exciton binding energy

[nrbsh]

Dear yambo users and developer,
Using GW0+BSE in yambo-4.0.2-rev.90, I am interested to calculate exciton binding energy of mono-layer phosphorene.
I converged the k-points, BndsRnXs, NGsBlk; and calculate the binding energy for both singlet and triplet excitons, I also consider the dark excitons (using ypp -e s). But my binding energy is 0.5 eV which is about 0.3 eV less than what I found in different literatures (~0.8eV) .

My question is that is there any thing else that maybe I didn't care about that?

For triplet excitons I used

BSEmod= "causal or coupling" # [BSE] resonant/causal/x_coupling/xc_coupling
BSKmod= "SEX" # [BSE] IP/Hartree/HF/ALDA/SEX
BSSmod= "d" # [BSS] (h)aydock/(d)iagonalization/(i)nversion/(t)ddft`
BSENGexx= 0 RL # [BSK] Exchange components
WehCpl


I really appreciate your help.
Zara
Tehran University

[Daniele Varsano]

Dear Zara,
The triplet input is correct.
There can be many reason why experiments are not reproduced. Anyway, did you check first that your GW gap is reasonable?
Next you are dealing with a 2D system, and convergence with the volume of the supercell is often very hard both for GW and BSE.
Did you check it?
In order to speed up convergence you can use a modified cutoff coulomb potential (Phys. Rev. B 73, 205119 (2006).).
Here you need to use the box geometry, even if not well documented, this has been discussed many times in the forum.
See e.g. here:
viewtopic.php?f=14&t=309&p=1793&hilit=truncated#p1793

If you use the box shape do not forget to activate the RIM, and to pay attention to the factor 1/2 in the definition of the box side.

Best,
Daniele

[nrbsh]

Dear Daniele,
Thank you very much for your quick response.

Actually my GW gap is 1.85 eV which is ~ 0.15 eV less than the predicted value.Then I increase the scissor parameter in my BSE calculations but it does not have any effect on the exciton binding energy. (I also check that scissor parameter and KfnQPdb= "E<./SAVE/ndb.QP" have a same results.)

The predicted binding energy of 0.8 eV for phosphorene is not only experimental, there are some theoretical papers that find this value too (For example phys. rev. b 89, 235319 (2014)).

About the volume, I do BSE for both z=17 Angstrom and z=32 A but the increasing of the volume does not have any remarkable effect on the binding energy of excitons.

Best regards,
Zara

[Daniele Varsano]

Dear Zara,
please fill your signature with your affiliation. It is rule of the forum, and you can do once for all in the setting of your profile.

About the volume, I do BSE for both z=17 Angstrom and z=32 A but the increasing of the volume does not have any remarkable effect on the binding energy of excitons.

That's sound a bit strange. Size effects use to play a role. In order to compare results with different volume, convergence parameter should be scaled accordingly unless you gives them in input in terms of energy, I do not know if you are doing that.
In addition, if you are not already doing, you can try to use the RIM, random integration method in order to evaluate correctly the elements of BSE matrix.
In general, If you post a report file we can take a look and see if there's something strange.
Best,
Daniele

[nrbsh]

Dear Daniele,
Thankyou for your response.

I am trying to use RIM, but one more question is that is it possible that this problem comes from pseudo-potential?
I am using P.pbe-n-nc.UPF in quantum-espresso.

Another point is that my optical gap is almost same as the experimental value but the GW gap and the exciton binding energy are less than the expected values.

Best regards,
Zara
Tehran University

[Daniele Varsano]

Dear Zara,

is it possible that this problem comes from pseudo-potential?

Sure, it is possible. Unfortunately I'm not at all expert on the system you are studying so I cannot say more.
Hopefully someone with experience on phosphorene can tell you more.
Best,
Daniele