Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Transmittance of absorbing multilayer thin-films on an absorbing substrate

  1. Jan 21, 2009 #1
    I have found a general expression for the amplitude transmittance [tex](t)[/tex] of multilayer film stacks in the literature [1], but the author does not explain how to obtain the transmittance [tex](T)[/tex]. I looked up other references, and the closest I could find was the description of "an absorbing film on a transparent substrate" [2].

    On page 756 of [2] there are expressions for transmittance:

    [tex]T = \frac{n_3 \cos \theta _3}{n_1 \cos \theta _1} \left| t \right| ^2 \qquad \qquad \mbox{(TE)}[/tex]

    [tex]T = \frac{(\cos \theta _3) / n_3}{(\cos \theta _1)/ n_1} \left| t \right| ^2 \qquad \qquad \mbox{(TM)}[/tex]

    In other words, I'm trying to find the transmittance (using the amplitude transmittance value I already know) for a system that consists of a semi-infinite incidence medium (dielectric), many thin-films (absorbing), and a semi-infinite substrate (absorbing). In comparison, the reflectance is easy to find, because you just multiply the reflectivity by its complex conjugate; this is not the case. If you use the expressions above, replacing [tex](n_3)[/tex] and [tex](\theta_3)[/tex] by the substrate complex refractive index and the complex angle on the exit side, respectively, the results will be complex as well.

    Any ideas? Thanks.

    [1] J. Eastman, Surface scattering in optical interference coatings. PhD thesis, University of Rochester, 1974.
    [2] M. Born and E. Wolf, Principles of Optics. Cambridge, UK: Cambridge University Press, 7th ed., 1999.
  2. jcsd
  3. Jan 21, 2009 #2
    PS: maybe it's simply 0, because of my semi-infinite assumption for an absorbing substrate.
  4. Jan 23, 2009 #3
    I've found an explanation in this reference [3]. This is not exactly what is in the book, but it's what I think is correct. Please correct me if I am wrong.

    In a system where the media are:

    # 0 = a semi-infinite dielectric (transparent incident medium)
    # 1 = the first thin film (absorbing)
    # m = the last thin film (absorbing)
    # m+1 = a semi-infinite substrate (absorbing)

    I assume:

    1. Oblique incidence
    2. I already know the amplitude transmittance [tex](t)[/tex]
    3. The coordinate system has the origin at the last interface, where +z points down.

    Beyond the last interface, the transmittance is

    [tex]T = \left| \frac{\hat{n}_{m+1} \cos \hat{\theta} _{m+1}}{n_0 \cos \hat{\theta} _0} \right| \left| t \right| ^2 \exp \left[ - \left( \frac{4\pi \, \kappa_{m+1}}{\lambda} \right) \frac{z}{\cos \theta _{m+1}} \right] \qquad \qquad \mbox{(TE)} \right|[/tex]

    [tex]T = \left| \frac{ (\cos \hat{\theta} _{m+1})/\hat{n}_{m+1}}{(\cos \hat{\theta} _0)/n_0} \right| \left| t \right| ^2 \exp \left[ - \left( \frac{4\pi \, \kappa_{m+1}}{\lambda} \right) \frac{z}{\cos \theta _{m+1}} \right] \qquad \qquad \mbox{(TM)} \right|[/tex]

    where [tex](\kappa)[/tex] is the imaginary part of the complex refractive index.

    This means I can generalize the two transmittance expressions (from the first post) for the case of complex media [3]. I also introduce a decay, because the semi-infinite substrate is absorbing---according to the Lambert law of absorption (in its oblique version) [3]. So, the better way to put it is that the transmittance approaches 0 very fast.

    [3] C. Mack, Fundamental Principles of Optical Lithography: The Science of Microfabrication. West Sussex, England: John Wiley & Sons, 2007.
    Last edited: Jan 23, 2009
  5. Jan 23, 2009 #4
    One last thought:

    I've found yet another reference [4] with a slightly different expression for the case of a multilayer stack:

    [tex] T = \Re \left\{ \frac{\hat{n}_{s} \cos \hat{\theta} _s}{n_a \cos \hat{\theta} _a} \right\} \left| t \right| ^2 [/tex]

    where "s" and "a" stand for substrate and ambient media, respectively. Of course, this author tells the reader to carry out separate computations for TE and TM-polarized light, but that's really the only formula presented. It appears that the only difference between the two cases is the amplitude transmittance, which does not sound right for a TM calculation. I would expect something like:

    [tex] T = \Re \left\{ \frac{\cos \hat{\theta} _s / \hat{n}_{s}}{\cos \hat{\theta} _a / n_a} \right\} \left| t \right| ^2 [/tex]

    Also, notice that he does not take the absolute value of the ratio; he uses the real part of the result instead. What confuses me even more is that he cites [2] as his reference, which does not present the same formalism.

    This takes me back to my original question:

    How do I compute the transmittance in (absorbing) stratified media? Any help is highly appreciated!

    [2] M. Born and E. Wolf, Principles of Optics. Cambridge, UK: Cambridge University Press, 7th ed., 1999.
    [4] D. L. Windt, IMD - Software for modeling the optical properties of multilayer films, Computers in Physics, 12, 360 (1998).
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook