Open Access
Add to BibSonomy
Abstract
Alignment is critical for efficient integration of photonic integrated circuits (PICs), and microelectromechanical systems (MEMS) actuators have shown potential to tackle this issue. In this work, we report MEMS positioning actuators designed with the ultimate goal of aligning silicon nitride (SiN) waveguides either to different outputs within a SiN chip or to active chips, such as lasers and semiconductor optical amplifiers. For the proof-of-concept, suspended SiN waveguides implemented on a silicon-on-insulator wafer were displaced horizontally in the direction of light propagation to close an initial gap of 6.92 µm and couple the light to fixed output waveguides located on a static section of the chip. With the gap closed, the suspended waveguides showed ∼ 345 nm out-of-plane misalignment with respect to the fixed waveguides. The suspended waveguides can be displaced laterally by more than ±2 µm. When the waveguides are aligned and the gap closed, an average loss of −1.6 ± 0.06 dB was achieved, whereas when the gap is closed with a ± 2 µm lateral displacement, a maximum average loss of ∼ −19.00 ± 0.62 dB was obtained. The performance of this positioner does not only pave the way for active chip alignment, but it could also be considered for optical switching applications.
© 2023 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
There are significant efforts currently being made to develop photonic integrated circuits (PICs) to provide high-performance and low power optical communication systems [1]. These PICs generally integrate several optical components such as laser sources, modulators, multiplexers, and detectors [1], into a single package. Silicon nitride (SiN) is a promising platform for developing photonic devices due several advantages [2]. SiN-based photonics platforms enable the implementation of low loss devices, and SiN as material is compatible with standard CMOS fabrication processes [3]. In addition, SiN has a broad transparency window, low sensitivity to temperature variation and high fabrication tolerance [2], and can be used to create efficient non-linear devices [4]. Despite the aforementioned advantages, SiN does not possess electro-optical properties and thus SiN photonic circuits are typically tunned using power consuming thermo-optic effects. Furthermore, essential parts of PICs, i.e., active components, such as laser sources and semiconductor optical amplifiers (SOAs), cannot be realized on SiN and silicon-based platforms. Making an efficient light source from silicon is impossible since it has an indirect bandgap [5]. The light amplification needed to compensate for optical losses caused by passive optical components such as couplers, interferometers, and filters is also not trivial to implement in SiN PICs. Therefore, direct bandgap III-V semiconductor materials [6], such as indium phosphide (InP) and gallium arsenide (GaAs) [7], need to be integrated with SiN and silicon-on-insulator (SOI) platforms. There are two main types of integration, hybrid and heterogeneous, which combine active components implemented on III-V platforms with passive devices fabricated on SiN and SOI platforms [1,8]. Hybrid integration combines two or more fully processed PICs into a single package. On the other hand, heterogeneous integration combines different materials into a single chip [9]. Additionally, heterogeneous integration allows for wafer-level integration of the III-V materials on silicon substrates, enabling the demonstration of efficient tunable lasers and SOAs [7,10].
Various techniques are used to bond different PICs. They include wafer bonding [11,12], suited for heterogeneous integration, die-to-die bonding [13], and flip-chip bonding, which is used primarily for hybrid integration [14]. Although these bonding techniques make the integration of the PICs components possible, sub-micron level alignment is still a significant challenge [15], where several fabrication dependent parameters determine the vertical and lateral alignment accuracies. For example, the vertical alignment accuracy depends on the accuracy of the etching processes, thickness tolerance of the deposited layers and bonding force. The effectiveness of lateral alignments depends on the offset angle due to rotation caused by thermal drift of the chip during bonding [16]. One of the existing solutions to tackle the alignment issues seen in heterogeneous and hybrid integrations is to use photonic wire bonding (PWB) for directly connecting waveguides between the different processed chips or platforms, as reported in [17] and [18]. However, although PWB is effective, its manufacturing process is complex since it requires several elaborate steps [18], and PWB has yet to reach manufacturability levels that will make it ubiquitous. Another promising solution to address the alignment challenge is Micro-Electro-Mechanical Systems (MEMS) actuators that have long been used in various sensing and telecommunications applications due to their compact size and low power consumption. MEMS actuators are used in photonic switches [19–22], reconfigurable ring resonators [23], phase shifters [24] and tunable optical couplers [25–27]. In [28], electrothermal bimorph actuators were proposed to compensate out-of-plane misalignment between InP active chip and silicon photonic chip.
MEMS tuning approaches have been contributing to the development of programmable PICs [29], where waveguide meshes of tunable couplers and phase shifters could be reconfigured in software to define diverse functions and arbitrary connectivity between the input and output ports [30]. For example, an electrostatically-actuated 1 × 2 optical MEMS switch with an extinction ratio of more than 23 dB over 70 nm of optical bandwidth was reported in [31]. In [32], disk and ring resonators with MEMS-movable waveguides showed a high loaded optical quality factor of up to 3.6 × 104 and more than 20 dB of extinction ratio. In [33,34], compact low-power comb-drive MEMS phase shifters were demonstrated where ∼ 3π phase shifts and a 3 dB bandwidth of over 1 MHz were achieved at a wavelength of 1550 nm. In [35], a suspended MEMS-actuated directional coupler with an insertion loss of 0.5 dB and a 1 dB bandwidth of 3 nm at a wavelength of 1550 nm was implemented. In addition, optical beam steering over 5.6° has also been demonstrated with a MEMS actuator stretching a surface grating coupler [36]. A comprehensive review of MEMS-actuated gratings can be found in [37].
In this work, we demonstrate a MEMS positioner based on electrothermal actuators that can align suspended silicon nitride (SiN) waveguides to others located on a fixed section of the chip as a first step to show the potential of MEMS actuators to enable hybrid integration or to switch between devices in a SiN PIC. The micropositioners were implemented in a unique fabrication platform enabling monolithic integrating of SiN waveguides and silicon MEMS. It is the first fully integrated positioner capable of moving SiN along two degrees of freedom with high precision thermal actuators. The suspended waveguides were horizontally displaced to close an initial gap of 6.92 µm between the suspended and fixed waveguides. An insertion loss of −1.6 ± 0.06 dB was achieved for the best alignment along the x and y axes, with an intrinsic average out-of-plane misalignment of 345 nm along the z axis, caused by mechanical stress.
This article is structured as follows: section 2 presents the design and simulations of the device; section 3 describes the layout and fabrication; section 4 presents mechanical and optical characterization setups; section 5 is dedicated to the experimental results; section 6 provides a discussion of the results; and section 7 presents conclusions drawn from the results and future work directions.
2. Device design and simulations
A schematic of the positioner is shown in Fig. 1. It consists of a platform supported by electrothermal chevron-type actuators through pulling arms. These arms are supported by side springs to prevent buckling.
The platform carries suspended waveguides routed over two dedicated optical path beams, to align them to fixed waveguides that simulate waveguides on an active chip such as a laser diode or a semiconductor optical amplifier. The lateral actuators are used to tune the light coupling into the fixed waveguides by finely displacing the platform laterally along the x-axis in both the positive or negative directions. The gap closing actuator is used to displace the platform in the direction of the positive y-axis to gradually close the gap and bring the suspended waveguides into close contact with the fixed ones to maximize coupling of the light. It is worth mentioning that when the gap is fully closed while both lateral and gap closing actuators are activated, the activation of the lateral actuator precedes in time that of the gap closing actuator to prevent damaging the contact surface of the waveguides.
To maximize the performance of the positioner, several parameters of the chevron actuators must be optimized. These parameters include the inclination angle θ, shown in Fig. 2(a), the length, the width, the thickness and the number of chevron beams. The different parameters were optimized using a finite element analysis software (CoventorWare), where a single parameter was varied while keeping the others fixed and observing the electrical actuation power and / or displacement. The proposed positioner was fabricated using a multi-project process with a fixed device layer thickness of 59 µm, a minimum feature size of 4 µm, and a pre-defined cavity size. The fabrication process was developed by AEPONYX inc. and more details are provided in section 3. Since the displacement provided by electrothermal actuators depends on the thermal expansion of the beams, which is directly related to the size of the actuator, the size of the cavity sets the largest possible displacement because it constraints the maximum size of our actuators. Therefore, the parameters optimization was mainly focused on the inclination angle and the number of chevron beams. Figure 2(b) shows the effect of the inclination angle on the displacement for a single chevron beam with length, width and thickness fixed at 297 µm, 4 µm and 59 µm, respectively. It is worth noting that the length of the beams is determined by the pre-defined cavity size, whereas the width used was the minimum allowed by the fabrication process (displacement is reversely proportional to the width of the beam). At 5 V of actuation voltage, which corresponds to ∼53 mW, the displacement was found to rapidly increase with increasing inclination angle. The displacement peaks at 2.77° then it gradually decreases beyond that angle. Thus, to determine the optimum number of chevron beams the angle was fixed at 2.77°.
Fig. 2. Parameters optimization of the chevron actuator: (a) schematic of a single chevron beam, (b) displacement versus inclination angle, and (c) displacement and actuation power versus the number of chevron beams.
Figure 2(c) shows the effect of chevron beams on both the displacement and the power consumption. Increasing the number of chevron beams slightly increases the displacement. However, it will linearly increase the power. As such, single chevron beams were used, as shown in Fig. 1. Table 1 lists the design parameters of the positioner.
Table 1. Specifications of electrothermal MEMS waveguide positioner
To investigate the thermal cross-talk between the gap closing and lateral actuators, electrothermomechanical simulations was conducted, where a single actuator is activated while the rest are off, and the displacement was monitored in all the three directions, i.e., along the x, y, and z-axes. Fig. 3(a) shows the displacements along the x and z-axes due to thermal cross-talk when the gap closing actuator in y-axis was activated. As can be seen, when the maximum actuation power of ∼ 234 mW was provided to the gap closing actuator, a displacement of over 6 µm was obtained along the y-axis, whereas the maximum thermal cross-talk displacements along the x and z-axes were found to be only 52 nm and -22 nm, respectively. Likewise, applying 268 mW to the lateral actuators along the x-axis displaced the platform by 2.4 µm with only 9 nm of thermal-cross talk displacement along the z-axis, as shown in Fig. 3(b). On the other hand, ∼ 400 nm of displacement in + y axis direction was recorded (see Fig. 3(b)). However, since this is the direction to close the gap, the lateral actuators can contribute to reduce the power required by the y-axis gap closing actuator.
Fig. 3. Simulated thermal cross-talk between the actuators: (a) displacements along the x and z-axes due to actuation along the y-axis, and (b) displacements along the y and z-axes due to lateral actuation along the x-axis.
The response time for the actuators was investigated by conducting FEA simulations, where a 15 ms-pulse of 6.1 V amplitude was applied as shown in Fig. 4. The time for both the gap closing and lateral actuators to reach 90% of the maximum displacement amplitude was found to be 3 ms as in the figure.
3. Device layout and fabrication
The layout of the device, shown in Fig. 5, was fabricated following a customized process developed by AEPONYX inc. [38]. The process uses a silicon-on-insulator wafer with a 59 µm thick silicon device layer. The wafer handle has predefined cavities to ease the release of the MEMS actuators of the positioner. Single-mode waveguides made of a stack of silicon dioxide-silicon nitride-silicon dioxide layers were used. The silicon dioxide (SiO2) cladding layers are 3.4 µm-thick each and 10 µm-wide, whereas the core is made of a 435 nm-thick and 850 nm-wide silicon nitride (SiN) layer. The core is tapered down to 400 nm near the gap over a length of 100 µm.
Fig. 5. Layout of the MEMS waveguide positioner showing (a) the routing waveguides, (b) input/output gratings for vertical coupling of the light signals, and (c) the cross section corresponding to the A—B line in (a).
The layout contains a reference waveguide (Fig. 5(a)) that has a total length equivalent to that of the waveguides in the device to de-embed the optical propagation loss of the waveguides and find the insertion loss due only to the gap between the fixed and suspended waveguides. The waveguides have surface gratings at their end (Fig. 5(b)) to vertically couple the light between the input/output and a fiber array. A cross-section of the device (identified by the line A—B) is shown in Fig. 5(c). To make the connecting pads, a 250 nm-thick aluminum copper (AlCu) alloy was deposited and patterned on the terminals of the actuators.
After fabricating the devices, scanning electron microscope (SEM) pictures were taken to investigate fabrication variations by comparing the measured parameters to their designed values, as shown in Fig. 6. Variations from the designed dimensions were observed. For instance, in Fig. 6(b), the gap was measured to be 5.86 ± 0.11 µm at the top of the silicon device layer compared to the designed value of 6 µm. It is worth noting that due to the etching sidewall slope, we could measure a larger gap of 6.87 ± 0.07 µm at the top of the waveguide. In Fig. 6(c), the width of the optical path beam was measured to be 9.89 ± 0.20 µm compared to its designed value of 10 µm, and in Fig. 6(d), the fabricated width of the gap closing actuator, which had a design value of 4 µm, was found to be 4.03 ± 0.05 µm. Fig. 6(e) shows an optical stack sidewall angle of 5.41°, with an average value for the three devices found to be 5.15 ± 0.45°. This angle prevents the total closure of the gap and leaves a space of 0.59∼0.6 µm between the core of the waveguides, as will be discussed in the results section of the mechanical characterization.
Fig. 6. SEM images of the fabricated MEMS waveguide positioner, (a) SEM image of the MEMS actuators, (b) gap between the suspended and fixed waveguides, (c) width of an optical path beam, (d) width of the gap closing actuator along the y-axis, (e) sidewall angle of the optical stack, and (f) platform with the suspended waveguides (WGs) along with the fixed waveguides.
4. Characterization procedures
To characterize the fabricated devices, two type of tests, namely electromechanical and optical, were conducted. The devices were wire-bonded on specially designed printed circuit boards (PCB), and connection wires were soldered to the PCB pads to provide terminals for external power sources and measurement equipments. In general, measurements were repeated five times and the average values with their standard error deviations are reported.
4.1 Electromechanical characterization
Electromechanical tests were conducted to characterize the gap closing and lateral actuators by finding their displacements as a function of actuation power. A schematic of the setup used for these tests is shown in Fig. 7(a). Before activating the actuators, the initial gaps around the platform (i.e., the separation between the movable and fixed waveguides and the gaps on the sides of the platform as shown in Fig. 6(f)) were measured. Then, the actuators were activated, and the gaps were re-measured at each actuation voltage to find the displacement of the platform. Voltages in the range of 0 –16 V were applied using a DC voltage source (Keithley 2260B-800 l, Cleveland, OH, USA), initially in increments of 2 V and then 1 V as the gap was nearing closure. Currents were measured using a digital multi-meter (Tektronix DMM 4050, Beaverton, OR, USA) and the corresponding power consumptions were calculated. A LEXT 3D confocal microscope (model OLS 4100) was used to capture and process the images to measure the gaps and the displacement of the platform.
Fig. 7. Schematic of the (a) mechanical and (b) optical testing setups used to characterize the MEMS devices under test (DUT).
4.2 Optical characterization
The optical tests were conducted using the setup shown in Fig. 7(b). The MEMS dies were mounted on XYZ micro positioning stages (PS) to align the grating couplers with an optical fiber array polished at 30° with the help of top and side view cameras. A tunable laser source (TLS) model (T100S-HP), and an optical component tester (CT440) from EXFO were used to optically address the structures. With the TLS, an optical signal with polarization aligned along the transverse magnetic direction was swept across a wavelength range from 1500 to 1630 nm and coupled through a polarization maintaining fiber array to one of the fixed waveguides of the positioner. The light thus passed through the gap and was collected from the suspended waveguide through an output grating coupler. Output signals were read by the optical detectors (OD) and saved for analysis. The distance between the device under test (DUT) and the optical fiber array was estimated to be between 10 to 20 µm. The devices were tested in different states, including deactivating all actuators or only some of them, as summarized in Table 2.
Table 2. Conditions under which the DUTs were tested
After measuring the optical signal while all the actuators are inactive and the gaps are fully open, the gap closing actuator was activated by applying voltages following the procedure outlined above. The optical signal was monitored to determine the transmitted power as a function of the electrical actuation power and to find the conditions when the optical output signal experienced the lowest loss. Measurements were also conducted to see the effect of laterally moving the platform either to the left or to the right when the gap was either open or closed.
Note that, to be concise, only the results of lateral actuation after the gap was closed are reported here. To precisely determine the coupling losses, and hence the alignment performance of the actuators, a reference loop, i.e., a fixed structure with the same length and number of bends as the waveguides on the device, was included, as can be seen in Fig. 5(a). Each test was repeated five times to investigate the repeatability of the measurements and to determine the measurement variability.
The experimental results are compared to 3-D FDTD simulations performed with Ansys Lumerical (Canonsburg, PA, USA) to calculate the transmission efficiency for different gaps and lateral displacements, and also while considering different out-of-plane misalignments.
5. Results
5.1 Electromechanical results
A total of three devices, referred to as Device 1, Device 2, and Device 3, were characterized, and the results were compared to finite element simulation carried out using the CoventorWare 10.5 software (Coventor, A Lam Research Company, Raleigh, NC, USA). The setup shown in Fig. 7(a) was used to characterize the devices. First, the actuators were individually characterized by activating a single actuator while others are maintained in the off-state to measure the platform displacement along a single axis, then simultaneous activation of the gap closing actuator along the y-axis with either the lateral left actuator along the –x-axis or the lateral right actuator along the + x-axis were performed to measure the platform displacement in the x and y axes.
Figure 8 shows the results, where in Fig. 8(a) the gap and out-of-plane misalignment between the suspended and fixed waveguides are given as a function of the actuating power of the gap closing actuator when the lateral actuators are inactive, whereas Fig. 8(b) shows the lateral displacement of the platform versus drive power of the lateral actuators in addition to the simulated out-of-plane misalignments.
Fig. 8. Mechanical characterizations results: (a) average gap size and average out-of-plane misalignment versus actuation power for Devices 1, 2 and 3; (b) average lateral displacement and simulated out-of-plane misalignment versus power supplied to the lateral actuators for Device 1.
As shown in Fig. 8(a), the gap size is plotted as a function of the gap closing actuator drive power. When no power is provided to the actuator, the gap (designed value of 6 µm) was measured to be 6.92 ± 0.01, 6.80 ± 0.28 and 6.92 ± 0.01 µm for Devices 1, 2 and 3, respectively. These gaps were gradually reduced to their minimum value of 0.59∼0.63 µm by increasing the actuation power to 189 mW, 215 mW and 194 mW for Devices 1, 2 and 3, respectively. This corresponds to a displacement of the platform of more than 6 µm. Increasing the actuation power beyond these values was found to have no effect on the remaining gap. This remaining gap is due to sidewall angles at the gap edges as will be discussed in section 6.
The average out-of-plane misalignment of the platform (i.e., suspended waveguides) along the z-axis with respect to the fixed waveguides was given as a function of the gap closing actuation power in Fig. 8(a). It was found that the misalignment slightly increases by closing the gap. In Device 1, it changed from an initial value of 273 ± 32 nm to 345 ± 52 nm when 189 mW was applied. For Device 2, the misalignment increased from 287 ± 45 nm to 482 ± 29 nm at 214 mW, whereas in Device 3, the misalignment increased from a starting value of 484 ± 44 nm to 521 ± 43 nm at 194 mW. On the other hand, the finite element analysis (FEA) simulations predicted that the misalignment would decrease from +55 nm initially to −51 nm when the gap is closed, which required 211 mW.
By activating the lateral actuators, the suspended waveguides could be displaced to the left or right along the x-axis, with respect to the fixed waveguides. Figure 8(b) shows the average lateral displacement of Device 1 as a function of actuation power. More than ±2 µm of displacement was achieved with an actuation power of 196 mW. The results were found to be close and follow the same trends predicted by the FEA simulations. As can be seen in Fig. 8(b), the impact of the simulated x-axis actuation on the out-of-plane misalignment is not significant.
Simultaneous activation of the gap closing and either the lateral left or lateral right actuators was also investigated since it would be required to align to different chips. Figure 9 shows optical images of the different actuation conditions, where Fig. 9(a) shows the initial gap between the suspended and fixed waveguides when all three actuators are inactive.
Fig. 9. Optical image showing different actuation states of the MEMS waveguide positioners: (a) the initial state when all the actuators are inactive and the gap between the fixed and suspended waveguides is fully open (GO); (b) the gap closing actuator is activated and the gap is fully closed (GC); (c) simultaneous activation of both gap closing and lateral left actuators, where the platform was moved to the left and then the gap closed, and (d) simultaneous activation of both gap closing and lateral right actuators, where the platform was moved to the right and then the gap closed.
Figure 9(b) shows the activation of only the gap closing actuator to displace the platform along the positive y-axis to close the gap and bring the suspended waveguides into close contact with the fixed ones. Figures 9(c) and (d) respectively show the lateral displacement of the platform to the left in the negative x-axis direction and to the right in the positive x-axis direction, while the gap closing actuator is activated.
5.2 Optical results
In this section, the optical results obtained from the test procedure described above (see Fig. 7(b)) are presented. The optical results were obtained for the mechanical states depicted in Fig. 9 above; where results were recorded for the gap opened and gap closed with or without lateral displacements. The average measured results for five cycles are reported and compared to optical simulations carried out using the same gaps and out-of-plane misalignments obtained during the mechanical characterization. Figure 10(a) shows the average normalized transmission over five cycles for each device over a broad wavelength range of 1555 nm to 1620 nm. The data normalization was carried out by subtracting the output signal of the device from the signal of the reference loop described in section 4.2. The data for wavelengths below 1555 nm were omitted due to high ripples associated with the grating reflections. The average normalized transmission was found when the gap is reduced to its minimum size obtained during the mechanical characterization. The average normalized transmission versus the gap size is shown by the primary y-axis of Fig. 10(b). At the initial gap of 6.92 ± 0.01 µm, the average normalized transmissions for Devices 1, 2 and 3 were found to be −7.69 ± 0.09 dB, −7.85 ± 0.19 dB and −8.45 ± 0.04 dB, respectively versus −7.06 ± 0.10 dB for the simulation. As the gap closing actuation power increases and the gap decreases, the average normalized transmission loss decreases to −1.60 ± 0.06 dB, −1.70 ± 0.03 dB and −1.88 ± 0.02 dB for Devices 1, 2 and 3, respectively, when the gap is closed compared to −1.52 ± 0.06 dB for simulations. Note that the simulations did not consider the optical stack sidewall angle, where right angled sidewalls are used. The out-of-plane misalignment of the suspended waveguides with respect to the fixed waveguides is given by the secondary y-axis in Fig. 10(b). While the out-of-plane misalignment was found to slightly increase when the gap was decreased, it is clear that this misalignment has a noticeable impact on the average normalized transmission since devices with higher misalignments had higher losses. At the minimum measured gaps of 0.63 µm, 0.67 µm and 0.59 µm for Devices 1, 2 and 3, respectively, the measured misalignments were found to be 345 ± 52 nm, 482 ± 29 nm and 521 ± 43 nm respectively. The out-of-plane misalignment used for the simulation was 400 nm.
Fig. 10. Optical characterization results showing: (a) the average normalized transmission over the wavelength range of 1555-1620 nm, and (b) the average normalized transmission and average out-of-plane misalignment versus the gap between the fixed and suspended waveguides.
The effect of lateral displacement on the transmission and the study of crosstalk between the two suspended waveguides were investigated for Device 1 as shown in Fig. 11(a) and (b), respectively. To evaluate the impact of lateral displacements on the response of the device, either the lateral left or the lateral right actuator was activated with different actuation powers. For each lateral displacement value, the transmission efficiency was measured after the gap was closed. The transmission is plotted as a function of the lateral displacement in Fig. 11(a). When the fixed waveguide (input) and suspended waveguide (output) are aligned, the average normalized transmission was −1.37 ± 0.33 dB for this device whereas the simulation predicted −1.52 ± 0.06 dB. Increasing the lateral displacement of the platform to ∼ ± 2 µm (−2.08 ± 0.17 µm for lateral left and 2.12 ± 0.37 µm for lateral right) increases the average transmission loss to −19.0 ± 0.62 dB. The measured results were found to follow the same trend observed in the simulations for the same conditions, where for ∼ ± 2.5 µm, a simulated transmission loss of −20.48 ± 0.07 dB was predicted. The curve in the figure indicates that the measured transmission saturates when displacement is close to 2 µm, contrary to simulations. This could be because we are reaching the limit of the lateral actuator, as shown in Fig. 8(b).
Fig. 11. Optical characterization results of Device 1 showing: (a) the effect of lateral displacement of the platform on the average normalized transmission, and (b) the study of crosstalk between the two suspended waveguides on the platform.
To investigate the crosstalk between the suspended waveguides, which are separated by 10 µm, light was coupled to fixed waveguide 1 shown in Fig. 9(a) and the output was measured on both suspended waveguides 1 and 2. When the gap is closed, the input waveguide 1 is aligned to output waveguide 1 whereas the output waveguide 2 is offset by 10 µm from the input light path. The transmission was measured when the gap is open (GO) and closed (GC), as shown in Fig. 11(b). For both the GO and GC conditions, output waveguide 2 shows almost no coupling of light from input waveguide 1, with less than −40 dB of crosstalk between the two suspended output waveguides when the gap is closed.
6. Discussion
The 59 µm device layer used to build the devices was previously reported to have a great resilience to residual stresses caused by waveguides, where significant improvement in out-of-plane misalignment between the fixed and suspended parts were achieved over structures with a 10 µm device layer [38]. For the three tested devices (Devices 1, 2 and 3), out-of-plane misalignments of 345 ± 52 nm, 482 ± 29 nm and 521 ± 43 nm were measured, respectively when the gap was closed. The corresponding average normalized optical transmission were −1.60 ± 0.06 dB, −1.70 ± 0.03 dB and −1.86 ± 0.02 dB, respectively. This result indicates a correlation between the out-of-plane misalignment and the coupling efficiency. This is also confirmed by simulations, as shown in Fig. 12, obtained with the same residual measured closed gap size of 0.63 µm with no lateral misalignments. Simulations were carried out using the three-dimensional finite-difference time-domain (3-D FDTD) method for vertically aligned waveguides and with vertical out-of-plane misalignments of 200 nm, 400 nm, and 600 nm. Figures 12(a) and (b) show the electric field propagation for vertically aligned and 600 nm of vertical (i.e., out-of-plane) misalignment, respectively. As expected, simulations of the gaps with minor vertical misalignment between the fixed and suspended waveguides show less scattering. The simulated average transmission over the wavelength range of 1550 to 1620 nm compared to that of the tested devices is shown in Fig. 12(c). The simulated loss increases from −0.66 dB for the vertically aligned waveguides to −2.56 dB for 600 nm vertical misalignment. This is in good agreement with the experimental data, where the devices with low out-of-plane misalignment, such as Device 1, have shown better performance (−1.60 ± 0.06 dB) compared to that with higher misalignment, such as Device 3 (−1.86 ± 0.02 dB).
Fig. 12. Effect of out-of-plane misalignment on the propagation of light from the fixed waveguide to the suspended one: simulated electric field propagation for (a) vertically aligned waveguides, (b) 0.6 µm out-of-plane misalignment, and (c) the effect of the out-of-plane misalignment on the transmission.
The effect of out-of-plane misalignments on optical losses was also reported by other authors. For instance, in [39], losses of −0.8 dB/µm of out-of-plane misalignment were mentioned when a silicon nitride waveguide of a tapered width of 120 nm and cladding of 8 µm was coupled to a fiber. This loss is smaller than that measured by our devices where 600 nm misalignment increases the loss by −1.9 dB. This significant difference could be due to the difference in design geometries (such as tapered waveguide width; 120 nm vs 400 nm in our case). In another study, simulation results on a silicon nitride waveguide with a 500 × 500 nm2 cross-section aligned to a InP active chip showed −1.78 dB of coupling loss for 200 nm of out-of-plane misalignment and 300 nm of lateral misalignment [40], compared to 0.9 dB for the case with ideal alignment. A second simulated result [41] obtained for the hybrid integration of a laser diode and a silicon photonic waveguide with the help of spot size converters, a 1 dB loss tolerance was reported for misalignments of ±1.69 µm in the lateral direction, of ±0.49 µm in the out-of-plane direction and of 3.8 µm in the direction of light propagation. These results support our findings about the impact of out-of-plane misalignment on the insertion loss.
The performance of the positioners could be improved by the use of spot size converters and coatings to minimize the optical modes mismatch and to reduce reflections and interferences [42]. Furthermore, it was also reported that the sidewall angles limits the minimum achievable gap size [43], and hence the minimal optical loss obtainable. For example in [38], a residual air gap of 1 µm was reported for sidewall angles of 7.86° and 8.97°, and improving the etching process to improve the sidewall angle was found to reduce the minimal gaps achievable, and enhance the optical losses. Likewise, our SEM results showed that the two sidewalls of the gap that separate the input and output waveguides are not etched vertically. Angles of 5.26 ± 0.04°, 5.60 ± 0.07° and 4.91 ± 0.01° were measured for the sidewalls of Devices 1, 2 and 3, respectively. Because of these angles, measuring the initial gap reveals two distinct topographies, as shown in Fig. 6(b). The first measurement shows a gap of 5.86 ± 0.11 µm at the level of the silicon layer, which is close to the designed value of 6 µm. However, the other measurement shows a gap in the range of 6.87 ± 0.07 µm, which is the one at the top of the optical stack. The measured etching angles lead to a residual gap at the level of the core of the waveguide of 0.63 µm, 0.67 µm and 0.59 µm for Device 1, 2 and 3, respectively, as depicted in Fig. 13.
Fig. 13. Residual gap sizes due to the measured sidewall angles for (a) Device 1 (b) Device 2 and (c) Device 3.
These measured residual gaps contribute to the minimal insertion losses obtained. Device 3, which has the lowest residual gap of 0.59 µm, is expected to show a lower loss compared to Devices 1 and 2. However, due to higher out-of-plane misalignment, Device 3 shows the highest loss nonetheless, in line with the previously discussed simulation results. A better loss profile thus requires right angled sidewalls along with precise vertical waveguide alignment.
The micropositioner demonstrated a flat transmission over the entire 1555-1620 nm wavelength range. For wavelengths below 1555 nm, the transmission was limited by the bandwidth of the surface grating coupler. The maximum wavelength was limited by our tunable laser. With over 70 nm of experimental optical transmission bandwidth, we consider that our device is broadband. Furthermore, since the waveguides are made of silicon nitride and silicon dioxide and that the geometry of the waveguides (i.e., the dimensions of the core and the thickness of the cladding) can be modified without changing the actuators, our design could be readily adapted to operate anywhere from the visible to the mid-infrared. The lower limit is set by the transparency of silicon nitride while the upper one is defined by the absorption of the silicon dioxide cladding. With the geometry used in the prototypes, the waveguide is single mode for wavelengths larger than 1425 nm. Therefore, with proper gratings or if light was launched onto the chip through edge coupling, the devices presented in this manuscript could operate from 1425 nm until approximately 2000nm, where losses due to absorption by the cladding would become significant.
Despite the high thermal stability of the SiN used to build the waveguides, electrothermal actuators generate heat, and temperature is expected to impact the waveguides characteristics, such as their effective refractive index. Nevertheless, since the micropositioner consists of a simple waveguide, variations in temperature will not affect its optical response.
Table 3 summarizes works involving active chips, such as laser diodes and semiconductor optical amplifiers (SOAs) that are integrated with silicon photonics. The comparison is focused on the resulting misalignment in the lateral direction (x), in the light propagation direction (y) and in the out-of-plane direction (z), in addition to the minimum insertion loss achieved.
Table 3. Comparison of state-of-art works using different integration techniques for active chip integration with silicon and silicon nitride photonics
The table shows that photonic wire bonding (PWB) is a promising technology to avoid misalignment related issues, where in [17,18], optical insertion losses were reduced to sub 1-dB levels. Nonetheless, the production in volumes of this process still requires further optimization [48], and its accessibility is still limited. Moreover, this technique does not provide post-assembly modification of the alignment profile or configuration. Monolithic integration shown in [45] resolves the misalignment issues due to the local growth of laser sources on silicon platforms. However, monolithic integration requires a tradeoff in material properties to cater for both III-V devices and silicon passive devices [49]. In addition, III-V devices grown on silicon will have a shorter lifetime due to defects caused by lattice mismatch between the grown III-V devices and the silicon substrates [50]. Thus, typically for best performance, III-V devices are fabricated separately and then integrated onto the silicon substrate using flip-chip bonding, as demonstrated in [44], [46] and [47]. Our proposed MEMS devices that demonstrate silicon waveguide-waveguide misalignments of 0 µm, 0.63 µm and 0.345 µm in x, y and z directions, respectively, and an insertion loss of −1.60 ± 0.06 dB, represent a promising solution for a feasible alignment system that could be performed not only during the chips integration, but also throughout its operation, allowing for readjustments or reconfiguration. Nevertheless, the current demonstration was performed with waveguides on the same chip, which avoids the challenges related to positioning the active chip. Creating a cavity to integrate the active chip on the SiN PIC will be the next step.
With regard to MEMS positioners for chip-to-chip alignment, the research field remains limited. Only a few studies were reported about active components alignment. For instance, in [51], MEMS electrothermal actuators were used to actuate flexible suspended waveguides to align an InP active chip with a silicon chip on a common carrier. The silicon chip was comprised of a 16 µm-thick stack of materials of the TriPleX platform (silicon dioxide-silicon nitride- silicon dioxide on a silicon substrate). A pick-and-place machine was used to make the initial coarse alignment, whereas the MEMS actuators were activated to fine tune the alignment. A maximum vertical deflection of 18.5 µm at a consumed power of 130 mW (12 V) was achieved. In another study [52], two-axis in-plane electrothermal actuators were used to align a discrete laser diode, flip-chip-bonded to a fiber placed inside a v-groove on a silicon substrate. A displacement of 50 µm was achieved at a voltage of 25 V with a positioning resolution of 0.1 µm. With the great success of MEMS devices in several PICs components, including tunable couplers [53], resonators [54], optical switches [38] and movable mirrors [55], MEMS actuators in combination with low loss silicon based waveguides, are expected to pave a way for a dynamic and low cost alignment method.
7. Conclusions
This work reported MEMS waveguide positioners proposed for alignment of active chips with passive silicon nitride photonic chips. The proof of concept was tested by aligning suspended and fixed silicon nitride waveguides on the same SOI substrate. The waveguides were composed of a stack of oxide-nitride-oxide deposited on top of a 59 µm device layer. Three devices were mechanically and optically tested, and the results were compared to FEA and 3-D FDTD simulations carried out with CoventorWare and Lumerical, respectively. Results showed that with 189 mW of actuation power, an initial gap of 6.92 µm between the suspended and fixed waveguides could be closed to 0.63 µm with an out-of-plane misalignment of 345 nm. When the gap is closed, an average insertion loss of −1.60 ± 0.06 dB was measured in the wavelength range between 1550 to 1620 nm. With 196 mW of actuation power, the suspended waveguides were laterally misaligned by ±2 µm, and when the gap is closed they provide an attenuation of up to 20 dB. Thus, the positioner can operate as an attenuator or an on/off switch. The positioner has two suspended output waveguides separated by 10 µm, and less than −40 dB of crosstalk was observed between the waveguides. With the great performance achieved, the next step will be to incorporate an out-of-plane actuator to compensate the out-of-plane misalignment, to realize the actual application of chip-to-chip alignment.
Funding
AEPONYX inc (CRDPJ 530551 - 18); PRIMA Quebec (R16-46-002 PSO); Natural Sciences and Engineering Research Council of Canada (CRDPJ 530551 - 18).
Acknowledgments
The authors would like to thank AEPONYX Inc. for access to their test facilities, device fabrication, and technical support.
Disclosures
AASR: AEPONYX (F,P), SS: AEPONYX (F,P), JP: AEPONYX (F), MM: AEPONYX (F,P), FN: AEPONYX (F,P)
Data availability
Data is available upon reasonable request.
References
1. P. Kaur, A. Boes, G. Ren, T. G. Nguyen, G. Roelkens, and A. Mitchell, “Hybrid and heterogeneous photonic integration,” APL Photonics 6(6), 061102 (2021). [CrossRef]
2. L. Wang, H. Peng, L. Zheng, H. Chen, Y. Zhang, J. Huang, X. Zhang, X. Feng, R. Wei, S. Wang, and M. Zhu, “Broadband and CMOS-compatible polarization splitter and rotator built on a silicon nitride-on-silicon multilayer platform,” Appl. Opt. 62(4), 1046–1056 (2023). [CrossRef]
3. M. Lelit, M. Słowikowski, M. Filipiak, M. Juchniewicz, B. Stonio, B. Michalak, K. Pavłov, M. Myśliwiec, P. Wiśniewski, A. Kaźmierczak, K. Anders, S. Stopiński, R. B. Beck, and R. Piramidowicz, “Passive Photonic Integrated Circuits Elements Fabricated on a Silicon Nitride Platform,” Materials 15(4), 1398 (2022). [CrossRef]
4. J. A. Smith, H. Francis, G. Navickaite, and M. J. Strain, “SiN foundry platform for high performance visible light integrated photonics,” Opt. Mater. Express 13(2), 458–468 (2023). [CrossRef]
5. Z. Hong, Y. Zhao, Y. Ma, Y. Shi, X. Wang, and X. Chen, “Multi-channel High Power Laser Array Chip for Silicon Photonic Integration,” in 2021 19th International Conference on Optical Communications and Networks (ICOCN), (2021), 1–3.
6. Z. Yao, K. Wu, B. X. Tan, J. Wang, Y. Li, Y. Zhang, and A. W. Poon, “Integrated Silicon Photonic Microresonators: Emerging Technologies,” IEEE J. Select. Topics Quantum Electron. 24(1), 1–11 (2018). [CrossRef]
7. H. Yada, N. Kamada, Y. Onuki, X. Han, G. K. Periyanayagam, K. Uchida, H. Sugiyama, M. Aikawa, N. Hayasaka, and K. Shimomura, “Successful Fabrication of GaInAsP Ridge Waveguide Laser Diode Using Hydrophilic Bonded InP/Si Substrate,” in 2018 Conference on Lasers and Electro-Optics Pacific Rim (CLEO-PR), (2018), 1–2.
8. J. M. Ramirez, H. Elfaiki, T. Verolet, et al., “III-V-on-Silicon Integration: From Hybrid Devices to Heterogeneous Photonic Integrated Circuits,” IEEE J. Select. Topics Quantum Electron. 26(2), 1–13 (2020). [CrossRef]
9. L. Di and E. B. John, Recent Progress in Heterogeneous III-V-on-Silicon Photonic Integration (2021), Vol. 2.
10. J. M. Ramírez, P. Fanneau de la Horie, J.-G. Provost, S. Malhouitre, D. Néel, C. Jany, C. Besancon, N. Vaissière, J. Decobert, K. Hassan, and D. Bitauld, “Low-Threshold, High-Power On-Chip Tunable III-V/Si Lasers with Integrated Semiconductor Optical Amplifiers,” Appl. Sci. 11(23), 11096 (2021). [CrossRef]
11. M. Krainak, M. Stephen, E. Troupaki, et al., Integrated photonics for NASA applications, SPIE LASE (SPIE, 2019), Vol. 10899.
12. A. W. Fang, H. Park, O. Cohen, R. Jones, M. J. Paniccia, and J. E. Bowers, “Electrically pumped hybrid AlGaInAs-silicon evanescent laser,” Opt. Express 14(20), 9203–9210 (2006). [CrossRef]
13. J. Zhang, H. Xu, G. Zhang, Y. Chen, H. Wang, K. H. Tan, S. Wicaksono, C. Wang, C. Sun, Q. Kong, C. C. W. Lim, S. F. Yoon, and X. Gong, “First InGaAs/InAlAs Single-Photon Avalanche Diodes (SPADs) Heterogeneously Integrated with Si Photonics on SOI Platform for 1550 nm Detection,” in 2021 Symposium on VLSI Circuits, (2021), 1–2.
14. M. S. Alias, M. Tangi, J. A. Holguín-Lerma, E. Stegenburgs, A. A. Alatawi, I. Ashry, R. C. Subedi, D. Priante, M. K. Shakfa, T. K. Ng, and B. S. Ooi, “Review of nanophotonics approaches using nanostructures and nanofabrication for III-nitrides ultraviolet-photonic devices,” J. Nanophotonics 12(04), 1 (2018). [CrossRef]
15. E.-H. Lee, S.-G. Lee, B. H. O, and S. G. Park, Integration of micro-/nano-/quantum-scale photonic devices: scientific and technological considerations, Photonics Europe (SPIE, 2004), Vol. 5451.
16. D. Fitsios, T. Alexoudi, G. T. Kanellos, K. Vyrsokinos, N. Pleros, T. Tekin, M. Cherchi, S. Ylinen, M. Harjanne, M. Kapulainen, and T. Aalto, “Dual SOA-MZI Wavelength Converters Based on III-V Hybrid Integration on a µm-Scale Si Platform,” IEEE Photonics Technol. Lett. 26(6), 560–563 (2014). [CrossRef]
17. M. Blaicher, M. R. Billah, J. Kemal, T. Hoose, P. Marin-Palomo, A. Hofmann, Y. Kutuvantavida, C. Kieninger, P.-I. Dietrich, M. Lauermann, S. Wolf, U. Troppenz, M. Moehrle, F. Merget, S. Skacel, J. Witzens, S. Randel, W. Freude, and C. Koos, “Hybrid multi-chip assembly of optical communication engines by in situ 3D nano-lithography,” Light: Sci. Appl. 9(1), 71 (2020). [CrossRef]
18. M. R. Billah, M. Blaicher, T. Hoose, P.-I. Dietrich, P. Marin-Palomo, N. Lindenmann, A. Nesic, A. Hofmann, U. Troppenz, M. Moehrle, S. Randel, W. Freude, and C. Koos, “Hybrid integration of silicon photonics circuits and InP lasers by photonic wire bonding,” Optica 5(7), 876–883 (2018). [CrossRef]
19. Q. Qiao, M. S. Yazici, B. Dong, X. Liu, C. Lee, and G. Zhou, “Multifunctional mid-infrared photonic switch using a MEMS-based tunable waveguide coupler,” Opt. Lett. 45(19), 5620–5623 (2020). [CrossRef]
20. S. Sharma, N. Kohli, J. Brière, M. Ménard, and F. Nabki, “Translational MEMS Platform for Planar Optical Switching Fabrics,” Micromachines 10(7), 435 (2019). [CrossRef]
21. J. Briere, M. Y. Elsayed, M. Saidani, M. Bérard, P.-O. Beaulieu, H. Rabbani-Haghighi, F. Nabki, and M. Ménard, “Rotating Circular Micro-Platform with Integrated Waveguides and Latching Arm for Reconfigurable Integrated Optics,” Micromachines 8(12), 354 (2017). [CrossRef]
22. B. Barazani, A. Gascon, C. Coia, F. Nabki, and M. Ménard, “Broadband 1 × 4 Silicon Nitride Photonic Switch Fabricated on a Hybrid Electrothermal and Electrostatic MEMS Platform,” Journal of Lightwave Technology, 1–14 (2023).
23. Q. Qiao, H. Sun, X. Liu, B. Dong, J. Xia, C. Lee, and G. Zhou, “Suspended Silicon Waveguide with Sub-Wavelength Grating Cladding for Optical MEMS in Mid-Infrared,” Micromachines 12(11), 1311 (2021). [CrossRef]
24. C. V. Poulton, M. J. Byrd, P. Russo, E. Timurdogan, M. Khandaker, D. Vermeulen, and M. R. Watts, “Long-Range LiDAR and Free-Space Data Communication With High-Performance Optical Phased Arrays,” IEEE J. Sel. Top. Quantum Electron. 25(5), 1–8 (2019). [CrossRef]
25. Z. K. Bishop, A. P. Foster, B. Royall, C. Bentham, E. Clarke, M. S. Skolnick, and L. R. Wilson, “Electro-mechanical control of an on-chip optical beam splitter containing an embedded quantum emitter,” Opt. Lett. 43(9), 2142–2145 (2018). [CrossRef]
26. A. Y. Takabayashi, H. Sattari, P. Edinger, P. Verheyen, K. B. Gylfason, W. Bogaerts, and N. Quack, “Continuously Tunable Silicon Photonic MEMS $2 \times 2$ Power Coupler,” in 2021 21st International Conference on Solid-State Sensors, Actuators and Microsystems (Transducers), (2021), 447–450.
27. X. Liu, Q. Qiao, B. Dong, W. Liu, C. Xu, S. Xu, and G. Zhou, “MEMS enabled suspended silicon waveguide platform for long-wave infrared modulation applications,” Int. J. Optomechatron. 16(1), 42–57 (2022). [CrossRef]
28. K. Wu and M. Tichem, “Post-Release Deformation and Motion Control of Photonic Waveguide Beams by Tuneable Electrothermal Actuators in Thick SiO2,” Micromachines 9(10), 496 (2018). [CrossRef]
29. X. Liu, W. Liu, Z. Ren, Y. Ma, B. Dong, G. Zhou, and C. Lee, “Progress of optomechanical micro/nano sensors: a review,” Int. J. Optomechatronics 15(1), 120–159 (2021). [CrossRef]
30. W. Bogaerts and A. Rahim, “Programmable Photonics: An Opportunity for an Accessible Large-Volume PIC Ecosystem,” IEEE J. Select. Topics Quantum Electron. 26(5), 1–17 (2020). [CrossRef]
31. A. Y. Takabayashi, H. Sattari, P. Edinger, P. Verheyen, K. B. Gylfason, W. Bogaerts, and N. Quack, “Broadband Compact Single-Pole Double-Throw Silicon Photonic MEMS Switch,” J. Microelectromechanical Syst. 30(2), 322–329 (2021). [CrossRef]
32. N. Quack, H. Sattari, A. Y. Takabayashi, Y. Zhang, P. Verheyen, W. Bogaerts, P. Edinger, C. Errando-Herranz, and K. B. Gylfason, “MEMS-Enabled Silicon Photonic Integrated Devices and Circuits,” IEEE J. Quantum Electron. 56(1), 1–10 (2020). [CrossRef]
33. P. Edinger, A. Y. Takabayashi, C. Errando-Herranz, U. Khan, H. Sattari, P. Verheyen, W. Bogaerts, N. Quack, and K. B. Gylfason, “Silicon photonic microelectromechanical phase shifters for scalable programmable photonics,” Opt. Lett. 46(22), 5671–5674 (2021). [CrossRef]
34. L. Van Iseghem, E. Picavet, A. Y. Takabayashi, P. Edinger, U. Khan, P. Verheyen, N. Quack, K. B. Gylfason, K. De Buysser, J. Beeckman, and W. Bogaerts, “Low power optical phase shifter using liquid crystal actuation on a silicon photonics platform,” Opt. Mater. Express 12(6), 2181–2198 (2022). [CrossRef]
35. H. Sattari, A. Y. Takabayashi, Y. Zhang, P. Verheyen, W. Bogaerts, and N. Quack, “Compact broadband suspended silicon photonic directional coupler,” Opt. Lett. 45(11), 2997–3000 (2020). [CrossRef]
36. C. Errando-Herranz, N. Le Thomas, and K. B. Gylfason, “Low-power optical beam steering by microelectromechanical waveguide gratings,” Opt. Lett. 44(4), 855–858 (2019). [CrossRef]
37. G. Zhou, Z. H. Lim, Y. Qi, F. S. Chau, and G. Zhou, “MEMS gratings and their applications,” Int. J. Optomechatronics 15(1), 61–86 (2021). [CrossRef]
38. S. Sharma, N. Kohli, J. Brière, F. Nabki, and M. Ménard, “Integrated 1x 3 MEMS silicon nitride photonics switch,” Opt. Express 30(12), 22200–22220 (2022). [CrossRef]
39. T.-H. Chang, X. Zhou, H. Tamura, and C.-L. Hung, “Realization of efficient 3D tapered waveguide-to-fiber couplers on a nanophotonic circuit,” Opt. Express 30(18), 31643–31652 (2022). [CrossRef]
40. D. Chatzitheocharis, D. Ketzaki, C. Calò, C. Caillaud, and K. Vyrsokinos, “Design of Si-rich nitride interposer waveguides for efficient light coupling from InP-based QD-emitters to Si3N4 waveguides on a silicon substrate,” Opt. Express 28(23), 34219–34236 (2020). [CrossRef]
41. Y. Zhang, S. Liu, W. Zhai, C. Peng, Z. Wang, J. Feng, and J. Guo, “Reconfigurable spot size converter for the silicon photonics integrated circuit,” Opt. Express 29(23), 37703–37711 (2021). [CrossRef]
42. R. Marchetti, C. Lacava, L. Carroll, K. Gradkowski, and P. Minzioni, “Coupling strategies for silicon photonics integrated chips [Invited],” Photonics Res. 7(2), 201–239 (2019). [CrossRef]
43. A. Brenes, B. Vysotskyi, E. Lefeuvre, and J. Juillard, “Reliability of etching angle estimation of MEMS resonators from electrical measurements,” in 2019 Symposium on Design, Test, Integration & Packaging of MEMS and MOEMS (DTIP), (2019), 1–4.
44. M. Theurer, M. Moehrle, A. Sigmund, K. O. Velthaus, R. M. Oldenbeuving, L. Wevers, F. M. Postma, R. Mateman, F. Schreuder, D. Geskus, K. Wörhoff, R. Dekker, R. G. Heideman, and M. Schell, “Flip-Chip Integration of InP to SiN Photonic Integrated Circuits,” J. Lightwave Technol. 38(9), 2630–2636 (2020). [CrossRef]
45. Y. Yang, H. Zhao, X. Ren, and Y. Huang, “Monolithic integration of laser onto multilayer silicon nitride photonic integrated circuits with high efficiency at telecom wavelength,” Opt. Express 29(18), 28912–28923 (2021). [CrossRef]
46. N. Hatori, T. Shimizu, M. Okano, M. Ishizaka, T. Yamamoto, Y. Urino, M. Mori, T. Nakamura, and Y. Arakawa, “A Hybrid Integrated Light Source on a Silicon Platform Using a Trident Spot-Size Converter,” J. Lightwave Technol. 32(7), 1329–1336 (2014). [CrossRef]
47. T. Matsumoto, T. Kurahashi, R. Konoike, K. Suzuki, K. Tanizawa, A. Uetake, K. Takabayashi, K. Ikeda, H. Kawashima, S. Akiyama, and S. Sekiguchi, “Hybrid-Integration of SOA on Silicon Photonics Platform Based on Flip-Chip Bonding,” J. Lightwave Technol. 37(2), 307–313 (2019). [CrossRef]
48. N. Lindenmann, “Photonic Wire Bonding as a Novel Technology for Photonic Chip Interfaces,” (Karlsruhe Institute of Technology (KIT), 2018).
49. F. Chollet, “Devices Based on Co-Integrated MEMS Actuators and Optical Waveguide: A Review,” Micromachines 7(2), 18 (2016). [CrossRef]
50. D. A. Miller, Dense two-dimensional integration of optoelectronics and electronics for interconnections, Optoelectronics and High-Power Lasers and Applications (SPIE, 1998), Vol. 10292.
51. K. Wu, T.-J. Peters, M. Tichem, F. Postma, A. Prak, K. Worhoff, and A. Leinse, “Bimorph actuators in thick SiO 2 for photonic alignment,” Proc. SPIE 9753, 975311 (2016). [CrossRef]
52. Q. X. Zhang, Y. Du, C. W. Tan, J. Zhang, M. B. Yu, W. G. Yeoh, G. Lo, and D. Kwong, “A Silicon Platform With MEMS Active Alignment Function and Its Potential Application in Si-Photonics Packaging,” IEEE J. Sel. Top. Quantum Electron. 16(1), 267–275 (2010). [CrossRef]
53. M. W. Pruessner, B. J. Roxworthy, D. A. Kozak, N. F. Tyndall, W. S. Rabinovich, and T. H. Stievater, “On-Chip Mode Conversion with MEMS-Tunable Phase-Matching,” in Optica Advanced Photonics Congress 2022, Technical Digest Series (Optica Publishing Group, 2022), IM3B.8.
54. T. Ikeda and K. Hane, “A microelectromechanically tunable microring resonator composed of freestanding silicon photonic waveguide couplers,” Appl. Phys. Lett. 102(22), 221113 (2013). [CrossRef]
55. D. Sadhukhan and G. Prasad Singh, “Design of electrostatic actuated MEMS biaxial scanning micro-mirror with serpentine structure,” Mater. Today: Proc. 65(Part 1), 229–234 (2022). [CrossRef]
