Open Access
Add to BibSonomy
Abstract
Ever-increasing complexity of communication systems demands the co-integration of electronics and photonics. But there are still some challenges associated with the integration of thin film lithium niobate (TFLN) electro-optic modulators with the standard and well-established silicon photonics. Current TFLN platforms are mostly not compatible with the silicon photonics foundry process due to the choice of substrate or complicated fabrication requirements, including silicon substrate removal and formation of radio-frequency (RF) electrodes on the top of the TFLN. Here, we report on a platform where all the optical and RF waveguiding structures are fabricated first, and then the TFLN is bonded on top of the silicon photonic chip as the only additional step. Hence, the need for substrate removal is eliminated, and except for the last step of TFLN bonding, its fabrication process is silicon foundry compatible and much more straightforward compared to other fabrication methods.
© 2020 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Optical modulation is an essential part of modern communication systems for both digital and analog applications. Among various platforms, lithium niobate (LN) modulators operating based on linear electro-optic or Pockels effect have been of great interest due to their superior optical properties, such as large electro-optic coefficient and broad transparency range in the electromagnetic spectrum (0.4 – 5 $\mu$m). Wide electro-optic modulation bandwidths up to 100 GHz and large extinction ratios of greater than 20 dB have been demonstrated and widely exploited specially for long-haul communication using bulk LN [1,2]. However, conventional LN electro-optic modulators are bulky and exhibit low efficiency in terms of power consumption and overall device footprint, and hence are not suitable for large scale integration. In conventional LN modulators, waveguides are traditionally formed by in-diffusion of dopants or proton exchange processes. They posses large cross-sections and their low index-contrast ($< 0.1$) yields weak optical confinement which causes high bending loss and low power efficiency in electro-optic modulators. These characteristics render them unattractive for large-scale integration demands of advanced optical networks.
Thin-film LN approaches have been recently pursued in order to overcome these drawbacks [3–5]. Particularly, exploiting heterogeneous integration techniques has led to the emergence of ultracompact thin-film LN electro-optic modulators on silicon substrates [6]. One of the key challenges have been to achieve low-loss sub-micron thin-film LN waveguides. Various methods such as rib-loading the thin-film LN with a refractive index-matched material [5,7–12], dry etching [13–16], and LN die-bonding [17–22], among few others, have been demonstrated and optical loss values as low as 0.027 dB/cm [13] have been achieved. Since first demonstration of thin-film LN electro-optic modulators on Si [6], several high-performance thin-film LN electro-optic modulators have been reported with 3-dB bandwidths up to $\sim$100 GHz [21–25]. Detailed modelling for reliable prediction of high-speed performance of such devices have been recently reported [26] and 3-dB bandwidths up to 400 GHz have been proved attainable for sub-terahertz applications [27].
Silicon Photonic technology has very attractive manufacturing benefits due to the infrastructure developed to build CMOS electronics. The integration capability and ease of manufacturing are very advantageous in mass production of integrated photonic circuits. Unfortunately silicon has non-ideal electro optical material properties. Monolithic silicon photonic modulators operate through the plasma dispersion effect, in which the refractive index of silicon is modulated by changing the carrier density within the waveguide core [28]. The modulation bandwidth of these modulators is fundamentally limited by the mobility of the carriers within silicon [29,30]. Also, nonlinear modulation and high loss due to two-photon absorption [31] limit the power handling capability of silicon waveguides. Adding lithium niobate to photonic integrated circuits boosts its electro optic performance. With the ever-increasing demands for co-integration of electronics and photonics, it is very desirable to integrate thin-film LN electro-optic modulators with the standard and well-established Silicon photonics technology. But, based on the current foundry regulations lithium niobate platforms are not compatible with standard foundry silicon photonics due to some contamination issues.
Here, we report on a platform for fabrication of thin film LN traveling wave modulator, where all the optical and RF components are fabricated first based on the standard silicon photonic foundry processes and then the thin-film lithium niobate is bonded on top of the silicon photonic chip as the only additional back end of the line step. Hence, the fabrication process has a very good silicon foundry compatibility.
2. Proposed structure
Figure 1(a) shows the schematic cross-section of the proposed TFLN Mach-Zehnder traveling-wave modulator in the push-pull configuration and Fig. 1(b) shows the three dimensional schematic of the proposed structure. The slab region of the optical waveguide is an Y-cut thin-film lithium niobate bonded to a layer of silicon oxide on silicon. The applied RF electric field is aligned along the $z$-axis of lithium niobate. Hence, the strongest electro optic coefficient of lithium niobate crystal ($r_{33} \approx$31 pm/V) [32], will be efficiently utilized. The extraordinary index which is along the z-axis of the lithium niobate crystal almost constant in $C$ band ($n_e\approx 2.21$).
Fig. 1. a) Cross-section of the proposed structure b)Three dimensional view of the proposed structure. In the following simulations typical dimensions are $W_g$ = 5 $\mu$m, $W_c$ = 8 $\mu$m, $W_l$ = 12 $\mu$m, $t$ = 1.8 $\mu$m, $h_{LN}$ = 400 nm, $h^{sup}_{Si}$ =$h^{sub}_{Si}$= 80 $\mu$m, and $h^{Sup}_{SiO2}$ =$h^{sub}_{SiO2}$= 2 $\mu$m, and $h^{gap}_{SiO2}$ = 30 nm (which is the silicon oxide layer between top of the RF electrodes and the TFLN layer. The gap between the silicon waveguide top and TFLN layer is also the same value) . Unless mentioned otherwise, these values are used in the following calculations at the frequency of 50 GHz.
Figure 2 shows the top view of the final modulator. Since RF and optical waveguides are in the same oxide layer, optical waveguides cannot be straight and they go around the input and output RF pads. Figure 3 shows the general overview of the fabrication process of the proposed modulator. At first, optical waveguides are fabricated on silicon-on-insulator (SOI) wafer using standard foundry techniques (Figs. 3(a) and (b)). The next step is fabrication of the RF wavegides inside the silicon oxide layer (Fig. 3(c)). After both optical and RF waveguides are fabricated, a thin oxide layer is deposited on top of the planarized structure is ready for the final step which is bonding the lithium niobate chip on top (Fig. 3(d)). The un-patterned thin-film of lithium niobate is oxide-bonded at room temperature to the patterned and planarized RF and optical waveguide circuits. An anneal step is necessary after bonding. No patterning or etching of the LN film is performed in our approach. Also since the RF waveguides are embedded in the SOI wafer, there is no need to remove the thin-film LN backside.
Fig. 2. Top overview of the proposed structure. TFLN is bonded on top of the RF and optical waveguides, between the input and output RF pads.
Fig. 3. General overview of the fabrication process of the proposed structure : a) Optical waveguide fabrication on SOI wafer b) RF waveguide fabrication inside the silicon oxide box c) Deposition and planarization of the top silicon oxide layer d) Bonding the thin-film LN on top of the structure.
Thin film lithium niobate wafer is commercially available in the form of 3,4 and 6 inch wafers from NANOLN company. Different layer configurations are available commercially but the one which is appropriate for our proposed structure has three layers. A 400 to 500 $\mu m$ thick silicon substrate, 1 to 4 $\mu m$ silicon oxide, and 300 to 900 $nm$ layer of lithium niobate on top. In Fig. 4, we have considered three different configurations for the bounded superstrate. In Fig. 4(a) thin film lithium niobate chip is bonded on top of the processed SOI wafer which includes the RF and optical waveguides. In this configuration no further processing is needed after the bonding. In Fig. 4(b), the silicon substrate of the bonded thin film lithium niobate is removed by adding an etching as a post processing step. In Fig. 4(c) both silicon and silicon oxide layers of the bonded thin film lithium niobate chip is removed. Obviously structures B and C need extra post processing steps which complicates the fabrication process.
In any travelling-wave Mach-Zehnder modulator, the three main parameters which determine the performance of the device are RF loss, effective index mismatch between the RF and optical guided modes and characteristic impedance of the RF electrodes ($Z_0$) [26]. In the following, RF and optical properties of the proposed device are studied. To optimize the electro-optic bandwidth of the modulator, we need to simultaneously tune the phase velocity of the RF waveguide modes such that it is matched to the optical group velocity utilizing a CPW design such that attenuation is minimized while maintaining a characteristic impedance of 50 ohms.
3. RF properties
In the proposed structure the RF transmission line is a symmetric coplanar waveguide (CPW) structure [33]. CPW is a low dispersion transmission line, so it is suitable for development of wide-band circuits and components. To implement a push pull system, the RF signal is applied to the middle conductor and the other two conductors are ground planes (Ground-Signal-Ground (GSG) configuration). By doing this the electric fields inside the gaps of the CPW line have opposite direction which produces index change with opposite signs in the thin film lithium niobate layer. The main transmission line parameters of a CPW structure are the characteristic impedance ($Z_0$), attenuation constant ($\alpha ^{RF}$) and effective phase constant ($\beta ^{RF}$). The characteristic impedance is determined by the ratio $W_c/(W_c+W_g) < 1$. Hence, size reduction is possible without limit. The main penalty is higher loss. In addition a ground plane exists between any two adjacent lines, hence cross talk effects between adjacent lines are very week. As a result, CPW circuits can be made denser than conventional microstrip circuits.
The key physical parameters that play an important role in defining the RF properties of the modulator are the thickness of the electrodes ($t$), the gap between the electrodes ($W_g$), and the middle electrode’s width ($W_c$) (as shown in Fig. 1(a)). These physical dimension parameters define the RF effective index, characteristic impedance and loss. The difference between the effective refractive indices of the optical ($n^{opt}_{eff}$) and RF ($n^{RF}_{eff}$) waveguides is an important parameter for bandwidth enhancement. As optical effective index is constant within the concerned RF frequency range, depending on the optical waveguide structure and its $n^{opt}_{eff}$, $n^{RF}_{eff}$ can be appropriately tuned in order to minimize the difference. Also, considering the standard 50 ohms transmission lines, the characteristic impedance of the CPW line should be close to 50 ohms to minimize the reflections in the input and output.
RF attenuation constant plays an important role in the frequency response of the electro optic modulator and its bandwidth. The attenuation constant of the CPW structure is generally defined as $\alpha ^{RF} = \alpha _c + \alpha _d + \alpha _r$, where $\alpha _c$ is the ohmic or metallic conductor attenuation constant, $\alpha _d$ is the dielectric attenuation constant and $\alpha _r$ is the radiation attenuation constant. With smooth metal deposition for electrodes, $\alpha _r$ can be neglected. The expressions for the attenuation constant due to dielectric loss in CPW structure is the same as that for a microstrip, which is [34]
The ohmic loss has been calculated by evaluating the power dissipated in the line through conformal mapping of the current density in the finite metal thickness structure [35] as $\alpha _c=\alpha _{c0}\sqrt {f}$, where $\alpha _{c0}$ is defined as
Figures 5(a), (b), and (c) show the RF effective index, real part of the characteristic impedance , and RF propagation loss with respect to the width of the center RF electrode ($W_c$) for different values of the gap between the RF electrodes ($W_g$). CPW line with smaller values of $W_c$ and $W_g$ has slightly lower RF effective index and higher characteristic impedance. Regarding the propagation loss, it is almost constant with the width of the center electrode ($w_c$) but smaller gap sizes ($w_g$) increases the RF loss. Considering 50 ohms standard for RF transmission lines and optical effective index of $\approx 2$, to have better impedance and effective index matching between RF and optical transmission lines, smaller values of $W_c$ or $W_g$ are desirable. But, smaller values of RF waveguide gap increases the RF loss which lowers the final bandwidth of the modulator. Generally, the effect of effective index matching is much higher than the RF loss on the final bandwidth of the modulator.
Fig. 5. a) RF effective refractive index b) RF Characteristic impedance for different values of $w_g$ and $w_c$.
Considering a ground-signal-ground (GSG) configuration for RF transmission line in Fig. 1(a), the two side electrodes are grounded. In theory, ground planes of a CPW lines are infinite. In practice, it is desirable for the ground planes be as large as possible. Based on the results in Fig. 6, as the width of the ground plane increases, changes in RF effective index ($n^{RF}_{eff}$) and characteristic impedance ($Re{Z_0}$) of the transmission line are minimal but propagation loss is lower for larger ground electrodes widths which is desirable for achieving higher modulator bandwidth.
Fig. 6. a) RF effective refractive index b)RF Characteristic impedance c) Propagation loss for different values of the ground electrode of the RF waveguide ($w_l$).
Thickness of the RF electrodes has considerable effect on the RF properties of the transmission line. Using conformal mapping technique the RF effective index of a CPW line with infinite ground planes and zero conductor thickness is [34,36]:
Fig. 7. a) RF effective refractive index b)RF Characteristic impedance c) Propagation loss for different values of the RF waveguide thickness $t$.
Figures 8 and 9 show the RF properties of the CPW transmission line for different thicknesses of LN and substrate silicon oxide layer for these three cases. Since the thickness of the silicon layer is constant in the standard SOI and TFLN wafers available in the market, it is not included in these simulations. In general, based on these simulation results, removing the silicon and silicon oxide layers of the bonded thin film lithium niobate decreases the RF effective index and characteristic impedance of the CPW line. On the other hand, RF loss for the case that the silicon substrate is present is much lower than the other two cases. Also it is clear that changing the thickness of the thin film lithium niobate layer or the silicon oxide layer of the SOI does not change the effective index, characteristic impedance and RF loss drastically. But since lower RF effective index and RF loss are desirable smaller values for the thickness of the thin film lithium niobate and silicon oxide layer of the substrate is preferable.
Fig. 8. a) RF effective refractive index b)RF Characteristic impedance c) Propagation loss for different values of $h_{LN}$.
Fig. 9. a) RF effective refractive index b)RF Characteristic impedance c) Propagation loss for different values of $h_{SiO_2}^{Sub}$.
Figure 10 shows the frequency dependence of the RF effective index and propagation constant with respect to frequency for different values of $W_c$, $W_g$, and $W_l$ for the proposed structure. In general, RF effective index decreases and RF propagation loss increases with frequency. Changing $W_l$ has the largest effect on $n_{eff}^{RF}$ and $\alpha ^{RF}$ as frequency changes.
Fig. 10. RF effective refractive index and RF Propagation loss of the proposed structure as a function of frequency for different values of a)$W_c$ b)$W_g$ c)$W_l$.
4. Optical properties
In the proposed hybrid structure, a thin-film LN layer is bonded on top of a silicon-on-insulator (SOI) wafer which makes it possible to build complex waveguiding circuits by multiple vertical transitions between silicon and LN waveguides. The most interesting property of this hybrid structure is that only standard silicon layer lithography is used to control the light path both outside the bonded region and also under the bonded LN area [19]. The optical waveguiding structure in the proposed structure is similar to the system reported in [21]. A hybrid TFLN-SOI integrated electro optic modulator chip as in Fig. 2, includes four types of optical waveguide structures: tapers for light input and output from the chip (edge couplers), normal single mode silicon waveguide with air cladding, transition from waveguide with air cladding to lithium niobate loaded waveguide, and adiabatic waveguide tapers for inter-layer transitions (Silicon-to-lithium niobate and vice-versa). Since the LN layer is neither patterned nor etched in our design, there is no alignment issue at the bonding step and the silicon waveguides determine the optical propagation path. In the proposed traveling-wave thin-film LN Mach-Zehnder modulator light enters the chip through an edge coupler and is transfered to a single mode silicon waveguide with air cladding. Before entering the lithium niobate cladded region, the waveguides are widened to minimize light scattering at the edge of the LN overlay. Vertical transitions to and from the hybrid silicon-LN region occur inside the bonded region by narrowing the silicon waveguide. It is very important to transfer maximum amount of optical energy from the silicon waveguide to the thin-film LN layer (and vice versa) by selecting optimum width of the silicon waveguide and thin-film LN.
In order to study the range of achievable values for n$^{opt}_{eff}$, the optical waveguide region of the compact electro optic modulator is simulated using Lumerical$^{TM}$. Optical effective index and the total energy inside silicon optical waveguide and LN layer are calculated for different values of the LN layer height ($h_{LN}$), silicon waveguide width, and silicon oxide layer thickness above the silicon waveguide ($h^{gap}_{SiO2}$). Figure 11(a) shows the total energy inside silicon waveguide and LN layer for different thicknesses of the LN layer and the optical effective index for different LN layer thicknesses. Optical effective index varies between 1.92 and 2.02 as the LN layer thickness increases from 300nm to 600nm. The energy inside LN layer is maximum at 450nm LN layer. Figure 11(b) shows the total energy inside silicon waveguide and LN layer for different width of the silicon waveguide below the LN layer and the optical effective index for different width of the silicon waveguide below the LN layer. For the waveguide with 240nm width the energy inside LN layer is maximum. Figure 11(c) shows the total energy inside silicon waveguide and LN layer for different thicknesses of the gap between silicon waveguide and LN layer and the optical effective index for different thicknesses of the gap between silicon waveguide and LN layer. As the size of the gap increases the amount of energy inside the LN layer decreases. Figure 12 shows the optical field distribution for three values of the gap between the waveguide and LN layer. The case that there is no gap between the silicon waveguide and LN layer (Fig. 12(a)) has the highest field confinement. Figure 13(a) shows the optical propagation loss coefficient inside silicon waveguide for different gap sizes between two RF electrodes. For smaller gap size between two RF electrodes (smaller $W_g$), part of the optical field will intercept with the electrodes which increases the optical loss (as shown in Fig. 13(b)).
Fig. 11. (a) Right:Total energy inside silicon waveguide and LN layer for different thicknesses of the LN layer ($h_{LN}$). Left: Optical effective index for different LN layer thicknesses ($h_{LN}$).b) Right: Total energy inside silicon waveguide and LN layer for different width of the silicon waveguide below the LN layer. Left: Optical effective index for different width of the silicon waveguide below the LN layer. c) Right: Total energy inside silicon waveguide and LN layer for different thicknesses of the gap between silicon waveguide and LN layer ($h^{gap}_{SiO2}$). Left: Optical effective index for different thicknesses of the gap between silicon waveguide and LN layer ($h^{gap}_{SiO2}$).
Fig. 12. Electric field distribution in silicon waveguide and LN layer for (a) $h^{gap}_{SiO_2}=0$, (a) $h^{gap}_{SiO_2}=50nm$, (a) $h^{gap}_{SiO_2}=100nm$ .
Fig. 13. a) Optical propagation loss coefficient inside silicon waveguide for different gap sizes between two RF electrodes. b) Optical field distribution for $W_g=6nm$
5. Electro-optic properties
The complex propagation constant of the RF transmission line is defined as,
where the real part of the propagation constant, $\alpha ^{RF}$, is the attenuation constant and the imaginary part, $\beta ^{RF}$, is the phase constant of the transmission line. In general, total phase shift of the optical field along the electrode length $L$ which the RF transmission line is presented is, where, $\Delta \beta ^{opt}$ is the change in the optical propagation constant and $\Delta \Phi ^{opt}$ is the optical phase change. The $x$-direction is considered as the propagation direction ($y$ and $z$ are along and perpendicular the TFLN layer). Neglecting the change in the imaginary part of the refractive index, where $k_0$ is the vacuum wave-number and $\Delta n^{ot}_{eff}$ is the change in the effective optical index of the silicon waveguide.In the presented structure the TFLN slab is a Y-cut thin-film of LN bonded to a layer of silicon oxide on silicon substrate. The applied RF electric field is aligned along the z-axis of TFLN layer. Hence, the strongest electro optic coefficient of LN crystal, i.e., $r_{33}$ = 31 pm/V, will be efficiently utilized. Effective optical refractive index along the $z$-direction is
where $n_e$ is the extraordinary refractive index of LN, and $E^{RF}_z$ is the total electric field applied along the $z$ direction. Since the RF field frequency is much lower than the optical field frequency, it can be considered as a DC field compared to the optical field. Therefore, $E^{RF}_z$ is effectively the amplitude of the RF field at the location of the optical field. The RF field can be written as Voltage $V$ is a time-varying signal for traveling-wave EO modulators and $f(y,z)$ is the normalized spatial distribution of the electric field applied at the optical waveguide transverse plane. Defining the overlap of the electrical and optical fields as,Figure 14 shows the overlap function for different values of electrode widths ($W_c$, $W_g$), LN thickness ($h_{LN}$), and electrode thickness ($t$). By increasing $W_c$, $h_{LN}$. and $t$, the value of the overlap function increases. On the other hand, overlap function is higher for smaller electrode gap ($W_g$). The value of $v_\pi L$ is calculated using the equation,
As depicted in Fig. 15, Smaller gap size between RF electrodes lowers the value of $v_\pi L$. Also $v_\pi L$ is lower for thinner LN layer and thicker RF electrodes.The electro optic transfer function ($H(\omega _{RF})$) can be defined as [26],
6. Conclusion
We have proposed a new structure for the hybrid Si-LN Mach-Zehnder modulator where RF waveguides are implemented on the same layer as the optical waveguides inside the SOI wafer. This makes the fabrication process more compatible to the foundry fabrication process and minimizes the back end of the line fabrication to single step which is bonding the thin-film LN on top of the SOI wafer which includes the RF and optical waveguides. Based on the simulation results, the trade off in using the proposed structure compared to typical structure in which the RF electrodes are not in the same layer as the optical waveguide is that presence of the silicon layer on top of the thin-film LN increases the RF effective index which lowers the performance of the device due to effective index mismatch between the RF and optical waveguieds. It is possible to further enhance the performance of the proposed structure by optimizing the thickness of the silicon and silicon oxide layers on top of the thin-film LN layer which complicates the fabrication process.
Disclosures
The authors declare that there are no conflicts of interest related to this article.
References
1. E. Wooten, K. Kissa, A. Yi-Yan, E. Murphy, D. Lafaw, P. Hallemeier, D. Maack, D. Attanasio, D. Fritz, G. McBrien, and D. Bossi, “A review of lithium niobate modulators for fiber-optic communications systems,” IEEE J. Quantum Electron. 6(1), 69–82 (2000). [CrossRef]
2. K. Noguchi, O. Mitomi, and H. Miyazawa, “Millimeter-wave ti:linbo3 optical modulators,” J. Lightwave Technol. 16(4), 615–619 (1998). [CrossRef]
3. J. Macario, P. Yao, S. Shi, A. Zablocki, C. Harrity, R. Martin, C. Schuetz, and D. Prather, “Full spectrum millimeter-wave modulation,” Opt. Express 20(21), 23623–23629 (2012). [CrossRef]
4. S. Jin, L. Xu, H. Zhang, and Y. Li, “Linbo3 thin-film modulators using silicon nitride surface ridge waveguides,” IEEE Photonics Technol. Lett. 28(7), 736–739 (2016).
5. A. Rao, A. Patil, J. Chiles, M. Malinowski, S. Novak, K. Richardson, P. Rabiei, and S. Fathpour, “Heterogeneous microring and mach-zehnder modulators based on lithium niobate and chalcogenide glasses on silicon,” Opt. Express 23(17), 22746–22752 (2015). [CrossRef]
6. P. Rabiei, J. Ma, S. Khan, J. Chiles, and S. Fathpour, “Heterogeneous lithium niobate photonics on silicon substrates,” Opt. Express 21(21), 25573–25581 (2013). [CrossRef]
7. S. Li, L. Cai, Y. Wang, Y. Jiang, and H. Hu, “Waveguides consisting of single-crystal lithium niobate thin film and oxidized titanium stripe,” Opt. Express 23(19), 24212–24219 (2015). [CrossRef]
8. A. Rao, M. Malinowski, A. Honardoost, J. Talukder, P. Rabiei, P. Delfyett, and S. Fathpour, “Second-harmonic generation in periodically-poled thin film lithium niobate wafer-bonded on silicon,” Opt. Express 24(26), 29941–29947 (2016). [CrossRef]
9. L. Chang, Y. Li, N. Volet, L. Wang, J. Peters, and J. E. Bowers, “Thin film wavelength converters for photonic integrated circuits,” Optica 3(5), 531–535 (2016). [CrossRef]
10. L. Chang, M. H. P. Pfeiffer, N. Volet, M. Zervas, J. Peters, C. Manganelli, E. Stanton, Y. Li, T. Kippenberg, and J. E. Bowers, “Heterogeneous integration of lithium niobate and silicon nitride waveguides for wafer-scale photonic integrated circuits on silicon,” Opt. Lett. 42(4), 803–806 (2017). [CrossRef]
11. A. Honardoost, G. F. C. Gonzalez, S. Khan, M. Malinowski, A. Rao, J.-E. Tremblay, A. Yadav, K. Richardson, M. C. Wu, and S. Fathpour, “Cascaded integration of optical waveguides with third-order nonlinearity with lithium niobate waveguides on silicon substrates,” IEEE Photonics J. 10(3), 1–9 (2018). [CrossRef]
12. G. F. C. Gonzalez, M. Malinowski, A. Honardoost, and S. Fathpour, “Design of a hybrid chalcogenide-glass on lithium-niobate waveguide structure for high-performance cascaded third-and second-order optical nonlinearities,” Appl. Opt. 58(13), D1–D6 (2019). [CrossRef]
13. M. Zhang, C. Wang, R. Cheng, A. Shams-Ansari, and M. Lonar, “Monolithic ultra-high-q lithium niobate microring resonator,” Optica 4(12), 1536–1537 (2017). [CrossRef]
14. Y. Wang, Z. Chen, and H. Hu, “Analysis of waveguides on lithium niobate thin films,” Crystals 8(5), 191 (2018). [CrossRef]
15. A. Rao, K. Abdelsalam, T. Sjaardema, A. Honardoost, G. F. Camacho-Gonzalez, and S. Fathpour, “Actively-monitored periodic-poling in thin-film lithium niobate photonic waveguides with ultrahigh nonlinear conversion efficiency of 4600%w−1cm−2,” Opt. Express 27(18), 25920–25930 (2019). [CrossRef]
16. M. Bahadori, Y. Yang, L. Goddard, and S. Gong, “High performance fully etched isotropic microring resonators in thin-film lithium niobate on insulator platform,” Opt. Express 27(15), 22025–22039 (2019). [CrossRef]
17. L. Chen, Q. Xu, M. Wood, and R. Reano, “Hybrid silicon and lithium niobate electro-optical ring modulator,” Optica 1(2), 112–118 (2014). [CrossRef]
18. L. Chen, J. Chen, J. Nagy, and R. Reano, “Highly linear ring modulator from hybrid silicon and lithium niobate,” Opt. Express 23(10), 13255–13264 (2015). [CrossRef]
19. P. Weigel, M. Savanier, C. DeRose, A. Pomerene, A. Starbuck, A. Lentine, V. Stenger, and S. Mookherjea, “Lightwave circuits in lithium niobate through hybrid waveguides with silicon photonics,” Sci. Rep. 6(1), 22301 (2016). [CrossRef]
20. S. Fathpour, A. Honardoost, and S. Khan, “Thin-film integration compatible with silicon photonics foundry production,” US Patent Application 6, 15/9328000–9328000 (2018).
21. P. Weigel, J. Zhao, K. Fang, H. Al-Rubaye, D. Trotter, D. Hood, J. Mudrick, C. Dallo, A. Pomerene, A. Starbuck, C. DeRose, A. Lentine, G. Rebeiz, and S. Mookherjea, “Bonded thin film lithium niobate modulator on a silicon photonics platform exceeding 100 ghz 3-db electrical modulation bandwidth,” Opt. Express 26(18), 23728–23739 (2018). [CrossRef]
22. M. He, M. Xu, Y. Ren, J. Jian, Z. Ruan, Y. Xu, S. Gao, S. Sun, X. Wen, L. Zhou, L. Liu, C. Guo, H. Chen, S. Yu, L. Liu, and X. Cai, “High-performance hybrid silicon and lithium niobate mach-zehnder modulators for 100 gb/s and beyond,” Nat. Photonics 13(5), 359–364 (2019). [CrossRef]
23. A. Rao, A. Patil, P. Rabiei, A. Honardoost, R. DeSalvo, A. Paolella, and S. Fathpour, “High-performance and linear thin-film lithium niobate mach–zehnder modulators on silicon up to 50 ghz,” Opt. Lett. 41(24), 5700–5703 (2016). [CrossRef]
24. C. Wang, M. Zhang, X. Chen, M. Bertrand, A. Shams-Ansari, S. Chandrasekhar, P. Winzer, and M. Loncar, “Integrated lithium niobate electro-optic modulators operating at cmos-compatible voltages,” Nature 562(7725), 101–104 (2018). [CrossRef]
25. X. Wang, P. Weigel, J. Zhao, M. Ruesing, and S. Mookherjea, “Achieving beyond-100-ghz large-signal modulation bandwidth in hybrid silicon photonics mach zehnder modulators using thin film lithium niobate,” APL Photonics 4(9), 096101 (2019). [CrossRef]
26. A. Honardoost, R. Safian, A. Rao, and S. Fathpour, “High-speed modeling of ultracompact electrooptic modulators,” J. Lightwave Technol. 36(24), 5893–5902 (2018). [CrossRef]
27. A. Honardoost, F. Juneghani, R. Safian, and S. Fathpour, “Towards subterahertz bandwidth ultracompact lithium niobate electrooptic modulators,” Opt. Express 27(5), 6495–9501 (2019). [CrossRef]
28. A. Lentine, C. Derose, P. Davids, J. D. Nicolas, W. Zortman, A. J. J.A. Cox, D. Trotter, A. Pomerene, A. Starbuck, M. W. D.J. Savignon, and P. Chu, “Silicon photonics platform for national security applications,” IEEE Aerosp. Conf. 20, 1–9 (2015). [CrossRef]
29. R. Soref and B. Bennett, “Electrooptical effects in silicon,” IEEE J. Quantum Electron. 23(1), 123–129 (1987). [CrossRef]
30. F. Gardes, G. Reed, N. Emerson, and C. Png, “A sub-micron depletion-type photonic modulator in silicon on insulator,” Opt. Express 13(22), 8845–8854 (2005). [CrossRef]
31. L. Yin and G. P. Agrawal, “Impact of two-photon absorption on self-phase modulation in silicon waveguides,” Opt. Lett. 32(14), 2031–2033 (2007). [CrossRef]
32. J. Liu, Photonic Devices (Cambridge Univ. Press, Cambridge, U.K., 2005).
33. D. M. Pozar, Microwave Engineering (Wiley, Hoboken, NJ, USA., 2012).
34. R. Garg, I. Bahl, and M. Bozzi, Microstrip Lines and Slotlines, 3rd Ed. (Artech House, Norwood, MA, USA., 2013).
35. G. Chione, “A cad-oriented analytical model for the losses of general asymmetric coplanar lines in hybrid and monolithic mics,” IEEE Trans. Microwave Theory Tech. 41(9), 1499–1510 (1993). [CrossRef]
36. E. Chen and S. Chou, “Characteristics of coplanar transmission lines on multilayer substrates: Modeling and experiments,” IEEE Trans. Microwave Theory Tech. 45(6), 939–945 (1997). [CrossRef]
37. H. Chung, W. Chang, and E. Adler, “Modeling and optimizationof traveling-wave linbo3interferometric modulators,” IEEE J. Quantum Electron. 27(3), 608–617 (1991). [CrossRef]
