Hi everyone, I am getting some questions regarding the calculation of near-field enhancement, which is basically the enhancement of the electric field in the vicinity of the nanoparticles over the incident field (in absence of the nanoparticles). It's quite simple to do it, so I am describing it here point by point (not making a new video, but I can make one if a lot of people need it): 1) Make a copy of the above simulation after completion (it's always good to save up to the point you have done correctly). 2) Now in the new file, go add new physics, and add another electromagnetic wave frequency domain module. (comsol will automatically assign number to it like ewfd1 and ewfd2. 3) Now in the newly added EMwave module, open one electromagnetic wave tab from the top and select the nanoparticles to change their refractive index to the refractive index of the medium (i.e. for water 1.33 or for air 1). This will virtually remove the nanoparticles. 4) Now compute. In the computation, COMSOL will solve the Maxwell's equations for both the modules: one with nanoparticles and one without nanoparticles. 5) Look on the top of the module what is the nomenclature for each module. For instance if the earlier EM wave module had the name ewfd, the newly added one may have ewfd2 or ewfd1 etc. 6) Now for the visualization of the near-field enhancement, one can make a cutplane across the nanoparticles, and then visualize as a 2D cutplane or 3D cutplane. You can also make 3D surface over the nanoparticles, it just looks fancy. 7) In the cutplane, the COMSOL variable for electric field has to be mentioned which is ewfd.normE, which ewfd is the module and the normE is the magnitude of electric field. For near-field enhancement, in my case it was ewfd.normE/ewfd2.normE. So the numerator ewfd.normE is the electric field for the system with nanoparticles, and the denominator ewfd2.normE is the electric field for the system without nanoparticles. 8) Dont forget to choose the wavelength where you want to probe the near-field enhancement. Hope it helps! Happy computing and modeling light matter interactions!
Here is some equation which is related with the example of n_Au. Units are in the brackets. E(J)=h(J.s) x f(1/s) ----> E=h⋅f-------(equation 1) 1 eV = 1.602176565⋅e-19 J------------------------------------->(e-19 means 1/10000000000000000000) h=6.62607015⋅e-34 J⋅s c=299792458 m/s c=lambda⋅f------------(equation 2) So, If we would like to calculate the frequency of photon ,which also has the wavelength, then we can calculate the frequency because we already know the energy of photon in terms of eV. After that, we can find the lambda because we know the c(speed of light). Here is an example. 6.6 eV is an energy value for gold in last row and column and also 1.28 is real part of refractive index for gold in last row and column. You can see below in the paper (the old one 1972). 6.6 (eV) = 10.574365329e-19 (J) 10.574365329e-19 (J) = 6.62607015⋅e-34 (J⋅s) ⋅ f(1/s)---------(it comes from equation 1) f=1.5955872831⋅e15(1/s) lambda = 299792458 (m/s) / 1.5955872831⋅e15(1/s)----------(e15 means 1000000000000000)----------(it comes from equation 2) lambda = 1.878884729⋅e-7 (m) --------FINAL------- lambda = 187.8884729 nm----------------------when n=1.28 (you can see in the 15:35 first row and column)
Here if someone want to get the Au txt of n and k,you can tend to the first paper given in the description ( I mean the 1972 one) in which there's a long list of n and k of many materials with the expansion of eV. You can turn to 1240/neV to calculate lambda and get the same result with the video
Hello Sir, I followed each step in Comsol 5.6 version but I am getting following error - "Undefined value found. - Detail: Undefined value found in the equation residual vector. There are 1 degrees of freedom giving NaN/Inf in the vector for the variable comp1.ewfd.S2x. at coordinates: (0,0,0), ..." Could you help me with where I am going wrong?
Hi! I’m simulating the scattering cross-section of two spherical nanoparticles with a 2 nm gap between them. I’ve set up the incident electric field as E=1 V/m⋅exp(−ik0z), and I can see the electric field outside the nanoparticles, but there’s no electric field in the gap. Could you suggest what might be causing this and how to capture the field in the gap between the particles? Thanks!
@@baseeratbibi630 if your solution converged and you set up the model correctly, then the solver solved for the E field everywhere in the computation domain. Make sure to use the right cut plane for vizualization and also make sure that in the EWFD setup you have selected all the nanoparticles and the surrounding as the domain. Seems more like a sementic error..
@@riturajborah382 Thank you for your helpful response. I have some additional questions regarding the simulation and would greatly appreciate your further guidance. Would you kindly share your email address so I can contact you for additional clarification?
Thank you, as I am writing the principles of a hypothetical new source of energy. It involves a Negative refractive indexed meta material, Bismuth ferrite nanowires and gold nano particles. I use AI GPT and Google Scholar to validate my assumptions. I have retired 13 years ago after a career in Analytical Chemistry. The laboratory was a “think-tank” for problem solving and advisory for product development. The energy is a non idea plasma from water splitting. The crux of the process is creating a water splitting ultraviolet field energy and your video is invaluable.
@@gene4094 Hi, First, amazing to hear that you are still passionately pursuing research after retirement. I don't have the full picture, but such metamaterial structures are gaining importance in the field of catalysis indeed. Engineered for better management of energy at nanoscale. Wish you all the best!
Hi can you explain about TM wave equation to be input in the port section. ? You didnt mention clearly about it. Bit confused because am trying to move from 2d to 3d simulations. So the question is what will be port e field equations for Te and TM mode. Your help is greatly appreciated
Sure. So you are giving the incoming wave in terms of the electric field or magnetic field (when we are also setting the impedance to be 1, E and H are perpendicular vectors of equal magnitude). So you can choose TM or TE to define the inlet wave. When it is TM, that means the magnetic field is perpendicular to the plane of incidence. So you can just give 1 in the x or y component of the Magnetic field depending upon the direction of propagation (but would not matter). If it is TE, you choose TE and just give 1 in the x or y component of E field.
@@riturajborah382 thank you for the clarification. however I have more more doubt. in the hexagonal grating example in the comsol website there are two components in E field for TE and @ H components in x and y for TM. so both way for defining is correct ?. to accurately model do one have to implement x and y components like this as seen in one of the comsol file I have , x component is emw.nx*(exp(-i*kax*x)[A/m])*emw.ny*2, Y component is -(exp(-i*kax*x)[A/m]-emw.ny*(exp(-i*kax*x)[A/m])*emw.ny*2). from where does this equation comes from and do you have any reference to share so that I can be sure of the simulation file I received from lab?.. thank you for replying
Thank you, really very nice work and helpful for us. Sir, I am newly working on COMSOL and have doubt, in the parameter you define I0=1[MW/m^2] may be this is for AM1.5G but for AM1.5G standard value is I0=1[KW/m^2]. If I am wrong please help me to clear my doubt and if possible please help me to give reference for I0=1[MW/m^2]?
Hi Sadhna, thank you for your feedback. So regarding the inlet intensity. You can actually use any intensity. You can just replace the I0 with your values. Accordingly, the electric field values will be determined. But if you are only concerned about the absorptance, reflectance and transmittance spectra, then the inlet intensity or power does not latter as it is calculated as a % or fraction of what is absorbed, reflected or transmitted with respect to the inlet intensity. How to calculate this is shown in the part 2 of this video. If you want to show the near-field enhancement, I think it is better to show with respect to the incident field in absence of the nanoparticlea. For that, one more time the model has to be solved without the nanoparticles. Hope it helps!
Thank you very much for the video, it helped me a lot to learn the basics of the electromagnetic waves module. One question I have left, is it possible to also define circular polarised or unpolarised light with port 1 and 2?
Yes. You can. Circularly polarized light is just the sum of two sine waves perpendicular to each other and separated by a phase difference of pi/2. So if you figure out the expression for it, you can work it out.
Thanks for the quick answer, then I would assume Comsol should understand the waves as: E0x*sin(2*pi/(lambda[nm])*z) and E0y*sin(2*pi/(lambda[nm])*z+pi/2). I think my main problem lies in understanding of the periodic port boundary condition. Does it actively change the position of my wave or is it always constant? E.g. if I would instead use the scattered field formulation in comsol and define the same background electric field, would the result then be the same? I would be very grateful for any comment from you.
@@MultiMetal31 if it is not a periodic structure, then scattered field simulations should be appropriate. In that case, one has to define a background field and what you have written above as the x- component and y-component of the electric field separated by a phase difference of pi/2 seems to be correct (put pi/2 in a bracket just to be sure). The approach is right, just put the expression for a circularly polarized wave as the background field. In the periodic case, I suppose you have to provide the inlet twice for the two waves making up the circularly polarized wave. On the periodic boundary condition I dont think anything specific has to be mentioned as it will be automatically taken care of. Hope it helps, please keep me updated I am curious to know ! 😊 Thank you for your comments!
Dear author, hello, I am a student, recently read your article, want to build a model, but encountered a problem, I wonder if you can send the comsol file,😊
Hello, I am getting the following error, can you help? I am using COMSOL version 6.1. The slit condition should be applied only on an interior boundary or the port should be placed on an exterior boundary. - Feature: Port 2 (port2) - Feature: Port 1 (port1)
This is kinda unrelated maybe but can your approach of calculating RAT be applied to tandem cells with textures? Since your using nanoparticles i was thinking like nanocones or pyramids as well
@@kenken6012 integrating poynting vector over a surface gives the energy flow rate through the surface. So you may try integrating poynting vector over specific layers to find the transmittance and reflectance. Or you could also decouple the problem by modeling each layer separately to find their RAT. And then couple them in one model for final estination.
@riturajborah382 wait for poynting dont we need the magnetic field? So that means we have to do one more frequency domain but iterate that through magnetic field?
@@kenken6012The magnetic field is basically perpendicular to the electric field. So COMSOL calculates it itself. When you are calculating the poynting vector on COMSOL with the Poav variable, COMSOL does the entire process itself. So no need to solve for the magnetic field.
@riturajborah382 can you share your email with me ? I have a comsol file in 3d and the file is having custom equation to solve RF simulation. I worked with 2 d models, grating and MTM previously. I want to clear the confusion through an expert in this field. IF you could help me It would be wonderful
Hi everyone, I am getting some questions regarding the calculation of near-field enhancement, which is basically the enhancement of the electric field in the vicinity of the nanoparticles over the incident field (in absence of the nanoparticles). It's quite simple to do it, so I am describing it here point by point (not making a new video, but I can make one if a lot of people need it):
1) Make a copy of the above simulation after completion (it's always good to save up to the point you have done correctly).
2) Now in the new file, go add new physics, and add another electromagnetic wave frequency domain module. (comsol will automatically assign number to it like ewfd1 and ewfd2.
3) Now in the newly added EMwave module, open one electromagnetic wave tab from the top and select the nanoparticles to change their refractive index to the refractive index of the medium (i.e. for water 1.33 or for air 1). This will virtually remove the nanoparticles.
4) Now compute. In the computation, COMSOL will solve the Maxwell's equations for both the modules: one with nanoparticles and one without nanoparticles.
5) Look on the top of the module what is the nomenclature for each module. For instance if the earlier EM wave module had the name ewfd, the newly added one may have ewfd2 or ewfd1 etc.
6) Now for the visualization of the near-field enhancement, one can make a cutplane across the nanoparticles, and then visualize as a 2D cutplane or 3D cutplane. You can also make 3D surface over the nanoparticles, it just looks fancy.
7) In the cutplane, the COMSOL variable for electric field has to be mentioned which is ewfd.normE, which ewfd is the module and the normE is the magnitude of electric field. For near-field enhancement, in my case it was ewfd.normE/ewfd2.normE. So the numerator ewfd.normE is the electric field for the system with nanoparticles, and the denominator ewfd2.normE is the electric field for the system without nanoparticles.
8) Dont forget to choose the wavelength where you want to probe the near-field enhancement.
Hope it helps! Happy computing and modeling light matter interactions!
Here is some equation which is related with the example of n_Au. Units are in the brackets.
E(J)=h(J.s) x f(1/s) ----> E=h⋅f-------(equation 1)
1 eV = 1.602176565⋅e-19 J------------------------------------->(e-19 means 1/10000000000000000000)
h=6.62607015⋅e-34 J⋅s
c=299792458 m/s
c=lambda⋅f------------(equation 2)
So, If we would like to calculate the frequency of photon ,which also has the wavelength, then we can calculate the frequency because we already know the energy of photon in terms of eV. After that, we can find the lambda because we know the c(speed of light).
Here is an example.
6.6 eV is an energy value for gold in last row and column and also 1.28 is real part of refractive index for gold in last row and column. You can see below in the paper (the old one 1972).
6.6 (eV) = 10.574365329e-19 (J)
10.574365329e-19 (J) = 6.62607015⋅e-34 (J⋅s) ⋅ f(1/s)---------(it comes from equation 1)
f=1.5955872831⋅e15(1/s)
lambda = 299792458 (m/s) / 1.5955872831⋅e15(1/s)----------(e15 means 1000000000000000)----------(it comes from equation 2)
lambda = 1.878884729⋅e-7 (m)
--------FINAL-------
lambda = 187.8884729 nm----------------------when n=1.28 (you can see in the 15:35 first row and column)
Here if someone want to get the Au txt of n and k,you can tend to the first paper given in the description ( I mean the 1972 one) in which there's a long list of n and k of many materials with the expansion of eV. You can turn to 1240/neV to calculate lambda and get the same result with the video
Thank you for posting!
Thank you very much, really helpful!!
Nice. Thank you. It's helpful.
Since I am getting a lot of queries asking for the optical constants of Au. They can be found in Johnson and Christy 1972 Physical Review B. But for the lazy ones, here are the constants :
n
Wavelength (nm) n
187.8548501 1.28
191.629368 1.32
195.2507103 1.34
199.331513 1.33
203.2527886 1.33
207.3314399 1.3
211.9388052 1.3
216.3773142 1.3
221.400359 1.3
226.2485421 1.31
231.313808 1.3
237.063482 1.32
242.6305305 1.32
248.9642592 1.33
255.1115248 1.33
261.5700444 1.35
268.9462062 1.38
276.1340781 1.43
284.3674336 1.47
292.4155685 1.49
300.9325269 1.53
310.737346 1.53
320.3726126 1.54
331.508559 1.48
342.497793 1.48
354.2405745 1.5
367.9056411 1.48
381.4898494 1.46
397.3852598 1.47
413.2806702 1.46
430.5006981 1.45
450.8516402 1.38
471.4228177 1.31
495.9368043 1.04
520.9420213 0.62
548.6026596 0.43
582.085451 0.29
616.8368212 0.21
659.4904312 0.14
704.4556879 0.13
756.001226 0.14
821.0874243 0.16
891.9726695 0.17
984.0015957 0.22
1087.580711 0.27
1215.531383 0.35
1393.080911 0.43
1610.184429 0.56
1937.253142 0.92
k
Wavelength (nm) k
187.8548501 1.188
191.629368 1.203
195.2507103 1.226
199.331513 1.251
203.2527886 1.277
207.3314399 1.304
211.9388052 1.35
216.3773142 1.387
221.400359 1.427
226.2485421 1.46
231.313808 1.497
237.063482 1.536
242.6305305 1.577
248.9642592 1.631
255.1115248 1.688
261.5700444 1.749
268.9462062 1.803
276.1340781 1.847
284.3674336 1.869
292.4155685 1.878
300.9325269 1.889
310.737346 1.893
320.3726126 1.898
331.508559 1.883
342.497793 1.871
354.2405745 1.866
367.9056411 1.895
381.4898494 1.933
397.3852598 1.952
413.2806702 1.958
430.5006981 1.948
450.8516402 1.914
471.4228177 1.849
495.9368043 1.833
520.9420213 2.081
548.6026596 2.455
582.085451 2.863
616.8368212 3.272
659.4904312 3.697
704.4556879 4.103
756.001226 4.542
821.0874243 5.083
891.9726695 5.663
984.0015957 6.35
1087.580711 7.15
1215.531383 8.145
1393.080911 9.519
1610.184429 11.21
1937.253142 13.78
For the thermal simulations, you just need to fill in the constants from the video.
Hello Sir, I followed each step in Comsol 5.6 version but I am getting following error -
"Undefined value found.
- Detail: Undefined value found in the equation residual vector.
There are 1 degrees of freedom giving NaN/Inf in the vector for the variable comp1.ewfd.S2x.
at coordinates: (0,0,0), ..."
Could you help me with where I am going wrong?
Hello sir hopefully you will be fine, please make some more videos about plasmonic effects, hotspots etc. Thanks
How do we calculate RTA? Do we have to add another specific solver apart from the electric field one?
Thanks
Hi! I’m simulating the scattering cross-section of two spherical nanoparticles with a 2 nm gap between them. I’ve set up the incident electric field as E=1 V/m⋅exp(−ik0z), and I can see the electric field outside the nanoparticles, but there’s no electric field in the gap. Could you suggest what might be causing this and how to capture the field in the gap between the particles? Thanks!
@@baseeratbibi630 if your solution converged and you set up the model correctly, then the solver solved for the E field everywhere in the computation domain. Make sure to use the right cut plane for vizualization and also make sure that in the EWFD setup you have selected all the nanoparticles and the surrounding as the domain. Seems more like a sementic error..
@@riturajborah382 Thank you for your helpful response. I have some additional questions regarding the simulation and would greatly appreciate your further guidance. Would you kindly share your email address so I can contact you for additional clarification?
Can a laser field increase the energy?
Yes. The absorptance is defined as %. So a high intensity laser illumination will also lead to higher absolute absorption.
Thank you, as I am writing the principles of a hypothetical new source of energy. It involves a Negative refractive indexed meta material, Bismuth ferrite nanowires and gold nano particles. I use AI GPT and Google Scholar to validate my assumptions. I have retired 13 years ago after a career in Analytical Chemistry. The laboratory was a “think-tank” for problem solving and advisory for product development. The energy is a non idea plasma from water splitting. The crux of the process is creating a water splitting ultraviolet field energy and your video is invaluable.
@@gene4094 Hi, First, amazing to hear that you are still passionately pursuing research after retirement.
I don't have the full picture, but such metamaterial structures are gaining importance in the field of catalysis indeed. Engineered for better management of energy at nanoscale.
Wish you all the best!
Hi can you explain about TM wave equation to be input in the port section. ? You didnt mention clearly about it. Bit confused because am trying to move from 2d to 3d simulations. So the question is what will be port e field equations for Te and TM mode. Your help is greatly appreciated
Sure. So you are giving the incoming wave in terms of the electric field or magnetic field (when we are also setting the impedance to be 1, E and H are perpendicular vectors of equal magnitude). So you can choose TM or TE to define the inlet wave. When it is TM, that means the magnetic field is perpendicular to the plane of incidence. So you can just give 1 in the x or y component of the Magnetic field depending upon the direction of propagation (but would not matter). If it is TE, you choose TE and just give 1 in the x or y component of E field.
@@riturajborah382 thank you for the clarification. however I have more more doubt. in the hexagonal grating example in the comsol website there are two components in E field for TE and @ H components in x and y for TM. so both way for defining is correct ?. to accurately model do one have to implement x and y components like this as seen in one of the comsol file I have , x component is emw.nx*(exp(-i*kax*x)[A/m])*emw.ny*2, Y component is -(exp(-i*kax*x)[A/m]-emw.ny*(exp(-i*kax*x)[A/m])*emw.ny*2). from where does this equation comes from and do you have any reference to share so that I can be sure of the simulation file I received from lab?.. thank you for replying
Thank you, really very nice work and helpful for us. Sir, I am newly working on COMSOL and have doubt, in the parameter you define I0=1[MW/m^2] may be this is for AM1.5G but for AM1.5G standard value is I0=1[KW/m^2]. If I am wrong please help me to clear my doubt and if possible please help me to give reference for I0=1[MW/m^2]?
Hi Sadhna, thank you for your feedback. So regarding the inlet intensity. You can actually use any intensity. You can just replace the I0 with your values. Accordingly, the electric field values will be determined. But if you are only concerned about the absorptance, reflectance and transmittance spectra, then the inlet intensity or power does not latter as it is calculated as a % or fraction of what is absorbed, reflected or transmitted with respect to the inlet intensity. How to calculate this is shown in the part 2 of this video. If you want to show the near-field enhancement, I think it is better to show with respect to the incident field in absence of the nanoparticlea. For that, one more time the model has to be solved without the nanoparticles. Hope it helps!
@@riturajborah382 Thank you for your reply, yes it's helpful.
thank you very much for your sharing, can we have your defined file of Au and Ag from your material library? Thanks
You can read my latest re under this video
Thank you very much for the video, it helped me a lot to learn the basics of the electromagnetic waves module.
One question I have left, is it possible to also define circular polarised or unpolarised light with port 1 and 2?
Yes. You can. Circularly polarized light is just the sum of two sine waves perpendicular to each other and separated by a phase difference of pi/2. So if you figure out the expression for it, you can work it out.
Thanks for the quick answer, then I would assume Comsol should understand the waves as: E0x*sin(2*pi/(lambda[nm])*z) and
E0y*sin(2*pi/(lambda[nm])*z+pi/2).
I think my main problem lies in understanding of the periodic port boundary condition. Does it actively change the position of my wave or is it always constant? E.g. if I would instead use the scattered field formulation in comsol and define the same background electric field, would the result then be the same? I would be very grateful for any comment from you.
@@MultiMetal31 if it is not a periodic structure, then scattered field simulations should be appropriate. In that case, one has to define a background field and what you have written above as the x- component and y-component of the electric field separated by a phase difference of pi/2 seems to be correct (put pi/2 in a bracket just to be sure). The approach is right, just put the expression for a circularly polarized wave as the background field.
In the periodic case, I suppose you have to provide the inlet twice for the two waves making up the circularly polarized wave. On the periodic boundary condition I dont think anything specific has to be mentioned as it will be automatically taken care of.
Hope it helps, please keep me updated I am curious to know ! 😊 Thank you for your comments!
Dear author, hello, I am a student, recently read your article, want to build a model, but encountered a problem, I wonder if you can send the comsol file,😊
Hello, I am getting the following error, can you help? I am using COMSOL version 6.1.
The slit condition should be applied only on an interior boundary or the port should be placed on an exterior boundary.
- Feature: Port 2 (port2)
- Feature: Port 1 (port1)
Hi, if you could share your email, I might send you the case file.
I miss same question with you, did you figure out?
This is kinda unrelated maybe but can your approach of calculating RAT be applied to tandem cells with textures? Since your using nanoparticles i was thinking like nanocones or pyramids as well
Yes. The structure wont matter of course.
@Rituraj Borah ah nice. Then another question would be how to get the RAT of a specific layer? Say you wanted to see the Gold nanoparticle RAT plot?
@@kenken6012 integrating poynting vector over a surface gives the energy flow rate through the surface. So you may try integrating poynting vector over specific layers to find the transmittance and reflectance. Or you could also decouple the problem by modeling each layer separately to find their RAT. And then couple them in one model for final estination.
@riturajborah382 wait for poynting dont we need the magnetic field? So that means we have to do one more frequency domain but iterate that through magnetic field?
@@kenken6012The magnetic field is basically perpendicular to the electric field. So COMSOL calculates it itself. When you are calculating the poynting vector on COMSOL with the Poav variable, COMSOL does the entire process itself. So no need to solve for the magnetic field.
Dada, Thank You so much very much helpful
I am also simulating in comsol but i am new in this field
Dada apunar email id Share koribo pariboni??
Thank you for your feedback. Yes, sure. It's rituraj.borah@uantwerpen.be
@riturajborah382 can you share your email with me ? I have a comsol file in 3d and the file is having custom equation to solve RF simulation. I worked with 2 d models, grating and MTM previously. I want to clear the confusion through an expert in this field. IF you could help me It would be wonderful
Thanks