Discussion Forum

Dear all,

Thanks in advance for any response.

I am following the way provided by model exchange

www.comsol.com/community/exchange/215/

to calculated scattering and absorption spectra of Au nanosphere. According to the definition inside that model, scattering cross-section is calculated by SCS = 4*int_SCS(scPoav)/S_in, where scPoav is the outward power flow of the scattered field with a definition as
scPoav = emw.relPoavx*onx+emw.relPoavy*ony+emw.relPoavz*onz (eq. 1)
and S_in is the incident light intensity.

So far, everything is fine and the definition is clear. But after running the model, an interesting problem came up: I was changing the phase of solution to check the field variations of the field plot emw.relEz, and unexpectedly, it turns out that the scattered spectra also changed (please see attachments). Physically speaking, as far as I know, scattering spectra should not be dependent on phase of solutions. Also, the way how (eq.1) computes the scattered power has already used time-averaged value of scattered power flow. I mean, quantities such as emw.relPoavx is called "Power flow of the relative fields, time average, x component". So, I do not understand why there is a phase-dependence on the scattering spectra, and I already checked carefully that scPoav is indeed dependent on phase.

I've already searched over the internet, and found that such a way to compute the scattering cross-sections is quite popular, but, if the spectra is dependent on phase, then, which one is the correct one to use? After all, inside a harmonic cycle, there are 0 -> 2*pi phase change and countless solutions to use.

Up to now, I do not have a conclusion over this, maybe I did something wrong when implementing this model, but anyone interested could just use that model and try. I really appreciate any comment and suggestions over this issue. (ps: there is no such problem for absorption cross-section)

All the best,
Song