Open Access
Add to BibSonomy
Abstract
An integrable on-chip spectrometer, based on a transversely-chirped-grating waveguide-coupler for the 400- to 700-nm visible spectral range is demonstrated. For a fixed angle of incidence, the coupling wavelength is dependent on the local grating period and the waveguide structure. The transversely-chirped-input grating is fabricated on a SiO2-Si3N4-SiO2 waveguide atop a Si substrate by interferometric lithography in two sections on a single silicon substrate. A uniform period grating, separated from the input coupler by a propagation region, is provided for out-coupling to a 2048 element CMOS detector array. The incident light with wavelength spanning 400- to 700-nm is coupled into waveguide at 33.5° through the chirped grating coupler. A resolution of ∼ 1.2 nm is demonstrated without any signal processing reconstruction.
© 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Traditional optical spectroscopy has been laboratory-centric as a consequence of the large size and environmental sensitivity (vibration, temperature, ambient environment) of traditional spectrometers. There has been substantial research towards compact, integrated spectrometer-on-a-chip approaches [1] aimed at expanding spectroscopic applications to a broader range of fieldable uses including: analytical chemistry [2]; manufacturing control [3]; agriculture and food safety [4]; and environmental monitoring [5], among others. Miniaturized spectroscopic instruments with high-resolution, low-cost and real-time detection capability, remain an important, as yet unfulfilled, requirement for many fields [6]. Functionally, spectrometer approaches can be classified into fiber/waveguide and free-space inputs, and into four categories depending on the dispersive mechanism [1,7]. These include: miniaturized dispersive spectrometers [8,9,10]; narrow-band filter spectrometers [11]; Fourier transform spectrometers [12,13], which are typically fiber-based; and computational spectrometer systems based on deconvolution of spectrally overlapping response functions [14].
Dispersion-based spectrometers generally have one input and multiple outputs. Light is separated by array waveguide channels, diffraction from gratings or in-plane propagation in superprisms [15–17]. The resolution is dependent on the path length from the dispersive element to the detector, resulting in a trade-off between resolution and miniaturization. In many cases, the integrated spectrometer is fabricated with micro-electro-mechanical systems (MEMS) [18,19], or based on waveguides and integrated optics [20].
Narrow-band filter spectrometers are based on number of non-overlapping spectral filters matched to a detector array. The filters can be dielectric, Fabry-Perot, either in an array or with MEM spacer control [11], or dielectric metasurfaces [21]. Typically, the fabrication complexity increases dramatically with the number of channels, which limits the resolution and wavelength coverage. A guided mode resonance filter, based on a longitudinally chirped grating providing a notch-filter function, has been previously demonstrated [22].
Waveguide-based Mach-Zehnder interferometers are the principal ingredients of Fourier transform systems (FTS) [23]. When two beams with different optical paths recombine, the phase difference leads to interference. The wavelength is calculated by final intensity and path-length difference. FTS have good signal-to-noise ratio and resolution. However, the small size available for FTS on a chip limits the optical path difference, which impacts the achievable resolution. FTS systems are mainly aimed at communications applications with fiber input. Ambient temperature compensation and active trimming of resonances are also issues for FTS [6].
Computational spectrometers are a new spectrometer paradigm, relying on an array of relatively broad, overlapping filter bands such as: quantum-dots [24], photonic crystal slabs [25], plasmonic resonances [26,27], or disordered photonic chips [28]. A computational reconstruction of the spectrum is necessary as individual spectral bands do not have high resolution (∼5-10 nm), making the resolution dependent on signal-to-noise levels [1].
In our previous work, a transversely chirped grating was fabricated atop a waveguide as a lab-on-a-chip spectrometer [29] The chirped grating waveguide coupler serves as a spatially dependent narrow-band filter. After a short propagation region to eliminate light not coupled into the waveguide, the light is outcoupled with a second, fixed period grating to a linear detector array. As opposed to dispersion-based spectrometers, the grating coupler functions primarily as a filter and there is no relationship between the resolution and the propagation distance. The resolution corresponds to a combination of the coupling strength, the transverse chirp rate, and the number of elements in the detector array (1170 active devices in our initial experiment; 2048 in the current implementation). The chirped grating, which was defined with a single interferometric lithography step, and the bottom waveguide cladding are composed of SiO2 above and below a Si3N4 waveguide core on a Si substrate. Incident free-space radiation is coupled into a slab waveguide by the chirped grating, propagates to an out-coupling grating at the other side of a propagation region that serves to eliminate any scattered light not coupling into a waveguide mode, and is detected by a linear CMOS detector array. The grating period was varied from 475 nm to 535 nm in a 16 mm width chip. Spectral range variation throughout the visible and the NIR was achieved by adjusting coupling angle. The demonstrated resolution was 0.3 nm at 633 nm.
In this paper, the extension of the chirped grating spectrometer to cover the entire visible spectrum (from 400- to 700-nm) in a single measurement at a fixed angle is described. The structure is comprised of a SiO2/Si3N4/SiO2 waveguide atop a Si wafer. The top SiO2 cladding is fabricated as a transversely chirped grating in the input coupling region. A schematic of the plenoptic spectrometer structure is shown in Fig. 1. The light to be analyzed is incident at a fixed angle, is selectively coupled into the waveguide at wavelength-dependent lateral positions where the grating coupling phase-matching condition is met along the 28 mm length of the collection area, chosen to match a commercial 2048-element detector array [30]. The coupled light transits the free propagation region, eliminating any light that is not coupled into the waveguide, and finally is coupled out with a uniform grating. A CMOS linear detector array is mounted directly above the grating out-coupler to avoid signal loss due to variations in the out-coupling angle across the visible spectrum, and to shield the detector array from incident light. The signal at each pixel corresponds to a continuously varying wavelength defined by the local grating period. In future iterations, the detector pixels can be integrated into the underlying Si to form a fully on-chip integrated spectrometer.
Fig. 1. Schematic of transversely chirped grating plenoptic device. Light is incident at a fixed angle on the chirped grating; is coupled into the Si3N4 waveguide depending on the local grating period; is filtered to eliminate near-field scattered light in the propagation direction; is outcoupled with a uniform grating, and is detected with a linear CMOS detector array (not shown) that is placed directly over the output grating. A narrow band source is depicted. The input illuminates the full length of the chirped grating. The coupling to the slab waveguide only occurs in a narrow region where the grating coupler is phase-matched to the waveguide. This light propagates in the waveguide and is out-coupled with a second uniform grating to the linear detector array.
2. Device design
The SiO2/Si3N4/SiO2 waveguide structure determines the coupling and propagation. The bottom SiO2 layer is 1-µm thick to eliminate propagation losses associated with the underlying silicon. The Si3N4 core layer is 160 nm thick to allow only zero-order mode propagation in the waveguide across the visible spectral region. Both TE and TM zero-order modes exist with different effective modal indices; in operation, the input light is TE polarized to eliminate excitation of the TM mode. The top cladding is 450 nm thick SiO2 and is etched to form both the input chirped and output gratings. Based on this structure, the zero-order modal index for TE polarization is 1.92 at 400 nm and 1.72 at 700 nm. The relationship of structure, wavelength and coupling angle is
The wavelength vs. position curve is shown in Fig. 2.
Assuming a plane wave input overfilling the grating area, the resolution is set by a combination of 1) the coupling length (Lc) into the waveguide determined by grating and waveguide parameters; 2) the physical width of the grating, 3) the transverse grating chirp rate, and 4) the pixelation of the detectors. For the present design, Lc dominates the resolution [29,31]. Approximately, this resolution is given by:
Two-dimensional finite-difference-time-domain (FDTD) simulations [32] were used to evaluate this resolution component. Figure 3 gives simulated wavelength scans evaluated at the two wavelength extremes (a) 400 nm and (b) 700 nm of the wavelength range. Figure 3(c) shows the numerical calculation of the Lc ∼ 40 µm, roughly constant across the entire wavelength range. This calculation was based on propagating a waveguide mode under the grating and observing the radiation loss. The coupling lengths are comparable across the spectrum, so the difference in the resolution is primarily due to the wavelength and the waveguide dispersion differences.
Fig. 3. Simulated wavelength scans evaluated at the wavelength extremes at (a) 700 nm, (b) 400 nm; (c) coupling length of 700-nm wavelength at 605-nm pitch, and 400-nm wavelength at 293 nm pitch.
An additional constraint on the resolution is the transverse variation of the input coupling grating along with the 14 µm width of each detector pixel. From Fig. 2, the chirp rate for the large period region is 15 nm/mm (0.2 nm/pixel) while the chirp rate for the small period region is 10 nm/mm (0.14 nm/pixel). For the ∼ 1 nm grating resolution set by the coupling length, this gives about 5-6 pixels, and does not significantly affect the final resolution. This modeling is consistent with the experimental result for a 633 nm HeNe laser source with an intrinsic linewidth much less than the instrumental resolution as shown in Fig. 4.
The diameter of the input beam was ∼1 mm. As the chirped grating period variation is 0.015 nm/µm, the linewidth can be calculated by Eq. (1) where each pixel width is 14 µm. The increment of chirped grating spectrometer at red light wavelength is 0.23 nm/pixel. The points in Fig. 4 are normalized data monitored by individual pixels directly from the detector readout with an integration time of 1s. The solid line is the data fitting by Lorentzian function to reject noise due to scattering by laboratory fabrication, no other averaging or smoothing method was used. The peak wavelength was calibrated with a HORIBA Micro-HR spectrometer. There are ∼6 pixels within the FWHM of the response and the demonstrated resolution is ∼1.2 nm, in good agreement with the simulated linewidth of 1.16 nm at this wavelength.
3. Chirped grating spectrometer results
For demonstrating the wavelength range of the spectrometer, three individual color diode lasers are used at 33.5° coupling angle and TE polarization. From Eq. (2) the resolution depends on the wavelength and is improved at shorter wavelengths. No post-acquisition averaging was used; the integration time was 1 s. Figure 5 shows the spectral results for the three lasers, beam spot size on the collection region was ∼1 mm. The black lines are reference results measured with a commercial spectrometer (offset for clarity). The three peak wavelengths are 655.5 nm, 531.7 nm, and 407.5 nm. The colored lines are the chirped grating spectrometer results. In each case, the peak wavelength is calibrated to the commercial spectrometer. The FWHM linewidth of these lasers measured by the chirped grating spectrometer are 1.6 nm (Red), 1.2 nm (Green) and 1.2 nm (Blue). The result for the blue and red lasers matches very well with reference data. The linewidth of the green laser is extended compared to the commercial spectrometer results as a result of the instrumental resolution. The FDTD simulation, including both input and output coupling, gives a throughput of ∼7.4% in the blue and ∼9.7% in the red. This result can be improved to ∼ 40% through tailoring the width of the collection region, by additional cladding on the top of waveguide to reduce fabrication-induced scattering losses, and by integrating the detection with the waveguide, eliminating the out-coupling grating loss.
Fig. 5. The comparison of spectral results for three diode lasers between chirped-grating spectrometer (color) and HORIBA MicroHR Spectrometer (black, offset for clarity) at (a) 655.5 nm, (b) 531.7 nm, and (c) 407.5 nm.
As is the case with all spectrometers based on filter arrays, any light incident on regions of the transversely chirped grating that do not couple the light into the waveguide is reflected and not detected. The overall efficiency of the transmission depends on the total area illuminated; on the collection length, Lc; and on the spectral content of the incident light, with higher overall efficiencies for broadband light. For example, for the narrow-band lasers reported in Fig. 5, the coupling extends over only about 10 pixels out of the 2048 (∼0.5%) so that only a small fraction of the incident light is propagated to the detector array for a fully illuminated collection area. Since these were narrowband spectral sources, the illumination was restricted to ∼1 mm (∼70 pixels) so about 10% of the incident light was collected. Coupling losses further limit the efficiency. For plane wave excitation, the irradiance is uniform across the grating; for illumination sources that are focused within the collection area, the intensity variation across the length of the grating has to be taken into account.
In order to verify the performance of the chirped grating spectrometer for more extended wavelength sources, the zero-order TE mode results for three primary color LEDs are shown in Fig. 6. A polarizer was used over the chirped grating region to eliminate any TM mode coupling. Beam spot size was 15 mm. The peak wavelengths are 635 nm, 519 nm and 465 nm. The angle of incidence of the three LED beams was again set to 33.5°. As the LED linewidths are much wider than the resolution of the chirped-grating spectrometer, a five-point moving average is used with 1 s integration time. From Fig. 6, the LED spectra data from chirped-grating spectrometer (color curves) matches the results from a commercial spectrometer (offset black curves) very well.
Fig. 6. The comparison of the spectral results for three LEDs between chirped-grating spectrometer(color) and Exemplar Plus Spectrometer (black, offset for clarity) for diode laser sources at (a) 635 nm, (b) 519 nm, and (c) 465 nm.
To check the entire visible region, the spectrum of a white LED, consisting of a blue LED along with a yellow phosphor, was measured and is shown in Fig. 7. Since the chirped grating region has two different period regions with different chirp rates and opposite chirp rate variations, the measurement was separated two parts for easier converting calculation from pixel number to wavelength. The beam spot size was 15 mm and was shifted between measurements. The coupling angle was fixed at 33.5°. This result could be obtained in a single measurement with additional optics. A five-point moving average is used with 1 s integration time. The coupling efficiency of the two grating sections are normalized to each other in the overlap region around 500 nm. Locally weighted scatterplot smoothing was used to further reduce the noise. The data was further corrected for the detector response which varies significantly across this large wavelength range, the small bump at ∼ 600 nm is likely due to window effects in the separately packaged detector array that were not characterized. The chirped grating spectrum is close to the commercial spectrometer result. For quantitative spectral measurements, a calibration of the overall system response across the full wavelength range will be required.
Fig. 7. Spectral result for the TE-polarized output of a white LED incident across the chirped grating period (color) with equivalent result from a commercial spectrometer (black).
4. Fabrication
The optical arrangement for fabrication of the chirped grating is based on a Lloyd’s mirror arrangement with the addition of a focusing lens and tilt of the sample at an angle β from the optical axis [33–35]. The experimental setup is show in Fig. 8(a) and Zemax optical ray tracing is show in Figs. 8(b),(c). The Lloyd’s mirror is perpendicular to the plane of the planoconvex lens. Half of the beam is incident on the Lloyd’s mirror with angle ϕ, is reflected and is incident at -ϕ on the lens, symmetric to the other half of the beam which is directly incident on the surface of the lens. The two halves of the beam result in two converging wavefronts after the planoconvex lens and produce an interference pattern at the sample surface. The sample is mounted with a tilt angle β from the optical axis. Without this tilt, the constant period contours of the IL patterned would be circular corresponding to the interference between two spherical waves with a small quadratic variation in period in all directions. The tilt adds a dominant linear variation along the tilt direction. D is the distance on the optical axis between the lens back surface and the sample. The plane of the lens is defined as the x-y plane, the plane of the sample surface is rotated to the u-v plane. The period on the sample surface at different position is decided by the foci location X and Z of the two sections of the beam after the lens, which is defined by optical ray tracing matrix, as shown in Eq. (3):
Fig. 8. (a) Photograph of the setup of interferometric lithography for a chirped grating; (b) top view of Zemax optical ray tracing; (c) side view of Zemax optical ray tracing.
Here n is refractive index of lens, t is center thickness of the lens, R1,2 is the radius of the convex surface. R2 = ∞ for the plane surface of the lens. To avoid high-NA errors due lens aberrations, xo should not be close to the edge of lens. The final period is given by Eq. (4):
Based on the equations above, the chirped grating period variation is determined by three parameters: the lens focal length; the sample tilt angle; and the incident IL angle. For the current experiments, the chirped grating collection region is separately formed in two sections, one is from 610- to 386-nm with IL incident angle of 13.5°, the other section is from 410- to 290-nm with an IL incident angle of 20°. The focal length of the N-BK7 planoconvex lens was 40 mm and the diameter was 30 mm. The length of the chirped grating region from 610 nm to 386 nm is 15 mm, while the length of the chirped grating region from 410 nm to 290 nm is 12 mm. An overlap between periods of 10 nm is provided to allow cross-calibration between the two sections. There is a ∼ 1 mm gap between the two chirped grating regions to avoid interference between the two exposures. The total length of chirped grating collection, including the gap, is 28 mm. Figure 9(a) is an image of the chirped grating spectrometer after fabrication, and Fig. 9(b) is the period variation comparison with calculation by Matlab and period measurement by diffraction.
Fig. 9. (a) Image of chirped grating spectrometer after lithography; (b) period variation comparison with calculation and measurement for both chirped grating regions; (c) SEM of fabricated chirped grating at 360 nm period; (d) SEM of fabricated chirped grating at 550 nm period.
It is difficult to reach the full period range from 610- to 290-nm in a single exposure. Changing the IL angle to shift the center period allows reaching the necessary variation in two exposures. As a result of the tilt, the intensity varies along the v direction; the power density at the small periods, closer to the focus, was higher than at the large period, limiting the allowable tilt. To optimize the tradeoff between the grating uniformity and chirp rate, β was set to 45°. A bottom antireflection coating, icon-7, and negative photoresist, NR7-250P were used [36]. Soft baking temperatures were 180°C and 150°C, respectively, for 1 min each. The dosage of the uniform grating is 0.44 J/cm2. The dosage in front of the plano-convex lens for the large (small) period range was 0.17 J/cm2 (0.25 J/cm2). It was necessary to adjust the duration of the exposure from top to bottom of the grating to avoid overexposure of the higher intensity region. For a uniform exposure time, the pattern at small periods was overexposed while the pattern at large periods was underexposed. This was done with a block set in front of chirped-grating setup. The block was adjusted vertically during the exposure to vary the exposure time from end to end of each grating. Additionally, the average exposure time was different in the two chirped grating regions for pattern quality. Post baking was at 100°C with MF-321 developer for 1 min after exposure. A 10 s oxygen plasma etch was applied for cleaning out any residual photoresist and ARC. Figures 9(c),(d) are SEMs of fabricated SiO2 gratings showing that well-defined, vertical-sidewall gratings were obtained across the entire grating period range.
5. Discussion and conclusion
A spectrometer on-a-chip based on a transversely chirped grating waveguide coupler has been designed and fabricated. The chirped grating device is composed of a chirped-grating collection region, a waveguide propagation region to eliminate uncoupled light and a uniform grating out-coupling region. A commercial 2048-pixel linear array CMOS detector [29] is mounted in soft contact over the out-coupling grating. With today’s integrated circuit capabilities, the CMOS detector and chirped grating device can be integrated to provide a fully CMOS-compatible spectrometer. Light from free space covering the full visible spectrum, 400- to 700-nm is coupled into a waveguide with a chirped grating at a fixed angle and is spectrally resolved. The chirped grating serves as a continuously varying narrow band filter. The footprint of the current spectrometer is ∼ 28 × 20 = 560 mm2; this can be dramatically reduced to ∼ 5 mm2 as discussed below. The grating period is varied from 290 nm to 610 nm. For a He-Ne laser at 633 nm, the spectrometer resolution is ∼1.2 nm, principally determined by the collection grating coupling length. Results of narrow band diode lasers and broadband LEDs at different wavelengths across the visible match well with commercial spectrometer results. Converting the diode array output to a spectrum requires only a single step and signal processing deconvolution is not required. The collection grating is simply defined by interferometric lithography. In future work, the CMOS detector array and signal processing electronics can be integrated along the with chirped grating on a silicon chip. This will result in an inexpensive, high-resolution free-space-input miniaturized spectrometer.
The dimensions of the current device were largely determined by limitations of the available IL capability. The use of a commercial CMOS detector array set the 28-mm transverse dimension. Since the detector was externally mounted, a ∼ 1-cm propagation distance was needed to avoid covering the collection grating with the packaged array. In an integrated device, fabricated with standard IC processing the system footprint can be greatly reduced. We need ∼ 150 µm for the length of the collection area, ∼ 4xLc, only ∼ 5 µm for the propagation distance to filter out any scattered light not coupled into waveguide modes, and ∼ 10 µm for the detector, giving a total system width of only ∼ 165- to 175-µm, or a footprint of only ∼ 28 × 0.175 = 4.9 mm2. The chirped grating and the detectors could be arranged in multiple rows of collection/detection to provide a more convenient package shape, smaller detector pixels and a Ge detector with higher absorption would allow further miniaturization. The collection gratings could be designed with a variable duty factor to allow uniform coupling efficiency and resolution throughout the spectrum. The detector could be elevated to be close to the waveguide, so that an output coupling grating is not needed, giving higher throughput, and the readout and signal processing can be co-located on the chip. The coupling grating periods of > 290 nm and the modest chirp are well within the capabilities of 193-nm ArF lithography tools. The wavelength range could be extended to the near IR with Ge detectors, and into the mid- to long-wave IR with different waveguide and detector materials. This chirped grating spectrometer concept is well suited for integration and within the reach of modest IC fabs.
Funding
Advanced Research Projects Agency - Energy (DE-AR0000942).
Acknowledgment
We thank Dr. R. Karlicek of Rensselaer Polytechnic Institute for bringing the need for spectral sensing for solid-state lighting to our attention and for insightful suggestions throughout the course of this work.
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. Z. Yang, T. Albrow-Owen, W. Cai, and T. Hasan, “Miniaturization of optical spectrometers,” Science 371(6530), 722–727 (2021). [CrossRef]
2. Y. Ozaki, “Near-infrared spectroscopy—Its versatility in analytical chemistry,” Anal. Sci. 28(6), 545–563 (2012). [CrossRef]
3. M. Inoue, I. Moreno, O. Uchino, Y. Miyamoto, T. Saeki, Y. Yoshida, T. Yokota, C. Sweeny, P. P. Tans, S. C. Biraud, T. Machida, J. V. Pittman, E. A. Kort, T. Tanaka, S. Kawakami, Y. Sawa, K. Tsuboi, and H. Matsueda, “Validation of XCH4 derived from SWIR spectra of GOSAT TANSO-FTS with aircraft measurement data,” Atmos. Meas. Tech. 7(9), 2987–3005 (2014). [CrossRef]
4. A. Hashimoto and T. Kameoka, “Applications of infrared spectroscopy to biochemical, food, and agricultural processes,” Appl. Spectrosc. Rev. 43(5), 416–451 (2008). [CrossRef]
5. T. Yokota, H. Oguma, I. Morino, A. Higurashi, T. Aoki, and G. Inoue, “Test measurements by a BBM of the nadir-looking SWIR FTS aboard GOSAT to monitor CO2 column density from space,” Proc. SPIE 5652, 182 (2004). [CrossRef]
6. L. Zhang, J. Chen, C. Ma, W. Li, Z. Qi, and N. Xue, “Research progress on on-chip Fourier transform spectrometer,” Laser Photonics Rev. 15(9), 2100016 (2021). [CrossRef]
7. S. A. Khan and A. K. E. Bowden, “Colloidal quantum dots for cost-effective, miniaturized, and simple spectrometers,” Clin. Chem. 62(4), 548–550 (2016). [CrossRef]
8. Y. M. Sabry, D. Khalil, and T Bourouina, “Monolithic silicon-micromachined free-space optical interferometers on chip,” Laser Photonics Rev. 9(1), 1–24 (2015). [CrossRef]
9. A. Y. Zhu, W-T. Chen, M. Khorasaninejad, J. Oh, A. Zaidi, I. Mishra, R C Devlin, and F apasso, “Ultra-compact visible chiral spectrometer with meta-lenses,” APL Photonics 2(3), 036103 (2017). [CrossRef]
10. G. Micó, B. Gargallo, D. Pastor, and P. Muñoz, “Integrated optic sensing spectrometer: concept and design,” Sensors 19(5), 1018 (2019). [CrossRef]
11. M. Ebermann, N. Neumann, K. Hiller, M. Seifert, M. Meinig, and S. Kurth, “Tunable MEMS Fabry-Pérot filters for infrared microspectrometers: a review,” Proc. SPIE 9760, 97600H (2016). [CrossRef]
12. M. C. M. M. Souza, A. Grieco, N. C. Frateschi, and Y. Fainman, “Fourier transform spectrometer on silicon with thermo-optic non-linearity and dispersion correction,” Nat. Commun. 9(1), 665 (2018). [CrossRef]
13. E. Heidari, X. Xu, C. Chung, and R. T. Chen, “On-chip Fourier transform spectrometer on silicon-on-sapphire,” Opt. Lett. 44(11), 2883–2886 (2019). [CrossRef]
14. Y. Tsur and A. Arie, “On-chip plasmonic spectrometer,” Opt. Lett. 41(15), 3523–3526 (2016). [CrossRef]
15. A. Rahnama, K. M. Aghdami, Y. H. Kim, and P. R. Herman, “Ultracompact lens-less ‘spectrometer in fiber’ based on chirped filament-array gratings,” Adv. Photonics Res. 1(2), 2000026 (2020). [CrossRef]
16. S. Xie, Y. Meng, J. Bland-Hawthorn, S. Veilleux, and M. Dagenais, “Silicon nitride/silicon dioxide echelle grating spectrometer for operation near 1.55 µm,” IEEE Photonics J. 10(6), 1–7 (2018). [CrossRef]
17. B. Momeni, E. S. Hosseini, M. Askari, M. Soltani, and A. Adibi, “Integrated photonic crystal spectrometers for sensing applications,” Opt. Commun. 282(15), 3168–3171 (2009). [CrossRef]
18. J. Chai, K. Zhang, Y. Xue, W. Liu, T. Chen, Y. Lu, and G. Zhao, “Review of MEMS Based Fourier Transform Spectrometers,” Micromachines 11(2), 214 (2020). [CrossRef]
19. K. Yu, D. Lee, U. Krishnamoorthy, N. Park, and O. Solgaard, “Micromachined Fourier transform spectrometer on silicon optical bench platform,” Sens. Actuators 130-131(A130), 523–530 (2006). [CrossRef]
20. B. Gao, Z. Shi, and R. W. Boyd, “Design of flat-band superprism structures for on-chip spectroscopy,” Opt. Express 23(5), 6491–6496 (2015). [CrossRef]
21. A. Tittl, A. Leitis, M. Liu, F. Yesilkoy, D. Choi, D. N. Neshev, Y. S. Kivshar, and H. Altug, “Imaging-based molecular barcoding with pixelated dielectric metasurfaces,” Science 360(6393), 1105–1109 (2018). [CrossRef]
22. Y.-J. Hung, X.-W. Kao, T.-C. Kao, C.-W. Huang, J.-J. Lin, and C.-C. Yin, “Optical spectrometer based on continuously chirped guided mode resonance filter,” Opt. Express 26(21), 27515–27527 (2018). [CrossRef]
23. H. Podmore, A. Scott, P. Cheben, A. V. Velasco, J. H. Schmid, M. Vachon, and R. Lee, “Demonstration of a compressive-sensing Fourier-transform on-chip spectrometer,” Opt. Lett. 42(7), 1440–1443 (2017). [CrossRef]
24. J. Bao and M. G. Bawendi, “A colloidal quantum dot spectrometer,” Nature 523(7558), 67–70 (2015). [CrossRef]
25. Z. Wang, S. Yi, A. Chen, M. Zhou, T.-S. Luk, A. James, J. Nogan, W. Ross, G. Joe, A. Shahsafi, K. X. Wang, M. A. Kats, and Z. Yu, “Single-shot on-chip spectral sensors based on photonic crystal slabs,” Nat. Commun. 10(1), 1020 (2019). [CrossRef]
26. U. Kurokawa, B. I. Choi, and C.-C. Chang, “Filter-based miniature spectrometers: spectrum reconstruction using adaptive regularization,” IEEE Sens. J. 11(7), 1556–1563 (2011). [CrossRef]
27. W.-Y. Jang, Z. Ku, J. Jeon, J. O. Kim, S. J. Lee, J. Park, M. J. Noyola, and A. Urbas, “Experimental demonstration of adaptive infrared multispectral imaging using plasmonic filter array,” Sci. Rep. 6(1), 34876 (2016). [CrossRef]
28. W. Hartmann, P. Varytis, H. Gehring, N. Walter, F. Beutel, K. Busch, and W. Pernice, “Waveguide-integrated broadband spectrometer based on tailored disorder,” Adv. Opt. Mater. 8(6), 1901602 (2020). [CrossRef]
29. S. Nezhadbadeh, A. Neumann, P. Zarkesh-Ha, and S. R. J. Brueck, “Chirped-grating spectrometer-on-a-chip,” Opt. Express 28(17), 24501–24510 (2020). [CrossRef]
30. https://www.hamamatsu.com/resources/pdf/ssd/s11639-01_kmpd1163e
31. J. C. Brazas and L. Li, “Analysis of input-grating couplers having finite lengths,” Appl. Opt. 34(19), 3786–3792 (1995). [CrossRef]
33. S. R. J. Brueck, “Optical and interferometric lithography – nanotechnology enablers,” Proc. IEEE 93(10), 1704–1721 (2005). [CrossRef]
34. S. Benoit and S. R. J. Brueck, “Design of chirped gratings using interferometric lithography,” IEEE Photonics J. 10, 1–14 (2018). [CrossRef]
35. X. He, S. Benoit, R. Kaspi, and S. R. J. Brueck, “Optically pumped continuously tunable mid-IR distributed-feedback semiconductor laser,” IEEE J. Quantum Electron. 52(10), 1–10 (2016). [CrossRef]
36. https://futurrex.com/en/products/negative-photoresists/negative-etch-resists.html
