Note: This discussion is about an older version of the COMSOL Multiphysics® software. The information provided may be out of date.

Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.

Electric Currents and Transport of a Dilute Species to Generate IV Sweep Plots

Please login with a confirmed email address before reporting spam

Hello community,

I've created a working and functional model that shows the migration of two oppositely charged species over time in response to an applied electric potential through the bulk material. I used electric currents and transport of a dilute species to model this.

The model works great and has no errors. I used a parametric sweep (-20V to 20V) of an applied voltage at an interface which affects the electric potential throughout the bulk material.

I chose to use use a -20 to 20V parametric sweep because in real life, my research group is researching the current-voltage plots of our model under these conditions. In the lab, we induce an electric potential, the ions in the solution migrate, and then we change the voltage to the next step (-20 to -19 to -18 etc.) All the while, we are using other instruments to calculate the current through a specific area. This current, is of course, affected by both the induced electric potential as well as the charged ions.

The issue that I'm experiencing, though, lies in how COMSOL solves for the parametric sweep. In the parametric sweep of the applied voltage at a specified interface, COMSOL solves for each value of parametric voltage independently of the other values. That is, it shows how the ions react to a voltage over time, and then when a new voltage is induced, it considers the concentration profile to have started back at 0, independent of where the ions were previously at the end of the time-dependent study for the previous parametric voltage value.

This is an issue because this is, of course, not how the ions react in real life. In the lab, when changing induced voltages, the ions begin at the concentration profile at which they were associated at the instant the new voltage was applied. They don't instantaneously return to a homogeneous solution in the bulk material.

My question: Is it possible to sort of "link" the parametric sweep somehow so that the concentration profile of the ions at the end of the time step for one voltage is carried over to the beginning concentration profile when changing the parametric voltage? Or maybe some other way so that the simulation is more realistic? Or is this just out of COMSOl's computational capabilities?

We can't simulate an actual IV curve for the parametric sweep in COMSOL, because currently the ions instantaneously return to a homogenous profile when the new parametric voltage step is computed.

Thank you for any help and responses,

Buck Bourdon

11 Replies Last Post Jun 17, 2015, 12:34 p.m. EDT

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Jul 17, 2013, 5:26 p.m. EDT
Hi,


Some thoughts...

Why can't you use continuation option so that the next step uses the previous step solution as its initial condition. But the only thing is that both your charge and concentration distributions are taken forward to the next step as initial conditions and not just concentration, I think. Is this what you want? If yes, you need to go to stationary, study extensions, select continuation and fill in the parameter range. When you select default solver, a parametric continuation solver will appear.

Out of curiosity, why can't you consider time dependent, in which case you have a time dependent voltage boundary condition?


Suresh
Hi, Some thoughts... Why can't you use continuation option so that the next step uses the previous step solution as its initial condition. But the only thing is that both your charge and concentration distributions are taken forward to the next step as initial conditions and not just concentration, I think. Is this what you want? If yes, you need to go to stationary, study extensions, select continuation and fill in the parameter range. When you select default solver, a parametric continuation solver will appear. Out of curiosity, why can't you consider time dependent, in which case you have a time dependent voltage boundary condition? Suresh

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Jul 19, 2013, 12:39 p.m. EDT
Hi Suresh,

Thank you for your response. Your input helped me for future trials - I didn't know of the continuation option within the stationary study step. However, my model requires a time-dependent study. I apologize for not being too explicit in my first inquiry about whether I was using a stationary or time-dependent study.

Could you please elaborate a bit more on your comment about the time-dependent study?

Thanks,

Buck
Hi Suresh, Thank you for your response. Your input helped me for future trials - I didn't know of the continuation option within the stationary study step. However, my model requires a time-dependent study. I apologize for not being too explicit in my first inquiry about whether I was using a stationary or time-dependent study. Could you please elaborate a bit more on your comment about the time-dependent study? Thanks, Buck

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Jul 19, 2013, 8:06 p.m. EDT
Hi,


To reflect your laboratory experiment wherein the voltage is varied in a particular fashion without restarting the experiment each time clearly means you are running a time depedent problem and not some sensitivity runs over a parameter range. So essentially you are running a coupled electrostatics (stationary) and ion transport problem (time dependent) in a time dependent mode for your electro-migration problem. You can therefore apply the voltage across the cell as a function of time via interpolation function (x-axis being time, y-axis the actual value of voltage). For this go to Functions, select interpolation and the default name will usually be int1. And under the boundary condition for voltage, type something like int1(t).

Please note I have not solved such problems in Comsol.


Suresh
Hi, To reflect your laboratory experiment wherein the voltage is varied in a particular fashion without restarting the experiment each time clearly means you are running a time depedent problem and not some sensitivity runs over a parameter range. So essentially you are running a coupled electrostatics (stationary) and ion transport problem (time dependent) in a time dependent mode for your electro-migration problem. You can therefore apply the voltage across the cell as a function of time via interpolation function (x-axis being time, y-axis the actual value of voltage). For this go to Functions, select interpolation and the default name will usually be int1. And under the boundary condition for voltage, type something like int1(t). Please note I have not solved such problems in Comsol. Suresh

Please login with a confirmed email address before reporting spam

Posted: 9 years ago Jun 9, 2015, 2:11 p.m. EDT
Hello,

I know this is quite an old thread. But I appreciate if anyone can help.

I've been trying for several days to do a similar simulation coupling the electric currents and transport of a dilute species modules. I'm not sure if I've set my physics correctly. In the original post Buck mentioned that he used a parametric sweep (-20V to 20V) to generate a potential gradient in the solution/bulk material. One of the questions I have is how would you set the "number of participating electrons" under "Electrode-Electrolyte Interface Coupling" when you do a parametric sweep(with positive and negative values) or applying a bi-phasic pulse to one electrode. Because the electrode changes from anode to cathode if I supply a bi-phasic pulse rather than a constant voltage.

Can anyone provide a sample model that answers this question. Thanks a bunch for any help or comment.

S.K.
Hello, I know this is quite an old thread. But I appreciate if anyone can help. I've been trying for several days to do a similar simulation coupling the electric currents and transport of a dilute species modules. I'm not sure if I've set my physics correctly. In the original post Buck mentioned that he used a parametric sweep (-20V to 20V) to generate a potential gradient in the solution/bulk material. One of the questions I have is how would you set the "number of participating electrons" under "Electrode-Electrolyte Interface Coupling" when you do a parametric sweep(with positive and negative values) or applying a bi-phasic pulse to one electrode. Because the electrode changes from anode to cathode if I supply a bi-phasic pulse rather than a constant voltage. Can anyone provide a sample model that answers this question. Thanks a bunch for any help or comment. S.K.

Please login with a confirmed email address before reporting spam

Posted: 9 years ago Jun 10, 2015, 1:51 a.m. EDT
Hi

First, if you are working in aqueous solutions voltages like ±20 V does not make very much sense due to water splitting.

It all comes down with the electrode reaction: how is the transport coupled to that? If you tell your system, I can reply in more detail.

br
Lasse
Hi First, if you are working in aqueous solutions voltages like ±20 V does not make very much sense due to water splitting. It all comes down with the electrode reaction: how is the transport coupled to that? If you tell your system, I can reply in more detail. br Lasse

Please login with a confirmed email address before reporting spam

Posted: 9 years ago Jun 12, 2015, 9:22 a.m. EDT
Hi Lasse,

Thank you for your response. I'm actually trying to model a scenario of charge balancing. So assume two electrodes dipped in a saline solution. One electrode is grounded and the 2nd electrode is supplying a bi-phasic voltage pulse to the system (Ex pulse: www.frontiersin.org/files/Articles/12119/fneng-04-00009-HTML/image_m/fneng-04-00009-g001.jpg). I would like to visualize the charges in the saline solution with time. The goal is to see some unbalanced charges around the electrodes if the positive area of the pulse is not equal to the negative area of the pulse. (www.nervestudy.com/wp-content/uploads/2013/03/image013.jpg)

I've attached a sample model just with the EC physics.
Hi Lasse, Thank you for your response. I'm actually trying to model a scenario of charge balancing. So assume two electrodes dipped in a saline solution. One electrode is grounded and the 2nd electrode is supplying a bi-phasic voltage pulse to the system (Ex pulse: http://www.frontiersin.org/files/Articles/12119/fneng-04-00009-HTML/image_m/fneng-04-00009-g001.jpg). I would like to visualize the charges in the saline solution with time. The goal is to see some unbalanced charges around the electrodes if the positive area of the pulse is not equal to the negative area of the pulse. (http://www.nervestudy.com/wp-content/uploads/2013/03/image013.jpg) I've attached a sample model just with the EC physics.


Please login with a confirmed email address before reporting spam

Posted: 9 years ago Jun 12, 2015, 11:55 a.m. EDT
I have some troubles with the display with your model as I do not have CAD link: graphics window turns black.

But you have no coupling of transfer to electrostatics. Try Electrochemistry module, Primary Current Distribution. Also, the relative permittivity of a saline solution is somewhere around 70, not 1200 as you have defined.

br
Lasse
I have some troubles with the display with your model as I do not have CAD link: graphics window turns black. But you have no coupling of transfer to electrostatics. Try Electrochemistry module, Primary Current Distribution. Also, the relative permittivity of a saline solution is somewhere around 70, not 1200 as you have defined. br Lasse

Please login with a confirmed email address before reporting spam

Posted: 9 years ago Jun 12, 2015, 4:14 p.m. EDT
I'm not sure I understand why your graphics window turns black. Here I've attached picture describing the model, it's nothing but three cubes. bottom surfaces of the electrodes touch the top surface of the cube that represent the saline. I know the dielectric values for saline are not correct. The values I used were some random values used for testing purposes.

Could you please help me set up the correct modules to do the simulation. I'm not sure I understand how to setup the primary current distribution. I was under the impression that 'electric current' module with 'diluted spices transfer module' would suffice to do the simulation.
I'm not sure I understand why your graphics window turns black. Here I've attached picture describing the model, it's nothing but three cubes. bottom surfaces of the electrodes touch the top surface of the cube that represent the saline. I know the dielectric values for saline are not correct. The values I used were some random values used for testing purposes. Could you please help me set up the correct modules to do the simulation. I'm not sure I understand how to setup the primary current distribution. I was under the impression that 'electric current' module with 'diluted spices transfer module' would suffice to do the simulation.


Please login with a confirmed email address before reporting spam

Posted: 9 years ago Jun 13, 2015, 1:45 a.m. EDT
I wish you can open this; I am using version 5.0. I did no see the solution as my graphics window is black, but please tell me how it looked :)

In the Transport of Diluted Species, activate migration, inactivate convection. Choose Electric potential as "Electric potential (ec)".

Notice my changes in the time stepping and relative tolerance, as well as the diffusion coefficients and charge numbers, they correspond to NaCl. Concentration is 0.3 M, ca. the ocean salinity.

br
Lasse
I wish you can open this; I am using version 5.0. I did no see the solution as my graphics window is black, but please tell me how it looked :) In the Transport of Diluted Species, activate migration, inactivate convection. Choose Electric potential as "Electric potential (ec)". Notice my changes in the time stepping and relative tolerance, as well as the diffusion coefficients and charge numbers, they correspond to NaCl. Concentration is 0.3 M, ca. the ocean salinity. br Lasse


Please login with a confirmed email address before reporting spam

Posted: 9 years ago Jun 13, 2015, 2:50 p.m. EDT
Hello Lasse,

The file did not open in version 4.4. Luckily I had access to a trial of version 5 and using that I was able to replicate the physics. Simulation works exactly as I expected and this solves most of the problems I had. Thank You so much for your help.

P.S. one of the resulting plots is attached.

Saliya K.
Hello Lasse, The file did not open in version 4.4. Luckily I had access to a trial of version 5 and using that I was able to replicate the physics. Simulation works exactly as I expected and this solves most of the problems I had. Thank You so much for your help. P.S. one of the resulting plots is attached. Saliya K.


Please login with a confirmed email address before reporting spam

Posted: 9 years ago Jun 17, 2015, 12:34 p.m. EDT
Hello Lasse,

You were really helpful in understanding the use of transport of diluted spices module in my simulation. Could you please give me any suggestions on how to get the double layer capacitance effects into the same simulation. It seems like this effect is not simulated in the current system. Thank You.
Hello Lasse, You were really helpful in understanding the use of transport of diluted spices module in my simulation. Could you please give me any suggestions on how to get the double layer capacitance effects into the same simulation. It seems like this effect is not simulated in the current system. Thank You.

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.