Measurement of the polarization-induced asymmetry in focal spots

Abstract

Polarization effects may have a strong influence on the shape of the focal spot. Measurements and theoretical cal(.'ulations for various input fields are presented. We present the first experiment.al evidence for an ariymmetry in the case of linear polarization and for a strong longitudinal field component on the axis in the case of radial polarization. © 2001 Optical Society of America OCIS codes: 110.0110 Imaging Systems, 110.0180 Microscopy, 260.0260 Physical opti("S, 260.5430 Polarization For strong focusing with high numerical apertures, the scalar diffraction theory is no longer valid and the full vectorial character of the electromagnetic field has t.o be taken into account. Fundamental calculations, which are based on the Debye-approximation, were performed by Richards and Wolf (1). They found that for a linearly polarized beam the focal spot is asymmetrically defonued along the direction of linear polar­ ization. This asymmetry can be revoked when an input field with a doughnut-shaped intensity distribut,ion is used, which is radially or azimut.hally polarized. Here we present measurements of intensity distributions in the focal region for input fields ""rith different polarization states and show the experimental verifica.tion of the predicted asynunetry for linear polarization. To mea.'lure the focal intensity distribution with high lateral resolution we have developed an experimental setup which is based on the knife-edge method combined with tomographic reconstruction (2). The incoming beam is fo(.'used by a microscope objective onto a specially designed photodiode where parts of the active are.a are covered with sharp edgt\d opaque st.ructures that are used as a knife edge. The measurement of the photocurrent a.'l a function of the displacement of the diode allows for the calculation of the projection of the intensity distribut.ion ont.o t.he translation axis. To recom,'truct the intensity distribution using the Radon t.ransformation several projections have to be measured by moving the edge through the foells Imder different angles. vVhen this process is repeate,d for different distances between the diode and the focal plane, a three dimensional intt\llsity distribution can be obtained. In the experiment the beron of a linearly polarized Helium-Neon laser (A=633nm) was expanded to allow for a homogeneous illumination of the entrance pupil of the focusing microscope objective (NA=O.9, l00x, poL). Intensity profiles were measured for various orientations of t.he polarizat.ion with respL'Ct t.o the diode and for different defocl.l.,> positions with a relative distance of 100nm. Fig. la and 1 b show the measured projections of the int.ensity distribution onto a plane parallel to the optical axis and the direction of displacement of the edge for per­ pendicular and parallel polarized input beams with respect to the edge. It is remarkable that in contra.<,'i; to the theoretical predictions (fig. lc and Id) t.he experimentally ob­ tained intensit.y distribution does not show forward-·backward symmetry with respect to t.he focal plane (z=O) which we yet do not undt'rstand. In the area between lens and focal plcUle one finds a qualitatively good agreement and a significant differenct' between