Open Access
Add to BibSonomy
Abstract
Here, we present the design and simulation of an ultrawide-bandwidth on-chip spectrometer that can be used in various applications, e.g. spectral tissue sensing. It covers 1200 nm wavelength range (400 nm-1600 nm) with 2 nm spectral resolution. The overall design size is only 3 × 3 cm2. The ultra-wide spectral range is made possible by using novel on-chip band-pass filters for the coarse wavelength division. The fine resolution is provided by the arrayed waveguide gratings. The band-pass filter is formed by using bend waveguides and adiabatic full-couplers. The additional loss caused by the band-pass filter is relatively small. The proposed spectrometer covers entire 400 nm-1600 nm range continuously with low crosstalk values. We envision that this design can be used in several different applications including food safety, agriculture, industrial inspection, optical imaging, and biomedical research.
© 2020 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Optical spectroscopy is an invaluable method that sheds light into the interaction between light and matter and is used in a remarkable range of applications from industrial inspection to biomedical research [1,2]. One application that recently gained attention is spectral tissue sensing, which is based on diffuse reflectance spectroscopy [3]. It provides information on the composition of the tissue directly from a needle tip over full visible, near infrared, and short-wave infrared range (400 nm – 1600 nm) (see Fig. 1). In the short wavelength range, tissue scattering and blood absorption can be monitored while at the high wavelength range different material composition of tissue such as lipid and water can be differentiated. It is also a valuable diagnostic tool for screening cancerous tissue [4,5].
Fig. 1. Schematic of a spectral tissue sensing system. The input light is sent to tissue via an optical fiber that is placed in a needle, and back reflected light is collected via another optical fiber, which is sent to an optical spectrometer to be analyzed.
The main component of a spectral tissue sensing system is a spectrometer, which ideally has to be compact, high-resolution and large-bandwidth [6]. On-chip spectrometers have the potential to be used in spectral tissue sensing; however, the most exciting designs have limited bandwidth ranges [7–9]. Recently an on-chip spectrometer based on cascaded arrayed waveguide gratings (AWG) was demonstrated to be used in spectral tissue sensing; however, the crosstalk and loss values were high while the resolution was low [10]. A novel spectral beam combiner was demonstrated by Stanton et al., which has a high potential to be used in broadband on-chip spectrometers [11]. In this work, we describe the design of a novel on-chip spectrometer that can cover 400 nm-1600 nm spectral range continuously with high resolution and low crosstalk values. It is a cascaded system, which is comprised of novel band-pass filters (for initial coarse wavelength separation) and several AWGs (for fine wavelength separation). The existing on-chip band-pass filters, i.e. apodized sub-wavelength grating coupler [12], multimode waveguides with sub-micron gratings [13], and multi-mode air-cladding ring-resonators [14], have very limited bandwidth range and therefore they are not suitable for this application. An interesting photonics crystal filter operating in the infrared range with ultrawide passband was demonstrated by Baldycheva et al. [15], which could have a potential in on-chip spectrometer applications.
The novelty of the proposed spectrometer design comes from the on-chip band-pass filter configuration, which is formed by using bend waveguides and adiabatic full-couplers. It is a passively tunable, broadband, and intrinsically lossless filter with very high extinction ratios. The coarsely divided spectrum by these on-chip band-pass filters are finely separated by the high-resolution AWGs. In this paper, we designed a proof-of-principle device with three band-pass filters with 200 nm bandwidth and two band-pass filters with 300 nm bandwidth using silicon nitride (Si3N4) platform. Five AWGs were designed to cover these desired spectral ranges. The outputs of these AWGs were removed and the continuous spectra formed in the end of second free propagation region (FPR) were imaged onto array detectors via focusing lenses. The optical resolution of this spectrometer configuration is 2 nm, which is sufficient for differentiating different tissue layers. The overall design size is only 3 × 3 cm2.
2. Working principle and design
2.1 Material system and waveguide geometry:
The proposed spectrometer design is simulated for the Si3N4-on-Si waveguide platform. The material system is 200-nm-thick low-pressure chemical vapor deposited (LPCVD) Si3N4 film on a thermally-oxidized silicon wafer. The oxide thickness is 3.5 µm, and the refractive index is 1.46 at 1550 nm. The refractive index of the Si3N4 layer is 2.00 at 1550 nm. A 4-µm-thick silicon dioxide (SiO2) layer is used as the top cladding, which has a refractive index of 1.47 at 1550 nm. Single mode rib waveguides with 0.1 µm of slab height and 0.5 µm, 0.7 µm, 0.9 µm, 1.1 µm, and 1.4 µm of waveguide widths, corresponding to five band-pass filters indicated by BP1-5 in Fig. 2(c), respectively, are designed. Transverse electric mode profiles of these single-mode waveguides at their central wavelengths are given in Fig. 3.
Fig. 2. (a) The schematic of a bend waveguide, an adiabatic full coupler and their combination (left to right). Relevant design parameters of each structure are given in each figure. (b) Typical transmission response of a bend waveguide, an adiabatic full-coupler and their combined responses forming a band-pass filter (left to right). (c) Five band-pass filters (denoted by BP1, BP2, BP3, BP4, and BP5) using five bend waveguides with different radii and five adiabatic full-couplers with different design parameters.
Fig. 3. Transverse electric mode profiles of single-mode waveguides with widths of 0.5 µm, 0.7 µm, 0.9 µm, 1.1 µm, and 1.4 µm at the central wavelengths of (a) 0.5 µm, (b) 0.7 µm, (c) 0.9 µm, (d) 1.15 µm, and (e) 1.45 µm, respectively.
2.2 Working principle and simulations of a band-pass filter
A band-pass filter can be formed by combining a short-pass filter with a long-pass filter. The band-pass filters that are used in this work comprise bend waveguides and adiabatic full-couplers (Fig. 2(a)). A bend waveguide with a small radius behaves like a short-pass filter, which radiates longer wavelengths (Fig. 2(b), cut-off wavelength λ2) while an adiabatic full-coupler behaves like a long-pass filter, which transmits longer wavelengths, if the separation between coupler arms is designed accordingly (Fig. 2(b), cut-off wavelength λ1). When several bend waveguides with different bending radii are combined with several adiabatic full-couplers with different coupling parameters, one can make a series of these band-pass filters. Such a structure can form a coarse wavelength division system as shown in Fig. 2(c). The bandwidth and the central wavelength of the band-pass filter can be adjusted by changing the bending radius of the bend waveguide, and the coupling parameters of the full-coupler. In a similar way, macro bending of optical fibers has been used to design optical band-pass filters by Muhd-Yassin et al. [16].
In this work, five band-pass filters were designed to cover 400 nm-1600 nm range. The 400 nm-1000 nm range was divided into three 200-nm-wide bands by three band-pass filters, whereas 1000 nm to 1600 nm range was divided into two 300-nm-wide bands by using two band-pass filters. Five bend waveguides with different radius of curvature values, i.e. R = 20 µm, 30 µm, 50 µm, 70 µm, and 130 µm, were used to build five short-pass filters [Fig. 4(a)]. Adiabatic full-couplers, which are used as the long-pass filters, consist of two tapered waveguides with starting widths of w1 and w2, which are interchanged at the end of the tapered coupling region [17]. This type of couplers have wavelength-flattened-performance compared to directional couplers. Five long-pass filters were designed using five adiabatic full-couplers with the following design parameters: difference between starting and ending waveguide widths; Δw = w1-w2, coupler length; L, and separation between waveguides; d [see Table 1 and Fig. 4(b)]. By combining these bend waveguides with adiabatic full-couplers, five band-pass filters were formed as shown in Fig. 4(c). The crossings of adjacent band-pass filters are arranged to be at 3 dB points. The extinction ratios of these filters were ≥ 70 dB.
Fig. 4. Transmission responses of (a) 6 cascaded bend waveguides with different bending radii, (b) 6 cascaded adiabatic full-couplers with different design parameters, and (c) band-pass filters that are formed using these bend waveguides and adiabatic full-couplers.
Table 1. Design parameters of band-pass (BP) filters.
Transmission responses of the band-pass filters were simulated by using 3D finite difference time domain method (FDTD, Lumerical Inc.), while for the mode calculations and AWG transmission simulations beam propagation method (BPM) (BeamProp, RSOFT Inc.) was used. Transverse electric (TE) polarization was used in all simulations. The optimum values of d, L, and Δw were obtained by scanning one of these parameters over the entire spectral range while keeping other two parameters constant. As an example, the initial simulation parameters of Coupler 5 (C5 in Table 1) was chosen as d=1 µm, L=500 µm, Δw=0.4 µm, and the optimum values were obtained as d=1.1 µm, L=400 µm, and Δw=0.4 µm after BPM simulations as shown in Fig. 5. Plan view of squared electric-field profiles, E2, are given in Fig. 6 for bend waveguides (a-e) and full-couplers (f-j).
Fig. 5. BPM simulation results of Coupler 5, C5, showing the wavelength dependency of (a) d, (b) L, and (c) Δw = w1-w2. The design parameters of C5 are indicated with white dashed lines.
Fig. 6. Plan view of normalized E2 profiles of bend waveguides (a-e), and full couplers (f-j) for two wavelengths on both sides of the cut-off wavelength.
The slope of these band-pass filters can get steeper by cascading several of these bend waveguides and full-couplers. As an example, the transmission spectra of a 100-µm-radius bend waveguide for n = 1, n=6, and n=12, where n is the number of cascaded units, are given in Fig. 7(a). In a similar manner, the transmission response of Coupler 5 for n = 1, n=6, and n=12 are shown in Fig. 7(b). As expected, a significant improvement on steepness of the band-pass slope is observed in both cases. The final transmission response of the cascaded bends and full-couplers can be formulated as Tf = (Ts)n, where Tf is the transmission response of the cascaded system and Ts is the transmission response of a single component.
Fig. 7. Effect of cascading several (a) bend waveguides and (b) adiabatic full-couplers on transmission response curve.
2.3 Spectrometer layout and AWG design
The overall spectrometer layout is given in Fig. 8. The on-chip part is comprised of band-pass filters and AWGs. The overall design size of this part is 3×3 cm2, which includes 0.6×2.8 cm2 of each AWG size, and 0.25×2 mm2 of the largest band-pass filter size. A more compact coupler design can be used for further size reduction [18]. Five AWG designs, i.e. AWG1, AWG2, AWG3, AWG4, and AWG5, were used to finely separate the bandpass-filtered spectrum. An improved AWG layout based on identical bends across the entire array was used [19]. This layout reduces the systematic phase errors of the conventional AWG designs. At the entrance of each AWG, a band-pass filter was placed. The input light was transferred to these filters via single mode optical waveguides.
Fig. 8. Schematic of the ultrawide-bandwidth spectrometer using on-chip band-pass filters and AWGs. Outputs of AWGs are imaged onto Si and InGaAs sensors via focusing lenses. The rectangular area indicated by black dashed lines shows the designed Si3N4 chip.
The spectral resolution of the AWGs were increased without increasing the device size by removing the output channels of the AWGs [20,21]. The output light was imaged onto two line array detectors (Si for AWG1, AWG2, and AWG3 and InGaAs for AWG4 and AWG5) 1024 pixels each, via focusing lenses. In this way, the number of sampling points can be increased from N (number of output waveguides) to M (number of detector pixels). Each pixel will be wavelength-separated by B/M where B is the 3-dB bandwidth of the AWG spectrum. The overall spectral resolution of this configuration can be defined by the combination of AW-limited resolution and the detector pixel size if the diffraction caused by the imaging lens is neglected. The convolution of these two factors yields the overall spectral resolution of the imaging system. The AW-limited resolution of an AWG is defined as ΔλAW = λ/Mm, where m is the grating order, M is the number of arrayed waveguides, and λ is the central wavelength [20]. Using this formula, the AW-limited resolution was calculated as 1.66 nm, 1.55 nm, 2 nm, 1.92 nm, and 1.8 nm for AWG1, AWG2, AWG3, AWG4, and AWG5, respectively (see Table 2). The overall spectral resolution i.e., the convolution of the diffraction pattern of the AW for a single frequency and the finite size of the detector pixel can be calculated by using Eq. (4) given in Ref. [21].
Table 2. Design parameters of the AWGs.
The design parameters (i.e. central wavelength (λc), free spectral range (FSR), spectral resolution (Δλ), path length increment between adjacent arrayed waveguides (ΔL), grating order (m), number of arrayed waveguides (M), number of output channels (N), and the length of the free propagation region (R)) of the AWGs are given in Table 2.
BPM simulation results of five AWGs are given in Fig. 9. The overall loss values of 2.3 dB, 1.3 dB, 1.4 dB, 1.1 dB, and 0.7 dB were obtained for the central waveguides of these AWGs, respectively. The non-adjacent crosstalk values were −25 dB, −23 dB, −25 dB, −22 dB, and −21 dB at the outer channels and −29 dB, −32 dB, −30 dB, −33 dB, and −42 dB at central channels of the AWGs, respectively.
3. Improved spectrometer design with a higher SNR
In the proposed spectrometer design, the input light is divided among five paths, which reduces the signal-to-noise ratio (SNR) by 7 dB. For some applications where the illumination power is limited due to safety regulations, e.g. live tissue imaging, this could be a significant drawback and therefore a single-path spectrometer design is necessary. In this regard, we designed an alternative band-pass filter scheme that can separate the spectral range of interest into five bands without reducing SNR. The schematic of this configuration is given in Fig. 10.
Fig. 10. Schematic of the single-path on-chip spectrometer design. The pre-filter is comprised of a bend waveguide and an adiabatic full-coupler to select the interested spectral band. In the following stage, adiabatic full-couplers with different coupling properties are used to divide this bandwidth range into five bands, which are then sent to AWGs.
The input light is pre-filtered with a broad band-pass filter which is comprised of a bend waveguide and an adiabatic full coupler (R5 = 130 µm and C1 in Table 1) to define the upper and lower wavelength ranges of the spectrometer in a more compact way. In the second stage, full-couplers with different coupling properties are used to create band-pass filters, which are connected to AWGs. This design will increase the SNR value by 7 dB at the expense of bigger design size (∼3×4 cm2) and lower extinction ratio. Moreover, compared to bend waveguides, full couplers’ performance is less tolerant to fabrication problems. If some light remains in the bar port of the full coupler, it will travel to the next stage where it will eventually couple to the next AWG and show up in one of its output ports. This will increase both crosstalk and loss values of the spectrometer and possibly create some ghost images. For medical applications, this design can be more beneficial over five-path spectrometer due to its higher SNR value, whereas for some applications where the input light power is not limited, five-path design can be more beneficial due to its compact size, higher fabrication tolerance, and better extinction ratios.
The expected loss value of a fabricated single-arm spectrometer is ∼3.5 dB including ∼1 dB of input coupling loss when a proper photonics lantern is used, ∼1 dB of output coupling optics loss, and ∼ 1.5 dB of propagation loss in the Si3N4 waveguides.
4. Conclusions
In summary, we designed a compact, low-loss and ultrawide-band spectrometer that covers the wavelength range of 400 nm-1600 nm, continuously. We proposed a novel band-pass filter design by combining two well-known concepts of integrated optics; bend waveguides and adiabatic full-couplers. These band-pass filters were used to separate the incoming wavelength band into five coarse bands while five AWGs were used for the fine wavelength separation. The overall design size is significantly small, considering the ultrawide-bandwidth range of the spectrometer. We also proposed an alternative band-pass filter design that uses only a single path to divide desired spectrum into five bands. This design provides higher SNR values at the expense of larger device size and lower extinction ratios. The proposed spectrometer can be a very useful tool in medical diagnostic applications; specifically in tissue spectral sensing. Moreover, its small size holds promise for implantable devices where real-time monitoring is crucial. We also envision that several other application fields such as food safety, water quality, agriculture, cosmetics, industrial inspection, and biomedical research can benefit from it.
Funding
Vrije Universiteit Amsterdam (Start-up Contribution).
Disclosures
The authors declare no conflicts of interest.
References
1. Nikolai V. Tkachenko, Optical Spectroscopy: Fundamentals and Advanced Applications, 1st ed. (Elsevier, 2006)
2. H. Ahn, H. Song, D.-M. Shin, K. Kim, and J. Choi, “Emerging optical spectroscopy techniques for biomedical applications—A brief review of recent progress,” Appl. Spectrosc. Rev. 53(2-4), 264–278 (2018). [CrossRef]
3. L. L. de Boer, J. W. Spliethoff, H. J. Sterenborg, and T. J. Ruers, “Review: In Vivo Optical Spectral Tissue Sensing—How to Go from Research to Routine Clinical Application?” Lasers Med. Sci. 32(3), 711–719 (2017). [CrossRef]
4. J. W. Spliethoff, L. L. de Boer, M. A. J. Meier, W. Prevoo, J. de Jong, T. M. Bydlon, H. J. C. M. Sterenborg, J. A. Burgers, B. H. W. Hendrik, and T. J. M. Ruers, “Spectral sensing for tissue diagnosis during lung biopsy procedures: The importance of an adequate internal reference and real-time feedback,” Lung Cancer 98, 62–68 (2016). [CrossRef]
5. E. J. M. Baltussen, S. G. Brouwer de Koning, J. Sanders, A. G. J. Aalbers, N. F. M. Kok, G. L. Beets, B. H. W. Hendriks, H. J. C. M. Sterenborg, K. F. D. Kuhlmann, and T. J. M. Ruers, “Tissue diagnosis during colorectal cancer surgery using optical sensing: an in vivo study,” J. Transl. Med. 17(1), 333–343 (2019). [CrossRef]
6. R. A. Crocombe, “Miniature optical spectrometers: The art of the possible, Part IV: New near-infrared technologies and spectrometers,” Spectroscopy 23(6), 26–32 (2008).
7. B. I. Akca, C. R. Doerr, G. Sengo, K. Wörhoff, M. Pollnau, and R. M. de Ridder, “Broad-spectral-range synchronized flat-top arrayed-waveguide grating applied in a 225-channel cascaded spectrometer,” Opt. Express 20(16), 18313–18318 (2012). [CrossRef]
8. A. van Wijk, C. R. Doerr, Z. Ali, M. Karabiyik, and B. I. Akca, “Compact ultrabroad-bandwidth cascaded arrayed waveguide gratings,” Opt. Express 28(10), 14618–14626 (2020). [CrossRef]
9. D. Pohl, M. Escalé, M. Madi, F. Kaufmann, P. Brotzer, A. Sergeyev, B. Guldimann, P. Giaccari, E. Alberti, U. Meier, and R. Grange, “An integrated broadband spectrometer on thin-film lithium niobate,” Nat. Photonics 14(1), 24–29 (2020). [CrossRef]
10. D. H. Geuzebroek, A. van Rees, E. J. Klein, and K. Lawniczuk, “Ultra-wide band (400–1700nm) integrated spectrometer based on arrayed waveguide gratings for spectral tissue sensing,” Proc. IEEE 14th Int. Conf. Group IV Photon, pp. 83–84, 2017.
11. E. J. Stanton, M. J. R. Heck, J. Bovington, A. Spott, and J. E. Bowers, “Multi-octave spectral beam combiner on ultra-broadband photonic integrated circuit platform,” Opt. Express 23(9), 11272–11283 (2015). [CrossRef]
12. B. Liu, Y. Zhang, Y. He, X. Jiang, J. Peng, C. Qiu, and Y. Su, “Silicon photonic bandpass filter based on apodized subwavelength grating with high suppression ratio and short coupling length,” Opt. Express 25(10), 11359–11364 (2017). [CrossRef]
13. P. Sah and B. Krishna Das, “Photonic bandpass filter characteristics of multimode SOI waveguides integrated with submicron gratings,” Appl. Opt. 57(9), 2277–2281 (2018). [CrossRef]
14. S. Samanta, P. K. Dey, P. Banerji, and P. Ganguly, “Development of micro-ring resonator-based optical bandpass filter using SU-8 polymer and optical lithography,” Opt. Mater. 77, 122–126 (2018). [CrossRef]
15. A. Baldycheva, V. A. Tolmachev, T. S. Perova, Y. A. Zharova, E. V. Astrova, and K. Berwick, “Silicon photonic crystal filter with ultrawide passband characteristics,” Opt. Lett. 36(10), 1854–1856 (2011). [CrossRef]
16. S. Z. Muhd-Yassin, S. D. Emami, M. S. Khairul Anuar, M. I. Zulkifli, N. Tamchek, C. J. Leong, and H. A. Abdul-Rashid, “All fiber optical band-pass filter using macro-bending approach,” IEEE 3rd Int. Conf. Photon., pp. 242–245, 2012.
17. R. Adar, C. Henry, R. Kazarinov, R. Kistler, and G. Weber, “Adiabatic 3-dB couplers, filters, and multiplexers made with silica waveguides on silicon,” J. Lightwave Technol. 10(1), 46–50 (1992). [CrossRef]
18. E. S. Magden, N. Li, M. Raval, C. V. Poulton, A. Ruocco, N. Singh, D. Vermeulen, E. P. Ippen, L. A. Kolodziejski, and M. R. Watts, “Transmissive silicon photonic dichroic filters with spectrally selective waveguides,” Nat. Commun. 9(1), 3009 (2018). [CrossRef]
19. N. Ismail, F. Sun, G. Sengo, K. Wörhoff, A. Driessen, R. M. de Ridder, and M. Pollnau, “Improved arrayed-waveguide-grating layout avoiding systematic phase errors,” Opt. Express 19(9), 8781–8794 (2011). [CrossRef]
20. B. I. Akca, L. Chang, G. Sengo, K. Wörhoff, R. M. de Ridder, and M. Pollnau, “Polarization-independent enhanced-resolution arrayed-waveguide grating used in spectral-domain optical low-coherence reflectometry,” IEEE Photonics Technol. Lett. 24(10), 848–850 (2012). [CrossRef]
21. Z. Hu, Y. Pan, and A. M. Rollins, “Analytical model of spectrometer- based two-beam spectral interferometry,” Appl. Opt. 46(35), 8499–8505 (2007). [CrossRef]
