The calculation did not converge to the requested tolerance!

[Megan]

Hi!
   I constructed a organic molecule device which have Co chain lead , and its central region is Benzene. I calculated the IVcurve,but the calculation did not converge to the requested tolerance!
   I think my basic parameter settings are right. the key point is the interface action of the Co chain lead and the Benzene.
I want to know how to cantact the Co chain lead and the Benzene.
  could you please tell me that how can i fix this problem?   Thank you very much!

[Jess Wellendorff]

Is the Co chain supposed to be spin-polarized? What does your script and structure look like?

[Megan]

Thanks for your reply.
I used spin in the Co chain. as you can see, I dont know how to decide the interface reasonably. Can you help me ?

[Jess Wellendorff]

From chemical intuition, I see no reason why the closed-shell molecule in the image should bond to the Co chain. Are you sure all of those carbon atoms should be hydrogen-saturated?

[Megan]

this is my constructure.
first I optimized it, then I used the optimized constructure to calculate the IV curve. when i increased the voltage to 0.2V, it cant converge.
 

[Anders Blom]

You could try to increase the XY box to have more vacuum around the molecule, but generally this doesn't seem to be a physically relevant system...?

[Megan]

I have do it .but it  have no effect.you mean that my structure is unreasonable ?can you heip me improve it ?I use this lead because of I read a paper .

[Umberto Martinez]

your structure is indeed quite different from this last picture.
You have to rotate by 90 degrees around the x axes your molecule.

out of curiosity, can you add the full reference of that paper?

[Megan]

My structure is a single molecule spin valve .This is that paper.but,this document  is big it can not  upload.so i list the paper nameand abstract :'Bias-Controlled Giant Magnetoresistance throughCyclopentadienyl−Iron Multidecker Molecules' .'We investigated giant magnetoresistance (GMR) across cyclopentadienyl−iron multidecker molecules [Fem(C5H5)n] using thedensity functional theory and nonequilibrium Green’s function techniques.'
 The reason why this paper use Co chain lead is :'we choose Co chain leads for the purpose of reducing the computational cost. The single wire channels,although not good models of 3D substrates, deserve investigations and can be used to understand some basic
mechanisms in molecular junctions.'