Silicon-based wavelength division multiplexer using asymmetric grating-assisted couplers

Lin Zhu; Junqiang Sun; Yong Zhou

doi:10.1364/OE.27.023234

1. Introduction

In recent years, silicon-based optical interconnect technology has developed rapidly with its unique advantages. On-chip wavelength division multiplexing [1–7], mode multiplexing [8–10], wavelength and mode hybrid multiplexing [11–13] and polarization multiplexing [14] have become research hotspots in the field of optical communications. As one of the key components of on-chip multiplexing system, WDM plays an important role in improving the communication capacity.

Previously, we proposed a WDM based on asymmetric directional couplers [15]. Although the performance of the proposed device has been verified theoretically and experimentally, there are still some limitations. Firstly, in asymmetric directional couplers, the increasing coupling length will lead to the continuous conversion of optical signals between the input waveguide and the bus waveguide, so the coupling length needs to be strictly controlled. Secondly, since the asymmetric directional coupler is not sufficiently sensitive to wavelength, the designed WDM is more suitable for the coarse wavelength division multiplexing systems. To overcome these limitations, we propose a WDM based on asymmetric grating-assisted couplers. Comparison to the asymmetric directional couplers, asymmetric grating-assisted couplers, coupling the light propagating in the opposite direction to the input light, can be called contra-directional couplers. When the optical signal is coupled from the input waveguide to the bus waveguide, the coupling length does not need to be strictly controlled in asymmetric grating-assisted couplers [16,17]. Moreover, the bandwidth of each wavelength channel can be designed flexibly because adjusting the corrugation width of the grating will change the coupling coefficient of different wavelengths.

In the paper, the wavelength signals multiplexed in the bus waveguide are converted into orthogonal modes based on asymmetric grating-assisted couplers. The signals of $λ_{0}, λ_{1}, λ_{2},$ and $λ_{3}$ are injected from input waveguides in the form of fundamental modes and respectively converted into different high-order modes TE₀, TE₁, TE₂, and TE₃ of the bus waveguide after passing through grating-assisted couplers. Like the asymmetric directional couplers (ADCs)-based WDM, multiplexed optical signals can travel steadily in the bus multimode waveguide owing to the orthogonal property of waveguide modes. Simulation results show that the WDM based on asymmetric uniform grating-assisted couplers has large sidelobes near the central wavelength, which leads to large crosstalk between wavelength channels and affects the performance of the WDM. In order to solve this problem, we use asymmetric unilateral amplitude apodization grating-assisted couplers and asymmetric bilateral amplitude apodization grating-assisted couplers to suppress the sidelobes. The inverse apodization in the bilateral asymmetric grating-assisted coupler concepts that are implemented for the first time to the best of our knowledge. According to the simulation results, WDMs based on different grating structures are fabricated and tested. The experimental results verify the feasibility of the proposed WDMs.

2. Principle and device design

We assume that the cladding material is silicon dioxide. Taking four channels as an example, WDM based on asymmetric uniform grating-assisted couplers is shown in Fig. 1(a), which is composed of input single-mode waveguides, asymmetric grating-assisted couplers with different periods, and tapered connectors. Figure 1(b) is the top view of the asymmetric grating-assisted coupler, where the waveguide thicknesses are 220 nm, the gap widths are 200 nm, the width of each input waveguide is $W$ , the width of each multimode waveguide is $W_{i} (i = 1, 2, 3)$ , the corrugation width of each grating is $D$ , the grating period is $Λ_{i}$ , and the coupling length is $L_{i}$ . The signals of $λ_{1}, λ_{2}, λ_{3}$ are injected from input waveguides in the form of fundamental modes and respectively converted into different high-order modes TE₁, TE₂, TE₃ of the bus waveguide after passing through grating-assisted couplers. In order to simplify the design process, the signal $λ_{0}$ with the wavelength of 1550 nm is input from the single-mode waveguide and directly transmitted in the form of the bus waveguide fundamental mode after passing through the adiabatic taper waveguide, thereby realizing the purpose of multiplexing different wavelengths in different modes in the bus waveguide.

Fig. 1 (a) Schematic structure of WDM based on asymmetric uniform grating-assisted couplers. (b) Top view of the asymmetric uniform grating-assisted coupler.

Download Full Size | PDF

Assuming that the operating wavelength of each channel is $λ_{i}$ , the period $Λ_{i}$ of the grating in each coupler is determined according to the phase matching principle [18]:

n_{e f f_0} + n_{e f f_i} = λ_{i} / Λ_{i} (i = 1, 2, 3...),

where

n_{e f f_0}

is the effective refractive index corresponding to the fundamental mode of the wavelength

λ_{i}

, and

n_{e f f_i}

is the effective refractive index corresponding to the higher-order mode of the wavelength

λ_{i}

. In the simulation process, we use the dispersive models for the refractive indices, which are provided by RSoft’s Material Editor. The relationship between the effective refractive index of different modes and the grating periods is shown in Fig. 2. The solid line and the dotted line are the left and right terms of Eq. (1), respectively, and the intersection points indicate that the phase matching condition are satisfied. The operating wavelength of each channel is

λ_{1} = 1545 n m, λ_{2} =1540 n m,

and

λ_{3} =1535 n m,

according to Eq. (1), the corresponding grating period is

Λ_{1} = 332 n m, Λ_{2} = 329 n m,

and

Λ_{3} = 328 n m,

respectively. The wavelength channels are chosen to be at 1550 nm, 1545 nm, 1540 nm and 1535 nm to meet the operating wavelength requirements of the device in our project. If we want to match special applications and standards, we can achieve such kind of wavelength multiplexers by changing the relevant parameters in our proposed approach. In principle, multiplexing any wavelengths can be implemented, but the device size, fabrication accuracy and manufacture technology should be taken into consideration.

Fig. 2 The relationship between the effective refractive index of different modes and the phase matching condition.

Download Full Size | PDF

Since the corrugation width of the grating influences the coupling coefficient, we next analyze the relationship between the coupling coefficient and the corrugation width $D$ , as shown in Fig. 3(a), where the corrugation depth is 220 nm. We find that when the corrugation width $D$ increases from 20 nm to 280 nm, the coupling coefficient is gradually increased because the overlap integral between the changing region of the refractive index and the distribution of the mode field in the coupled system increases, thus the appropriate corrugation width $D$ should be selected in the design.

Fig. 3 (a) The relationship between the coupling coefficient and the corrugation width. (b) The relationship between coupling length and power.

Download Full Size | PDF

It can be seen from the above analysis that when the signals of different wavelengths are coupled from the input waveguide in the form of fundamental modes to the higher-order modes of the bus waveguide, their coupling coefficients are different, that is to say, the coupling lengths of different wavelengths in the mode conversion are different. Figure 3(b) shows the variation of power with the coupling length, where the corrugation width $D$ is 100 nm. As the coupling length increases, the energy coupling from the input waveguide to the bus waveguide becomes more complete, which is a monotonically increasing relationship. Therefore, if we ensure that the coupling length is long enough, the coupling coefficient will not be reduced theoretically, thereby eliminating the effects of process errors. This advantage is also one of reasons why we design the WDM using grating-assisted couplers. From Fig. 3(b), the coupling lengths of $λ_{1}, λ_{2},$ and $λ_{3}$ channels are chosen as 250 μm, 300 μm, and 300 μm, respectively.

Firstly, RSoft GratingMod is used to calculate the spectrum of each wavelength channel reversely coupled to the bus waveguide. We choose the width of the input waveguide is $W = 450 n m,$ the width of the multimode waveguide is $W_{1} = 0.82 μ m, W_{2} = 1.27 μ m,$ and $W_{3} = 1.7 μ m,$ the corrugation width of the grating is $D = 100 n m$ . It should be noted that Bragg reflection exists in the input waveguide itself for the asymmetric grating-assisted coupler, which may interfere with other signals in the WDM system. Some solutions have been reported so far [19]. During the simulation, we find that the Bragg reflection peak of the input waveguide itself exceeds the C-band in the structure composed of the above parameters. Since we only consider the transmission of optical signals in the C-band, we will not discuss the influence of the Bragg reflection on the WDM.

The signals of wavelengths $λ_{1}, λ_{2}, λ_{3}$ are multiplexed and output from the bus multimode waveguide, as shown in Fig. 4. As can be seen from the figure, the insertion loss and 3dB bandwidth of the $λ_{1}, λ_{2}, λ_{3}$ channels are 0.27 dB and 5.7 nm, 0.16 dB and 4.6 nm, 0.15 dB and 4.1 nm, respectively. As can be seen from Fig. 4(b), since the corrugation width of the grating affects the coupling coefficient, this further changes the 3dB bandwidth. The 3dB bandwidth of each channel can be flexibly designed if the corrugation width is adjusted as needed.

Fig. 4 (a) The spectral responses of the $λ_{1}, λ_{2},$ and $λ_{3}$ channels. (b) The relationship between the 3dB bandwidth and the corrugation width.

Download Full Size | PDF

Based on the simulation results of RSoft GratingMod, the 3D FDTD simulation of the WDM based on asymmetric grating-assisted couplers using the FDTD Solutions software of Lumerical is further performed. In the simulation process, we use the Sampled data model to create a lossless material. We assume that the cladding material is silicon dioxide. Here, the waveguide thicknesses are 220 nm, the width of the input waveguide is $W = 450 n m,$ the corrugation width is $D = 100 n m$ , the spacing between the input waveguide and the multi-mode waveguide is $g a p = 300 n m,$ the number of grating period is $N,$ and other parameters used in the simulation are listed in Table 1. The simulated light propagation and spectral responses of the WDM based on asymmetric uniform grating-assisted couplers are shown in Fig. 5 and Fig. 6, respectively. Since the computer memory required for simulation increases dramatically as the coupling length increases, the coupling length chosen here cannot fully couple the energy from the input waveguide to the bus waveguide, but it can still be used for analysis.

Table 1. Simulation parameters of WDM based on asymmetric uniform grating-assisted couplers.

View Table | View all tables in this article

Fig. 5 Simulated light propagation of the $λ_{1}, λ_{2},$ and $λ_{3}$ channels.

Download Full Size | PDF

Fig. 6 Simulated spectral responses of WDM based on asymmetric uniform grating-assisted couplers.

Download Full Size | PDF

Figures 5(a)-5(c) show the simulated light propagation of the $λ_{1},$ $λ_{2},$ and $λ_{3}$ channels, respectively. It can be seen from Fig. 5 that signals of wavelengths $λ_{1},$ $λ_{2},$ and $λ_{3}$ are injected in the form of fundamental modes and converted into high-order modes TE₁, TE₂, and TE₃ of the bus multimode waveguide after passing through asymmetric uniform grating-assisted couplers, which is the same as the theoretical design. Figure 6 is basically consistent with the simulation results of Rsoft software. Although the WDM based on asymmetric uniform grating-assisted couplers can achieve multiplexing of different wavelengths, there are intra-waveguide Bragg reflections in grating-based directional couplers, which will cause interference to adjacent channels and the crosstalk between adjacent wavelength channels is large due to the presence of sidelobes, which affects the performance of the WDM systems. Therefore, how to suppress or eliminate sidelobes is the next problem we need to solve.

3. Improved design and simulation

Sidelobe is the inherent characteristic of the uniform grating [20–22]. Due to the presence of sidelobes, the crosstalk between different wavelength channels increases, which interferes with the signal transmission. Therefore, methods must be taken to suppress or eliminate sidelobes. In general, we can suppress sidelobes by apodization technique [23–25]. At present, some methods to inhibit sidelobes have been reported [26–31]. The key to suppress sidelobes is to adjust the coupling coefficient [32–35]. We assume that the cladding material is silicon dioxide. In this section, we propose two schemes to suppress sidelobes of the uniform grating. Considering that the 3D FDTD algorithm has high requirements on computer performance, we first use the Var FDTD algorithm to make a preliminary analysis. Then, 3D FDTD algorithm is used to design the structural parameters of the WDM that can effectively suppress sidelobes. In the simulation process, we use the Sampled data model to create a lossless material.

3.1 WDM based on asymmetric unilateral amplitude apodization grating-assisted couplers

The first improved WDM is based on asymmetric unilateral amplitude apodization grating-assisted couplers, as shown in Fig. 7(a). Figure 7(b) is the top view of the asymmetric unilateral amplitude apodization grating-assisted coupler, which shows the parameters used in the design. The waveguide thicknesses are 220 nm, the gap widths are 200 nm, the width of the input single-mode waveguide is $W,$ the width of each multi-mode waveguide is $W_{G},$ the corrugation width of each grating is $D,$ the grating period is $Λ_{i},$ and the coupling length of each channel is $L_{i}$ . The corrugation width $D$ varies with the grating propagation direction according to Eq. (2), the upper boundary of the multimode waveguide is horizontal, and the lower boundary changes with the grating propagation direction according to Eq. (3). The variation of the multimode waveguide is opposite to that of the corrugation width, thereby realizing the control of the refractive index and the coupling coefficient.

D = H e^{- a \frac{{(z - L_{i} / 2)}^{2}}{L_{i}^{2}}},

W_{G} = W_{i} + b (1 - e^{- a \frac{{(z - L_{i} / 2)}^{2}}{L_{i}^{2}}}),

where a is the apodization coefficient of the corrugation width, b presents the apodization coefficient of the multimode waveguide width, and

H

is the maximum corrugation width. The variation of the corrugation width

D

and the multimode waveguide width

W_{G}

along the grating propagation direction when a and b take different values are calculated as shown in Fig. 8, and in Fig. 8(b), the value of a is 5. The changes of the corrugation width and the multimode waveguide width are

(0, H]

and

[W_{i}, W_{G}]

over the entire length of the grating.

Fig. 7 (a) Schematic structure of WDM based on asymmetric unilateral amplitude apodization grating-assisted couplers. (b) Top view of the asymmetric unilateral amplitude apodization grating-assisted coupler.

Download Full Size | PDF

Fig. 8 (a) The variation of $D$ along the propagation direction when a takes different values. (b) The variation of $W_{G}$ along the propagation direction when b takes different values.

Download Full Size | PDF

In order to verify the performance of the asymmetric unilateral amplitude apodization grating-assisted coupler to suppress sidelobes, the $λ_{1}$ channel is taken as an example. We set the apodization coefficient $a = 5$ to highlight the advantage of the Gaussian functions in suppressing side lobes and reduce the difficulty of production, and the maximum corrugation width $H$ is 100 nm to compare the apodization grating with the uniform grating. The period number $N$ is selected to be 600, which is the tradeoff result between the simulation time and the coupling efficiency. Figure 9 is the reflection spectra of the $λ_{1}$ channel when b takes different values. From the simulation results, the reflection spectrum in Fig. 9(a) still has sidelobes at short wavelengths, but the sidelobes have been suppressed compared with that in Fig. 6. This indicates that the width variation of the multimode waveguide introduced at this time compensates the refractive index at both ends of the coupler, but it is still in the state of under-compensation, so the sidelobes are not eliminated. As the apodization coefficient b increases, the sidelobes on both sides of the reflection spectrum in Fig. 9(b) are almost suppressed below −20dB, indicating that the width variation of the multimode waveguide introduced by the apodization coefficient b can compensate the refractive index at both ends of the coupler so that the resonant wavelength falls exactly at the central wavelength. When the apodization coefficient b continues to increase, as shown in Fig. 9(c) and 9(d), the reflection spectra gradually start to have sidelobes at long wavelengths, indicating that the apodization coefficient b is too large at this time, so that the refractive index at both ends of the coupler is overcompensated. In summary, for the given apodization coefficient a, the multimode waveguide has an optimal apodization coefficient b, which can control the refractive index while adjusting the coupling coefficient. Therefore, the sidelobes can be effectively suppressed and the sidelobe suppression ratio can be improved.

Fig. 9 Reflection spectra of the asymmetric unilateral amplitude apodization grating-assisted coupler in the $λ_{1}$ channel when b takes different values.

Download Full Size | PDF

Next, we design the WDM based on asymmetric unilateral amplitude apodization grating-assisted couplers. Parameters used in simulation are listed in Table 2,where the values of the waveguide width $W_{i}$ and the grating period $Λ_{i}$ are calculated and optimized according to the phase matching condition, the apodization coefficient b is the optimization parameter, the number of grating periods $N$ is the result of balancing the simulation time and the coupling efficiency, and the coupling length $L_{i}$ is the product of the number of grating periods $N$ and the grating period $Λ_{i} .$ The spectral responses are shown in Fig. 10. From Fig. 10, we can see that the sidelobes of each channel are suppressed below −20dB, that is, the crosstalk between the wavelength channels of the WDM is significantly reduced, and the performance of the device is effectively improved.

Table 2. Simulation parameters of WDM based on asymmetric unilateral amplitude apodization grating-assisted couplers.

View Table | View all tables in this article

Fig. 10 Simulated spectral responses of WDM based on asymmetric unilateral amplitude apodization grating-assisted couplers.

Download Full Size | PDF

3.2 WDM based on asymmetric bilateral amplitude apodization grating-assisted couplers

The second improved WDM is based on asymmetric bilateral amplitude apodization grating-assisted couplers, as shown in Fig. 11(a). Figure 11(b) is the top view of the asymmetric bilateral amplitude apodization grating-assisted coupler, which shows the parameters used in the design. The waveguide thicknesses are 220 nm, the gap widths are 200 nm, the width of the input single-mode waveguide is $W,$ the width of each multi-mode waveguide is $W_{i},$ the grating period is $Λ_{i},$ and the coupling length of each wavelength channel is $L_{i},$ a presents the apodization coefficient of the corrugation width, $H_{1}$ and $H_{2}$ are the maximum corrugation width of the inner and outer grating, respectively. The corrugation widt $D_{1}$ of the inner grating and the corrugation width $D_{2}$ of the outer grating vary with the propagation direction according to Eqs. (4) and (5), respectively. That is to say, the variation of the corrugation width on both sides of the multimode waveguide is opposite, so that coupling coefficient can be adjusted.

Fig. 11 (a) Schematic structure of WDM based on asymmetric bilateral amplitude apodization grating-assisted couplers. (b) Top view of the asymmetric bilateral amplitude apodization grating-assisted coupler.

Download Full Size | PDF

D_{1} = H_{1} e^{- a \frac{{(z - L_{i} / 2)}^{2}}{L_{i}^{2}}},

D_{2} = H_{2} (1 - e^{- a \frac{{(z - L_{i} / 2)}^{2}}{L_{i}^{2}}}),

According to the Eqs. (4) and (5), the variation of $D_{1}$ and $D_{2}$ along the grating propagation direction when a and $H_{2}$ take different values are calculated, as shown in Fig. 12. The changes of the inner and outer corrugation widths are $(0, H_{1}]$ and $[0, H_{2})$ respectively, over the entire length of the grating.

Fig. 12 (a) The variation of $D_{1}$ along the propagation direction when a takes different values. (b) The variation of $D_{2}$ along the propagation direction when $H_{2}$ takes different values.

Download Full Size | PDF

In order to verify the performance of the asymmetric bilateral amplitude apodization grating-assisted coupler to suppress sidelobes, the $λ_{1}$ channel is taken as an example. Set the apodization coefficient $a = 5,$ the maximum corrugation width $H 1$ of the inner grating is 100 nm, and the period number $N$ is 600. Figure 13 is the reflection spectra of the $λ_{1}$ channel when $H_{2}$ takes different values. From the simulation results, the reflection spectra of Fig. 13(a) and 13(b) still have sidelobes at short wavelengths, and the sidelobes in Fig. 13 (b) are smaller than those in Fig. 13(a). This shows that the variation of the corrugation width $H_{2}$ plays a role in compensating the refractive index at short wavelengths, but is still in the state of under-compensation, so the sidelobes are not eliminated. As $H_{2}$ continues to increase, the sidelobes of the reflection spectrum in Fig. 13(c) are suppressed below −20dB, which indicates that the change in the corrugation width of the outer grating introduced at this time can control the refractive index at both ends of the coupler, so that its resonant wavelength falls exactly at the central wavelength. If $H_{2}$ is further increased, the sidelobes of the reflection spectrum in Fig. 13(d) begin to have at long wavelengths, indicating that the corrugation width of the outer grating changes too much, resulting in the over-compensation of the refractive index at both ends of the coupler. In summary, for the given apodization coefficient a and the maximum corrugation width $H_{1}$ of the inner grating, the maximum corrugation width $H_{2}$ of the outer grating has an optimum value within a certain range, so that the refractive index can be controlled while adjusting the coupling coefficient. Therefore, the sidelobes can be effectively suppressed and the sidelobe suppression ratio of each channel can be improved.

Fig. 13 Reflection spectra of the asymmetric bilateral amplitude apodization grating-assisted coupler in the $λ_{1}$ channel when $H_{2}$ takes different values.

Download Full Size | PDF

Next, we design the WDM based on asymmetric bilateral amplitude apodization grating-assisted couplers. The maximum corrugation widths on both sides of the multimode waveguide are $H_{1} = 100 n m$ and $H_{2} = 60 n m$ , respectively. Parameters used in simulation are listed in Table 3, and the spectral responses are shown in Fig. 14. From the Fig. 14, we can see that the sidelobes of each channel are suppressed below −23dB. Therefore, the crosstalk between the wavelength channels of the WDM is significantly reduced, and the performance of the device is further improved.

Table 3. Simulation parameters of WDM based on asymmetric bilateral amplitude apodization grating-assisted couplers.

View Table | View all tables in this article

Fig. 14 Simulated spectral responses of WDM based on asymmetric bilateral amplitude apodization grating-assisted couplers.

Download Full Size | PDF

4. Device fabrication and experimental results

We fabricate the WDMs on a silicon-on-insulator (SOI) substrate with a top silicon layer of 220 nm and a buried oxide layer of 2 μm. The fabrication process of waveguides mainly includes the electron beam lithography (EBL) and the inductively coupled plasma dry-etching (ICP). The etching depths of the waveguides and the grating couplers are 220 nm and 70 nm, respectively. Finally, the SiO₂ layer with 1 μm thick is deposited on the devices by the plasma enhanced chemical vapor deposition (PECVD). Scanning electron microscope (SEM) pictures of the WDMs are shown in Fig. 15. The spectral responses of the WDMs are obtained after subtracting the transmission loss of the reference straight waveguide, as shown in Fig. 16. It should be noted that since the signal of the $λ_{0}$ channel is not pass through the asymmetric grating-assisted coupler but is directly transmitted in the form of the fundamental mode of the bus waveguide through the tapered connector, its insertion loss is comparable to that of the reference straight waveguide and can be ignored after normalization, so the test results of the $λ_{0}$ channel are not analyzed here.

Fig. 15 SEM pictures of the fabricated (a) wavelength division (de)multiplexer, (b) asymmetric uniform grating-assisted coupler, (c)-(d) asymmetric unilateral amplitude apodization grating-assisted coupler, and (e)-(g) asymmetric bilateral amplitude apodization grating-assisted coupler.

Download Full Size | PDF

Fig. 16 Experimental spectral responses of the WDMs based on (a) asymmetric uniform grating-assisted couplers, (b) asymmetric unilateral amplitude apodization grating-assisted couplers, and (c) asymmetric bilateral amplitude apodization grating-assisted couplers.

Download Full Size | PDF

The test results of the $λ_{1}, λ_{2},$ and $λ_{3}$ channels as shown in Fig. 17, where Figs. 17(a)-17(c) are the spectral responses of asymmetric uniform grating-assisted couplers, Figs. 17(d)- 17(f) are the spectral responses of asymmetric unilateral amplitude apodization grating-assisted couplers, and Figs. 17(g)-17(i) are the spectral responses of asymmetric bilateral amplitude apodization grating-assisted couplers. Parameters such as the central wavelength, insertion loss, 3dB bandwidth and sidelobe suppression ratio of each channel can be obtained according to Fig. 17, and the test results are compared, as listed in Table 4.

Fig. 17 Experimental spectral responses of (a)-(c) asymmetric uniform grating-assisted couplers, (d)-(f) asymmetric unilateral amplitude apodization grating-assisted couplers, and (g)-(i) asymmetric bilateral amplitude apodization grating-assisted couplers.

Download Full Size | PDF

Table 4. Test results of each wavelength channel of asymmetric grating-assisted couplers.

View Table | View all tables in this article

According to Fig. 16, Fig. 17 and Table 4, the test results are summarized as follows. Firstly, the central wavelength of theoretical design is $λ_{1} = 1545 n m, λ_{2} =1540 n m,$ $λ_{3} =1535 n m,$ and the measured central wavelength deviation is about 0.6 nm~1.6 nm. Secondly, the insertion loss of each channel is ideal overall, ranging from 0.23dB to 0.58dB. Thirdly, the target bandwidth for the prototypes is 300 GHz, that is, 2.4 nm, and the measured bandwidth of each wavelength channel is between 2~3 nm. Finally, compared with the asymmetric uniform grating-assisted coupler, the sidelobes of the apodized gratings are suppressed to some extent, and the sidelobe suppression ratio is above 10 dB, which reduces the channel crosstalk. However, the sidelobe suppression ratio is lower than that of the simulation results, and some channels have better unilateral sidelobe suppression, such as Figs. 17(e) and 17(i).

The main reason for the above experimental results is that the existing process errors cannot meet the high requirements of our design for the manufacturing accuracy. Since the maximum corrugation width of the grating is 100 nm, although the high precision is selected as much as possible in the manufacturing process, some grating teeth have trapezoidal shapes due to the influence of etching process. For amplitude apodized gratings, since the corrugation width of gratings has a process of decreasing gradually from the maximum to zero, the small-sized grating teeth are no longer rectangular but approximately sinusoidal shapes. Therefore, the above experimental results can be explained as follows. Firstly, since the shapes and sizes of the fabricated grating teeth slightly deviate from the ideal value, this will change the original design of the grating period, resulting in a shift in the central wavelength. Secondly, because the coupling length of each coupler manufactured is long enough to achieve a high coupling efficiency of light energy at the central wavelength, the test results indicate that each wavelength channel has a low insertion loss. Thirdly, the bandwidths of wavelength channels are changed because the fabricated gratings are different from the designed gratings. Finally, due to the sizes and shapes deviation caused by the manufacturing process, the optimal adjustment of the refractive index and coupling coefficient of the coupler cannot be realized, which will make the sidelobe suppression effect of each channel worse. When the refractive index is over-compensated or under-compensated, there is a case where the unilateral sidelobes can only be effectively suppressed.

The experimental results demonstrate that the asymmetric unilateral amplitude apodization grating-assisted coupler and the asymmetric bilateral amplitude apodization grating-assisted coupler can effectively suppress sidelobes. Although the sidelobe suppression ratio of the experimental results is slightly different from that of the simulation results, it is mainly limited by the minimum size of the devices and the manufacturing accuracy of the existing process, which makes the fabricated device deviated from the design to some extent. In a word, the experimental test results verify the feasibility of our design ideas and WDMs based on asymmetric grating-assisted couplers.

5. Conclusion

We demonstrate a WDM based on asymmetric grating-assisted couplers, which can flexibly adjust the bandwidth by changing the corrugation width of the grating. Since the coupling efficiency of the grating-assisted coupler varies monotonously with the coupling length, it reduces the requirement of the coupling length during the manufacturing process and makes the device easier to manufacture. WDM based on asymmetric uniform grating-assisted couplers, WDM based on asymmetric unilateral amplitude apodization grating-assisted couplers and WDM based on asymmetric bilateral amplitude apodization grating-assisted couplers are designed and fabricated, respectively. The simulation results show that, the sidelobes can be suppressed below −20dB and −23dB respectively by the asymmetric unilateral amplitude apodization grating-assisted coupler and the asymmetric bilateral amplitude apodization grating-assisted coupler compared with the asymmetric uniform grating-assisted coupler. The experimental results show that the insertion loss of each wavelength channel is between 0.23dB and 0.58dB, and the sidelobe suppress ratio of both unilateral amplitude apodization grating-assisted couplers and bilateral amplitude apodization grating-assisted couplers is larger than 10dB, which reduces the channel crosstalk and proves the feasibility of the WDMs. Although the corrugation width of the apodized grating and the width of the multimode waveguide need to be adjusted by Gaussian functions, which requires many parameters to be calculated during the design process and increases the computational complexity, the scheme has the following advantages. Firstly, the proposed wavelength division multiplexer realizes the multiplexing of different wavelength signals in different orthogonal modes through converting the fundamental mode into different high-order modes simultaneously by asymmetric grating-assisted couplers. Thus, the multiplexed signals can travel steadily in the bus multimode waveguide owing to the orthogonal property of the waveguide modes. Secondly, the asymmetric unilateral amplitude apodization grating-assisted couplers and asymmetric bilateral amplitude apodization grating-assisted couplers are employed to suppress the sidelobes and reduce the crosstalk between the adjacent wavelength channels. Finally, by increasing more asymmetric grating-assisted couplers in the architecture to support the higher order mode conversion, this wavelength division multiplexer can be expanded conveniently. Consequently, the proposed approach enhances the flexibility and scalability of the wavelength division multiplexers for integrated optoelectronic devices. Therefore, the work presented in this paper demonstrates that the WDM based on asymmetric grating-assisted couplers holds for the expectation to be applied to on-chip large-capacity wavelength and mode hybrid multiplexing communication systems in the future.

Funding

National Natural Science Foundation of China (NSFC) (Grant No. 61435004).

References

1. Y. Hu, R. M. Jenkins, F. Y. Gardes, E. D. Finlayson, G. Z. Mashanovich, and G. T. Reed, “Wavelength division (de)multiplexing based on dispersive self-imaging,” Opt. Lett. 36(23), 4488–4490 (2011). [CrossRef] [PubMed]

2. P. De Heyn, J. De Coster, P. Verheyen, G. Lepage, M. Pantouvaki, P. Absil, W. Bogaerts, J. Van Campenhout, and D. Van Thourhout, “Fabrication-tolerant four-channel wavelength-division-multiplexing filter based on collectively tuned Si microrings,” J. Lightwave Technol. 31(16), 3085–3092 (2013). [CrossRef]

3. F. Horst, W. M. Green, S. Assefa, S. M. Shank, Y. A. Vlasov, and B. J. Offrein, “Cascaded Mach-Zehnder wavelength filters in silicon photonics for low loss and flat pass-band WDM (de-)multiplexing,” Opt. Express 21(10), 11652–11658 (2013). [CrossRef] [PubMed]

4. S. Chen, X. Fu, J. Wang, Y. Shi, S. He, and D. Dai, “Compact dense wavelength-division (de) multiplexer utilizing a bidirectional arrayed-waveguide grating integrated with a Mach-Zehnder interferometer,” J. Lightwave Technol. 33(11), 2279–2285 (2015). [CrossRef]

5. D. Malka, Y. Sintov, and Z. Zalevsky, “Design of a 1× 4 silicon-alumina wavelength demultiplexer based on multimode interference in slot waveguide structures,” J. Opt. 17(12), 125702 (2015). [CrossRef]

6. A. Y. Piggott, J. Lu, K. G. Lagoudakis, J. Petykiewicz, T. M. Babinec, and J. Vučković, “Inverse design and demonstration of a compact and broadband on-chip wavelength demultiplexer,” Nat. Photonics 9(6), 374–377 (2015). [CrossRef]

7. D. Mu, H. Qiu, J. Jiang, X. Wang, Z. Fu, Y. Wang, X. Jiang, H. Yu, and J. Yang, “A Four-Channel DWDM Tunable Add/Drop Demultiplexer Based on Silicon Waveguide Bragg Gratings,” IEEE Photonics J. 11(1), 1–8 (2019). [CrossRef]

8. D. Dai, J. Wang, and S. He, “Silicon multimode photonic integrated devices for on-chip mode-division-multiplexed optical interconnects (invited review),” Prog. Electromagnetics Res. 143, 773–819 (2013). [CrossRef]

9. Z. Zhang, Y. Yu, and S. Fu, “Broadband on-chip mode-division multiplexer based on adiabatic couplers and symmetric Y-junction,” IEEE Photonics J. 9(2), 1–6 (2017). [CrossRef]

10. S. Li, Y. Zhou, J. Dong, X. Zhang, E. Cassan, J. Hou, C. Yang, S. Chen, D. Gao, and H. Chen, “Universal multimode waveguide crossing based on transformation optics,” Optica 5(12), 1549–1556 (2018). [CrossRef]

11. D. Dai, J. Wang, S. Chen, S. Wang, and S. He, “Monolithically integrated 64-channel silicon hybrid demultiplexer enabling simultaneous wavelength-and mode-division-multiplexing,” Laser Photonics Rev. 9(3), 339–344 (2015). [CrossRef]

12. D. Melati, A. Alippi, A. Annoni, N. Peserico, and A. Melloni, “Integrated all-optical MIMO demultiplexer for mode- and wavelength-division-multiplexed transmission,” Opt. Lett. 42(2), 342–345 (2017). [CrossRef] [PubMed]

13. S. Wang, X. Feng, S. Gao, Y. Shi, T. Dai, H. Yu, H.-K. Tsang, and D. Dai, “On-chip reconfigurable optical add-drop multiplexer for hybrid wavelength/mode-division-multiplexing systems,” Opt. Lett. 42(14), 2802–2805 (2017). [CrossRef] [PubMed]

14. J. Wang, S. He, and D. Dai, “On-chip silicon 8-channel hybrid (de) multiplexer enabling simultaneous mode-and polarization-division-multiplexing,” Laser Photonics Rev. 8(2), 18–22 (2014). [CrossRef]

15. L. Zhu and J. Sun, “Silicon-based wavelength division multiplexer by exploiting mode conversion in asymmetric directional couplers,” OSA Continuum 1(1), 92–103 (2018). [CrossRef]

16. H. Qiu, H. Yu, T. Hu, G. Jiang, H. Shao, P. Yu, J. Yang, and X. Jiang, “Silicon mode multi/demultiplexer based on multimode grating-assisted couplers,” Opt. Express 21(15), 17904–17911 (2013). [CrossRef] [PubMed]

17. H. Qiu, Y. Su, F. Lin, J. Jiang, P. Yu, H. Yu, J. Yang, and X. Jiang, “Silicon Add-Drop Filter Based on Multimode Grating Assisted Couplers,” IEEE Photonics J. 8(6), 1–8 (2016). [CrossRef]

18. A. Yariv, “Coupled-mode theory for guided-wave optics,” IEEE J. Quantum Electron. 9(9), 919–933 (1973). [CrossRef]

19. W. Shi, H. Yun, C. Lin, M. Greenberg, X. Wang, Y. Wang, S. T. Fard, J. Flueckiger, N. A. Jaeger, and L. Chrostowski, “Ultra-compact, flat-top demultiplexer using anti-reflection contra-directional couplers for CWDM networks on silicon,” Opt. Express 21(6), 6733–6738 (2013). [CrossRef] [PubMed]

20. T. Erdogan, “Fiber grating spectra,” J. Lightwave Technol. 15(8), 1277–1294 (1997). [CrossRef]

21. H. Qiu, G. Jiang, T. Hu, H. Shao, P. Yu, J. Yang, and X. Jiang, “FSR-free add-drop filter based on silicon grating-assisted contradirectional couplers,” Opt. Lett. 38(1), 1–3 (2013). [CrossRef] [PubMed]

22. W. Shi, X. Wang, C. Lin, H. Yun, Y. Liu, T. Baehr-Jones, M. Hochberg, N. A. Jaeger, and L. Chrostowski, “Silicon photonic grating-assisted, contra-directional couplers,” Opt. Express 21(3), 3633–3650 (2013). [CrossRef] [PubMed]

23. A. D. Simard, N. Belhadj, Y. Painchaud, and S. LaRochelle, “Apodized silicon-on-insulator Bragg gratings,” IEEE Photonics Technol. Lett. 24(12), 1033–1035 (2012). [CrossRef]

24. A. D. Simard and S. LaRochelle, “Complex apodized Bragg grating filters without circulators in silicon-on-insulator,” Opt. Express 23(13), 16662–16675 (2015). [CrossRef] [PubMed]

25. J. Jiang, H. Qiu, G. Wang, Y. Li, T. Dai, D. Mu, H. Yu, J. Yang, and X. Jiang, “Silicon lateral-apodized add-drop filter for on-chip optical interconnection,” Appl. Opt. 56(30), 8425–8429 (2017). [CrossRef] [PubMed]

26. M. J. Strain and M. Sorel, “Integrated III–V Bragg Gratings for Arbitrary Control Over Chirp and Coupling Coefficient,” IEEE Photonics Technol. Lett. 20(22), 1863–1865 (2008). [CrossRef]

27. W. Shi, H. Yun, C. Lin, J. Flueckiger, N. A. Jaeger, and L. Chrostowski, “Coupler-apodized Bragg-grating add-drop filter,” Opt. Lett. 38(16), 3068–3070 (2013). [CrossRef] [PubMed]

28. H. Qiu, J. Jiang, P. Yu, J. Yang, H. Yu, and X. Jiang, “Broad bandwidth and large fabrication tolerance polarization beam splitter based on multimode anti-symmetric Bragg sidewall gratings,” Opt. Lett. 42(19), 3912–3915 (2017). [CrossRef] [PubMed]

29. X. Zhao, Y. Wang, Q. Huang, and J. Xia, “Two-mode contra-directional coupler based on superposed grating,” Opt. Express 25(3), 2654–2665 (2017). [CrossRef] [PubMed]

30. J. Jiang, H. Qiu, G. Wang, Y. Li, T. Dai, X. Wang, H. Yu, J. Yang, and X. Jiang, “Broadband tunable filter based on the loop of multimode Bragg grating,” Opt. Express 26(1), 559–566 (2018). [CrossRef] [PubMed]

31. M. Ma, Z. Chen, H. Yun, Y. Wang, X. Wang, N. A. F. Jaeger, and L. Chrostowski, “Apodized Spiral Bragg Grating Waveguides in Silicon-on-Insulator,” IEEE Photonics Technol. Lett. 30(1), 111–114 (2018). [CrossRef]

32. Y. Shibata, T. Tamamura, S. Oku, and Y. Kondo, “Coupling coefficient modulation of waveguide grating using sampled grating,” IEEE Photonics Technol. Lett. 6(10), 1222–1224 (1994). [CrossRef]

33. A. D. Simard, N. Ayotte, Y. Painchaud, S. Bedard, and S. LaRochelle, “Impact of Sidewall Roughness on Integrated Bragg Gratings,” J. Lightwave Technol. 29(24), 3693–3704 (2011). [CrossRef]

34. X. Wang, W. Shi, H. Yun, S. Grist, N. A. F. Jaeger, and L. Chrostowski, “Narrow-band waveguide Bragg gratings on SOI wafers with CMOS-compatible fabrication process,” Opt. Express 20(14), 15547–15558 (2012). [CrossRef] [PubMed]

35. X. Wang, Y. Wang, J. Flueckiger, R. Bojko, A. Liu, A. Reid, J. Pond, N. A. F. Jaeger, and L. Chrostowski, “Precise control of the coupling coefficient through destructive interference in silicon waveguide Bragg gratings,” Opt. Lett. 39(19), 5519–5522 (2014). [CrossRef] [PubMed]

Channel	λ₁	λ₂	λ₃
W_i (µm)	0.82	1.27	1.7
Λ_i (µm)	331.9	329.5	328.4
N	600	800	800
L_i (µm)	199.14	263.6	262.72

Channel	λ₁	λ₂	λ₃
W_i (µm)	0.82	1.27	1.7
b	0.021	0.03	0.03
Λ_i (µm)	331.5	329.6	328.2
N	600	800	800
L_i (µm)	198.9	263.68	262.56

Channel	λ₁	λ₂	λ₃
W_i (µm)	0.82	1.27	1.7
Λ_i (µm)	331.5	329.5	328.4
N	600	800	800
L_i (µm)	198.9	263.6	262.72

Type	Uniform grating			Unilateral amplitude apodization grating			Bilateral amplitude apodization grating
Channel	λ₁	λ₂	λ₃	λ₁	λ₂	λ₃	λ₁	λ₂	λ₃
Central wavelength (nm)	1543.6	1538.5	1533.4	1543.7	1538.9	1534.1	1544	1538.4	1534.4
Insertion Loss (dB)	0.32	0.27	0.23	0.33	0.25	0.43	0.47	0.58	0.43
3dB bandwidth (nm)	2.4	2.2	2.7	2.6	3	2.4	2	2.1	2.5
Sidelobe suppression ratio (dB)	2.1	2	3.5	10.6	12.9	12	11.6	11.2	12.5

Channel	λ₁	λ₂	λ₃
W_i (µm)	0.82	1.27	1.7
Λ_i (µm)	331.9	329.5	328.4
N	600	800	800
L_i (µm)	199.14	263.6	262.72

Silicon-based wavelength division multiplexer using asymmetric grating-assisted couplers

Abstract

1. Introduction

2. Principle and device design

3. Improved design and simulation

3.1 WDM based on asymmetric unilateral amplitude apodization grating-assisted couplers

3.2 WDM based on asymmetric bilateral amplitude apodization grating-assisted couplers

4. Device fabrication and experimental results

5. Conclusion

Funding

References

Cited By

Figures (17)

Tables (4)

Equations (5)

Optics Express