Top Page > Browsing

about Charge doping ? Date: 2017/06/30 16:59 Name: Viet Hung Nguyen   <viet-hung.nguyen@uclouvain.be> Dear developers, I found in the manual of OPENMX that we can introduce a charge doping by using the keyword "scf.system.charge". This allows for introducing a uniform background charge in the system. If I understand correctly, by introducing this charge doping, we can consider doped materials. Is this correct? For example, if now we consider 2D materials doped by applying a uniform back gate voltage, we should, in principle, add a uniform potential energy to the Hartree potential, similarly to when we model the effects of a local gate, i.e., using the keyword "NEGF.gate.voltage" in the NEGF calculation. Actually, the main effect of a uniform gate voltage is to tune the Fermi level and then the charge density. Hence, I wonder if I can consider this effect by introducing a charge doping? additionally, can I do it for both cases of band structure and NEGF calculations? I thank you very much and look forward to hearing from you. Yours sincerely, Viet-Hung Nguyen

Page: [1]

Re: about Charge doping ? ( No.1 ) Date: 2017/07/01 00:26 Name: Kylin Dear Viet Hung Nguyen I didn't get up your idea about charge doping. I think either you specifically fix a net charge (electrons + ionic cores) for a system to achieve the SCF condition. Or you shift the position of chemical potential (gate voltage) to introduce certain charge doping in the system. And this two conditions was connected via the so-called quantum capacitance where delta_Q = C_quantum(U) * delta_U. but, what is the meaning of "both cases of band structure and NEGF calculations?" Cheers Kylin Re: about Charge doping ? ( No.2 ) Date: 2017/07/03 18:02 Name: Viet Hung Nguyen  <viet-hung.nguyen@uclouvain.be> Thank Kylin very much for prompt reply. I would like to explain again my ideas and questions. - Regarding the idea of charge doping, my main question is Could we consider doped systems by introducing a uniform net charge, i.e., by using the keyword "scf.system.charge"? In particular, I am taking about the case of 2D materials doped by a (uniform) back gate voltage. As discussed, applying a uniform gate voltage can tune the position of chemical potential and hence can change the charge doping in the system. Because I did not find any method to introduce a uniform gate voltage in OPENMX calculations, I hence would like to use this charge doping to model this problem. As you explained, I understand that the answer is yes. - About the applications of this method, i) Regarding the band structure calculations, for a single material, it is, of course, very simple. The effect of this gate voltage/charge doping is simply to shift the position of chemical potential. There will be no effect on the bandstructure. However, the situations could be very different for heterosystems made of two or more materials. In particular, the position of bottom (top) of conduction (valence) bands originated from different materials members with respect to the Fermi level of the whole system could be differently dependent on the charge doping. This has been reported in the literature. What I am wondering is can I use the method mentioned to model this problem? ii) regarding NEGF calculations, the transport properties through a doped channel is an important problem. Similarly, I am wondering if I can use the same method mentioned in these NEGF calculations. For a simple example, if I want to compute I-V characteristics of a graphene device with a uniform gate voltage/ uniform doping, I will introduce the same net charge in all regions of leads and device? Can I do that and is it correct? Could you please confirm again all my points/questions? I would like to thank you in advance very much and I look forward to hearing from you. Yours sincerely, Viet-Hung Nguyen Re: about Charge doping ? ( No.3 ) Date: 2017/07/04 14:20 Name: Kylin Dear Viet Hung Nguyen I still not very sure about your idea and problem. For 1), it would be attributed to the intrinsic property in Density Functional Theory based on the target of the minimal energy state. If I understand it correctly, let's take a example, for two isolated molecules, one is doped and another is not which is possible in physics. Then in the DFT calculation, both molecules were doped in the SCF loops. I think you can give a penalty function to keep one molecule neutral. But if I understand it wrong, could you give me some papers about your heterosystems? For ii), I ever talked with the OZAKI about the NEGF with more than two leads. He told me in mathematically, it is possible to deal with multiply leads (e.g. an extra gate electrode) in DFT calculation. But it is not be implemented in openmx. So I think potential shift is possible, but not too much idea about charge/hole injection. Cheers Kylin Re: about Charge doping ? ( No.4 ) Date: 2017/07/04 21:21 Name: Viet Hung Nguyen  <viet-hung.nguyen@uclouvain.be> Dear Kylin, Here is one of example regarding the effect of doping in heterosystems http://pubs.acs.org/doi/abs/10.1021/acs.jpcc.5b05580. In this paper, the authors studied the doping effects on the band offset, schottky barrier ... in graphene/MoS2 vdW heterostructures. I am sorry if I did not present clearly my questions. To make it more clear, I would like to simplify them as follows: 1 - What is the meaning of charge doping introducing by the keyword "scf.system.charge" in OPENMX? I understand simply that introducing this charge doping allows for considering doped materials? Is this correct? 2 - and if above, I understand correctly, by introducing a charge doping everywhere (two leads and device center) in NEGF calculations, I can consider the transport in a doped channel? Do I understand correctly? Thank you in advance very much for your explanations. Yours sincerely, Viet-Hung Nguyen Re: about Charge doping ? ( No.5 ) Date: 2017/07/04 23:46 Name: T. Ozaki Hi, > 1 - What is the meaning of charge doping introducing by the keyword "scf.system.charge" in OPENMX? > I understand simply that introducing this charge doping allows for considering doped > materials? Is this correct? With scf.system.charge, one can control the number of electrons in the system, where the chemical potential is adjusted so that the total number of electrons can be the specified one. As you considered, the treatment corresponds to charge doping. The keyword is valid for the cluster and band methods, but not valid for NEGF. If you like to see effects of doping for isolated or bulk systems, the method will be useful. > 2 - and if above, I understand correctly, > by introducing a charge doping everywhere (two leads and device center) in NEGF > calculations, I can consider the transport in a doped channel? > Do I understand correctly? In the NEGF calculations, the chemical potentials for the left and right leads are predefined by the band structure calculations for the leads. Thus, one cannot simply control the chemical potential like in the cluster and band methods. If you want to apply gate voltage to scattering region, NEGF.gate.voltage can be utilized, see http://www.openmx-square.org/openmx_man3.8/node116.html In this case, the potential for the scattering region will be shifted. Regards, TO Re: about Charge doping ? ( No.6 ) Date: 2017/07/05 02:43 Name: Viet Hung Nguyen  <viet-hung.nguyen@uclouvain.be> Thank both of you (Ozaki and Kylin) very much, The problem is now clear for me. best regards, Viet-Hung

Page: [1]