1
General Questions and Answers / Re: General Questions
« on: October 16, 2012, 20:46 »
And with regards to 7.?
Show Posts
This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.
Pages: [1] 2
2
General Questions and Answers / Re: General Questions
« on: October 10, 2012, 15:37 »
Thank you zh for your thorough and meaningful response. It cleared up a lot of issues.
I still have trouble with points 4. and 5.
4. Indeed, the applied voltage shifts the fermi level up/down and the energy range needs to be adjusted. My questions is what general guideline should I follow in order to adjust the energy range appropriately for a given bias?
5. As per this post: http://quantumwise.com/forum/index.php?topic=21.0
The voltage drop is calculated by:
voltage_drop = calculateEffectivePotential(finite_bias) - calculateEffectivePotential(zero_bias)
With the electrode constraint 'DensityMatrix'.
Is it possible to find the effective potential if my self-consistent calculations were done using the semi-empirical Huckel method?
I also have one more point to add to this discussion:
7. Even though the workfunction of the gate was discussed in a previous post: http://quantumwise.com/forum/index.php?topic=976.0, I still don't understand what work function is used in the metallic region to carry out the self-consistent calculation. Is it 0 eV? Perhaps, an example would help clarify:
Suppose I have a:
channel workfunction of 4eV.
I'd like to specify a gate workfunction of 5eV (which I can't pre-simulation).
Simulation running at various gate voltage range [-2,2] V.
a)If I can't specify a work function pre-simulation, how do I know that what is happening at -2V is not in reality happening at -20V?
b)In this example, what must I shift my x-axis by?
Thanks again,
gM
I still have trouble with points 4. and 5.
4. Indeed, the applied voltage shifts the fermi level up/down and the energy range needs to be adjusted. My questions is what general guideline should I follow in order to adjust the energy range appropriately for a given bias?
5. As per this post: http://quantumwise.com/forum/index.php?topic=21.0
The voltage drop is calculated by:
voltage_drop = calculateEffectivePotential(finite_bias) - calculateEffectivePotential(zero_bias)
With the electrode constraint 'DensityMatrix'.
Is it possible to find the effective potential if my self-consistent calculations were done using the semi-empirical Huckel method?
I also have one more point to add to this discussion:
7. Even though the workfunction of the gate was discussed in a previous post: http://quantumwise.com/forum/index.php?topic=976.0, I still don't understand what work function is used in the metallic region to carry out the self-consistent calculation. Is it 0 eV? Perhaps, an example would help clarify:
Suppose I have a:
channel workfunction of 4eV.
I'd like to specify a gate workfunction of 5eV (which I can't pre-simulation).
Simulation running at various gate voltage range [-2,2] V.
a)If I can't specify a work function pre-simulation, how do I know that what is happening at -2V is not in reality happening at -20V?
b)In this example, what must I shift my x-axis by?
Thanks again,
gM
3
General Questions and Answers / General Questions
« on: October 4, 2012, 06:24 »
Dear QuantumWise team,
I've been exploring ATK and have come up with a few general questions:
1. As I understand, the electron concentration can only be calculated with the DFT package. If I'm using the semi-emprical Huckel model I cannot determine the concentration. What is the reason for that?
2. I've looked in the manual, and tutorials, but I still have trouble understanding the meaning of ElectronDifferenceDensity. Could someone please elaborate on that or refer me to a detailed explanation? Could I extract electron concentration from this?
3. In the "ballistic coherent tunnelling current" regime, an electron accelerates indefinitely in the central region (due to a finite voltage bias between the two terminals) until it reaches an electrode where it is absorbed or is scattered in the central region if there are deformities from periodicity, there is no drift or diffusion currents...true or false? In this way, does it matter what material type one is using if its perfectly periodic?
4. When the transmission coefficient is calculated, the user has the option to define the energy range. Later, the product of the transmission coefficient and the fermi function is integrated over that range to give an electric current. If one computes for larger energy range, you would get a larger current and if one computes for narrower range, smaller current. How can I determine the correct range that will give accurate current?
5. How can I plot the electric field strength in the XY plane in the scattering region?
6. Can the physical temperature of a device configuration be adjusted? How?
Kind Regards,
gM
I've been exploring ATK and have come up with a few general questions:
1. As I understand, the electron concentration can only be calculated with the DFT package. If I'm using the semi-emprical Huckel model I cannot determine the concentration. What is the reason for that?
2. I've looked in the manual, and tutorials, but I still have trouble understanding the meaning of ElectronDifferenceDensity. Could someone please elaborate on that or refer me to a detailed explanation? Could I extract electron concentration from this?
3. In the "ballistic coherent tunnelling current" regime, an electron accelerates indefinitely in the central region (due to a finite voltage bias between the two terminals) until it reaches an electrode where it is absorbed or is scattered in the central region if there are deformities from periodicity, there is no drift or diffusion currents...true or false? In this way, does it matter what material type one is using if its perfectly periodic?
4. When the transmission coefficient is calculated, the user has the option to define the energy range. Later, the product of the transmission coefficient and the fermi function is integrated over that range to give an electric current. If one computes for larger energy range, you would get a larger current and if one computes for narrower range, smaller current. How can I determine the correct range that will give accurate current?
5. How can I plot the electric field strength in the XY plane in the scattering region?
6. Can the physical temperature of a device configuration be adjusted? How?
Kind Regards,
gM
4
General Questions and Answers / Re: Bias loop question
« on: September 18, 2012, 19:15 »
Perfect. Thanks.
5
General Questions and Answers / Bias loop question
« on: September 18, 2012, 16:18 »
Dear All, I have a small question about creating a bias loop. What difference does it make (if any) between creating a bias loop like this:
for bias in numpy.arange(0.,1.,0.1)*Volt:
device_configuration.setCalculator(
calculator(electrode_voltages=(0.5*bias,-0.5*bias)),
initial_state=device_configuration
)
versus one like this:
for bias in numpy.arange(0.,1.,0.1)*Volt:
device_configuration.setCalculator(
calculator(electrode_voltages=(bias,0)),
initial_state=device_configuration
)
Aka. varying both terminals versus varying one terminal and keeping the other grounded with an equivalent potential difference in both cases.
Regards.
for bias in numpy.arange(0.,1.,0.1)*Volt:
device_configuration.setCalculator(
calculator(electrode_voltages=(0.5*bias,-0.5*bias)),
initial_state=device_configuration
)
versus one like this:
for bias in numpy.arange(0.,1.,0.1)*Volt:
device_configuration.setCalculator(
calculator(electrode_voltages=(bias,0)),
initial_state=device_configuration
)
Aka. varying both terminals versus varying one terminal and keeping the other grounded with an equivalent potential difference in both cases.
Regards.
6
General Questions and Answers / Re: The calculation of Indium Arsenide
« on: September 14, 2012, 19:40 »
Thanks Dr. Blom. That fixed the issue.
7
General Questions and Answers / Re: The calculation of Indium Arsenide
« on: September 13, 2012, 02:34 »
Hello, I have convergence issues when trying to do an SD scan of a 5nm long, 1nm wide InAs nanowire. The simulation reaches 200 iterations and warns that it finished without convergence. I use the following:
Extended Huckel Calculator
k-point sampling 3x3x100
CerdaHuckelParameters.Arsenic_InAs_GW_Basis
CerdaHuckelParameters.Indium_InAs_GW_Basis
I have attached the script. Could anyone tell me what I'm doing wrong with it? Should I increase the k-point sampling?
Extended Huckel Calculator
k-point sampling 3x3x100
CerdaHuckelParameters.Arsenic_InAs_GW_Basis
CerdaHuckelParameters.Indium_InAs_GW_Basis
I have attached the script. Could anyone tell me what I'm doing wrong with it? Should I increase the k-point sampling?
8
General Questions and Answers / Ballistic transport
« on: August 24, 2012, 00:55 »
Hello and thank you for reading. With respect to ballistic transport, if the length of a metallic wire is increased passed the mean free path of the metal, then ballistic regime is no longer dominant as there is scattering. I was wondering if ATK can capture this phenomenon and if there have been experiments/scripts on this?
Regards.
Regards.
9
General Questions and Answers / Re: regarding gate design
« on: August 24, 2012, 00:38 »
Ok thanks. So regarding Ash's question, what is the workfunction of the metal gate and/or how is it determined exactly?
Regards.
Regards.
10
General Questions and Answers / Re: regarding gate design
« on: August 17, 2012, 23:15 »
So is the default value for the metal workfunction of the gate 4eV?
11
General Questions and Answers / Re: Gold simulation
« on: July 25, 2012, 17:55 »
Thanks Dr. Blom. Is it necessary for the x-component to by twice its current value or just a little bigger is fine as long as the bonds of the end atoms do not show anymore? Referring to the attached, is there any difference between simulating B or C, except the fact that C has more vacuum in the X direction?
12
General Questions and Answers / Re: Gold simulation
« on: July 25, 2012, 15:45 »
Thank you Dr. Blom for your quick response. Perhaps I'm not setting my system up correctly. As noted above I'd like to setup a gold nanowire with certain width/length/height. As I understand correctly, the program looks at the periodicity in the C direction around the edges of my "bulk configuration" and duplicates that several times to create electrodes which extend infinitely (in the C direction only) approximating bulk properties at both ends. But, the X and Y dimensions in the central region, and electrodes should be limited to the width/height of the wire without extending further (infinitely periodic in those directions). How do I achieve exactly that?
Thanks again
!
Thanks again
!
13
General Questions and Answers / Re: Gold simulation
« on: July 20, 2012, 23:34 »
Thank you for your informative response. I would like to pose a few more questions:
1. So, should I expect any change if I were to simulate a structure similar to the one simulated only 10nm long? Why / why not?
2. What exactly are dielectric / metallic regions in terms of this program? If I were to build a metallic region with ATK and set a certain voltage on it, near a gold wire (with dielectric in between), how does it impact the conductance in the gold wire?
3. How do I determine which way the electric field from that metallic region is facing? Is it "away from the metallic region" when positive voltage is applied, and "towards the metallic region" when negative voltage is applied (treatment of electric field due to infinite charged sheet)?
4. What parameters do I need to adjust in the calculator in order to be able to simulate with metallic/dielectric regions? (I've come across changing the Poisson solver to have a Neumann boundary condition instead of Periodic on this forum) Would you be able to elaborate here?
5. Will an influence of a strong electric field, enhance/degrade the transport in the channel under ballistic regime?
1. So, should I expect any change if I were to simulate a structure similar to the one simulated only 10nm long? Why / why not?
2. What exactly are dielectric / metallic regions in terms of this program? If I were to build a metallic region with ATK and set a certain voltage on it, near a gold wire (with dielectric in between), how does it impact the conductance in the gold wire?
3. How do I determine which way the electric field from that metallic region is facing? Is it "away from the metallic region" when positive voltage is applied, and "towards the metallic region" when negative voltage is applied (treatment of electric field due to infinite charged sheet)?
4. What parameters do I need to adjust in the calculator in order to be able to simulate with metallic/dielectric regions? (I've come across changing the Poisson solver to have a Neumann boundary condition instead of Periodic on this forum) Would you be able to elaborate here?
5. Will an influence of a strong electric field, enhance/degrade the transport in the channel under ballistic regime?
14
General Questions and Answers / Re: Gold simulation
« on: July 12, 2012, 21:39 »
Thank you for your response. I was looking at conductance = dI/dV which is the inverse of resistance. Both should change with length and width because ballistic transport doesn't seem to dominate even in 5nm devices such as MOSFETs which are routinely fabricated. Is this an artifact of some approximation where you assume the lattice is infinite along that direction that's why the simulation results don't change with length?
Invoking theory rather than experiment to support the accuracy of your program is not reassuring. The program simply solves numerically a set of theoretical equations. According to you, what should be the length of gold wire beyond which non-ballistic behaviour (standard observed experimental behaviour) occurs (5nm, 10nm 100nm)? Can this program simulate that? Do you expect that the conductance should change with width of the wire at these dimensions?
Your answer so far does not give me confidence that this program can be used to simulate practical 5-10nm transistors which many academic and industrial labs have demonstrated experimentally to behave classically like 1um long devices.
I have attached the conductance plot of the above mentioned gold wires.
Invoking theory rather than experiment to support the accuracy of your program is not reassuring. The program simply solves numerically a set of theoretical equations. According to you, what should be the length of gold wire beyond which non-ballistic behaviour (standard observed experimental behaviour) occurs (5nm, 10nm 100nm)? Can this program simulate that? Do you expect that the conductance should change with width of the wire at these dimensions?
Your answer so far does not give me confidence that this program can be used to simulate practical 5-10nm transistors which many academic and industrial labs have demonstrated experimentally to behave classically like 1um long devices.
I have attached the conductance plot of the above mentioned gold wires.
15
General Questions and Answers / Gold simulation
« on: July 10, 2012, 18:46 »
Hi all and thank you for your attention 
. I simulated the IV characteristics from (-0.3V to 0.3V) of three gold wires with dimensions:
Contact Length (z direction) Width of wire (y direction) Central region length (z direction)
8.16A 20A 26.4A
8.16A 20A 30.6A
8.16A 20A 51.2A
For all three I used k-point sampling of 1,1,100 for left/right electrode accuracy.
The IV characteristics of all three structures look nearly identical when they should theoretically change slope as the resistance of a longer piece of metal should increase and hence less current should pass through. Would you be able to tell me what I may be doing wrong here? I have attached the three scripts for the devices as well as the result I got.
. I simulated the IV characteristics from (-0.3V to 0.3V) of three gold wires with dimensions:Contact Length (z direction) Width of wire (y direction) Central region length (z direction)
8.16A 20A 26.4A
8.16A 20A 30.6A
8.16A 20A 51.2A
For all three I used k-point sampling of 1,1,100 for left/right electrode accuracy.
The IV characteristics of all three structures look nearly identical when they should theoretically change slope as the resistance of a longer piece of metal should increase and hence less current should pass through. Would you be able to tell me what I may be doing wrong here? I have attached the three scripts for the devices as well as the result I got.
Pages: [1] 2