Applications: Structured and accurate forward calculation
A point source (spatial deltafunction) emits a spherical wave that is imaged by the optical system. The coherent point source produces a field distribution in the nominal image plane, called the (complex) pointspread function. In the case of an incoherent quadratic detector, only the intensity distribution can be obtained in the image plane.
Schematic layout of an optical imaging system The original NijboerZernike theory produces the amplitude and intensity distribution in the focal plane in the presence of an aberrated imaging system. Both the aberration of the system and the defocusing of the image plane should remain relatively small in the NijboerZernike theory. In the Extended NijboerZernike theory, both the aberration and the defocusing can assume larger values; as a result, we can use analytic formulas that describe the amplitude and intensity distribution in an image volume. Using the Extended NijboerZernike formalism, we obtain explicit expressions for axial, radial or azimuthal crosssections of the throughfocus intensity distribution. This is advantageous with respect to computational burden. Other numerical schemes rely on pointbypoint evaluation of numerical integrals or on Fourier transforms that only yield a solution for a single plane. In this sense, the numerical results from the Extended NijboerZernike analysis can serve as a benchmark with respect to other purely numerical techniques in diffraction theory like Fourier transform methods. An example of a forward calculation based on the ENZtheory is the intensity pattern below. The pupil function was moderately aberrated (Strehl ratio typically 0.80) by contributions from relatively highorder Zernike polynomials. In the corresponding infocus intensity pattern we observe a widely spread lowlevel intensity contribution that resembles a laser speckle pattern. Note that all higherorder diffracted energy is roughly contained within two circles given by v=n_{min}+3 and v=n_{max}+3 with n_{min }and n_{max} equal to the lowest and highest radial order of the Zernike polynomials in the aberration function; the central part of the diffraction image has retained its appearance of a classical Airy disc.
An ENZcalculated diffraction pattern derived from a pupil function with a relatively large content of highorder Zernike aberration terms (up to the 23^{rd} radial order) that emulates the effect of lowangle scattering in the optical system. The contour levels in the figure have to be multiplied by a factor of 10^{2} . The central peak intensity (Strehl intensity) is approximately 0.80 (just diffractionlimited). The lateral (x,y)coordinates are in units of v=2πr with the radial coordinate r expressed in the diffraction unit λ/NA in the image plane HighNA image formation The ENZdiffraction theory has been extended to highNA imaging by executing the following steps:
The colourful picture that was presented on the home
page is an example of a highNA forward diffraction calculation using
ENZtheory; the numerical aperture was 0.85 in this case. More field distributions
(modulus of the total field in the focal region) are shown in the figures
below, now at the extremely high numerical aperture of 0.95. The first
two pictures present a hypothetical classical Airy disc distribution (absolute
amplitude) and, at the same scale, the corresponding picture at high numerical
aperture (NA=0.95). The incident light is linearly polarized along the
horizontal direction. It is clearly seen that the highNA amplitude distribution
has become elliptical. In the horizontal crosssection, the full width
at half maximum (FWHM) is smaller than for the hypothetical Airy disc;
in the horizontal crosssection, a substantial broadening occurs.
