1. INTRODUCTION

Integrated silicon nitride (SiN) waveguides and resonators are an attractive platform for nonlinear optics [1]. SiN waveguides combine the material’s large bandgap ($\sim 5\mathrm{eV}$) and wide transparency range with CMOS fabrication methods and a large effective nonlinearity. Upon launching a femtosecond laser pulse inside a SiN waveguide, the interplay of anomalous group velocity dispersion (GVD) and high effective nonlinearity ($\gamma \approx 1.4{\mathrm{W}}^{-1}{\mathrm{m}}^{-1}$) leads to efficient supercontinuum generation [2]. Harnessing four-wave-mixing processes inside SiN waveguides, low-noise optical amplification of weak signals can be achieved [3]. Moreover, the fabrication of high-$Q$ SiN microresonators with anomalous GVD at telecom wavelengths has allowed for the observation of parametric oscillations and optical frequency comb generation in integrated SiN microresonators [4], based on the Kerr frequency comb generation mechanism first reported in 2007 [5]. The spectral bandwidth of the generated frequency combs can attain an octave [6].

Following this work, a significant advance in SiN nonlinear photonics was the generation of low-noise frequency combs [7,8,9]. Due to the comparatively large ratio of cavity linewidth to dispersion parameter in SiN microresonators, the nonlinear comb formation process takes place via the formation of subcombs that can lead to the loss of spectral coherence and high comb noise. It was shown that low phase noise comb states can still be attained via synchronization of the generated subcombs via $\delta -\mathrm{\Delta }$ matching [7,10] and have also been observed to allow ultrashort optical pulse generation [8]. Using such phase locked comb states, the high application potential of SiN microresonator based frequency combs was demonstrated in experiments on Tb/s coherent communication [11] and ultrafast optical waveform generation [9]. Most recently, it has been demonstrated that temporal dissipative Kerr soliton (DKS) formation and DKS Cherenkov radiation are accessible in SiN microresonators [12]. This allows the generation of frequency combs that are deterministically fully coherent, numerically predictable [13,14], and broadband (2/3 of an octave), and that exhibit a smooth spectral envelope, underscoring the potential of SiN based microresonators to serve as compact, mass producible, on-chip optical frequency comb generators.

The above advances relied on engineering of the waveguide’s GVD through tailoring the dimensions of the waveguide cross section. The GVD for a given waveguide cross section can be approximated as the sum of the material GVD [Fig. 1(b)] and its waveguide GVD [Fig. 1(d)]. While the material GVD is typically normal at short wavelengths and anomalous at longer wavelengths, the waveguide GVD exhibits the opposite behavior. Thus by adjusting the waveguide cross section the waveguide dispersion can be tailored to overcompensate the material dispersion. At telecom wavelengths the material GVD of SiN is normal, but by employing a high confinement waveguide with a height in excess of 0.7 μm, an effective anomalous GVD is tailored. Consequently even larger waveguide cross sections are needed in order to enable SiN based nonlinear integrated optics, especially Kerr comb generation, in the mid-infrared [15].

 

Fig. 1. Nonlinear wavelength conversion in integrated waveguides and principles of their dispersion engineering. (a) Upon launching a short light pulse into the waveguide (top), its initial spectrum is nonlinearly broadened. Coupling the light of a single frequency, continuous wave laser into the ring resonator (bottom) produces short soliton pulses with broad spectral envelope. (b) Most materials exhibit anomalous GVD at longer wavelengths but not necessarily at telecom wavelengths. However, for efficient nonlinear wavelength conversion, the GVD of the highly confining waveguides must be anomalous. (c), (d) By tailoring the waveguide cross section to highly confine the optical mode, the anomalous waveguide dispersion can overcompensate the normal material dispersion. Thus dispersion engineering determines the required waveguide dimensions for a given operating wavelength.

Download Full Size | PDF

2. PROBLEMS OF SUBTRACTIVE WAVEGUIDE FABRICATION

The fabrication of such high confinement waveguides with heights larger than 0.7 μm from stoichiometric SiN is challenging. So far, fabrication typically relies on a subtractive process approach: the waveguide structures are etched into a previously deposited thin film. The fabrication challenges posed by this process approach are summarized in Figs. 2(a)–2(c).

 

Fig. 2. Problems in subtractive SiN waveguide fabrication. (a) Schematic top view of a microresonator (blue) with cracks formed by the highly stressed SiN film. (b) Cleaved waveguide cross section showing the limited aspect ratio between adjacent waveguides (a geometry used for resonator waveguide coupling) and void formation of the low temperature oxide (LTO) cladding layer. (c) Transmission electron microscope (TEM) image showing the thin oxide layers formed during multistep deposition of SiN.

Download Full Size | PDF

While plasma enhanced chemical vapor deposition (PECVD) allows for crack-free SiN thin film deposition, high-Q optical microresonators were only achieved based on low pressure chemical vapor deposited (LPCVD), stoichiometric SiN thin films. The high temperatures during film deposition ($\sim 770°\mathrm{C}$) and for subsequent annealing (up to 1200°C) have been shown to lead to low material absorption losses [16]. But they also lead to highly tensile stressed films (typically $>800\mathrm{MPa}$), which are prone to crack formation [Fig. 2(a)]. Such cracks cause high scattering losses in the waveguide and are thus unacceptable for high-$Q$ microresonators. So far only a multistep growth approach, which partially relaxes film stress through thermal cycling during deposition, can reduce the crack density per wafer to an acceptable level [17]. Additional crack stop structures have been introduced as the device yield was still reduced by the remaining cracks [18].

Furthermore, the gap between two closely spaced, thick waveguides can present a critical aspect ratio for several processing steps. The dry etch process that defines the waveguide structures often has a limited anisotropy and aspect ratio dependent etch rates (ARDE). It is typically optimized to produce smooth waveguide sidewalls, to limit scattering losses, but often does not accurately transfer the resist mask pattern. Figure 2(b) shows an example: the measured waveguide gap of 450 nm deviates strongly from the design value of 200 nm. This is due to the limited anisotropy of the etch process applied, which is primarily designed for low sidewall roughness. Moreover, voids can form in between closely spaced waveguides during the ${\mathrm{SiO}}_2$ cladding deposition, due to the limited conformality of the deposition processes.

Finally, it was found that nanometer thin silicon oxide layers form between the individual SiN layers when using multistep deposition with thermal cycling [Fig. 2(c)], which may be undesirable in certain applications.

3. PHOTONIC DAMASCENE PROCESS

Inspired by the additive patterning process (commonly referred to as the “Damascene process” or “dual-Damascene process”) for microprocessor copper interconnects [19], we present here a novel “photonic Damascene process” that overcomes the aforementioned challenges. The process relies on substrate prepatterning prior to core material deposition and a subsequent planarization step [Figs. 3(a)–3(f)]. Using this process we are able to reliably fabricate SiN waveguides for the first time, to the best of our knowledge, as well as high-$Q$ microresonators, with unprecedentedly large waveguide dimensions, due to a novel stress control technique based on dense substrate prepatterning. Previously, this process approach has been demonstrated for waveguides based on amorphous silicon [20] and recently also for limited height SiN waveguides [21].

 

Fig. 3. Photonic Damascene process for integrated SiN waveguides. (a)–(f) Schematic process flow of the photonic Damascene process. (g) Optical image of a SiN microresonator surrounded by the stress release structure (rectangle dimensions $5\mathrm{\mu m}\times 5\mathrm{\mu m}$). The zoomed inset reveals the 10 μm wide area to each side of the waveguide that has no stress release structure to avoid scattering losses. The crack formed due to incomplete removal of excess SiN but does not penetrate the waveguide. (h), (i) Focused ion beam (FIB) cross section of the coupling region between the ring resonator and the bus waveguide, revealing a waveguide resonator separation below 200 nm. The individual SiN waveguides (blue) are 1.5 μm wide and 0.85 μm high and homogenously filled with SiN. The coupling region is free of voids, and no effect of the waveguide proximity on the waveguide shape is observed.

Download Full Size | PDF

In a first step the waveguide pattern is defined using electron beam lithography [Fig. 3(a)]. The resist mask is transferred via dry etching into an amorphous silicon (a-Si) hardmask. Then the dense stress release pattern is defined using photolithography and transferred into the hardmask layer as well [Fig. 3(b)]. During the subsequent preform etch the a-Si hardmask is thus structured with both patterns, which are dry etched into the wet thermal oxide layer of the substrate [Fig. 3(c)]. The $10:1$ etch selectivity between the a-Si hardmask and the thermal oxide avoids geometry limitations and sidewall roughness due to mask erosion. Before the SiN thin film can be deposited, the a-Si hardmask layer is stripped. Due to the stress release effect of the densely prestructured substrate, the LPCVD SiN thin film can now be deposited in one run up to the desired thickness [Fig. 3(d)]. The excess SiN material is removed using chemical mechanical planarization (CMP), and a polished substrate exposing only the waveguide’s top surface is obtained [Fig. 3(e)]. Next, thermal annealing is used to densify the SiN film, reducing its content of residual hydrogen. Finally, a cladding layer of low temperature oxide (LTO) is deposited and the chip side facets are dry etched before separating the wafer into individual chips [Fig. 3(f)].

Figure 3(g) shows an optical image of a ring resonator fabricated with the photonic Damascene process, as well as a focused ion beam (FIB) cross section of the coupling region. The waveguides are surrounded by a dense pattern, which relaxes the SiN film stress and thus ensures high yield fabrication. The FIB cross section [Figs. 3(h) and 3(i)] reveals SiN waveguides with 1.5 μm width and 0.85 μm height. The waveguides have nearly vertical sidewalls and a flat top surface, indicating that the planarization was successful and no top surface topography was transferred into the waveguide. The deposited SiN uniformly fills the preform, and no boundaries are observed within the SiN core.

One of the inherent advantages of the process is its ability to produce closely spaced waveguides comprising narrow aspect ratios between them. As seen in Fig. 3(i) the aspect ratio of the coupling gap between resonator and bus waveguide is better than $4:1$ (with a gap below 200 nm) while being free of voids or proximity-related effects, which is challenging to achieve with subtractive processing.

4. STRESS CONTROL BY DENSE PRESTRUCTURING

Furthermore, the photonic Damascene process approach allows the elegant use of substrate topography for stress control and crack prevention in the deposited thin film. Crack formation in high stress LPCVD SiN thin films is a longstanding problem that has so far often limited film thickness to at most 250 nm [22]. However, as shown in Ref. [23], crack development in SiN thin films strongly depends on the substrate topography. Here we show that the high tensile stress of micrometer thick, stoichiometric LPCVD SiN films can be efficiently relaxed through dense prestructuring of the substrate, allowing for crack-free high confinement waveguide processing with high yield.

The dense prestructuring consists of introducing a checkerboard structure (rectangle dimensions $5\mathrm{\mu m}\times 5\mathrm{\mu m}$) around the waveguides [Fig. 3(g)]. The checkerboard covers the full wafer, but the waveguides are surrounded by a 10 μm wide checkerboard free area to prevent scattering losses. While the pattern used here is a simple checkerboard, the stress release effect is not strongly dependent on the actual pattern. It was found that the stress control originates mainly from the surface modulation, which prevents the formation of a continuous thin film.

After deposition, no cracks are observed if the deposited film thickness is not exceeding the preetched depth by more than 100 nm. Importantly the dense prepatterning prevents crack formation also during the CMP. Additionally no wafer bending, due to asymmetric wafer coating with highly stressed thin films, is observed. Yet, if the planarization step does not remove all excess SiN, cracks can still occur during the post-deposition processing, as seen in the inset of Fig. 3(g). Advantageously, they do not penetrate the optical waveguide but are stopped at the pattern-free area around the waveguide.

Using the dense prepatterning, we deposit and process 1.2 μm thick, high stress SiN thin films, deposited without thermal cycling, and achieve a sample yield of more than 95%. Figure 4(a) shows a cleaved cross section of a resulting waveguide with a record thickness of 1.35 μm.

 

Fig. 4. Example of thick, high-$Q$ SiN resonators. (a) Cleaved cross section through a SiN waveguide (blue) with 1.35 μm height and 1.15 μm width fabricated using the photonic Damascene process. (b) Resonance linewidth at $\lambda =1598\mathrm{nm}$ of a microresonator with a radius of 238 μm ($\mathrm{FSR}\approx 100\mathrm{GHz}$) built from 1.35 μm high and 1.5 μm wide waveguides. A resonance linewidth of 50 MHz is extracted, which corresponds to a quality factor of $Q=3.7\times {10}^6$. The sidebands of the 400 MHz phase modulated beam for calibration are visible.

Download Full Size | PDF

5. DEVICE CHARACTERIZATION

We characterize microresonator devices fabricated using the photonic Damascene process with respect to their loss and dispersion. Figure 4(b) shows the linewidth of a resonator mode at $\lambda =1598\mathrm{nm}$. The resonator has a radius of 238 μm (free spectral range: $\mathrm{FSR}={\mathrm{D}}_1/2\pi \approx 100\mathrm{GHz}$), and its waveguide is 1.5 μm wide and 1.35 μm high. The resonance is strongly undercoupled so that the total linewidth $\kappa ={\kappa }_0+{\kappa }_{\mathrm{ex}}$ is dominated by the internal loss rate ${\kappa }_0$ and the external coupling rate ${\kappa }_{\mathrm{ex}}$ is small. An electro-optic modulator produced 400 MHz sidebands around the central resonance dip used to frequency calibrate the laser scan. By fitting the central transmission dip with a Lorentzian, a linewidth of $\kappa /2\pi =50\mathrm{MHz}$ is extracted, which corresponds to a quality factor of $Q=3.7\times {10}^6$.

Anomalous GVD and an undistorted mode structure are essential for the nonlinear performance of integrated microresonators, in particular the generation of frequency combs using DKS [24]. Here we use frequency comb assisted diode laser spectroscopy [25] to assess the dispersion and mode spectrum of a resonator with 238.2 μm radius, 1.75 μm waveguide width, and 0.85 μm waveguide height.

Figure 5(a) schematically shows the measurement setup used. The light of an external cavity diode laser (ECDL) is coupled into the device under test (DUT) using lensed fibers. By scanning the laser over 10 THz, the device’s transmission amplitude is recorded using photodiodes (PDs). The frequency calibration of the laser scan is achieved by simultaneously recording the beat signal of the ECDL and the teeth of a fiber laser frequency comb. The beat signal is filtered by two bandpass filters that produce reference marker peaks. For wavelength calibration the beat signal of the ECDL with a stable, narrow linewidth fiber laser is recorded, the wavelength of which is measured with an optical spectrum analyzer (OSA).

 

Fig. 5. Dispersion characterization of microresonators. (a) Schematic of the setup (adapted from [26]) used to record a frequency calibrated transmission amplitude of the device under test (DUT). The bandpass filters (BP) transform the beat signal between the scanning ECDL and the frequency comb (resp. the fiber laser) into marker peaks for relative (resp. absolute) frequency calibration. (b) Mode structure of a microresonator with 238.2 μm radius, 1.75 μm waveguide width, and 0.85 μm waveguide height. The plot shows the deviation of each detected mode from a FSR of 95.65 GHz. The $E_{11}^x$ mode family is underlaid in orange, and the $E_{11}^y$ mode family in yellow. (c) Magnified view of the local parabolic dispersion of the $E_{11}^y$ mode family. The detected resonances are fitted with a parabola (red dashed line), and a value of $D_2/2\pi =0.49\mathrm{MHz}$ is extracted (anomalous GVD).

Download Full Size | PDF

Due to the waveguide and material dispersion, the resonances of the microresonator are not equidistantly spaced, and the free spectral range (FSR) changes with frequency. This can be expressed as an integrated deviation $D_{\mathrm{int}}$ of the resonance frequency from an equidistant grid defined by the central FSR: $D_{\mathrm{int}}(\mu )\equiv {\omega }_{\mu }-({\omega }_0+D_1\mu )=D_2\frac{{\mu }^2}{2!}+D_3\frac{{\mu }^3}{3!}+\dots$, where $\mu \in \mathbb{Z}$ is the relative mode number counted from the central mode $\mu =0$, $D_1$ is the FSR of the central mode, and $D_2$ and $D_3$ are the second- and third-order dispersion coefficients.

Figure 5(b) shows an Echelle plot representation of the device’s mode spectrum. The plot shows every identified resonance in the transmission trace as one point and plots their individual deviation from a central FSR of 95.65 GHz. Resonances belonging to the same mode family are lined up, showing different slopes depending on their individual FSR. The fundamental mode families can be identified by comparing with finite element simulations of the resonator’s mode spectrum. Resonances belonging to the $E_{11}^x$ (${\mathrm{TE}}_{00}$) mode family are underlaid in orange, while the $E_{11}^y$ (${\mathrm{TM}}_{00}$) family is underlaid in yellow. Modes belonging to the higher-order mode families are only partly detected and cannot be clearly identified due to their very similar FSRs.

Figure 5(c) shows the parabolic deviation of the $E_{11}^y$ mode family from the central FSR of 95.65 GHz. By fitting with a polynomial the local dispersion of $D_2/2\pi =0.49\mathrm{MHz}$ is extracted, corresponding to anomalous GVD. This dispersion value is in approximate agreement with the value of $D_2/2\pi =0.26\mathrm{MHz}$ extracted from simulations (we attribute deviations to inaccuracy in the values of material dispersion). Moreover, avoided modal crossings of the $E_{11}^y$ mode family with the $E_{11}^x$ family as well as higher-order mode families are observed.

6. FREQUENCY COMB GENERATION IN THE SINGLE SOLITON REGIME

Next we demonstrate nonlinear frequency conversion with our resonators fabricated using the photonic Damascene process. Figure 6(a) shows an optical frequency comb generated upon tuning a 6 W laser into a resonance of a 95.7 GHz FSR resonator. The resonator has a radius of 238 μm, waveguide width of 1.5 μm, and height of 0.9 μm. As explained in Ref. [27], the resonator includes a 100 μm long, tapered single-mode waveguide section [Fig. 6(c)]. Within this section the resonator waveguide tapers down to only 0.5 μm width and 0.7 μm height in order to allow only propagation of the fundamental mode families. The device showed nearly critically coupled resonances with $\kappa /2\pi \approx 150\mathrm{MHz}$. The single-mode waveguide section efficiently suppressed modal crossings due to higher-order mode families while not deteriorating the device’s $Q$-factor [27].

 

Fig. 6. Frequency comb generation in the single soliton regime. (a) Dissipative Kerr soliton based frequency comb generation inside a 95.7 GHz SiN microresonator pumped with a 6 W laser at 192.9 THz. The red line is a fit of the spectral envelope with a ${\mathrm{sech}}^2$ function. A 3 dB bandwidth of 6.6 THz is extracted, corresponding to a soliton pulse duration of approximately 50 fs. (b) Statistics of step distribution in the converted comb light power based on 20 consecutive scans of a 3 W pump laser across the cavity resonance. The color-coded histogram reveals millisecond timescales for the steps and a higher occurrence probability for multisoliton states. (c) Schema of the device layout that comprises a 100 μm long single-mode waveguide section for higher-order mode suppression.

Download Full Size | PDF

Upon frequency scanning the pump laser from shorter to longer wavelengths across the resonance, we monitor the converted comb light, which includes all comb teeth except the central pump line. Distinct steps in the converted light power, as well as in the total transmitted light power, were observed. Previously such steps have been identified to be a characteristic sign for DKS formation inside a microresonator [12,14]. We obtain statistics of the step formation by repeatedly tuning a 3 W pump laser across the resonance with approximately 300 GHz/s scan rate, while recording the converted light power with an oscilloscope. Figure 6(b) shows a histogram representation (yellow and red denote higher occurrence rate) of 20 consecutive scans. The horizontal lines correspond to different multisoliton Kerr comb states with more than one pulse circulating in the microresonator: the more pulses circulate, the higher the converted light intensity. Notably, the observed step length in the millisecond timescale is significantly longer than previously observed in SiN microresonators [12], facilitating tuning into the single soliton states.

The frequency comb shown in Fig. 6(a) could thus be generated by tuning the pump laser inside the lowest step of the cascade using a laser tuning method, previously applied for crystalline resonators [14]. The frequency comb has a smooth spectral ${\mathrm{sech}}^2$ shaped envelope, characteristic for a single temporal dissipative soliton circulating inside the resonator. By fitting with a ${\mathrm{sech}}^2$ envelope function (shown in red), a spectral 3 dB bandwidth of 6.6 THz is extracted. Neglecting its dependence on pump detuning, the temporal full width at half maximum (FWHM) of dissipative solitons in critically coupled resonators is given as $\mathrm{\Delta }{t}_{\min }^{\mathrm{FWHM}}\approx 2\sqrt{\frac{-{\beta }_2}{\gamma \mathcal{F}{P}_{\mathrm{in}}}}$, where $\mathcal{F}$ is the resonator finesse, $P_{\mathrm{in}}$ is the coupled pump power, ${\beta }_2=-(n_0/c)(D_2/D_1^2)$ is the GVD, and $\gamma$ is the effective nonlinearity; see supplementary information of [14] for a complete theoretical treatment of the pulsewidth including the detuning. Based on the experimental parameters (pump power in waveguide $P=3.28\mathrm{W}$, finesse $\mathcal{F}=638$, effective nonlinearity $\gamma =1.41{\mathrm{W}}^{-1}{\mathrm{m}}^{-1}$, dispersion parameter ${\beta }_2=1.95{\mathrm{fs}}^2{\mathrm{m}}^{-1}$) we calculate a temporal width of 51.3 fs. This value is in close agreement with the pulse duration of 47.7 fs calculated for a transform-limited ${\mathrm{sech}}^2$ pulse using the measured spectral 3 dB bandwidth. Furthermore we note the offset between the pump line at 192.9 THz and the maximum of the hyperbolic secant envelope due to the Raman induced soliton self-frequency shift [28].

7. CONCLUSION

In summary, we have demonstrated a novel photonic Damascene process including a method for efficient thin film stress control. We fabricate SiN microresonators with so far unattainable waveguide dimensions and aspect ratios with close to unity yield. We show $Q$-factors of $3.7\times {10}^6$, on par with state of the art $Q$-factors obtained using subtractive processes [12,18,29], and achieve broadband temporal DKS based frequency comb generation with a 3 dB bandwidth of 6.6 THz. In the future the high yield and planar top surface of our process will enable integration of nonlinear waveguides with other photonic building blocks, e.g., via flip chip bonding integration [30] or novel optoelectronic 2D materials, such as graphene or ${\mathrm{MoS}}_2$ [31]. Additionally the large waveguide dimensions attainable are required for dispersion engineering of integrated SiN waveguides and microresonator frequency comb generation in the mid-infrared spectral region.