QuantumATK Forum

QuantumATK => General Questions and Answers => Topic started by: anjana on September 16, 2012, 23:38

Title: calculate IV curve for charged molecules
Post by: anjana on September 16, 2012, 23:38
I want to calculate the IV curve for a molecule which is charged. In the two probe calculation, I did not see any options for charge insertion. I even tried to include some charge inside the LCAO calculator, but it generated errors. Does that mean I cannot obtain the IV curve for a system wherein the  molecule in the device region has some charge???

Title: Re: calculate IV curve for charged molecules
Post by: kstokbro on September 17, 2012, 18:17
The charge of the central region is determined selfconsistently, thus, only the charge of the electrodes can be set.
To control the charge of the central region you may insert a gate, but note that the charge will not change in integer steps, since you have an interacting system.
For a weekly interacting system you may not be in the coherent regime and need to model the molecule incoherently in which case it will change charge in integer steps.
Please check the tutorial
http://www.quantumwise.com/documents/tutorials/latest/BenzeneSET/index.html/
for further discussion and examples of these aspects.

Title: Re: calculate IV curve for charged molecules
Post by: anjana on September 17, 2012, 19:41
Ok. That seems to make some sense. So if my molecules in the device region has three extra electrons, can I assign the gate voltage to be -3 eV, and do the device calculation??
Title: Re: calculate IV curve for charged molecules
Post by: Anders Blom on September 17, 2012, 22:22
In this case it is important to consider the real physics of the problem. How do the extra electrons end up in the central region, what is the physical mechanism that keeps them there? If it's some local potential, you can apply a local shift to individual atoms (or a group of atoms) to simulate this effect, by constructing a potential well for the electrons, localized in space. This can also be achieved by inserting a metallic gate and applying a potential to it.
Title: Re: calculate IV curve for charged molecules
Post by: anjana on September 18, 2012, 02:06
Actually I am trying to calculate the current through carborane cage with dithiophene .I have attached the structure here. With gaussian calculation, it came out that the structure has stable state with charge=-3 and spin =0; And now I want to calculate current through it using atomistix. I want to use the optimize structure from gaussian calculation, including the charge, which makes the molecule stable. This is why i needed to add some charge in the central region (and not even in the contacts). What can I do to realize this system  What is the use of externalpotential() function, that is included in atomistix?? will that be helpful??
Title: Re: calculate IV curve for charged molecules
Post by: Anders Blom on September 18, 2012, 11:29
The question is, would the charge stay on the molecule when it's attached to infinite reservoirs of electrons (i.e. the electrodes in a transport system)?

If you think they do, they must be localized to certain atoms/side groups, and in this case you would need to shift the potential locally there. There is a (somewhat overcomplicated) usage example in the reference manual (http://quantumwise.com/documents/manuals/latest/ReferenceManual/index.html/ref.atomicshift.html).