Author Topic: Problem regarding Monolayer BP Double Gate MOSFET IV characterestics  (Read 2112 times)

0 Members and 1 Guest are viewing this topic.

Offline mashru

  • Regular QuantumATK user
  • **
  • Posts: 9
  • Country: bd
  • Reputation: 0
    • View Profile
Hi,

I am trying to simulate a Black phosphorus channel double gate junctionless mosfet. For junctionless mosfet S/D and channel have the same doping concentration. So based on the concept of junctionless mosfet i considered the same doping concentration whole through the device. Unfortunately i am getting the iv curve which is not showing the any on off behaviour also i have some confusion between TCAD and Atomistic simulation.

Here are the specs of  device and parameters i used
Gate length = 6nm
Doping =1e14
Gate Metal = OV (for both gate metals)
Density Mesh cutoff =100 Hartree
K-point sampling (c-direction) =200
Gate voltage = -0.5 to 0V
Drain Voltage = 0.05V
Poisson solver= PCG and Neuman boundary condition in A direction and Periodic in B direction
Device Algorithm = Greens function is used with recursion sellf energy calculator

When i try to dope the channel region with the doping widget, it shows that channel region corresponds to left electrode

So please help me out about this problem whether i am making any mistake and also in TCAD we usually define workfunction of metal gate explicitly but in QuantumATK do i need to explicitly declare the gate metal workfunction or it will calculate the WF for a specific applied bias.

For the convenience i am attaching the log file of the simulation and alos the py file and screenshots of some of the settings.


« Last Edit: August 10, 2020, 17:03 by mashru »

Offline Petr Khomyakov

  • QuantumATK Staff
  • Supreme QuantumATK Wizard
  • *****
  • Posts: 1290
  • Country: dk
  • Reputation: 25
    • View Profile
One thing that I have noticed that the device is virtually undoped because compensation charge per atom is too small. For 2D material, one should set doping in units of charge per atom, not charge per volume.

The best way is to decide on doping concentration in units of charge per area as it would be normally defined in experimental work, and then recalculate it to charge per atom, and set it in the GUI or in the script directly. In the GUI, there is an option to charge units from charge per volume to charge per atom.

Regarding work function, one does not need to set it explicitly, as it is implicitly defined/computed for given potentials of source, drain and gate electrodes. If one uses the IVCharacteristics study object, then source potential is set to zero by default, and bias voltage V_ds is essentially drain potential. The gate potentials are then defined with respect to zero potential of the source electrode. For example, if gate potential is zero and V_ds=0 too, then work function of the metal gate = -E_Fermi, i.e., given by the Fermi energy of the system, which is computed and outputted into the log file after QuantumATK device calculation is finished.

If one uses IVCurve  analysis object, then the source and drain potentials, as well as the gate potentials, can be set separately with respect to the common zero-energy reference level of the system. 

 
« Last Edit: August 21, 2020, 16:32 by Petr Khomyakov »