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 The information provided may be out of date.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.
Electron energy in DC Discharge Plasma
Posted 05.02.2014, 02:33 GMT-5 Plasma Physics Version 4.3a 21 Replies
Please login with a confirmed email address before reporting spam
Hej,
I am working with the DC discharge Plasma interface. In my plasma model electrons are produced due the decay of tritium (T=>e+T+), in the real plasma case the main part of the electrons are produced due to secondary ionization. the secondary electrons have a mean energy of about 100meV. To make the model easier I decided to produce a larger amount of electrons from the T-decay reaction.
And here comes my question, which is the energy the electrons (from the decay reaction) are produced with in COMSOL? (As I would like to set this enegy to 100meV) Is it the mass difference of the involving species?
Would be nice, if somebody can help me.
Thanks
Laura
I am working with the DC discharge Plasma interface. In my plasma model electrons are produced due the decay of tritium (T=>e+T+), in the real plasma case the main part of the electrons are produced due to secondary ionization. the secondary electrons have a mean energy of about 100meV. To make the model easier I decided to produce a larger amount of electrons from the T-decay reaction.
And here comes my question, which is the energy the electrons (from the decay reaction) are produced with in COMSOL? (As I would like to set this enegy to 100meV) Is it the mass difference of the involving species?
Would be nice, if somebody can help me.
Thanks
Laura
21 Replies Last Post 25.10.2014, 01:40 GMT-4
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Laura,
It depends on the manner in which you've implemented the decay reaction. If you have used a Reaction node or an Electron Production Rate node, then the software does not add a source term to the electron energy density equation (i.e. the electrons are assumed to be produced with zero energy). If you are using an Electron Impact Reaction node, then it will depend on the settings you have specified under the "Collision type" section for this node.
The easiest way to specify a 100 meV initial electron energy for a decay reaction would perhaps be the following:
1) Add an Electron Impact Reaction node
2) Specify the reaction equation (e.g. T=>e+T+) under the "Reaction formula" section
3) Under the "Collision type" section, set the collision type to Ionization and set the energy loss to -100[mV] (note the minus sign)
4) Under the "Collision" section, choose either Arrhenius parameters or Rate constant and specify the required values.
Alternatively, you could use the General Power Deposition node to add a source term for the electron energy density that is a function of the decay reaction rate.
--
Luke Gritter
AltaSim Technologies
It depends on the manner in which you've implemented the decay reaction. If you have used a Reaction node or an Electron Production Rate node, then the software does not add a source term to the electron energy density equation (i.e. the electrons are assumed to be produced with zero energy). If you are using an Electron Impact Reaction node, then it will depend on the settings you have specified under the "Collision type" section for this node.
The easiest way to specify a 100 meV initial electron energy for a decay reaction would perhaps be the following:
1) Add an Electron Impact Reaction node
2) Specify the reaction equation (e.g. T=>e+T+) under the "Reaction formula" section
3) Under the "Collision type" section, set the collision type to Ionization and set the energy loss to -100[mV] (note the minus sign)
4) Under the "Collision" section, choose either Arrhenius parameters or Rate constant and specify the required values.
Alternatively, you could use the General Power Deposition node to add a source term for the electron energy density that is a function of the decay reaction rate.
--
Luke Gritter
AltaSim Technologies
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hello Luke, thanks for your answear.
The model works now more or less. But there is one question left. Can you tell me what the checkbox "use wall for electron density" ( and use wall for electron energy) in the wall settings meens as it influeces the model solution. The help module doesm't answear this question...
Thanks
Laura
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Laura,
These checkboxes simply turn on and off the "Wall" boundary condition for the electron density and electron energy density equations. By default (both boxes checked), the software is applying a flux term to account for losses due to diffusion and (optionally) migration and gains due to secondary emission. If you uncheck a box, the flux term is no longer applied for the corresponding equation, and the software will impose a no flux boundary condition for that dependent variable. The checkboxes allow for independent control of the electron density and electron energy density equations, but you will generally want to have them both activated on a wall.
--
Luke Gritter
AltaSim Technologies
These checkboxes simply turn on and off the "Wall" boundary condition for the electron density and electron energy density equations. By default (both boxes checked), the software is applying a flux term to account for losses due to diffusion and (optionally) migration and gains due to secondary emission. If you uncheck a box, the flux term is no longer applied for the corresponding equation, and the software will impose a no flux boundary condition for that dependent variable. The checkboxes allow for independent control of the electron density and electron energy density equations, but you will generally want to have them both activated on a wall.
--
Luke Gritter
AltaSim Technologies
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Dear Mr. Gritter,
When I use the COMSOL to model a DC plasma, I don't known how to get the Townsend Coefficients in the lookup table. Always, in the literatures, they provide us with an expression of the Townsend Coefficients, but in the COMSOL, we must enter in the lookup table with specific data, how do deal with this problem?
Hoping for your reply. Thanks a lot!
Zhiming Huang
When I use the COMSOL to model a DC plasma, I don't known how to get the Townsend Coefficients in the lookup table. Always, in the literatures, they provide us with an expression of the Townsend Coefficients, but in the COMSOL, we must enter in the lookup table with specific data, how do deal with this problem?
Hoping for your reply. Thanks a lot!
Zhiming Huang
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Zhiming,
If you have an expression from the literature, you can simply generate a table using this expression over a suitable range of mean electron energies and load this table into the model. Typically, the best way to get Townsend coefficients is to generate them for your specific plasma using a Boltzmann solver, which COMSOL has conveniently included in the Plasma Module (the "Boltzmann Equation, Two-Term Approximation" interface). To get familiar with the operation of this interface, I suggest looking at the two example models in the Model Library ("boltzmann argon" and "boltzmann oxygen".
--
Luke Gritter
AltaSim Technologies
If you have an expression from the literature, you can simply generate a table using this expression over a suitable range of mean electron energies and load this table into the model. Typically, the best way to get Townsend coefficients is to generate them for your specific plasma using a Boltzmann solver, which COMSOL has conveniently included in the Plasma Module (the "Boltzmann Equation, Two-Term Approximation" interface). To get familiar with the operation of this interface, I suggest looking at the two example models in the Model Library ("boltzmann argon" and "boltzmann oxygen".
--
Luke Gritter
AltaSim Technologies
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Dear Mr.Gritter，
Thanks very much for your replying!
And I have just taken a look at the Plasma Use's Guide and the Model Library ("boltzmann argon" and "boltzmann oxygen"), in the the "Boltzmann Equation, Two-Term Approximation" interface, we have to provide the cross-section data for the COMSOL to solve the Boltzmann Equation then we get the rate coefficients(k) and the drift velosity(ω), then we can get the Townsend coefficient by α=k/ω. Because the rate coefficients(k) vary by the mean electron energe, so we get a lookup table of the Townsend coefficients that vary by the mean electron energe too, is this right?
After all, no matter we use the Townsend coefficients or not, we have to get the cross-section data first. We can use this data in the Plasma Interface directly for plasma modeling. But if we want to improve the accurate of the model by using the Townsend coefficients, we should put the cross-section data into the "Boltzmann Equation, Two-Term Approximation" interface first to get the Townsend coefficients, then come back to the Plasma Interface for modeling. Have I understood correctly?
zhiming huang
Thanks very much for your replying!
And I have just taken a look at the Plasma Use's Guide and the Model Library ("boltzmann argon" and "boltzmann oxygen"), in the the "Boltzmann Equation, Two-Term Approximation" interface, we have to provide the cross-section data for the COMSOL to solve the Boltzmann Equation then we get the rate coefficients(k) and the drift velosity(ω), then we can get the Townsend coefficient by α=k/ω. Because the rate coefficients(k) vary by the mean electron energe, so we get a lookup table of the Townsend coefficients that vary by the mean electron energe too, is this right?
After all, no matter we use the Townsend coefficients or not, we have to get the cross-section data first. We can use this data in the Plasma Interface directly for plasma modeling. But if we want to improve the accurate of the model by using the Townsend coefficients, we should put the cross-section data into the "Boltzmann Equation, Two-Term Approximation" interface first to get the Townsend coefficients, then come back to the Plasma Interface for modeling. Have I understood correctly?
zhiming huang
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Because the Townsond expressions vary by the pressure and the electric field(P/E), so we can not use the expression to generate a table of Townsend coefficients that vary by the mean electron energies directly.
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Zhiming,
Yes, an appropriate set of electron collision cross-sections is necessary to generate the reaction rate coefficients or Townsend coefficients. The COMSOL plasma interfaces (e.g. DC Discharge) can automatically generate reaction rate coefficients from the cross-sections if you assume the EEDF is Maxwellian, Druyvesteyn, or somewhere in between. A Boltzmann solver can generate reaction rate coefficients or Townsend coefficients for Maxwellian and non-Maxwellian EEDFs, as well as the diffusivity and mobility of the electrons and the electron energy.
--
Luke Gritter
AltaSim Technologies
Yes, an appropriate set of electron collision cross-sections is necessary to generate the reaction rate coefficients or Townsend coefficients. The COMSOL plasma interfaces (e.g. DC Discharge) can automatically generate reaction rate coefficients from the cross-sections if you assume the EEDF is Maxwellian, Druyvesteyn, or somewhere in between. A Boltzmann solver can generate reaction rate coefficients or Townsend coefficients for Maxwellian and non-Maxwellian EEDFs, as well as the diffusivity and mobility of the electrons and the electron energy.
--
Luke Gritter
AltaSim Technologies
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Dear Mr. Gritter,
Sorry to disturb you again. But, in the cross-section data sheet for COMSOL, do we must input the ratio of the statistical weights of the final state to the initial state for EXCITATION reaction? How to get this radio?
From the website(www.lxcat.laplace.univ-tlse.fr/cross_sec_download.php ), it provides plentiful data of cross-section, but in most case it doesn't have the weights ratio data...
Sorry to disturb you again. But, in the cross-section data sheet for COMSOL, do we must input the ratio of the statistical weights of the final state to the initial state for EXCITATION reaction? How to get this radio?
From the website(www.lxcat.laplace.univ-tlse.fr/cross_sec_download.php ), it provides plentiful data of cross-section, but in most case it doesn't have the weights ratio data...
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Zhiming,
If you want the software to automatically calculate the superelastic de-excitation cross-section for the excited state, then the third number in the third line of the cross-section data for the excitation reaction should be set to 1. In this case, the ratio of statistical weights is necessary to calculate the superelastic collision cross-section. If the superelastic collision cross-section is not being calculated (i.e. if the third number in the third line is set to zero), the ratio of statistical weights will not be used in the simulation. These ratios are sometimes included in the literature when the excited states of an atom or molecule are described.
--
Luke Gritter
AltaSim Technologies
If you want the software to automatically calculate the superelastic de-excitation cross-section for the excited state, then the third number in the third line of the cross-section data for the excitation reaction should be set to 1. In this case, the ratio of statistical weights is necessary to calculate the superelastic collision cross-section. If the superelastic collision cross-section is not being calculated (i.e. if the third number in the third line is set to zero), the ratio of statistical weights will not be used in the simulation. These ratios are sometimes included in the literature when the excited states of an atom or molecule are described.
--
Luke Gritter
AltaSim Technologies
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hello,
I want to input the mobility results of the Boltzman-2-term-model for Argon into a microwave plasma model.
But: the export format from the Boltzmann model is far from what the mwp model accepts as input (import) in the look-up table for the DeN or other mobility parameters..
My Comsol hangs for 20 minutes whenever I want to read in 300 pairs of values .. (after modifying the export file of the boltzman model)
Is there a limit for those imports for the mobility look-up tables?
Why is it so mendokusai? troublesome? (i am in v44)
thanks
lukas
I want to input the mobility results of the Boltzman-2-term-model for Argon into a microwave plasma model.
But: the export format from the Boltzmann model is far from what the mwp model accepts as input (import) in the look-up table for the DeN or other mobility parameters..
My Comsol hangs for 20 minutes whenever I want to read in 300 pairs of values .. (after modifying the export file of the boltzman model)
Is there a limit for those imports for the mobility look-up tables?
Why is it so mendokusai? troublesome? (i am in v44)
thanks
lukas
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Lukas,
I'm not aware of a limit for the table size, though I have occasionally seen the software have difficulty loading multiple large tables. Three hundred pairs of values is probably somewhat excessive - you should be able to adequately describe the transport coefficients with much less than that.
If you are also loading rate coefficients from the COMSOL Boltzmann solver, beware of the units. Last I checked, the MWP interface expects rate coefficients in m^3/(mol*s) while the Boltzmann solver outputs rate coefficients in m^3/s, so be sure to do the conversion.
--
Luke Gritter
AltaSim Technologies
I'm not aware of a limit for the table size, though I have occasionally seen the software have difficulty loading multiple large tables. Three hundred pairs of values is probably somewhat excessive - you should be able to adequately describe the transport coefficients with much less than that.
If you are also loading rate coefficients from the COMSOL Boltzmann solver, beware of the units. Last I checked, the MWP interface expects rate coefficients in m^3/(mol*s) while the Boltzmann solver outputs rate coefficients in m^3/s, so be sure to do the conversion.
--
Luke Gritter
AltaSim Technologies
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Dear Mr. Gritter
Thanks for that..
my PC was not healthy, that is why it could not import the data..
(registry clean and so on helped)
I hope I do not have to change anything for the mobilties/diffusivities? it is not clear where to look for that?
the exports are documented (see file)
I believe these are also the import units..
arigatou..
may you have a look in my model (model exchange) and tell me why it is possible for a mwp model to have DC-plasma densities above 10E17 [1/m^3] (study1 of MVP16-upload using 250V instead of the 200V set in the variables)
Thanks for that..
my PC was not healthy, that is why it could not import the data..
(registry clean and so on helped)
I hope I do not have to change anything for the mobilties/diffusivities? it is not clear where to look for that?
the exports are documented (see file)
I believe these are also the import units..
arigatou..
may you have a look in my model (model exchange) and tell me why it is possible for a mwp model to have DC-plasma densities above 10E17 [1/m^3] (study1 of MVP16-upload using 250V instead of the 200V set in the variables)
Attachments:
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Conversion from [m^3/s] to [m^3/s.mol] ?
this would give a rate coefficient for 1.5E-15 [m^3/s] of 3E-15/6.02e23= 2.6E-39 [m^3/mol.s] ???
how do I do that?
same for the arrhenius laws..
kf = Af.Te(hoch(nf)).exp(-Ef/Te) in ?? [m^3/s] ??
confused !!
lukas
this would give a rate coefficient for 1.5E-15 [m^3/s] of 3E-15/6.02e23= 2.6E-39 [m^3/mol.s] ???
how do I do that?
same for the arrhenius laws..
kf = Af.Te(hoch(nf)).exp(-Ef/Te) in ?? [m^3/s] ??
confused !!
lukas
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Dear Mr. Gritter，
A reaction rate always contains the gas temperature(Tg), for example: kf=8.9E-15*(Tg/Te)^1.5, and we know that the electron temperature comes from the DC Discharge Interface by typing dc.Te, but what about the Tg ? Is it the same value of the temperature that defined in the Plasma Model 1(e.g. T=300K)? If so, that the reaction rate becames kf=8.9E-15*(300/Te)^1.5, is that right?
Another qustion is that the unit of the electron temperature is eV in the COMSOL, but in some papers that using COMSOL for modeling, the unit is K, which on is correct? (see the attach file)
A reaction rate always contains the gas temperature(Tg), for example: kf=8.9E-15*(Tg/Te)^1.5, and we know that the electron temperature comes from the DC Discharge Interface by typing dc.Te, but what about the Tg ? Is it the same value of the temperature that defined in the Plasma Model 1(e.g. T=300K)? If so, that the reaction rate becames kf=8.9E-15*(300/Te)^1.5, is that right?
Another qustion is that the unit of the electron temperature is eV in the COMSOL, but in some papers that using COMSOL for modeling, the unit is K, which on is correct? (see the attach file)
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi Zhiming
for the temperatures , I would say , YES you are right.
for the eV units, you can convert to K (with the Boltzmann constant somewhere,divide?)
1 [eV] = 11600 [K]
en.wikipedia.org/wiki/Electronvolt
but what is your unit for kf ?
lukas
for the temperatures , I would say , YES you are right.
for the eV units, you can convert to K (with the Boltzmann constant somewhere,divide?)
1 [eV] = 11600 [K]
en.wikipedia.org/wiki/Electronvolt
but what is your unit for kf ?
lukas
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi, Lukas,
The unit of Kf depends on the order of the reaction, answer of your question(Conversion from [m^3/s] to [m^3/s.mol] ) is: 1[m^3/s] =1*6.02e23[m^3/s.mol] .
For the arrhenius laws, I do know clearly. I have taken a look at the expression, we just need to specify the Af, nf anEf.
By the way, can you give me your Skype or MSN, that we can contact more conveniently.
The unit of Kf depends on the order of the reaction, answer of your question(Conversion from [m^3/s] to [m^3/s.mol] ) is: 1[m^3/s] =1*6.02e23[m^3/s.mol] .
For the arrhenius laws, I do know clearly. I have taken a look at the expression, we just need to specify the Af, nf anEf.
By the way, can you give me your Skype or MSN, that we can contact more conveniently.
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Dear Mr.Gritter,
Whne setting the Secondary Emission Parameters in the Dielectric Barrier Discharge in the model library, why the parameter of the high voltage side (γi =0.01) is larger than the grounding side (γi =1E-6)?
Whne setting the Secondary Emission Parameters in the Dielectric Barrier Discharge in the model library, why the parameter of the high voltage side (γi =0.01) is larger than the grounding side (γi =1E-6)?
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi,
DBD is a pulsed-DC discharge.. on electrode is on -Voltage, the other grounded.
one electrode gets the ion bombardement, the other not..
so different secondary electron emissions..
but 0.01 seems to me very low.. 0.04 for copper.. (Y. Raiser book)
up to 1 for quarz surfaces ?
lukas
DBD is a pulsed-DC discharge.. on electrode is on -Voltage, the other grounded.
one electrode gets the ion bombardement, the other not..
so different secondary electron emissions..
but 0.01 seems to me very low.. 0.04 for copper.. (Y. Raiser book)
up to 1 for quarz surfaces ?
lukas
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi! Lukas,
In this case, the DBD is applied alternating voltage, and the secondary electron emission is of the Dielectic, not the metal, so it is very low.
In this case, the DBD is applied alternating voltage, and the secondary electron emission is of the Dielectic, not the metal, so it is very low.
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
if I do not mistake myself, the secondary emission on dielectrics is not always low.. (higher than for metals for glass or quarz: Gas discharge Physic of Yuri P. Raizer..end of page 71)