Bow tie nano-antenna extinction efficiency



Hi all,
I am trying to reproduce the experimental results in the paper attached

kumar2008fabrication.pdf (492.3 KB)

Namely, I need to reproduce figure 2 where the extinction efficiency is calculated based on measurements.
The dimensions, according to the paper, are as stated below:
L (tip-to-base height)= 240 nm (Equilateral triangles)
Gap = 10 nm
Gold thickness = 50 nm

The problem is extinction efficiency I get is not identical to the paper’s result as shown below:

Can anyone help me with that, please?
Attached is the design file together with the script file.
BNA_gap10nm.fsp (256.8 KB) test.lsf (805 Bytes)

Bowtie nano-antenna
Scattering cross section of a nanoparticle on a substrate
Split ring resonator Tranmission
Calculating the extinction cross section of a nanodisk with a removed wedge

After an initial look at your simulation file, I have a couple of follow up question:

  • Is the X,Y boundary correct? You are currently using PML boundaries, but I think the structure is intended to be periodic.
    Note: To use periodic and symmetry boundaries at the same time, apply the symmetry boundary to both the min and max boundary.
  • Is the source propagation direction correct? From the paper, it looks like the structure should be illuminated from the air, not from the substrate.

I’ve made the following updates to your file:

  • added a mesh override region to cover the entire antenna structure plus the extinction monitors
  • reversed the order of the source
  • used periodic type boundary conditions
  • reverted to using standard cross section monitors
  • adjusted the size of the source, simulation region and monitors
  • updated script to work with standard cross section monitors
    BNA_gap10nm_2.fsp (294.4 KB)
    test.lsf (811 Bytes)

The simulation setup may not be perfect yet, but I’m sure that with a bit more work we will be able to get good agreement with the paper, or at least understand where the discrepancies come from.


Well, regarding the periodicity of the structure, I didn’t use periodic boundary condition since it was stated that at this large period of 3 um, the interaction between the individual antennas is almost negligible.

Thanks for your help!


Although the overall extinction efficiency is slightly changed, using the standard cross section monitors shows much less absorbing cross section and higher scattering. This makes more sense, I think.



Here are the results from some additional tests that I ran.

  1. Smaller mesh
    I ran a few more simulations using smaller mesh sizes (around 1nm for the tip), but didn’t see any significant change in the results, as shown below.

  2. PML boundaries rather than periodic in XY direction
    Out of curiosity, I re-ran the simulation using PML boundaries in the XY directions. The results are quite similar to the periodic case, but the curves tend to be much smoother. I think this is expected, since there is no possibility of interference between neighbouring antennas when PML boundaries are used.


I don’t understand why in your simulation file you use anti/symmetric B.C but you wrote that you changed it to PERIODIC ??

Thanks in advance


Hi mis,

You can refer to this post that explains the periodicity with symmetric and anti-symmetric boundary conditions.


I don’t understand why you didn’t extend TFSF source through simulation boundaries ? Does anti/symmetric B.C belong to periodic B.C?

Thanks for help!


Yes, with periodic boundaries or symmetric boundaries on both min and max boundaries the TFSF source should be extended into simulation boundaries. Here is a good link to learn more about TFSF with periodic boundaries: