next up previous contents
Next: E.2 Circular Aperture Up: E. Diffraction in Far Previous: E. Diffraction in Far


E.1 Linear Approximation

For the following discussion, Figure E.1 shows the situation.
Figure E.1: Standard geometry of theory of diffraction
\begin{figure}\begin{center}
\par
\input{figures/optical_standard_geometry.pstex_t}
\par\end{center}\end{figure}
The aperture is positioned in the $ \tilde{x}\tilde{y}$ plane, which is normal to the z-axis and at the position of point $ \tilde{O}$. The point source is positioned in the $ xy$ plane which is parallel to the $ \tilde{x}\tilde{y}$ plane in a distance D in the negative z direction. The coordinates of the point source are therefore $ (x,y,-D)$. The projection point P' with the coordinates $ (x',y',D')$ is in the projection plane $ x'y'$, which is parallel to the aperture plane $ \tilde{x}\tilde{y}$ also. The positions $ \tilde{P}$ in the aperture are defined by the coordinates $ (\tilde{x},\tilde{y},0)$. The directions of the incoming and outgoing light waves are given with respect to the center of the aperture plane (see Figure E.2).
Figure E.2: Viewing angles of source point and projection point as seen from the center of the aperture.
\begin{figure}\begin{center}
\par
\input{figures/optical_standard_geometry_angles.pstex_t}
\par\end{center}\end{figure}
The directional vectors of the light waves form the angles $ \Theta_x$, $ \Theta_y$ and $ \Theta_z$ with their respective axes x, y and z. Here we define the functions:

$\displaystyle \alpha = \cos \Theta_x = \frac{-x}{R_0} = \sin \Theta$ (E.1)
$\displaystyle \beta = \cos \Theta_y = \frac{-y}{R_0}$ (E.2)
$\displaystyle \gamma = \cos \Theta_z = \sqrt{1-\alpha^2-\beta^2} = \frac{D}{R_0}$ (E.3)
$\displaystyle \alpha' = \cos \Theta'_x = \frac{-x'}{R'_0} = \sin \Theta'$ (E.4)
$\displaystyle \beta' = \cos \Theta'_y = \frac{-y'}{R'_0}$ (E.5)
$\displaystyle \gamma' = \cos \Theta'_z = \sqrt{1-\alpha'^2-\beta'^2} = \frac{D'}{R'_0}$ (E.6)
$\displaystyle \Theta = \frac{\pi}{2}-\Theta_x$ (E.7)
$\displaystyle \Theta' = \frac{\pi}{2}-\Theta'_x$ (E.8)

If one defines the transmission function:

$\displaystyle \tilde{\tau}(\tilde{x},\tilde{y})= \left\{\begin{array}{r@{\quad:...
...\sigma_0}  0 & (\tilde{x},\tilde{y}) \in \tilde{\sigma_c} \end{array} \right.$ (E.9)

with $ \sigma=\sigma_0\vee\sigma_c$ and $ \tilde{\sigma_0}$ and $ \tilde{\sigma_c}$ being the aperture area which are transmitting and not transmitting light respectively. The resulting electric field on the projection side after diffraction at an aperture given by the transmission function $ \tilde{\tau}$ yields

$\displaystyle E'=C \int\!\!\!\int\limits_{\tilde{\sigma}}\tilde{\tau}\tilde{E}\frac{e^{-ikR'}}{R'} d\tilde{\sigma}$ (E.10)

taking the solution of a spherical wave for diffraction of a planar wave at an infinite small aperture opening

$\displaystyle \tilde{E}=\frac{A}{R} e^{i (\omega t - kR)}$ (E.11)

with A as the Amplitude of the incident planar wave. Equation (E.10) yields

$\displaystyle E'=C A e ^ {i \omega t} \int\limits_{\tilde{\sigma}}\tau\frac{e^{-ik(R+R')}}{RR'} d\tilde{\sigma}$ (E.12)

The lengths $ R$ and $ R'$ are given by

$\displaystyle R = \sqrt{(\tilde{x}-x)^2+(\tilde{y}-y)^2+D^2}$    
$\displaystyle R' = \sqrt{(x'-\tilde{x})^2+(y'-\tilde{y})^2+D'^2}$ (E.13)

For the far field approximation $ R$ and $ R'$ are linear functions of $ \tilde{x}$ and $ \tilde{y}$. Therefore we calculate the series expansion of $ R$ and $ R'$ around $ R_0$ and $ R'_0$, with $ R_0$ and $ R'_0$ given by

$\displaystyle R_0 = \sqrt{x^2+y^2+D^2}$    
$\displaystyle R'_0 = \sqrt{x'^2+y'^2+D'^2}$ (E.14)

These distances are no functions of $ \tilde{x}$ and $ \tilde{y}$. Using (E.14) in (E.13) yields

$\displaystyle R=\sqrt{R_0^2-2(x\tilde{x}-y\tilde{y})+\tilde{x}^2+\tilde{y}^2}=R...
...1-\frac{2}{R_0^2}(x\tilde{x}+y\tilde{y})+\frac{\tilde{x}^2+\tilde{y}^2}{R_0^2}}$ (E.15)

assuming that the second and third term is much smaller than the first one, the squareroot in (E.15) can be written as

$\displaystyle R=R_0\sqrt{1+\varepsilon}\simeq R_0(1+\frac{1}{2}\varepsilon-\frac{1}{8}\varepsilon^2)$ (E.16)

and

$\displaystyle \varepsilon=-\frac{2}{R_0^2}(x\tilde{x}+y\tilde{y})+\frac{\tilde{x}^2+\tilde{y}^2}{R_0^2}$ (E.17)

Finally R is calculated by the series expansion to the second order in $ \tilde{x}$ and $ \tilde{y}$

$\displaystyle R \simeq R_0(1-\frac{x\tilde{x}+y\tilde{y}}{R_0^2}+\frac{\tilde{x}^2+\tilde{y}^2}{2 R_0^2}-\frac{(x\tilde{x}+y\tilde{y})^2}{2 R_0^4})$ (E.18)

This result is only valid if $ \tilde{x} \ll R_0$ and $ \tilde{y} \ll
R_0$. This holds especially for the far field where $ R_0$ is big in relation to the aperture size

$\displaystyle \vert\tilde{x}\vert\ll\sqrt{R_0\lambda}$    
$\displaystyle \vert\tilde{y}\vert\ll\sqrt{R_0\lambda}$ (E.19)

Therefore the third term in (E.18) is negligible and this gives

$\displaystyle R \simeq R_0-(\frac{x}{R_0})\tilde{x}-(\frac{y}{R_0})\tilde{y}$    
$\displaystyle R' \simeq R'_0-(\frac{x'}{R'_0})\tilde{x}-(\frac{y'}{R'_0})\tilde{y}$ (E.20)

Using the approximation

$\displaystyle (RR')^{-1} \simeq (R_0R'_0)^{-1}$ (E.21)

which is valid for the far field, and the transformation of the coordinates

$\displaystyle u \equiv -(\frac{x}{R_0}+\frac{x'}{R'_0})\frac{1}{\lambda}=\frac{(\alpha-\alpha')}{\lambda}=\frac{\cos{\theta_x}-\cos{\theta'_x}}{\lambda}$    
$\displaystyle v \equiv -(\frac{y}{R_0}+\frac{y'}{R'_0})\frac{1}{\lambda}=\frac{(\beta-\beta')}{\lambda}=\frac{\cos{\theta_y}-\cos{\theta'_y}}{\lambda}$ (E.22)

the total optical path can be written as

$\displaystyle R+R' = (R_0+R'_0)+(u\tilde{x}+v\tilde{y})\lambda$ (E.23)

substituting this optical path in the phase factor of the diffraction integral in (E.12) yields

$\displaystyle e^{i \omega t} e^{-i k (R+R')} = e^{i \phi_0} e^{-i 2 \pi (u\tilde{x}+v\tilde{y})}$ (E.24)

with

$\displaystyle \phi_0 = \omega t - k (R_0+R'_0)$ (E.25)

$ \phi_0$ is not a function of $ \tilde{x}$ and $ \tilde{y}$. The coordinates $ (x,y)$ of the source and $ (x',y')$ of the projection point are now included in $ (u,v)$. Finally the electrical field at the projection point P' emanating from the source point P and diffracted at the aperture with the transmission function $ \tilde{\tau}(\tilde{x},\tilde{y})$ can be given as

$\displaystyle E'(u,v) = \frac{C A e^{i \phi_0}}{R_0 R'_0} \int\!\!\!\int\limits...
...tilde{x},\tilde{y}) e^{i 2 \pi (u\tilde{x}+v\tilde{y})} d\tilde{x} d\tilde{y}$ (E.26)

The integral of (E.26) is exactly the FOURIER transform of the transmission function $ \tilde{\tau}(\tilde{x},\tilde{y})$

$\displaystyle T(u,v) \equiv \int\!\!\!\int\limits_{-\infty}^{+\infty} \tilde{\t...
...tilde{x},\tilde{y}) e^{i 2 \pi (u\tilde{x}+v\tilde{y})} d\tilde{x} d\tilde{y}$ (E.27)

This important result implies, that under the given assumptions, the electrical field distribution after a diffracting aperture is proportional to the FOURIER transform of the transmission function of the aperture.
If as shown in Figure E.3 the points $ P$,$ \tilde{O}$ and $ P'$ are on a line, then $ (u,v)=(0,0)$ and (E.26) together with (E.27) reduces to

$\displaystyle E'(0,0)=\frac{C A e^{i \phi_{00}}}{R_{00}R'_{00}} T(0,0)$ (E.28)

Figure E.3: Direct light propagation through aperture
\begin{figure}\begin{center}
\par
\input{figures/optical_standard_geometry_direct.pstex_t}
\par\end{center}\end{figure}
With (E.28), (E.26) can be normalized to $ E'(0,0)$

$\displaystyle E'(u,v)=E'(0,0)\left[\frac{R_{00}R'_{00}}{R_0R'_0}\right]\frac{T(u,v)}{T(0,0)}e^{i(\phi_0-\phi_{00})}$ (E.29)


next up previous contents
Next: E.2 Circular Aperture Up: E. Diffraction in Far Previous: E. Diffraction in Far

R. Minixhofer: Integrating Technology Simulation into the Semiconductor Manufacturing Environment