QuantumATK Forum

QuantumATK => General Questions and Answers => Topic started by: rahulprajesh on May 20, 2011, 11:01

Title: geometric constraint
Post by: rahulprajesh on May 20, 2011, 11:01
Sir,
i've a two probe system with 36(left electrode)+(18+32+18)-central Region + 36(right electrode)
i want to constraint the electrode as well as electrod element in scattering region(as recommended in tutorials). i tried with
 geometric_constraints = range(0,53)+range(86,139)
it shows error ------ saying indices can't be more than 68.
--------------------------
tell me if my syntax is right or wrong ??
--------------------------

and one more thing is the indexing order same how they appear in script ?
--------------------------
Thank you.
Title: Re: geometric constraint
Post by: Anders Blom on May 20, 2011, 11:41
You shouldn't count the electrode atoms. Since you are only relaxing the central region, the constraints refer to these atoms only. Yes, the indices correspond to the atoms, in the order they appear in the geometry definition.
Title: Re: geometric constraint
Post by: rahulprajesh on May 20, 2011, 11:56
You shouldn't count the electrode atoms. Since you are only relaxing the central region, the constraints refer to these atoms only. Yes, the indices correspond to the atoms, in the order they appear in the geometry definition.
so should i put it like--- range(36,53)+range(86,139) ?
----
one more doubt if leave the geometric constraint(VNL) text box blank what does it mean.?
----
Title: Re: geometric constraint
Post by: Anders Blom on May 20, 2011, 14:02
No, there are only 68 atoms in the central region that you can constrain.

But wait, is this for ATK 11.2? The system can't be valid in 11.2, if you have fewer atoms in the central region than in the two electrodes combined (see here (http://quantumwise.com/documents/tutorials/latest/UpgradeGuide/index.html/chap.upgrade.html#sect1.upgrade.twoprobe)).

So something is wrong here.
Title: Re: geometric constraint
Post by: rahulprajesh on May 20, 2011, 15:43
No, there are only 68 atoms in the central region that you can constrain.

But wait, is this for ATK 11.2? The system can't be valid in 11.2, if you have fewer atoms in the central region than in the two electrodes combined (see here (http://quantumwise.com/documents/tutorials/latest/UpgradeGuide/index.html/chap.upgrade.html#sect1.upgrade.twoprobe)).

So something is wrong here.

but sir the indices i mentioned here belong to the central region so why it's not correct ?
Title: Re: geometric constraint
Post by: Forum Administrator on May 21, 2011, 11:42
Yes but how many atoms in the central region? It seems to be 18+32+18=68, and that's why the index is not allowed to be larger than 67. But it's weird, because the electrodes have 32, so there is some mismatch somewhere. I can't say anything more without actually seeing the geometry.