We present the design, fabrication, and characterization results of a compact, widely tunable laser realized on an indium phosphide membrane-on-silicon (IMOS) platform. The laser features a compact Mach–Zehnder interferometric structure as the wavelength-selective intracavity filter with a footprint of 0.13 mm2. The filter design is optimized to ensure narrow filter transmission and high side-mode-to-main-mode-ratio, enabling single-mode operation for the laser. The high optical confinement on the IMOS platform can support tight waveguide bends. Leveraging this, the laser achieves a short cavity length, further enhancing the single-mode operation. Measurement results indicate a threshold current of 29 mA and a maximum on-chip output power of approximately 3.6 dBm and wall plug efficiency of 1.8%. The side-mode suppression ratio ranges from 30 to 44 dB, with a tuning range spanning 40 nm, from 1555 to 1595 nm. A complete tuning lookup table is generated via an automated setup incorporating a stochastic search algorithm.

Widely tunable lasers serve as crucial optical sources across various applications, such as dense wavelength division multiplexing,1 optical coherence tomography,2 light-based detection and ranging (LiDAR) systems,3 and gas sensing.4 Some of these applications demand compact lasers and high-density integration to pave the way for further integration with other components. For example, in LiDAR applications, the optical phased array (OPA)-based solutions hinge on the integration of tunable lasers, semiconductor optical amplifiers, and phase modulators to develop fast, high-resolution, and high-power on-chip beam steerers. Likewise, the integration of a laser, optical phased array, and photodetector on a chip facilitates the miniaturization of gas sensing systems.4 

Tunable lasers realized on photonic integrated circuits (PICs), specifically on the monolithic InP platform,5,6 are a great candidate for such applications due to its native gain medium and mature active-passive fabrication process to combine components with different bandgaps. Various configurations of widely tunable lasers have been demonstrated, with high output power devices (>10 mW), wide tuning range (>30 nm), and good side-mode suppression ratio (SMSR) values (>30 dB). State-of-the art includes sampled-grating distributed-Bragg-reflector (SGDBR) tunable lasers,7 digital supermode DBR grating (DSDBR),8 Y-branch with gratings,9 Vernier ring resonators,10 and asymmetric Mach–Zehnder interferometric (MZI) filter with nested11 and parallelized design configurations.12 However, the low index contrast in the monolithic InP platform prevents the achievement of low bend radii for waveguides, which presents a limitation in realizing compact devices and further increase in integration density with additional circuitry. The silicon on insulator or SOI platform stands as another mature and established PIC platform, providing a high refractive index contrast (Δn ≈ 2) that enables compact circuitry.13 To realize lasers, heterogeneous integration of III–V materials via molecular wafer bonding has been used to achieve on-chip gain functionality.14 With this, the hybrid Si platform has showcased widely tunable lasers with a tuning range of 40 nm and SMSR exceeding 35 dB.15 However, it becomes more complex in integrating lasers alongside other active components such as photodetectors and high-speed modulators.

A promising solution for achieving device miniaturization and high integration density16 is wafer-bonding a few-micron-thin InP membrane onto a Si substrate.17–19, Figure 1 shows an example of such an approach. This method combines the native gain of III–V materials with the high confinement properties similar to silicon-on-insulator (SOI). All functionalities, both active and passive, occur within the monolithic InP layer, which is sandwiched between low-index materials. Therefore, this configuration results in high confinement and consequently, compact devices. This platform also benefits from the highly efficient modulators available on InP, which offer significant advantages over those on Si platforms.20 While state-of-the-art on the membrane platforms have previously demonstrated low threshold, high speed, and energy-efficient single-wavelength lasers,19,21,22 there has been limited progress in developing widely tunable laser. Taking advantage of the high optical confinement, different demonstrations of compact widely tunable lasers have recently been presented on the InP membrane-on-Si (IMOS) platform.17 Vernier ring laser achieving a tuning range of 25 nm and an SMSR of over 30 dB was demonstrated.23 More recently, this range was extended to 50 nm using an asymmetrical nested MZI-type structure.24 Here, we propose a filter design and present fabrication and measurement results incorporating a parallelized MZI intracavity structure. The filter design is very compact with a footprint of 0.13 mm2. This design outperforms previous IMOS demonstrations in maximum continuous-wave (CW) output power, wall-plug efficiency, and device footprint and demonstrates comparable performance metrics to lasers based on Si-InP and monolithic platform. The laser reaches a maximum CW power of 3.6 dBm (on-chip) with a threshold current of 29 mA. It supports a robust single-mode operation with a maximum side-mode suppression ratio (SMSR) of 44 dB and a tuning range of 40 nm, spanning 1555–1595 nm and encompassing critical sections of the C and L communication bands. Enhancements in filter design have significantly reduced the required chip area at a footprint of 0.49 mm2. This design demonstrates a size reduction of one-fourth compared to equivalent designs on the InP monolithic platform,25,26 half compared to the nested design,24 and one-third compared to the Vernier ring design on the IMOS platform.23 It also offers half the size in comparison to ring resonator-based designs on heterogeneous III–V on Si platforms.27 The laser's wall-plug efficiency of 1.8% marks a significant improvement over previous designs on the IMOS platform, which achieved efficiencies of 0.23%23 and 0.35%.24 These efficiency results are comparable to those of monolithic designs at 1.9%25 and III–V on Si platforms at 1.9% and 0.26%.27 The full tuning mapping of the laser was performed using an automated optimization method. The paper is organized as follows: the design and tuning mechanism of the laser is detailed; the fabrication process is briefly described; the experimental setup and tuning algorithm are briefly summarized; the tuning characterization and its results are discussed and followed by conclusions of the paper.

FIG. 1.

Schematic cross section of a simplified active-passive InP membrane-on-Si (IMOS) configuration. Image from Ref. 17.

FIG. 1.

Schematic cross section of a simplified active-passive InP membrane-on-Si (IMOS) configuration. Image from Ref. 17.

Close modal

A monolithic tunable laser design necessitates three primary components: an (active) gain medium, mirrors or equivalent for feedback, and optionally, a cavity phase section and a wavelength-selective filter. The design schematic is shown in Fig. 2(a). A 500-μm long 4-quantum-well semiconductor optical amplifier (SOA) section acts as the gain section, cascaded with a parallelized Mach–Zehnder interferometer (MZI) intracavity wavelength selection filter. The laser cavity is established by positioning the SOA section and the intracavity filter within broadband photonic crystal (PC) mirrors with a bandwidth exceeding 100 nm.28 The chosen mirror reflectivity is 0.82 for the front mirror and 0.99 for the back mirror. To extract light from the PIC, a focusing grating coupler is used. The peaks of the SOA gain and the grating coupler spectral envelope coincide around 1560 nm. The intracavity filter is split into four arms using a low-reflection butterfly multi-mode interferometers (MMIs) tree with a reflection level of −35 dB.29 The footprint of the filter is 0.28 × 0.45 mm2, with the resulting footprint of laser being 1.11 × 0.45 mm2.

FIG. 2.

(a) Schematic of a laser design with MZI intracavity filter marked by a gray dashed line. (b) The transmission spectra of the designed filter in blue solid lines, overlapped with gain bandwidth28 in cyan. The dashed red line depicts a linear increase for arm length difference ΔL for a four-arm configuration.

FIG. 2.

(a) Schematic of a laser design with MZI intracavity filter marked by a gray dashed line. (b) The transmission spectra of the designed filter in blue solid lines, overlapped with gain bandwidth28 in cyan. The dashed red line depicts a linear increase for arm length difference ΔL for a four-arm configuration.

Close modal
The wavelength selection filter comprises four arms, each containing thermo-optic phase shifters (TOPSs) with a length of 220 μm, resulting in a round trip 2π phase shift for electrical power of 7–8 mW. The electrical signal applied to the TOPSs serve to individually adjust the effective index of each arm. With this adjustment, constructive interference of fields in the four arms of the filter for a specific wavelength is achieved and, consequently, the lasing wavelength is selected. Each arm has an added waveguide section of N·ΔL that is increasing linearly as a multiple of the arm number N of [0, 1, 2, 3]. This results in a filter transmission with a sinc profile, as shown in red dashed line in Fig. 2(b). The width of the filter transmission is determined by the number of arms, and the free spectral range (FSR) is determined by the value of ΔL. Increasing the number of arms allows for a narrower filter, albeit with a larger circuit and more complicated tuning requirements.30 The optimization of the values of ΔL was performed through a constrained search, which could output only combinations made of multiple integers of ΔL. This choice turns into a direct benefit for the laser tuning scheme if the control is performed by means of a digital logic (i.e., only discrete bias values can be provided). This is detailed in Ref. 31. The best trade-off between a reduced 3-dB bandwidth for the mainlobe and a filter sidelobe suppression of more than 20% was obtained for the configuration N·ΔL, where N = [0, 1, 11, 13]. From this, the waveguide arm length differences were 8.24, 90.64, and 107.12 μm. The resulting shape with the optimized increase in ΔL is depicted in Fig. 2(b) with blue solid lines. The relevant performance metric of the filter to the SMSR of the laser is the side mode to main mode transmission ratio (SMTR). The following equation determines the required filter SMTR to achieve a certain laser SMSR:32 
(1)
where r1 and r2 are the reflectance values of the front and back mirrors, Pout is the output power, and Psp the spontaneous emission power coupled to the lasing mode. With reflectance values of 0.82 and 0.99, respectively, and values for output power Pout = 3 mW and spontaneous emission power into the lasing mode Psp = 0.5 μW. The SMTR value achieved from the filter simulation is 0.06, which gives an SMSR value exceeding 30 dB.

The tunable laser is fabricated with a double-sided process to realize active and passive components in the same process.16 The cross sections of components are as drawn in Fig. 3(a). The active part (gain section) is an S-cross-sectioned SOA, with its two sidewalls etched on different sides of the membrane.33 Metallization of the SOA contacts (M0) is done before bonding, and the metal pads are opened by wet-etching after wafer flipping and bonding to the Si wafer with BCB. Active-passive transition is achieved through evanescent coupling between the core and the i-InP waveguide by tapering the SOA claddings and the core.33 The passive components including the waveguide, the MMI, the photonic crystal reflector, the grating couplers, and the TOPS are fabricated after wafer bonding, yielding a cross-sectional structure resembling that of silicon-on-insulator (SOI) photonics. Therefore, a similar high optical confinement is achieved with a miniature waveguide cross section of 400 × 300 nm2 (W × H). The grating couplers are realized subsequently on top of the passive waveguide with a shallow 120-nm etch into the i-InP layer. After waveguide fabrication, the wafer is planarized using polyimide (PI), on top of which the final metallization (M1) is deposited. Through-polymer vias (TPV) with sloped sidewalls are utilized here to form electrical connections between M0 and M1. The heating element is also fabricated in M1 for the TOPS, as seen in Fig. 3(a). A microscope image of the final fabricated tunable laser is shown in Fig. 3(b).

FIG. 3.

(a) Cross-sectional structure of the passive waveguides, SOA, and TOPS. (b) Microscope image of a fabricated tunable laser.

FIG. 3.

(a) Cross-sectional structure of the passive waveguides, SOA, and TOPS. (b) Microscope image of a fabricated tunable laser.

Close modal

The measurement setup is shown in Fig. 4. The DUT was placed on a copper chuck held-on by a vacuum suction. The chuck has a water cooler connection, a thermistor, and a Peltier cell to ensure accurate temperature control. The temperature for measurements was set at 10 °C. The fiber-coupled output of the laser goes to a 90–10 splitter: 10% toward a tunable bandpass optical filter (EXFO-XTA50) coupled with power meter (Agilent 8614B) to be used as wavelength meter and 90% toward an optical spectrum analyzer (Yokogawa AQ6375) to collect the full spectral response of the DUT. Current is applied using multi-channel current sources (Keithley 2600) to tune the intracavity filter over the desired wavelength range, and the SOA current is supplied from a current source (PRO8000).

FIG. 4.

Experimental setup for an automated tuning measurement. Output from the laser DUT is split 10% into a bandpass filter onto a power meter and 90% into an optical spectrum analyzer.

FIG. 4.

Experimental setup for an automated tuning measurement. Output from the laser DUT is split 10% into a bandpass filter onto a power meter and 90% into an optical spectrum analyzer.

Close modal

The best way to achieve a reliable tuning scheme is to involve an automated optimization method. This way, the TOPSs can be operated to align the intracavity filter to a target wavelength avoiding the issue of stagnation into a local maxima, which could be the result of the sequential alignment of the TOPSs one after the other. The stochastic search optimizer involved in this work is based on the Globalized Bounded Nelder–Mead method.34 It was set to scatter (with a random initial condition) bias current configurations to the TOPS controllers of the intracavity filter within a four-dimensional search domain restricted to [0, 20] mA. The optimization routine is fed back by a power meter, which collects light through the external tunable filter. The target function to optimize is then optical power of the DUT at the wavelength the external tunable filter is set to. The fine-tuning of the algorithm allowed it to converge within 200 iterations with a tolerance of 10−2 on the absolute error of the target function. For a single wavelength point, the search was repeated three times supplying every time a different initial condition to limit the probability of the optimizer being trapped into local maxima. This procedure has been repeated over multiple instances for different wavelengths, eventually leading to the completion of the full tuning lookup-table of the laser.

The device's light-current-voltage (LIV) characteristics are shown in Fig. 5(a). The IV curve shows a turn-on voltage of 0.7 V. The LIV, which was taken after optimized filter settings, exhibits a maximum in-fiber output power of −3.9 dBm, meaning 3.6 dBm on-chip (with 7.5 dB of coupling loss). This indicates a wall plug efficiency of 1.8%. The threshold current is identified to be 29 mA. This threshold current is significantly lower than previously measured tunable lasers with a nested MZI configuration on this platform.24 The nested configuration had a threshold current of 105 mA attributed to added losses from the longer cavity and the six MMI couplers used. Overall internal losses are reduced in the presented design because of the parallelized design with two MMI couplers and shorter cavity resulting in less propagation losses.

FIG. 5.

(a) LIV curve of the laser. (b) Overlap of multiple spectra recorded on the OSA at a resolution of 0.01 nm, depicting a 40 nm tuning range.

FIG. 5.

(a) LIV curve of the laser. (b) Overlap of multiple spectra recorded on the OSA at a resolution of 0.01 nm, depicting a 40 nm tuning range.

Close modal

The first step to build up the tuning lookup-table involves an initial optimization search, targeting a wavelength around the gain peak of 1560 nm. During this step, two main tasks are performed. First, a stochastic search comprising only three of the TOPSs of the intra-cavity filter is run. This operation yields a bias configuration for which the intra-cavity filter peak is aligned to the initial target wavelength. The reason why not all four TOPSs are operated is because the filter working principle withstands phase difference relations among the MZI arms and not absolute phases. Therefore, we can keep one of the TOPSs fixed, which acts as a phase reference for the other arms. Second, the electrical power required to switch the TOPs by π, denoted as Pπ, for all the arms is measured by sweeping the bias current on each TOP from 0 mA to the current limit of 20 mA. The measured values of current to reach Pπ of the arms are in the range 17–18 mA. These values are used as boundary conditions for the next optimization searches, taken at steps of 0.5 nm from the previous one. Still, such a condition does not ensure the capability to hit any possible wavelength along the laser tuning range. This is because the intracavity filter will span over a full FSR only when the relative phase in the arms is swept multiple times from 0 to 2π.31 Selective results of different optimization searches are shown in Fig. 5(b), which shows a tuning range of 40 nm. Figure 6(a) illustrates the results of the full tuning map, with the target peak wavelengths determined by the algorithm alongside the achieved peak wavelength, with points obtained at a step size of 0.375 nm. While the obtained wavelengths mostly align with the target wavelengths, some gaps are apparent, particularly in the range of 1565–1568 and 1578–1583 nm, occurring at a period of 12.9 nm. Figure 6(b) exhibits the amplified spontaneous emission (ASE) at a current density of 1.5 kA/cm2, which illustrates a super-modulation with an FSR of 12.9 nm. This suggests the presence of a cavity at 25.4 μm, which is identified to be at the output. This indicates that residual reflections from the grating coupler and the reflective PC mirror create an extra cavity, thus creating this super-modulation in the spectrum, and as a result limiting the tuning range and leaving gaps in the spectrum. The kinks observed on the LIV curve at the specified current values are attributed to super-modulation in the laser cavity, caused by the grating coupler envelope, which leads to parasitic reflections within the laser cavity. These issues can be mitigated by fabricating grating couplers with reduced reflections.35–37 

FIG. 6.

(a) Target peak wavelength and wavelength attained with search algorithm, with SOA current set at 50 mA. (b) Amplified sponteneous emmission (ASE) spectra at 15 mA SOA current. (c) Comparison of SMSR values from a linear filter design and an optimized filter design. The box plots represent subsets of SMSR data clustered over steps of 5 nm, according to the x-axis scale. Each box carries the average value, the standard deviation, and the maximum data spread for the different laser design (in red, the optimized filter design; and in blue, the other design). On average, each box represents a cluster of 10 or more SMSR values.

FIG. 6.

(a) Target peak wavelength and wavelength attained with search algorithm, with SOA current set at 50 mA. (b) Amplified sponteneous emmission (ASE) spectra at 15 mA SOA current. (c) Comparison of SMSR values from a linear filter design and an optimized filter design. The box plots represent subsets of SMSR data clustered over steps of 5 nm, according to the x-axis scale. Each box carries the average value, the standard deviation, and the maximum data spread for the different laser design (in red, the optimized filter design; and in blue, the other design). On average, each box represents a cluster of 10 or more SMSR values.

Close modal

In Fig. 6(c), a comparison of SMSR performance is presented, comparing the results of the optimized filter design with the laser embedding a linearly increasing arm-length filter design. The plot clusters the experimental results over steps of 5 nm. The box plots represent the standard deviation for each data subset, and the error bars account for the maximum data spread. The comparison reveals that the optimized design with adjusted lengths (depicted in red) achieves notably higher SMSR values, with a mean value of 32 dB. The maximum SMSR attained with this optimized design reaches 44 dB. Instead, the other design (shown in blue) displays a mean SMSR value of 27 dB, with the maximum SMSR attained with the linear design being 35 dB. This comparison clearly underscores the reason to opt for the optimized design in enabling a single-mode operation.

In this paper, we presented a parallelized Mach–Zehnder interferometer (MZI) intracavity filter-based widely tunable laser on the indium phosphide membrane-on-silicon (IMOS) platform, featuring a compact (0.13 mm2) filter design. The circuit design was optimized to achieve a narrow filter shape without increasing complexity, resulting in excellent side-mode suppression ratio (SMSR) values. Notably, we achieved in-fiber power levels of −3.9 dBm, which translates to 3.6 dBm on-chip, considering 7.5 dB coupling losses. The results demonstrate that this laser, with a threshold current of 2.9 kA/cm2, a wide tuning range of 40 nm, and a mean SMSR value of 32 dB peaking at 44 dB. The tunability range spans significant portions of the C and L bands, covering 1555–1595 nm, which is ultimately limited by gain-bandwidth of SOA and the filter FSR. The filter FSR can be easily increased by decreasing the arm length difference ΔL. This design occupies only one-fourth of the area compared to similar designs on the monolithic InP platform and achieves two times reduction in size compared to lasers presented on the heterogeneous Si/InP platform. The tuning process was fully automated, facilitated by a stochastic search algorithm. This laser shows, for example, potential for gas-sensing applications due to its stable emissions at wavelengths corresponding to the absorption spectra of acetylene, carbon dioxide, and methane.

The authors acknowledge the support from Synergia—SYstem change for New Ecology-based and Resource efficient Growth with high tech In Agriculture, the NWO Zwaartekracht Project Research Centre for Integrated Nanophotonic, and European Union's Horizon 2020 programme under GA No. 899987.

The authors have no conflicts to disclose.

Tasfia Kabir: Conceptualization (equal); Data curation (lead); Formal analysis (lead); Investigation (lead); Methodology (lead); Project administration (lead); Software (equal); Validation (lead); Visualization (lead); Writing – original draft (lead); Writing – review & editing (lead). Yi Wang: Conceptualization (equal); Investigation (equal); Methodology (lead); Resources (lead); Validation (equal); Writing – review & editing (equal). Stefano Tondini: Conceptualization (equal); Investigation (equal); Software (lead); Validation (equal); Visualization (equal); Writing – original draft (equal); Writing – review & editing (equal). Kevin Williams: Funding acquisition (equal); Project administration (equal); Supervision (equal); Writing – review & editing (equal). Yuqing Jiao: Funding acquisition (equal); Investigation (equal); Project administration (equal); Supervision (equal); Validation (equal); Writing – review & editing (equal). Martijn J. R. Heck: Conceptualization (equal); Funding acquisition (lead); Investigation (equal); Project administration (equal); Supervision (lead); Validation (equal); Writing – original draft (equal); Writing – review & editing (equal).

The data that support the findings of this study are available from the corresponding author upon reasonable request.

1.
K.
Tsuzuki
,
Y.
Shibata
,
N.
Kikuchi
,
M.
Ishikawa
,
T.
Yasui
,
H.
Ishii
, and
H.
Yasaka
, “
Full C-band tunable DFB laser array copackaged with InP Mach–Zehnder modulator for DWDM optical communication systems
,”
IEEE J. Sel. Top. Quantum Electron.
15
,
521
527
(
2009
).
2.
V. J.
Srinivasan
,
R.
Huber
,
I.
Gorczynska
,
J. G.
Fujimoto
,
J. Y.
Jiang
,
P.
Reisen
, and
A. E.
Cable
, “
High-speed, high-resolution optical coherence tomography retinal imaging with a frequency-swept laser at 850 nm
,”
Opt. Lett.
32
,
361
363
(
2007
).
3.
M. J. R.
Heck
, “
Highly integrated optical phased arrays: Photonic integrated circuits for optical beam shaping and beam steering
,”
Nanophotonics
6
,
93
107
(
2017
).
4.
A.
Hänsel
and
M. J. R.
Heck
, “
Feasibility of telecom-wavelength photonic integrated circuits for gas sensors
,”
Sensors
18
,
2870
(
2018
).
5.
M.
Smit
,
X.
Leijtens
,
H.
Ambrosius
,
E.
Bente
,
J.
van der Tol
,
B.
Smalbrugge
,
T.
de Vries
,
E.-J.
Geluk
,
J.
Bolk
,
R.
van Veldhoven
,
L.
Augustin
,
P.
Thijs
,
D.
D'Agostino
,
H.
Rabbani
,
K.
Lawniczuk
,
S.
Stopinski
,
S.
Tahvili
,
A.
Corradi
,
E.
Kleijn
, and
D.
Dzibrou
, “
An introduction to InP-based generic integration technology
,”
Semicond. Sci. Technol.
29
(
8
),
083001
(
2014
).
6.
L. A.
Coldren
,
S. C.
Nicholes
,
L. A.
Johansson
,
S.
Ristic
,
R. S.
Guzzon
,
E. J.
Norberg
, and
U.
Krishnamachari
, “
High performance InP-based photonic ICs—A tutorial
,”
J. Lightwave Technol.
29
(
4
),
554
570
(
2011
).
7.
V.
Jayaraman
,
Z.-M.
Chuang
, and
L. A.
Coldren
, “
Theory, design, and performance of extended tuning range semiconductor lasers with sampled gratings
,”
IEEE J. Quantum Electron.
29
,
1824
1834
(
1993
).
8.
A. B.
Ward
,
D. J.
Robbins
,
G.
Busico
,
E. N.
Barton
,
L.
Ponnampalam
,
J. P.
Duck
,
N. D.
Whitbread
,
P.
Williams
,
D.
Reid
,
A. A.
Carter
, and
M. J.
Michael Wale
, “
Widely tunable DS-DBR laser with monolithically integrated SOA: Design and performance
,”
IEEE J. Sel. Top. Quantum Electron.
11
(
1
),
149
156
(
2005
).
9.
J.-O.
Wesstrom
,
S.
Hammerfeldt
,
J.
Buus
,
R.
Siljan
,
R.
Laroy
, and
H.
de Vries
, “
Design of a widely tunable modulated grating Y-branch laser using the additive Vernier effect for improved super-mode selection
,” in
IEEE 18th International Semiconductor Laser Conference
(
IEEE
,
2003
).
10.
B.
Liu
,
A.
Shakouri
, and
J. E.
Bowers
, “
Passive microring-resonator-coupled lasers
,”
Appl. Phys. Lett.
79
,
3561
(
2001
).
11.
S.
Latkowski
,
A.
Hansel
,
N.
Bhattacharya
,
T.
de Vries
,
L.
Augustin
,
K.
Williams
,
M.
Smit
, and
E.
Bente
, “
Novel widely tunable monolithically integrated laser source
,”
IEEE Photonics J.
7
(
6
),
1
9
(
2015
).
12.
Q.
Chen
,
Q.
Lu
, and
W.
Guo
, “
Theory and simulation of multi-channel interference (MCI) widely tunable lasers
,”
Opt. Express
23
,
18040
18051
(
2015
).
13.
T.
Komljenovic
,
M.
Davenport
,
J.
Hulme
,
A. Y.
Liu
,
C. T.
Santis
,
A.
Spott
,
S.
Srinivasan
,
E. J.
Stanton
,
C.
Zhang
, and
J. E.
Bowers
, “
Heterogeneous silicon photonic integrated circuits
,”
J. Lightwave Technol.
34
(
1
),
20
35
(
2016
).
14.
M. J. R.
Heck
,
J. F.
Bauters
,
M. L.
Davenport
,
J. K.
Doylend
,
S.
Jain
,
G.
Kurczveil
,
S.
Srinivasan
,
Y.
Tang
, and
J. E.
Bowers
, “
Hybrid silicon photonic integrated circuit technology
,”
IEEE J. Sel. Top. Quantum Electron.
19
(
4
),
6100117
(
2013
).
15.
J. C.
Hulme
,
J. K.
Doylend
, and
J. E.
Bowers
, “
Widely tunable Vernier ring laser on hybrid silicon
,”
Opt. Express
21
,
19718
19722
(
2013
).
16.
Y.
Wang
,
Y.
Jiao
, and
K.
Williams
, “
Scaling photonic integrated circuits with InP technology: A perspective
,”
APL Photonics
9
(
5
),
050902
(
2024
).
17.
Y.
Jiao
,
N.
Nishiyama
,
J.
van der Tol
,
J.
van Engelen
,
V.
Pogoretskiy
,
S.
Reniers
,
A. A.
Kashi
,
Y.
Wang
,
V. D.
Calzadilla
,
M.
Spiegelberg
,
Z.
Cao
,
K.
Williams
,
T.
Amemiya
, and
S.
Arai
, “
InP membrane integrated photonics research
,”
Semicond. Sci. Technol.
36
(
1
),
013001
(
2021
).
18.
T.
Fujii
,
K.
Takeda
,
N.-P.
Diamantopoulos
,
E.
Kanno
,
K.
Hasebe
,
H.
Nishi
,
R.
Nakao
,
T.
Kakitsuka
, and
S.
Matsuo
, “
Heterogeneously integrated membrane lasers on Si substrate for low operating energy optical links
,”
IEEE J. Sel. Top. Quantum Electron.
24
(
1
),
1500408
(
2018
).
19.
S.
Yamaoka
,
N.-P.
Diamantopoulos
,
H.
Nishi
,
R.
Nakao
,
T.
Fujii
,
K.
Takeda
,
T.
Hiraki
,
T.
Tsurugaya
,
S.
Kanazawa
,
H.
Tanobe
,
T.
Kakitsuka
,
T.
Tsuchizawa
,
F.
Koyama
, and
S.
Matsuo
, “
Directly modulated membrane lasers with 108 GHz bandwidth on a high-thermal-conductivity silicon carbide substrate
,”
Nat. Photonics
15
(
1
),
28
35
(
2021
).
20.
G.
Sinatkas
,
T.
Christopoulos
,
O.
Tsilipakos
, and
E. E.
Kriezis
, “
Electro-optic modulation in integrated photonics
,”
J. Appl. Phys.
130
(
1
),
010901
(
2021
).
21.
S.
Matsuo
,
T.
Fujii
,
K.
Hasebe
,
K.
Takeda
,
T.
Sato
, and
T.
Kakitsuka
, “
Directly modulated buried heterostructure DFB laser on SiO2/Si substrate fabricated by regrowth of InP using bonded active layer
,”
Opt. Express
22
(
10
),
12139
(
2014
).
22.
T.
Fujii
,
K.
Takeda
,
H.
Nishi
,
N. P.
Diamantopoulos
,
T.
Sato
,
T.
Kakitsuka
,
T.
Tsuchizawa
, and
S.
Matsuo
, “
Multiwavelength membrane laser array using selective area growth on directly bonded InP on SiO2/Si
,”
Optica
7
(
7
),
838
838
(
2020
).
23.
V.
Pogoretskiy
,
J.
Van Der Tol
,
M.
Smit
, and
Y.
Jiao
, “
Monolithically integrated widely tunable laser on an InP membrane circuits
,” in
24th OptoElectronics and Communications Conference (OECC) and 2019 International Conference on Photonics in Switching and Computing (PSC)
(
2019
).
24.
N.
Takahashi
,
Y.
Wang
,
N.
Nishiyama
,
K.
Williams
,
S.
Latkowski
, and
Y.
Jiao
, “
Compact widely-tunable laser on an InP membrane on silicon
,” in
Opto-Electronics and Communications Conference (OECC)
(
IEEE
,
2023
).
25.
T.
Kabir
,
S.
Tondini
,
J.
Hazan
, and
M. J. R.
Heck
, “
40-nm widely tunable laser with Mach-Zehnder interferometer based intra-cavity filter
,” in
IEEE Photonics Conference (IPC)
(
IEEE
,
2023
).
26.
C.
Jiang
,
Q.
Chen
,
K.
Wang
,
Q.
Lu
,
M.
Lu
, and
W.
Guo
, “
Narrow-linewidth thermally tuned multi-channel interference laser integrated with a SOA and spot size converter
,”
Opt. Express
29
,
13246
13255
(
2021
).
27.
M. A.
Tran
,
D.
Huang
,
J.
Guo
,
T.
Komljenovic
,
P. A.
Morton
, and
J. E.
Bowers
, “
Ring-resonator based widely-tunable narrow-linewidth Si/InP integrated lasers
,”
IEEE J. Sel. Top. Quantum Electron.
26
(
2
),
1500514
(
2020
).
28.
S. F. G.
Reniers
,
Y.
Wang
,
K. A.
Williams
,
J. J. G. M.
Van Der Tol
, and
Y.
Jiao
, “
Characterization of waveguide photonic crystal reflectors on indium phosphide membranes
,”
IEEE J. Quantum Electron.
55
(
6
),
6400107
(
2019
).
29.
E.
Kleijn
,
D.
Melati
,
A.
Melloni
,
T.
de Vries
,
M. K.
Smit
, and
X. J. M.
Leijtens
, “
Multimode interference couplers with reduced parasitic reflections
,”
IEEE Photonics Technol. Lett.
26
(
4
),
408
410
(
2014
).
30.
C.
Castellan
,
S.
Tondini
,
M.
Mancinelli
,
C.
Kopp
, and
L.
Pavesi
, “
Low crosstalk silicon arrayed waveguide gratings for on-chip optical multiplexing
,”
Proc. SPIE
10686
,
106860V
(
2018
).
31.
S.
Tondini
,
J.
Hazan
,
T.
Kabir
, and
M. J. R.
Heck
, “
Self-aligning Mach-Zehnder interferometer based filter for intra-cavity tuning of InP integrated lasers
,” in
Proceedings of the 2023 Conference on Lasers and Electro-Optics (CLEO) (
2023
).
32.
L. A.
Coldren
,
S. W.
Corzine
, and
M. L.
Mashanovitch
,
Diode Lasers and Photonic Integrated Circuits
(
Wiley
,
2012
).
33.
Y.
Wang
,
Y.
Wei
,
V.
Dolores-Calzadilla
,
D.
Dai
,
K.
Williams
,
M.
Smit
, and
Y.
Jiao
, “
Efficiency-boosted semiconductor optical amplifiers via mode-division multiplexing
,”
Optica
10
(
9
),
1153
1153
(
2023
).
34.
S.
Tondini
,
C.
Castellan
,
M. A.
Medina
, and
L.
Pavesi
, “
Automatic initialization methods for photonic components on a silicon-based optical switch
,”
Appl. Sci.
9
(
9
),
1843
(
2019
).
35.
G.
Roelkens
,
D.
Van Thourhout
, and
R.
Baets
, “
High efficiency silicon-on-insulator grating coupler based on a poly-silicon overlay
,”
Opt. Express
14
(
24
),
11622
(
2006
).
36.
V.
Dolores-Calzadilla
,
D.
Heiss
, and
M.
Smit
, “
Highly efficient metal grating coupler for membrane-based integrated photonics
,”
Opt. Lett.
39
,
2786
2789
(
2014
).
37.
A. A.
Kashi
,
J. J. G. M.
van der Tol
,
K. A.
Williams
, and
Y.
Jiao
, “
Efficient and fabrication error tolerant grating couplers on the InP membrane on silicon platform
,”
Appl. Opt.
61
,
9926
9936
(
2022
).