Miniature Plasmonic Wave Plates

Linear birefringence, as implemented in wave plates, is a natural way to control the state of polarization of light. We report on a general method for designing miniature planar wave plates using surface plasmons. The resonant optical device considered here is a single circular aperture surrounded by an elliptical antenna grating. The difference between the short and long axis of each ellipsis introduces a phase shift on the surface plasmons which enables the realization of a quarter wave plate. Furthermore, the experimental results and the theoretical analysis show that the general procedure used does not influence the optical coherence of the polarization state and allows us to explore completely the surface of the unit Poincaré sphere by changing only the shape of the elliptical grating.

Figure 1

(a) Scanning electron microscopy image of the elliptical bull’s eye structure. The scale bar is $2\mu \mathrm{m}$ long. The green arrow is an excited point dipole ${\mathbf{P}}_M$ located in $\mathbf{M}$. SPP (red arrow) launched in $\mathbf{M}$ propagates along the radial direction $\mathbf{M}\mathbf{O}$ to reach the hole located in $\mathbf{O}$. $w$ is the angle between $\mathbf{M}\mathbf{O}$ and ${\mathbf{P}}_M$. (b) Comparison between the experimental white light transmission spectrum (red curve) associated with the structure shown in (a) and the theoretical prediction (black curve) obtained with the 2D dipoles model. The transmission scale is dimensionless and corresponds to the ratio between the transmitted and the incoming power at the level of the aperture.

Figure 2

(a) Sketch of the optical polarization tomography setup. $P$, $\Lambda /4$, $\Lambda /2$ are, respectively, polarizers, quarter wave plates, and half wave plates located in the input and output beams. The image is recorded with a complementary metal oxide semiconductor (CMOS) camera and the light source $S$ is a laser diode emitting at $\lambda =785\mathrm{nm}$. $L_1$ and $L_2$ are two objective lenses ($\times 50$, $\mathrm{NA}=0.55$) and ($\times 40$, $\mathrm{NA}=0.6$), respectively. (b) Typical camera image of the light transmitted through our structure (the scale bar is $2\mu \mathrm{m}$). (c) Cross cut of the intensity profile along the white dotted line shown in (b).

Figure 3

(a) SOP analysis of the output beam for a linearly polarized input beam. The polarization angle is measured relative to the $x$ axis. Data points are compared to Eq. (1) (continuous curves) and to predictions of Eq. (3) (dotted curves). The colors black, green, orange, cyan, magenta, red, and blue correspond, respectively, to $I_{\mathrm{total}}$, $I_{\widehat{\mathbf{x}}}$, $I_{\widehat{\mathbf{y}}}$, $I_{\widehat{\mathbf{p}}}$, $I_{\widehat{\mathbf{m}}}$, $I_{\widehat{\mathbf{L}}}$, and $I_{\widehat{\mathbf{R}}}$. (b) Image of the input Poincaré sphere through the transformation ${\mathcal{M}}^{\mathrm{expt}}$ (yellow sphere). The black and red data points are, respectively, the input and output Stokes vectors associated with the experiment shown in (a). The circle represents the predictions deduced from ${\mathcal{M}}^{\mathrm{expt}}$. The device also converts an input $L$ state (black) star into a state shown by a red star.