Light polarization management via reflection from arrays of sub-wavelength metallic twisted bands
- 561 Downloads
With constant progress of nano- and microfabrication technologies, photolithography in particular, a number of sub-wavelength metallic structures have been demonstrated that can be used to manipulate light polarization. Numerical simulations of light propagation hint that helical twisted bands can have interesting polarization properties. We use three-dimensional two-photon photolithography (direct laser writing) to fabricate a few-micrometer-thick arrays of twisted bands and coat them uniformly with metal. We demonstrate that circular polarization can be generated from linear polarization upon reflection from such structures over a broad range of frequencies in the mid infrared.
Control of polarization state of light is vital for many industrial and scientific applications. In entertainment, stereoscopic projections and 3D movies may be watched with polarization glasses, in manufacturing quality control the material stress can be assessed via the mechanically induced birefringence and in polarized light microscopy the image contrast can be enhanced.
Most elements and devices used to manipulate light polarization are based on birefringence, dichroism or optical activity in either natural or man-made materials. New concepts are constantly explored, involving surface plasmons , plasmonic metasurfaces , or novel birefringent designs , and offer solutions to polarization control in a broad range of frequencies from visible up to terahertz. Among them, recent developments in three-dimensional sub-micron-resolution two-photon laser photolithography (direct laser writing) techniques [4, 5], combined with electrochemical metal deposition on a positive resist molds led to spectacular demonstration of high-aspect-ratio helical spring-like all metal structures . These were verified to perform as transmission broadband circular polarizers, covering one octave in the mid-infrared between 3.2 and 6.4 μm. This concept was further developed toward more complex geometries, and the tapered helices were demonstrated with extended bandwidth and improved extinction . Further theoretical (numerical) studies followed, focusing on the parameter optimization in terms of transmission, contrast and bandwidth [8, 9, 10]. In , upon comparing metal and dielectric helices, it was discovered that a trade-off exists between the transmission efficiency and bandwidth in this class of circular polarizers and other all-dielectric structures with interesting polarizing properties have also been demonstrated [12, 13, 14, 15, 16].
In cholesteric (also known as chiral nematic) liquid crystals, molecules tend to organize in layers with the director axis twisted by a certain angle between consecutive layers. The resulting mesoscopic helical structure has some unique optical properties; among them is the structural coloration that has already found practical application in, e.g., bistable displays . This has inspired us to test experimentally whether the chiral geometry of cholesteric liquid crystals can be re-created artificially in the microscale and what optical properties such structures would exhibit. High-resolution 3D laser photolithography allowed us to fabricate arrays of twisted bands (TBs) with the characteristic dimensions in the range of micrometers, suitable for applications in the mid infrared spectral band.
First, we have recently discovered that arrays of dielectric twisted band structures reveal polarization-dependent diffraction . Now, further developing this concept, in this communication we demonstrate sub-wavelength metallic structures, composed of a square grid of twisted band elements covering a metal mirror, capable of polarization manipulation in the mid-infrared. In particular, we demonstrate a device capable of converting a linear polarization of the incident infrared light into circular polarization upon normal reflection.
2 Light reflection from arrays of metal twisted bands
Oscillator strengths (f), inverse lifetimes (Γ), and frequencies (ω) for gold used in the Lorentz–Drude model.
The electric field components of the reflected wave are recorded on the monitor (MON) and by Fourier transforming them with respect to time we retrieve the amplitude and phase of each electric field component as a function of frequency. By normalizing these to the values obtained for an empty simulation box (without the structure inside) we determine the amplitude reflection coefficient for each reflected wave component (E x and E y ) and their relative phase shift.
3 Reflecting structure design and fabrication
Following the standard procedure of 3D laser writing , the structure was fabricated by photopolymerization of a liquid polymer photoresist (IP-L, Nanoscribe GmbH) on a microscope cover glass surface using a photolithographic workstation (Photonic Professional, Nanoscribe GmbH). After development in isopropanol bath for 30 min, the structure (including the base glass surface) was uniformly sputter coated with a 50-nm-thick layer of gold in a low-vacuum glow discharge sputtering coater (Q150R S, Quorum Technologies).
The polarizing properties of the structures were characterized with a Fourier transform infrared spectrometer (Tensor 37, Bruker) coupled to an optical microscope (Hyperion, Bruker). A low NA = 0.4 IR objective was used to limit the maximum incidence angles on the sample to 24°. Two broadband wire grid polarizers (5-2825, Optometrics) were used: the first one to generate linearly polarized incident beam and the second one (analyzer) to analyze the polarization state of the reflected light. Several spectra were measured in the range between 1000 and 1300 cm−1 for different orientations of the analyzer and were normalized using the reference spectrum for the light reflected from a plain, gold-coated glass substrate. Periodic functions were fitted to the data to retrieve the polarization state of the reflected beam at each wavelength, as described below. We have also printed several test structures with different geometries: the height h of 4,6,8 and 10 μm, the lattice constant a of 4, 5, 6, 7, and 8 μm, the twisted band radius r of 1.5, 1.7, 2.0, 2.4, and 3.0 μm and various orientations of the first rod with respect to the array orientation; one of them, exhibiting the most promising polarizing properties, is described in the following part (h = 8 μm, a = 6 μm, 2r = 4 μm, 13.5 ° μm−1 pitch). From the SEM measurements, we estimate the fabrication tolerances to be of the order of 0.2 μm for r and h.
4 Results and discussion
While comparing the simulation results with the experimental data, some discrepancies are visible within the overall qualitative agreement. Similarly to the experimental results, the calculated orientation angle as well as the ellipticity are nearly constant across the 1100–1200 cm−1 band; however, in the simulation the latter reaches 0.5, compared to almost unity (i.e., circular polarization) in the measurements. Outside this range, in both cases, the ellipticity decreases monotonically at around 0.5/100 nm rate, while the orientation angle increases/decreases toward the long/short wavelengths, modulo the sign. We attribute the discrepancies to the fact that the absolute height of the printed twisted bands may differ from the designed value of 8 μm—careful analysis of the simulation results revealed that minute changes of this parameter lead to significant variations in both the ellipticity and the orientation angle of the reflected waves. This sensitivity of the polarizing properties to the TB arrays to the height, also present in the dielectric, diffracting structures , is among the known drawbacks of this design. Decreasing the TB height from 8 to 7 μm (Fig. 2, dashed lines) leads to visibly different ellipticity and orientation angle spectral characteristics.
In conclusion, we have presented how metallic arrays of twisted bands can be used for light polarization manipulation. In particular, a structure has been designed, fabricated, and characterized that acts as a broadband quarter waveplate (although with moderate reflectivity so far) in the mid-infrared upon normal light reflection. The structure features simple fabrication, as compared to previously demonstrated designs  and potential for straightforward integration into opto-electronic and photonic devices. The best broadband achromatic retarders currently available for the MIR range (made of several stacked MgF2 plates) offer the retardation uniformity as high as λ/100 over the 1700–3300 cm−1 range. They are, however, very expensive and, more importantly, operate in certain spectral bands only, due to limited selection of transparent, birefringent materials. Our design offers simple scalability, e.g., toward longer wavelengths, as well as high manufacturing flexibility—the concept of twisted bands is not limited to uniform arrays, and “patterned waveplates” can be designed, with the polarizing characteristics varying across the surface.
JH acknowledges the use of the TeraACMIN computer cluster at the Academic Centre for Materials and Nanotechnology, AGH-UST, Krakow, Poland.
Compliance with ethical standards
National Science Centre (Poland) (DEC-2012/05/E/ST3/03281), European Regional Development Fund Operational Programme (POIG.02.01.00-14-122/09-00).
- 16.A. Passaseo, M. Esposito, M. Cuscuna, V. Tasco, Materials and 3D designs of helix nanostructures for chirality at optical frequencies. Adv. Mater. 5(16), 1601079 (2017)Google Scholar
- 23.M. Born, E. Wolf, Principles of Optics (Pergamon Press, Pergamon, 1970)Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.