Open Access
Add to BibSonomy
Abstract
We present novel polarization independent, high-quality monolithic spectral filters based on the guided-mode resonance (GMR) effect with orthogonal linear gratings on either side of the substrate operating in the longwave infrared (LWIR) spectral region. We employ high-spatial resolution e-beam lithography and reactive-ion etching (RIE) nanofabrication techniques to achieve large-area (10×10 mm2) notch filters with subwavelength features. We fabricated prototype filters and characterized their polarization independent spectral performance with both coherent and incoherent incident light using a tunable quantum cascade laser (QCL) system that spans the ∼8–12 µm spectral band as well as a Fourier transform infrared (FTIR) spectrometer with collimated incident beam.
© 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Development of polarization independent filters in the 8 to 12 µm LWIR spectral region is important as this region corresponds to an atmospheric window as well as to the peak terrestrial emission and is widely used for day/night sensing and imaging applications. Some of these applications require use of compact spectrally tunable notch or bandstop filters, which reflect a narrow band of incident light while transmitting the rest. In 1992, Magnusson and Wang [1] showed that small compact notch filters can be designed based on the GMR effect in dielectric waveguide-grating (WGG) and since then such filters have been demonstrated from the visible to LWIR spectral regions [1–12]. A GMR filter (GMRF) based on one-dimensional (1-D) gratings only on one side of substrate is polarization sensitive [1–10].
Recently, we demonstrated polarization sensitivity of an LWIR GMRF with zero-contrast (ZC) 1-D germanium (Ge) WGG on only one side of a zinc selenide (ZnSe) substrate by measuring its transmittance using a tunable QCL at the normal incidence of light [10]. To achieve polarization independent performance at the normal incidence, we used a combination of two separate 1-D LWIR GMRFs with linear gratings in orthogonal orientations [9]. Similar polarization independence using two orthogonally stacked 1-D GMRFs was also demonstrated in the shortwave infrared (SWIR) [13,14]. However, stacking individually fabricated filters can face challenges not only in alignment, but also uniformity over the filter surface. By fabricating 1-D WGG on both sides of the substrate with orthogonal orientation of linear gratings, a compact monolithic polarization independent GMRF can be realized without the alignment and uniformity issues. A polarization independent double-sided (DS) GMRF with orthogonal WGGs on either side of glass substrate was proposed in the SWIR using more conventional cleanroom materials [15].
In this paper, we report on fabrication and experimental demonstration of a novel LWIR monolithic DS filter that is polarization-independent for normal incidence of light. To the best of our knowledge, this is the first time such monolithic filters are reported. We designed, fabricated, and characterized such polarization independent GMR notch filters from 8 to 12 µm using high-refractive index (n) transparent dielectric materials, i.e., Ge with n = 4.0 and ZnSe with n = 2.41. The metamaterial filter consists of a 1-D Ge subwavelength ZC WGG structure on either side of a ZnSe substrate where the gratings at the top and bottom are oriented orthogonally to each other. The filter reflects the incident broadband light at two narrow spectral bands while fully transmitting the rest independent of polarization. We carefully chose the device parameters—grating period and fill factor, and grating and waveguide thicknesses—for this spectral range of operation. The notch filtering happens at two sharp narrowband reflection notches independent of incident polarization because of the GMR phenomenon. We implemented the rigorous coupled-wave analysis (RCWA) algorithm with particle swarm optimization to carry out the modeling and design of the filters to obtain the best diffraction efficiency. Physical vapor deposition, e-beam lithography, and RIE techniques were used to deposit and fabricate high-spatial resolution high-quality 1-D ZC WGG on either side of a ZnSe substrate yielding prototype DS GMRFs. In Fig. 1(a), we show the concept of the GMR effect on a DS GMRF device based on ZC grating geometry. The ZC WGG is composed of a grating on the sublayer waveguide guiding lateral Bloch modes coupled by evanescent diffraction orders [1–4]. Due to the leaky mode excitation in the periodic WGG, the notch filtering response occurs reflecting a narrow spectral region of the polarized incident light while transmitting the rest. By sandwiching a substrate between two identical ZC WGG oriented orthogonally, polarization independent GMR filtering is achieved in a compact manner. Figure 1(b) shows the optical image of the fabricated DS GMRF (back side is reflected by a mirror). Large diffraction patterns arising from a 10 ×10 mm2 metamaterial structure on the front and back sides confirm the grating patterns on 25.4-mm round substrate. We characterized the filtering performance of such filters using two separate experimental setups using coherent and incoherent incident light—(i) an automated room-temperature QCL system tunable from 8 to 12 µm and a thermal detector, and (ii) a modified commercial FTIR spectrometer with a collimated incident beam and a high-sensitivity, high-speed liquid nitrogen (LN2) cooled mercury cadmium telluride (MCT) detector. We obtain an excellent agreement between the theoretical and experimental results.
Fig. 1. A DS GMRF device. (a) Schematic illustration of the DS GMRF reflecting the incident light at two notch wavelengths. (b) Optical image of the DS GMRF. Top of image shows front of the device and bottom shows the reflection by a mirror of the back side of the device.
2. Materials and methods
2.1 Modeling and simulation
To predict the wideband electromagnetic properties of the DS GMRFs, we implemented the RCWA to obtain the design parameters using Ge and ZnSe [16]. In our model, the input parameters are refractive indices of incident/exit regions, grating layers, waveguide (WG) layers, and the substrate; incidence angle; period (P), thickness (t1), and fill factor (ff) of the grating; thickness of the planar WG (t2); and thickness of the substrate (t3) as shown in Fig. 1(a). Our design used same parameter values for both the top and bottom WGG structures. We chose design parameters for the DS GMRF—P = 3.1 µm; ff = 0.37; t1 = 370 nm; t2 = 1 µm, with t3 = 1 mm. In Fig. 2, we show computed contour color maps that display the spectral transmittance as a function of P, ff, t1, and t2 for both the transverse electric (TE) and transverse magnetic (TM) incident polarizations. It is clear our chosen parameter values correspond to the 8–12 µm range. The noise in the contour plots is due to the multiple reflections of coherent light in the 1-mm-thick substrate.
Fig. 2. Modeling and parameter selection for DS-GMRFs. Contour plots showing spectral transmittance as a function of (a) period P with ff = 0.37, t1= 0.37 µm and t2 = 1.03 µm; (b) fill factor ff with P = 3.1 µm, t1= 0.37 µm and t2 = 1.03 µm; (c) grating thickness t1 with P = 3.1 µm, ff= 0.37 and t2 = 1.03 µm; and (d) waveguide thickness t2 with P = 3.1 µm, ff = 0.37 and t1= 0.37 µm for both the TE and TM polarizations at normal incidence of light.
2.2 Device fabrication and characterization
For the fabrication of a DS GMRF, a Ge ZC WGG is deposited on the top and bottom of a ZnSe substrate using the e-beam lithography and RIE, as shown in Fig. 3. First, a 1-mm-thick polished ZnSe substrate with a diameter of 25.4 mm was sonicated in acetone and rinsed using deionized water (DIW) and isopropyl alcohol (IPA). After cleaning and surface roughening by an ion-miller for adhesion promotion, we deposited a 1.4-µm-thick Ge film using the e-beam evaporation technique. An argon-ion beam was used for the ion-milling process and allowed us to avoid creating an adhesion layer, which can deteriorate the optical performance of the GMRF. The deposition rate for Ge was 1.3 Å/s.
Fig. 3. DS GMRF fabrication steps. Fabrication of the top and bottom sides of the DS GMRF: scheme of the fabrication process. (i) Cleaning of the ZnSe substrate and ion-milling for surface adhesion promotion. (ii) Ge film deposition on top side using e-beam evaporation. (iii) e-beam lithography for etch-mask patterning and fluorine-based ICP-RIE etching for grating patterning. (iv) Repeat the same steps on the back side. (v) Completed DS GMRF filter. Alignment marks are used to align the relative grating orientations on the top and bottom of the substrate.
Next, an etch mask was patterned using the e-beam lithography process, which is a maskless direct writing process. A grating layout design was prepared using computer-aided design software (KLayout). An e-beam resist (ZEON ZEP520A) was spin-coated onto the Ge layer grown on a ZnSe substrate for 60 s at 1000 RPM, and then baked for 10 min at 180 °C. To avoid charging effects during e-beam lithography, an electrification dissipating material (Showa Denko America Espacer) was spun for 60 s at 1500 RPM and baked for 90 s at 80 °C. After e-beam writing the 1-D grating structure, the sample was rinsed with DIW, developed using hexyl acetate for 2 min, and rinsed with IPA. Next, followed by a plasma residue cleaning process, the 1-D grating structure was patterned using an inductively coupled plasma etcher for the pseudo-Bosch process with sulfur hexafluoride (SF6) and octafluorocyclobutane (C4F8) gases, which yields a vertical sidewall profile. After the etching process, the etch mask was removed using a resist stripper (Microchem Remover PG) and rinsed with IPA. To deposit the Ge WGG on the bottom of the substrate, alignment marks were made by a direct laser writer to provide the correct relative orientation of the bottom grating. After completing the top side fabrication, for protection, it was spin coated with polymethyl-methacrylate (PMMA) at 3000 RPM and baked at 180 °C for 10 min and then the same procedure described for the top side was used for the bottom of the substrate. Afterward to complete the fabrication process, PMMA layer is removed yielding the filter shown in Fig. 1. The fabricated filter size 10×10 mm2 on the top and bottom is quite large for nanofabrication.
Figure 4(a) shows an atomic force microscopy (AFM) image of the GMRF, and confirms the grating period and thickness. We also measured top-down and cross-sectional scanning electron microscopy (SEM) images, showing the detailed shape of the grating structures. The SEM images shown in Fig. 4(b) and (c) confirms that the fabricated GMRF has straight 1-D grating lines and vertical sidewall profiles. We obtained the fabricated filter parameters for two different filters—DS#1 and DS#2 from these measurements. Values of measured parameters for the two filters on the front and back side of substrate are listed in Table 1 and are close to the design values within measurement and fabrication uncertainties.
Fig. 4. AFM/SEM device measurements. Measurements of one side of the 1-D DS GMRF. (a) AFM image of the fabricated device. (b) SEM images showing both front and back sides. (c) cross-sectional view SEM images of the GMRF. Platinum (Pt) was deposited for protection during focused ion-beam (FIB) milling.
Table 1. Measured DS filter parameters
In Fig. 5, we show the simulated transmittance spectrum for either TE or TM polarization for the normal incidence of light with the measured filter parameters for DS #2 as listed in Table 1. We used the measured value n = 4.1 for a thin film of e-beam-grown amorphous Ge on a ZnSe substrate using an LWIR ellipsometer. The simulated spectrum has two deep notches at 8.63 and 10.38 µm and it shows a lot of noise due to inclusion of a 1-mm-thick substrate in the calculation. RCWA assumes a coherent plane wave at each wavelength and the multiple reflections from the parallel facets of the substrate give rise to this noise. Simulated transmittance spectrum for DS #1 is similar to Fig. 5 except the two notches are at 8.7 and 10.39 µm.
3. Filter characterization measurements and results
3.1 Coherent light measurements
We measured the spectral transmittance of the DS GMRFs using an air-cooled, quasi-continuous-wave QCL system running at 100 KHz (Daylight Solutions model 2300). There are three separate QCLs in this system to cover the LWIR range from 6.6 to 12.9 µm. Each QCL is single mode with very small sidelobes. The Labview code used to automate the QCL only covers from 8 to 12 µm corresponding to the most stable region of QCL. We can rotate the filter about the Z axis to precisely line up the grating-groove direction with the polarization orientation of the incident light along the Y axis as well as rotate about Y axis to change the angle of incidence. We used an Ophir-Spiricon model Vega-B power meter with a 10A-PPS uncooled broadband thermopile detector with a 16-mm diameter to measure the intensity of light. We used a collimated incident laser beam to be consistent with the theory that assumes an incident plane wave on the sample. Measured QCL spectra is shown in Fig. 6 where we list the tuning range of each QCL. A schematic of the experimental setup is shown in Fig. 7.
Fig. 7. Direct transmission measurement. Schematic drawing of the QCL measurement setup with uncooled thermal detector.
The direct transmission measurement using a QCL is a two-step process. First, the intensity of the laser and then the laser intensity transmitted through the filter are measured. The ratio of these two measurements at each spectral interval yields the filter spectral transmittance. We used a 10 nm spectral interval in the measurements. In our setup, the QCL is vertically polarized in the Y direction with respect to the incident beam propagating along the Z axis; hence, we can carry out measurements with incident TE polarizations at the normal angle of incidence by aligning the grating grooves on the top WGG along the Y axis. To take measurement for the TM polarization, we rotate the filter such that the grating grooves are oriented perpendicular to the laser polarization. To get the correct alignment between the grating grooves and polarization orientation, we view the diffraction pattern for a visible laser from the GMRF. The TE and TM transmittance spectra for DS #2 are shown in Fig. 8 and similar results were obtained for DS #1. The measured transmittance spectra are noisy due to Fabry-Perot resonance from reflections within the 1-mm thick ZnSe substrate with parallel facets.
3.2 Incoherent light measurements
To measure the transmittance measurements with incoherent incident light, we used a modified commercial FTIR (Bruker 70) spectrometer with a collimated incident beam as shown in Fig. 9. Since in a traditional FTIR spectrometer, the incident beam on the sample is focused it cannot be used for our measurements as our theoretical analysis assumes a plane wave incidence on the sample. To achieve this, we use a 45° turning mirror to reflect the light from the interferometer inside the FTIR to get a collimated incident beam propagating along the Z axis coming out through a ZnSe side window of the spectrometer. After the collimated beam is incident on the sample and gets transmitted through, it is focused by a parabolic mirror on a LN2-cooled 1-mm-diameter MCT detector with 20 KHz scan speed. We use an LWIR wire grid polarizer before the sample to change the polarization of the incident light along and perpendicular to the Y axis by rotating the polarizer while the top grating is oriented along the Y direction. Using this modified FTIR spectrometer we can take transmittance measurements with both the TE and TM polarized incident light at both the normal and non-normal incidence of light. In Fig. 10, we show the measured spectra for the DS #2 GMRF taken with a 1.5 cm−1 spectral resolution. Similar results were obtained for DS #1.
4. Discussion
The simulated and measured spectra shown in Figs. 5, 8 and 10 for both TE and TM polarizations show two notches as expected. The reason for this is that a monolithic DS GMRF is equivalent to two independent 1-D GMRFs in tandem with gratings in the WGG on the top and bottom of the substrate oriented orthogonally to each other. When the TE polarized light is incident on the top WGG, it gets transmitted with the first notch around 10.3 µm and looks like TM polarized light to the bottom WGG which provides the second notch around 8.7 µm. For the incident TM polarized light, the first notch from the top WGG is at 8.7 µm while the second notch from the bottom WGG is at 10 µm. The measurements with both the coherent light and incoherent incident light are close. The measured and predicted spectra and the values of the resonant notch wavelengths as listed in Table 2 for the two filters agree within the experimental uncertainties due to fabrication and measurement errors. There is much closer agreement for the TE transmittance between the simulated and measured data than for the TM transmittance. The measured notch transmittance is close to zero for the 8.7-µm notch, while it is a bit higher at the longer wavelength notch for both the QCL and FTIR measurements. The main discrepancy from simulation is in the measured longer wavelength notch for the incident TM polarization ∼10 µm for both the QCL and FTIR measurements. The AFM and SEM measured values listed in Table 1 are for a very small area of these large GMRFs and fabrication over the large area may not be uniform. Also, there may be slight misalignment in the relative orientations of the top and bottom gratings. These variations could explain the discrepancies in the measured spectra. The simulated transmittance spectrum shown in Fig. 5 and measured QCL spectra in Fig. 8 are noisier than those in Fig. 10 due to Fabry-Perot resonances for a coherent source within the 1-mm-thick ZnSe substrate which is absent for an incoherent source. We would also like to note that the sidelobes are not flat as the layer thicknesses used in the filter design did not satisfy a required antireflection (AR) condition [3].
Table 2. Theoretical and experimental values of notch wavelengths
5. Summary and conclusion
We designed and fabricated two novel LWIR monolithic DS GMRFs and experimentally demonstrated their polarization independent behavior for the normal incidence of light. To the best of our knowledge, this is the first time such filters are reported. We used RCWA algorithm with optimization to design the monolithic DS filters. We used a room-temperature tunable QCL system to carry out direct transmittance measurement as well as a modified FTIR spectrometer typically used to carry out filter transmittance measurements. Based on close agreement between the theoretical and experimental spectral transmittance for the fabricated DS GMRFs, it is clear that a monolithic DS GMRF with a very compact form factor is an important optical component in providing polarization independent filtering operation instead of using a cumbersome combination of two separate 1-D linear grating GMRFs oriented orthogonally. Nanofabrication of large DS GMRFs requiring precise alignment of the Ge WGGs on the top and bottom of the ZnSe substrate is more complex than fabricating regular GMRFs. We developed and reported here new fabrication procedures to accomplish this. We are exploring a few different schemes for AR coatings to reduce the Fresnel reflection loss and flatten the sidebands. We expect that fine-tuning of the fabrication processes will enable DS-GMRFs with matched TE and TM spectra thus delivering wideband polarization independence.
Funding
Army Research Laboratory (W911NF-18-0217).
Disclosures
The authors declare no conflicts of interest.
Data Availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request
References
1. R. Magnusson and S. S. Wang, “New principle for optical filters,” Appl. Phys. Lett. 61(9), 1022–1024 (1992). [CrossRef]
2. R. Magnusson, “Wideband reflectors with zero-contrast gratings,” Opt. Lett. 39(15), 4337 (2014). [CrossRef]
3. S. S. Wang and R. Magnusson, “Multilayer waveguide-grating filters,” Appl. Opt. 34(14), 2414 (1995). [CrossRef]
4. D. Rosenblatt, A. Sharon, and A. A. Friesem, “Resonant grating waveguide structures,” IEEE J. Quantum Electron. 33(11), 2038–2059 (1997). [CrossRef]
5. J.-N. Liu, M. V. Schulmerich, R. Bhargava, and B. T. Cunningham, “Optimally designed narrowband guided-mode resonance reflectance filters for mid-infrared spectroscopy,” Opt. Express 19(24), 24182 (2011). [CrossRef]
6. B. Hogan, S. P. Hegarty, L. Lewis, J. Romero-Vivas, T. J. Ochalski, and G. Huyet, “Realization of high-contrast gratings operating at 10 µm,” Opt. Lett. 41(21), 5130 (2016). [CrossRef]
7. N. Gupta and M. S. Mirotznik, “Performance characterization of tunable longwave infrared notch filters using quantum cascade lasers,” Opt. Eng. 57(12), 1 (2018). [CrossRef]
8. D. J. Carney and R. Magnusson, “Fabrication methods for infrared resonant devices,” Opt. Lett. 43(21), 5198 (2018). [CrossRef]
9. K. J. Lee, Y. H. Ko, N. Gupta, and R. Magnusson, “Unpolarized resonant notch filters for the 8–12 µm spectral region,” Opt. Lett. 45(16), 4452 (2020). [CrossRef]
10. N. Gupta and J. Song, “High-quality large-scale electron-beam-written resonant filters for the long-wave infrared region,” Opt. Lett. 46(2), 348 (2021). [CrossRef]
11. Y. Zhong, Z. Goldenfeld, K. Li, W. Streyer, L. Yu, L. Nordin, N. Murphy, and D. Wasserman, “Mid-wave infrared narrow bandwidth guided mode resonance notch filter,” Opt. Lett. 42(2), 223 (2017). [CrossRef]
12. A. S. Lal Krishna, V. Mere, S. K. Selvaraja, and V. Raghunathan, “Polarization-independent angle-tolerant mid-infrared spectral resonance using amorphous germanium high contrast gratings for notch filtering application,” OSA Continuum 3(5), 1194 (2020). [CrossRef]
13. K. Kawanishi, A. Shimatani, K. J. Lee, J. Inoue, S. Ura, and R. Magnusson, “Cross-stacking of guided-mode resonance gratings for polarization-independent flat-top filtering,” Opt. Lett. 45(2), 312 (2020). [CrossRef]
14. A. Monmayrant, S. Aouba, K. Chan Shin Yu, P. Arguel, A.-L. Fehrembach, A. Sentenac, and O. Gauthier-Lafaye, “Experimental demonstration of 1D crossed gratings for polarization-independent high-Q filtering,” Opt. Lett. 39(20), 6038 (2014). [CrossRef]
15. A.-L. Fehrembach, K. Chan Shin Yu, A. Monmayrant, P. Arguel, A. Sentenac, and O. Gauthier-Lafaye, “Tunable, polarization independent, narrow-band filtering with one-dimensional crossed resonant gratings,” Opt. Lett. 36(9), 1662 (2011). [CrossRef]
16. M. G. Moharam, D. A. Pommet, E. B. Grann, and T. K. Gaylord, “Stable implementation of the rigorous coupled-wave analysis for surface-relief gratings: enhanced transmittance matrix approach,” J. Opt. Soc. Am. A 12(5), 1077 (1995). [CrossRef]
