Tailorable and Broadband On-Chip Optical Power Splitter
Round 1
Reviewer 1 Report
In their manuscript, H. Kim et al. propose an improved design of an on-chip optical splitter for the photonic signal processing applications reported earlier in Ref.[15]. By means of the numerical simulations, the authors report a broadband in the telecommunication range power dividing ability of the proposed device with low insertion losses and tailorable splitting ratio.
I conclude that the paper is scientifically sound, and the results may deserve the publication in Applied Science journal providing that minor changes to the manuscript will be properly addressed by the authors.
Comments:
The MS lacks the details of the numerical simulations methodic (FDTD, FEM, ?) as well as the information about spatio-temporal meshes, simulation domain configuration, computing equipment used. In fig.2 it will be more instructive to draw the arrows plots visualizing energy flows in the splitter cross-section (Poynting vector) in addition to the E-field intensity distribution. 4 is low informative; it will be better to present the deviation (in per cents) of achieved splitting ratios from the ideal ones. Splitting tolerance to the nonuniform heating of the proposed device caused by different optical power absorption in the upper and lower waveguides (e.g., 90:10 switching ratio) should be discussed.Author Response
Response to Referee Reports
******************************************************************************************
Referee #1 (COMMENTS TO THE AUTHOR(S)):
In their manuscript, H. Kim et al. propose an improved design of an on-chip optical splitter for the photonic signal processing applications reported earlier in Ref.[15]. By means of the numerical simulations, the authors report a broadband in the telecommunication range power dividing ability of the proposed device with low insertion losses and tailorable splitting ratio.
I conclude that the paper is scientifically sound, and the results may deserve the publication in Applied Sciences journal providing that minor changes to the manuscript will be properly addressed by the authors.
[Author]
Dear Referee #1,
Thank you very much for your time to read our manuscript and for providing us with your thoughts and invaluable feedback. As requested by the editor, we respond to your comments and queries in a point-by-point fashion.
******************************************************************************************
[Referee 1: Query 1] The MS lacks the details of the numerical simulations methodic (FDTD, FEM, ?) as well as the information about spatio-temporal meshes, simulation domain configuration, computing equipment used.
[Author]
We thank you for the suggestion and completely agree. We added simulation details as follow: (page 2, 63-68)
“The simulations are performed by Lumerical’s MODE solution, which offers the 2.5D FDTD method. The mesh accuracy used in the simulation is the predefined value 6 in Lumerical, which is small enough for the narrow tips in the tapered region in Fig. 1. The time step is about 0.04 fs. The boundary conditions are the perfect matched layer condition in the x- and y-directions and the metal boundary condition in the z-direction. This metal boundary condition can reduce the simulation time and does not affect the simulation results as light is tightly confined in the waveguides.”
******************************************************************************************
[Referee 1: Query 2] In fig.2 it will be more instructive to draw the arrows plots visualizing energy flows in the splitter cross-section (Poynting vector) in addition to the E-field intensity distribution.
[Author]
We thank you for the suggestion. We enlarge the Fig. 2 (f)-(h) and draw energy flows and 15 μm scale bar.
******************************************************************************************
[Referee 1: Query 3] Fig. 4 is low informative; it will be better to present the deviation (in per cents) of achieved splitting ratios from the ideal ones.
[Author]
We appreciate this comment. We tried to add the deviation in Figure 4, but it becomes too complicated in one figure. Therefore, we added the explanation of power splitting deviation in range 1.5-1.6 μm and 1.2-1.8 μm in the manuscript as follows: (page 6, 184-191)
“The coupling ratio deviation is defined by (abs(P_(down,target)-P_down )/P_(down,target) ×100)%, and the maximum deviations in wavelength range between 1.5 μm and 1.6 μm (1.2 μm – 1.8 μm) are 0.49% (2.54%) for 50:50, 2.85% (12.30%) for 30:70, 0.27% (7.11%) for 10:90, and 0.86% (1.20%) for 1:99. Due to the slowly varying material refractive index in wavelength, the evanescent profile, as well as the splitting ratio, change slowly in wavelength. However, even in 1.2 μm – 1.8 μm range, the maximum deviation is about 10%, showing ultra-broadband tailorability compared to recent methods of broadband power splitters [14, 19].”
******************************************************************************************
[Referee 1: Query 4] Splitting tolerance to the nonuniform heating of the proposed device caused by different optical power absorption in the upper and lower waveguides (e.g., 90:10 switching ratio) should be discussed.
[Author]
We appreciate this comment. We believe that our proposed design will work in a linear response working regime. Any non-uniform heating is not considered in this paper. We think that it would be an interesting story for future work. Therefore, we added a comment about this issue in the discussion section as follow: (page 7, 219-221)
“In addition, we consider the linear optical properties of the proposed design by excluding the nonlinear absorption, power-dependent refractive index, and thermal heating. Further research about nonlinear optical behavior of this device is needed.”
Reviewer 2 Report
In this manuscript, the authors report a new design of on-chip waveguide-based 1×2 power splitter with a flexible control on the splitting ratio. The concept of interleaved tapered waveguides is simple, yet efficient and technically sound. The authors carried out comprehensive numerical simulation studies to investigate the properties of splitters.
However, I believe further technical details and analyses should be included to make this work stand out and easy for readers to appreciate. Before I can recommend this paper for acceptance, there are several points which require revisions.
On page 3, the top-view mode propagation profiles in Figure 2f-h are too small to look into the technical details. Zoomed-in views at the coupling regions (i.e., the waveguide tapering regions with different spacing) should be presented. The authors claim that “the coupling length should be at least 15 μm”. Notating axis or scale bar will be very helpful to indicate the adopted coupling length and also the varying spacing. For Figure 4 on Page 5, the small variation of splitting ratio over the 100 nm wavelength range is impressive. Are they all random fluctuations or following any kind of distributions (e.g., any periodicity)? In some other designs of splitter (e.g., doi.org/10.1364/OL.41.002053), wavelength dependence has been examined over a very wide range (e.g., 1200 – 1700 nm). Therefore, I wonder the exact operation wavelength window that the current designs for each splitting ratio can be extended to. The authors should at least comment on this. The authors should discuss more on the simulation details, such as the method, the software, the meshing size, the boundary condition, etc. Nowadays, silicon nitride (SiN) on silica material platform with ultra-low propagation loss emerges as a highly promising candidate for future photonic integrated circuits especially 3D photonic integration (1109/JPROC.2018.2861576), quantum photonics (10.1109/JSTQE.2018.2846047) and biophotonics. I wonder whether this design can be applied on SiN-based waveguides with a refractive index contrast of ~ 2:1.46. If so, what is the corresponding minimum feature size at the tapering tips? More quantitative simulation results will be very helpful for readers (but I also understand this requires some additional work). Otherwise, some rough estimations and comments on this will also be very helpful. Only TE fundamental mode has been considered in this study. Is there any intrinsic limitation for applying the design to TM mode? As silicon photonics is really a rapidly advancing field, I suggest the authors update the reference and include more literature reported in the recent 5 years for background information and reviewing other developed configurations of power splitters (e.g., 10.1364/OL.41.003041).Author Response
Response to Referee Reports
******************************************************************************************
Referee #2 (COMMENTS TO THE AUTHOR(S)):
In this manuscript, the authors report a new design of on-chip waveguide-based 1×2 power splitter with a flexible control on the splitting ratio. The concept of interleaved tapered waveguides is simple, yet efficient and technically sound. The authors carried out comprehensive numerical simulation studies to investigate the properties of splitters.
However, I believe further technical details and analyses should be included to make this work stand out and easy for readers to appreciate. Before I can recommend this paper for acceptance, there are several points that require revisions.
[Author]
Dear Referee #2,
Thank you very much for your time and effort to read our manuscript and for providing us with your thoughts and invaluable feedback. As requested by the editor, we respond to your comments and queries in a point-by-point fashion
******************************************************************************************
[Referee 2: Query 1] On page 3, the top-view mode propagation profiles in Figure 2f-h are too small to look into the technical details. Zoomed-in views at the coupling regions (i.e., the waveguide tapering regions with different spacing) should be presented. The authors claim that “the coupling length should be at least 15 μm”. Notating axis or scale bar will be very helpful to indicate the adopted coupling length and also the varying spacing.
[Author]
We thank you for the suggestion. We enlarge the Fig. 2 (f)-(h) and draw energy flows and 15 μm scale bar.
******************************************************************************************
[Referee 2: Query 2] For Figure 4 on Page 5, the small variation of splitting ratio over the 100 nm wavelength range is impressive. Are they all random fluctuations or following any kind of distributions (e.g., any periodicity)? In some other designs of the splitter (e.g., doi.org/10.1364/OL.41.002053), wavelength dependence has been examined over a very wide range (e.g., 1200 – 1700 nm). Therefore, I wonder the exact operation wavelength window that the current designs for each splitting ratio can be extended to. The authors should at least comment on this.
[Author]
We thank you for the suggestion. We updated power splitting deviation in the wavelength range 1.5 - 1.6 μm and 1.2 - 1.8 μm as follow: (page 6, 185-191)
“The coupling ratio deviation is defined by (abs(P_(down,target)-P_down )/P_(down,target) ×100)%, and the maximum deviations in wavelength range between 1.5 μm and 1.6 μm (1.2 μm – 1.8 μm) are 0.49% (2.54%) for 50:50, 2.85% (12.30%) for 30:70, 0.27% (7.11%) for 10:90, and 0.86% (1.20%) for 1:99. Due to the slowly varying material refractive index in wavelength, the evanescent profile, as well as the splitting ratio, change slowly in wavelength. However, even in 1.2 μm – 1.8 μm range, the maximum deviation is about 10%, showing ultra-broadband tailorability compared to recent methods of broadband power splitters [14, 19].”
******************************************************************************************
[Referee 2: Query 3] The authors should discuss more on the simulation details, such as the method, the software, the meshing size, the boundary condition, etc.
[Author]
We thank you for the suggestion and completely agree. We added simulation details as follow: (page 2, 63-68)
“The simulations are performed by Lumerical’s MODE solution, which offers the 2.5D FDTD method. The mesh accuracy used in the simulation is the predefined value 6 in Lumerical, which is small enough for the narrow tips in the tapered region in Fig. 1. The time step is about 0.04 fs. The boundary conditions are the perfect matched layer condition in the x- and y-directions and the metal boundary condition in the z-direction. This metal boundary condition can reduce the simulation time and does not affect the simulation results as light is tightly confined in the waveguides.”
******************************************************************************************
[Referee 2: Query 4] I wonder whether this design can be applied on SiN-based waveguides with a refractive index contrast of ~ 2:1.46. If so, what is the corresponding minimum feature size at the tapering tips? More quantitative simulation results will be very helpful for readers (but I also understand this requires some additional work). Otherwise, some rough estimations and comments on this will also be very helpful.
[Author]
We appreciate this comment. We added the case for low refractive index contrast as follows: (page 7, 222-226)
“Recently, other materials such as silicon nitride, aluminum nitride, and lithium niobate are intensively investigated in the fields of photonic integrated circuits. Our suggested structure can be employed not only in silicon but also in other platforms. Due to the different refractive index, mode size, and evanescent profile on other platforms, the parameters in equation 1 differ from those in silicon than silicon, and further research is required.”
******************************************************************************************
[Referee 2: Query 5] Only TE fundamental mode has been considered in this study. Is there any intrinsic limitation for applying the design to TM mode?
[Author]
We appreciate this comment. Some comments for TM mode is added as follows: (page 7, 212-219)
“In this simulation, only the TE mode is considered. In the TM mode, the evanescent electric field exists at the top and bottom of the waveguides, and the mode profile in the coupling region is complicated and very sensitive to wavelength and geometry dimension variance. In relevant simulations, we observe that the relationship between the gap size and power splitting ratio in the TM mode is different from equation 1 and is not as simple as that in the TE mode case. We could not find a meaningful and straightforward expression of the relationship as the ratio is too sensitive to the gap size.”
******************************************************************************************
[Referee 2: Query 6] As silicon photonics is really a rapidly advancing field, I suggest the authors update the reference and include more literature reported in the recent 5 years for background information and reviewing other developed configurations of power splitters (e.g., 10.1364/OL.41.003041).
[Author]
We appreciate this comment. We added recent research results and references as follows: (page 1, 34-41 and references)
“Recently, curved directional coupler and its combination with straight directional couplers show small footprints, low loss, and ultra-broadband power splitting, but they need careful design in coupling regions [14, 15].”
“Since the subwavelength grating can engineer refractive index and dispersion properties, a compact, low loss, and ultra-broadband directional couplers are proposed using its design flexibility [18 - 20], but the gratings are intrinsically sensitive to fabrication errors.”
“14. Chen, G.F.R.; Ong, J.R.; Ang, T.Y.L.; Lim, S.T.; Png, C.E.; Tan, D.T.H. Broadband Silicon-On-Insulator directional couplers using a combination of straight and curved waveguide sections. Sci. Rep. 2017, 7, 4–11.
Chen, S.; Shi, Y.; He, S.; Dai, D. Low-loss and broadband 2 × 2 silicon thermo-optic Mach-Zehnder switch with bent directional couplers. Opt. Lett. 2016, 41, 836–9. Yun, H.; Wang, Y.; Zhang, F.; Lu, Z.; Lin, S.; Chrostowski, L.; Jaeger, N.A.F. Broadband 2 × 2 adiabatic 3 dB coupler using silicon-on-insulator sub-wavelength grating waveguides. Opt. Lett. 2016, 41, 3041. Yun, H.; Chrostowski, L.; Jaeger, N.A.F. Ultra-broadband 2 × 2 adiabatic 3 dB coupler using subwavelength-grating-assisted silicon-on-insulator strip waveguides. Opt. Lett. 2018, 43, 1935. Xu, L.; Wang, Y.; Kumar, A.; El-Fiky, E.; Mao, D.; Tamazin, H.; Jacques, M.; Xing, Z.; Saber, M.G.; Plant, D. V. Compact high-performance adiabatic 3-dB coupler enabled by subwavelength grating slot in the silicon-on-insulator platform. Opt. Express 2018, 26, 29873.”
