and down transmission spectrum? does the new version give up these function?
Thank you very much.

34

General Questions and Answers / Re: what do the three SCF calculations mean?

« on: November 14, 2009, 11:23 »

thank you very much, Norland.

the second SCF that convergence trick is quite interesting.

your explanation is quite detailed. great job!!!! 

Come on!

35

General Questions and Answers / resolved

« on: November 14, 2009, 08:49 »

resolved

36

General Questions and Answers / what do the three SCF calculations mean?

« on: November 14, 2009, 08:34 »

Dear everyone,

    I have not the complete understanding about the two-probe calculations. In the output file(Voltage=0), there is three SCF calculations, as following,
the first one is about the Electrodes Calculation, second is Equivalent Bulk Calculation (Initial Density for TwoProbe), last one is TwoProbe Calculation.

anyone could give an explanation,respectively?

thanks a billion!

37

General Questions and Answers / Re: calculations are finished partialy

« on: July 23, 2009, 03:08 »

Dear Zh,
    Oh, now I have uploaded the out file(the end of this file). It shows that the self-consistent calculation for the bias voltage of 1.3 V is converged at 48th step. Another, I think if the DFT SC calculation is not converged it will not give any physical properties in the out file including the current, DOS, Energy Spectrum, and Total energy, etc.
   Thank you for you attention and patient. We should keep it warm and give any possible reason.

38

General Questions and Answers / Re: calculations are finished partialy

« on: July 22, 2009, 10:28 »

Dear zh,
    thank you for your replies!
    Maybe, these problems of disk space (the available space on my local hard disks is about 8.9GB) and power shut-off or the kill command can be eliminated.
   I have uploaded this script produced by VNL (I have juggled with it  ).
you can give the diagnosis of this script and tell me the errors if some parameters are not appropriate. Thank you very much!

39

General Questions and Answers / calculations are finished partialy

« on: July 22, 2009, 02:39 »

Dear everyone,

when I calculate my two-probe system, one strange thing happens. I have successfully obtained the transport properties from 0 to 1.2V, including Total energy, Transmission coefficients, Transmission eigenstates, Transmission eigenvalues, and Transmission spectrum. While at 1.3V,  
the job is finished without calculating the Transmission eigenstates, Transmission eigenvalues, and Transmission spectrum.(all the parameters are the same from 0 to 2.0V). The end of the .out file is as follows.  thank you very much!

# ----------------------------------------------------------------
# Total Energy           = -57955.8133 eV
# Potential Energy       = -57955.8133 eV
# Ion Kinetic Energy     = 0.0000 eV
# ----------------------------------------------------------------
# NumOrb: Energy of system
# ----------------------------------------------------------------
# Charge                  = 636.5473
# Band Structure Energy   = -4711.1567 eV
# ----------------------------------------------------------------
# El. Kinetic Energy      = 33326.2244 eV
# NL Energy               = -27236.4266 eV
# dEna                    = -547.2404 eVTotal
# dUscf                   = 4715.1972 eV
# XC Energy               = -29262.3599 eV
# ENA                     = 9545.7728 eV
# Eions                   = -48496.9808 eV
# ----------------------------------------------------------------
# Electron kinetic energy = 33326.2244 eV
# Electrostatic energy    = -62019.6777 eV
# Exchange-corr. energy   = -29262.3599 eV
# ----------------------------------------------------------------
# Total potential energy  = -57955.8133 eV
# ----------------------------------------------------------------
Total energy
-57955.8132739 eV
[ 0.04129895  0.30779887  0.30747487  0.04129895  0.23893055  0.41008104
  0.27410268  0.23893055  0.23916114  0.2739539   0.41003894  0.23916114
  0.04129895  0.30779887  0.30747

40

General Questions and Answers / Re: Error in removing file？

« on: July 15, 2009, 06:21 »

Ok,Thank you very much.
Maybe like you said.

41

General Questions and Answers / Error in removing file？

« on: July 12, 2009, 02:45 »

Dear everyone，

   When I calculate the transport properties of my two-probe system, one error happens.
when the bias voltage is increased to 0.7V, and "Restore self consistent calculation from check point file " using the .nc file that of 0.6V,the errro happens like that,

# Error in removing file [au-Mol-au07.nc]: Permission denied
#  Warning : Could not rename file D:\AYP\QuantumWise\ATK 2008.10\bin\_au-Mol-a
u07_HCOIJC.nc to au-Mol-au07.nc !

Surprisingly, all the calculated physical properties have been obtained successfully!
thanks!

42

General Questions and Answers / a question about the source code to define the broadening function in ATK

« on: June 1, 2009, 02:24 »

Dear Dr.
     when I try to understand the NEGF+DFT of ATK deeply,  I notice some different definitions on the broadening function.  following Eq.1 (Phys. Rev. B 65, 165401 (2002)) is same to Eq.2(J. Phys. Chem. C 2008, 112, 1685 1693), another Eq.3(PHYSICAL REVIEW B, VOLUME 64, 115411) is different. I want to konw which Eq is correct for the source code compiled in ATK.
thanks.

43

General Questions and Answers / Re: setting the quantum number to calculate the HOMO and LUMO

« on: May 13, 2009, 02:58 »

thank you Anders Blom  for your answer!

44

General Questions and Answers / Re: strange result

« on: May 13, 2009, 02:46 »

yes, like you saied, I found that it's more slow to run the job by using the hpmpich. thank you very much for  your remind!

45

General Questions and Answers / setting the quantum number to calculate the HOMO and LUMO

« on: May 12, 2009, 03:16 »

Dear everyone,

I have some confusion  about setting the quantum number to calculate the HOMO and LUMO of two-probe system. for example, for the molecule line C4BPH12, there are 40 outer-shell electrons, usually based on that two electrons occupy one orbit, so the 40 outer-shell electrons will occupy 20 orbits, and the quantum number of HOMO should be 19 (the quantum number of the first one lowest energy orbit is 0 ), for LUMO is that 20. however, it seems not like that through calculating the energy spectrum, as following,

0   -4.64
1   -4.36
2   -4.14
3   -3.8
4   -3.77
5   -3.76
6   -3.65
7   -3.06
8   -2.94
9   -2.7
10   -2.45
11   -2.19
12   -2.09
13   -2.07
14   -1.86
15   -1.67
16   -1.38
17   -1.31
18   -1.24
19   -0.71
20   -0.27
21   0.77
22   1.54
23   2.04
24   2.12
25   2.26
26   2.76
27   2.84
28   3.01

is it possible that two special electrons occupy two orbits, so, the quantum number of HOMO should be setted to 20(-0.27eV) .
or, the energy -0.27 could be considered as 0, so, the quantum number of HOMO and LUMO should be setted to 19 (-0.71eV) and 21(0.77eV).
how to explain it resonablely?
Another, for the molecule owning odd outer-shell electrons, ie. Si5Li cluster. there are total 23 electron, do they should occupy 12 orbits, and which orbit is occupied with only electron, HOMO ?
Thank you!