Patent Application
20250147383
The present invention relates to photonic integrated circuits. More specifically, certain embodiments of the invention relate to improved performance and control of frequency conversion elements realized as resonators in photonic integrated circuits.
A photonic integrated circuit (PIC) or integrated optical circuit is a device that integrates multiple photonic functions and as such is analogous to an electronic integrated circuit. The major difference between the two is that a photonic integrated circuit provides functions for information signals imposed on optical carrier waves.
Depending on the application, a PIC might operate at various frequencies or wavelengths. In the case of a datacom application, the PIC might for instance operate around 1.31 μm, while telecom applications might prefer PICs operating around 1.55 μm. Similarly, LIDAR application might utilize PICs operating around 940 nm, or 1.55 μm, while various quantum applications might utilize PICs operating in extremely broadband wavelength ranges from <400 nm to 1700 nm and beyond. Other applications might have different operating wavelengths. In many cases, it is beneficial if such wide spans can be coherently bridged using frequency conversion, such as e.g., frequency doubling or frequency tripling.
One example would be the frequency doubling (generating a second harmonic of) a narrow-linewidth InP laser operating around 1550 nm, to provide a similarly narrow linewidth source emitting at around 775 nm. A second example is addressing the green-yellow gap in available sources emitting between 535 nm and 630 nm. This wavelength region is very challenging to support with direct emission elements, with only LED sources (as opposed to laser sources) commercially available, due to the challenges associated with shifting the bandgap of GaN-based gain media to longer wavelengths. However, the wavelength “gap” region can be accessed by frequency doubling the output of GaAs sources operating between 1070 nm and 1260 nm, where direct emission from GaAs occurs with very high efficiency. Various other similar combinations can be envisioned, including cases where harmonics of higher order than the second are generated.
Recently, high efficiency PIC-based non-linear conversion has been demonstrated utilizing resonant elements that facilitate quasi-phase matched frequency conversion leveraging a photoinduced photo-galvanic effect (see e.g. Bohan Li et al., “High-coherence hybrid-integrated 780 nm source by self-injection-locked second-harmonic generation in a high-Q silicon-nitride resonator”, https://arxiv.org/abs/2306.10660 and Marco Clementi et al., “A chip-scale second-harmonic source via injection-locked all-optical poling” https://arxiv.org/abs/2307.00163). The effect leverages a doubly resonant configuration in which there is power buildup in a resonator both at the pump (fundamental) and at a doubled (second harmonic) or tripled (third harmonic) frequency. The all-optical poling effect was achieved via a photo-galvanic field-induced second-order non-linearity that facilitates efficient conversion. This approach is very powerful, as it enables efficient non-linear conversion (frequency doubling or tripling) using just high-Q resonators, without a need for precise phase matching of the two frequencies/wavelengths.
Historically, frequency doubling on PICs relied on either precise phase matching of modes using two polarizations, or quasi-phase matching by fabricating periodically poled (PP) structures; both of these approaches are very sensitive to fabrication tolerances. The resonator approach described in Bohan Li and Marco Clementi references is powerful, but in practice it has some limitations. One of the key limitations is that it can only provide a resonant condition at both fundamental and higher order harmonic frequencies at some, but not all, wavelengths (see Marco Clementi
The present invention is directed to improving functionality (so that conversion can be achieved at any arbitrary optical wavelength), performance (optimizing conversion efficiency) and yield (tuning control to account for fabrication variations). In particular, embodiments described below are concerned with the detailed design of photonic devices and with methods to control them to create high-performance, high-yield resonant based frequency converters, by utilizing at least two de-coupled tuning mechanisms for matching the resonances at the fundamental frequency and a higher order harmonic frequency at any wavelength inside the design range, which can be as wide or wider than 100s of GHz (typically but not always defined by resonators couplers).
Described herein are embodiments of a platform for the realization of photonic integrated circuits comprising frequency conversion elements with improved performance and yield, utilizing at least two de-coupled tuning mechanisms for the resonator at fundamental and harmonic frequency.
In the following detailed description, reference is made to the accompanying drawings which form a part hereof, wherein like numerals designate like parts throughout, and in which are shown by way of illustration embodiments in which the subject matter of the present disclosure may be practiced. It is to be understood that other embodiments may be utilized, and structural or logical changes may be made without departing from the scope of the present disclosure. Therefore, the following detailed description is not to be taken in a limiting sense, and the scope of embodiments is defined by the appended claims and their equivalents.
The description may use perspective-based descriptions such as top/bottom, in/out, over/under, and the like. Such descriptions are merely used to facilitate the discussion and are not intended to restrict the application of embodiments described herein to any particular orientation. The description may use the phrases “in an embodiment,” or “in embodiments,” which may each refer to one or more of the same or different embodiments. Furthermore, the terms “comprising,” “including,” “having,” and the like, as used with respect to embodiments of the present disclosure, are synonymous.
For the purposes of the present disclosure, the phrase “A and/or B” means (A), (B), or (A and B). For the purposes of the present disclosure, the phrase “A, B, and/or C” means (A), (B), (C), (A and B), (A and C), (B and C), or (A, B and C).
The term “coupled with,” along with its derivatives, may be used herein. “Coupled” may mean one or more of the following. “Coupled” may mean that two or more elements are in direct physical, electrical, or optical contact. However, “coupled” may also mean that two or more elements indirectly contact each other, but yet still cooperate or interact with each other, and may mean that one or more other elements are coupled or connected between the elements that are said to be coupled with each other. The term “directly coupled” means that two or more elements are in direct contact in at least part of their surfaces.
The non-linear element 110 can be realized in several ways as illustrated in views 150, 160 and 170. The approach illustrated in view 150 utilizes phase matching of two modes in the waveguide, which is commonly done by phase matching one TE mode to one TM mode (due to a large difference in wavelength between the modes) as described in e.g., Eric J. Stanton et al., Efficient second harmonic generation in nanophotonic GaAs-on-insulator waveguides”, Optica, 28, 9521-9532 (2020). Phase matching is generally challenging, as small fabrication variations of the waveguide dimensions can significantly reduce the efficiency of non-linear conversion and negatively impact yield.
The approach illustrated in 160 utilizes periodic polling (PP). PP can be created using a range of techniques such as pulsed electric field, electron bombardment, thermal pulsing, or other methods. A common material that is periodically polled for non-linear generation is lithium niobate (LiNbO3), see e.g., Lin Chang et al., “Thin film wavelength converters for photonic integrated circuits”, Optica, 3, 531-535 (2016), but other materials such as potassium titanyl phosphate (KTP), or lithium tantalate (LiTaO3) can be utilized. A general challenge with PP and materials used with PP is the lack of compatibility with established fabrication processes, in particular those of the silicon-based complementary metal-oxide semiconductor (CMOS) technology which is widely adopted by the electronics and silicon photonics market. This limits their deployment. Even in non-CMOS technologies, the PP process is generally challenging and, combined with waveguide dimension variations of these less developed processes, results in larger variation in the efficiency of non-linear conversion. and correspondingly impacts both the efficiency and the yield of such system.
The approach illustrated in 170 utilizes a resonant structure for power buildup where high quality factor and finesse of the resonator can increase intra-cavity powers to facilitate more efficient conversion. As described in e.g. Bohan Li et al., and Marco Clementi et al., (full references given above) the process can be very efficient if the resonator is configured such that both the fundamental and the higher order harmonic of interest are resonant (and therefore both exhibit a power buildup inside the resonator). This in turn leads to a photogalvanic-induced optical poling that is inherently phase matched and inherently compensates for fabrication variations (in contrast to prior art methods discussed above). A similar approach has also been demonstrated to result with frequency tripling (see e.g., Leiran Wang et al., “Frequency comb generation in the green using silicon nitride microresonators”, Laser & Photonics Review, 10, 631-638 (2016)). While the intrinsic self-alignment of optical poling significantly relaxes the fabrication tolerances of the waveguide dimensions, the need to match the resonances at both the fundamental and the second harmonic limits the technique as described in more details in Marco Clementi et al (full citation given above) if precise wavelengths are required (as in e.g., quantum applications utilizing atomic transitions), or if such systems are to be scaled in complexity because simply tuning the resonator utilizing e.g., a heater does not allow resonances to be matched at any arbitrary wavelength.
The views in
View 200 shows a cross-sectional view of an illustrative waveguide resonator geometry in which the optical mode 230 (operating at the fundamental frequency) is guided by a waveguide formed by layer 202 and layer 204. Layer 202 provides the core of the waveguide, and layer 204 provides the cladding for the optical mode. The optical mode propagates in the x-direction, while the lateral spread of the optical mode in the y-z plane is suggested by the dashed line 230. Layer 202, in some embodiments, comprises at least one of the following materials: silicon nitride (SiN), silicon oxynitride (SiNOx), titanium dioxide (TiO2), tantalum pentoxide (Ta2O5), (doped) silicon dioxide (SiO2), LiNbO3, LiTaO3 and aluminum nitride (AlN), while layer 204, in some embodiments, comprises at least one of the following materials SiO2, SiNOx, SiN. In all embodiments, the refractive index of layer 202 is larger than the refractive index of layer 204. The illustrative cross-section includes a substrate 205 that can be any suitable substrate for semiconductor and dielectric processing, such as silicon (Si), indium phosphide (InP), gallium arsenide (GaAs), quartz, sapphire, glass, gallium nitride (GaN), silicon-on-insulator or other materials known in the art.
Tuner element 220 positioned at a relatively large distance from waveguide core 202 in the view shown has no significant spatial overlap with the fundamental optical mode, meaning that it directly interacts with very little if any of the guided optical power in that mode. In some embodiments, the overlap might be <0.1%, meaning that “no significant” would be defined to be <0.1%; in others it might be larger, e.g., <5%. Two or more electrical connections 225 are used to provide control signals such as current and/or voltage to drive the tuner 220. In some embodiments, tuner 220 is a heater element or resistor that generates heat when current is passed through it. Examples of heater/resistor elements can be various metals, or doped semiconductors. Although there is no significant spatial overlap between the tuner 220 and the optical mode 230, the heat generated at tuner 220 when it is electrically driven propagates through cladding 204 and reaches waveguide core 202, changing the effective refractive index of both the cladding and the core, and correspondingly changing the optical path length for any guided mode, tuning the resonator as will be described in more detail with the help of
Tuner element 210, positioned significantly closer to the waveguide core 202 does have significant spatial overlap with the fundamental optical mode 230, meaning that, in some embodiments, it interacts with >0.1% of the guided optical power in that mode. In other embodiments, it interacts with >5% of guided optical power in that mode. Element 210 is a material that is optically transparent at the fundamental frequency, and that changes its own refractive index in response to an applied voltage or current. Examples of such materials are electro-optic materials (LiNbO3, LiTaO3, KTP, etc.) or transparent conductive oxide (TCO) materials such as indium-tin-oxide (ITO), zinc oxide (ZnO), aluminum-doped zinc oxide (AZO). In electro-optic materials, the refractive index of a medium can change with the application of an electric field, while in TCO materials, the refractive index can be changed by changing the carrier concentration either by doping and applying an electric field, or by applying a current that changes the carrier concentration. The voltage or current is applied through two or more electrical connections 215. The key difference between second tuner element 210 and first tuner element 220 is that the refractive index change caused by second tuner element 210 is localized to the tuner region, while in the case of first tuner element 220 the refractive index changes are indirect, as they are thermally induced, and can affect cladding 204 and core 202, even though there is no significant spatial overlap between tuner 220 and the mode. Other effects that can change the refractive index of a material locally, within the volume occupied by the material, while also providing optical transparency at the fundamental frequency can be used for second tuner element 210.
View 250 shows the same cross-sectional view of the illustrative waveguide geometry as shown in view 200, but now for an optical mode operating at the 2nd harmonic of the fundamental frequency. As the 2nd harmonic mode has a higher frequency than the fundamental, and a correspondingly shorter wavelength, it is more tightly confined around the waveguide core 202. The smaller lateral spread of the 2nd harmonic mode is indicated by the smaller size of the region bounded by the dotted line 280. In contrast to the fundamental optical mode at the fundamental frequency, the 2nd harmonic mode has no significant spatial overlap with either of the tuner elements 210 and 220, meaning that in some embodiments neither tuner element directly interacts with >0.1% of the guided optical power in the 2nd harmonic mode. In other embodiments, neither tuner element directly interacts with >5% of the guided optical power in the 2nd harmonic mode. This results in tuner 210 having negligible impact on the 2nd harmonic mode, as its tuning effect—directly changing refractive index—is confined to the region tuner 210 itself occupies, while tuner 220 can indirectly change the refractive index of the 2nd harmonic mode, as the heat it generates can reach and influence cladding 204 and core 202.
The end result is that as tuner 220 can have an impact on both the fundamental and the 2nd harmonic modes, while tuner 210 can impact only the fundamental mode, we can achieve de-coupled control of resonances at the two frequencies, as will be further explained with the help of
View 300 shows effective mode index calculated as a function of waveguide (WG) width at the fundamental and 2nd harmonic frequencies. This illustrative simulation, just as those discussed below for views 320, 340 and 360, assumes a waveguide geometry including a 100 nm thick SiN waveguide core layer in a SiO2 cladding, and fundamental mode calculations are performed for a wavelength of 1550 nm, while 2nd harmonic mode calculations are performed for a wavelength of 775 nm. It is immediately obvious that the effective index of the mode at the fundamental frequency is much lower than at the 2nd harmonic as the fundamental mode is less confined in the SiN core, and more of the optical mode resides in the SiO2 cladding, which of course has a lower refractive index than that of the core. The significant difference of index experienced by the fundamental and 2nd harmonic frequencies would be expected to make direct phase matching challenging.
View 320 shows thermo-refractive coefficient (dn/dT), corresponding to the change of the refractive index with temperature, calculated as a function of waveguide (WG) width, at the fundamental and 2nd harmonic frequencies. The thermo-refractive coefficient is a measure of thermal tunability of the mode and is determined by the materials used to make the waveguide. Si, for example, has a relatively large dn/dT (˜1.84e-4), while SiN has a significantly smaller dn/dT (˜2.5e-5), and SiO2 has an even lower dn/dT (˜1.1e-5). As the WG width is changed, the dn/dT of the mode changes as the confinement in SiN and SiO2 changes. While fundamental and 2nd harmonic modes have different dn/dT values, which allows some tunability to align both of the resonances, meaning that the optical path length for each mode can be adjusted, the two cannot be independently adjusted, so efficient non-linear frequency conversion is not possible at all temperatures and wavelengths as e.g. shown in FIG. 2 of Marco Clementi et al, cited above, which shows large regions with no efficient harmonic generation despite tuning the resonators through a relatively large temperature range, from ˜20° C. to ˜90° C.
View 340 shows mode sizes for fundamental and 2nd harmonic modes as a function of waveguide (WG) width. The two modes clearly have significantly different mode sizes, with the 2nd harmonic mode size being much smaller than the fundamental mode size. This means that a PIC can be designed with two tuner elements as shown in
View 360, relevant to more complex resonator embodiments such as those discussed below with respect to
Second resonator 404 is coupled to first resonator 403 using third coupler 433. Third coupler 433 is designed such that the fundamental frequency is coupled reasonably efficiently between the first resonator 403 and the second resonator 404, while the 2nd harmonic frequency is either not coupled at all or is coupled with an efficiency that is at least one order of magnitude lower. In some embodiments, reasonable efficiency of coupling can be defined as meaning a coupling value that is greater than 0.1%. In other embodiments reasonable coupling efficiency can be a different value (either larger or smaller), depending on the roundtrip loss of the second resonator 404 to provide substantial loading of the first resonator 403 at the fundamental frequency. This can easily be achieved by controlling the gap (pitch) between the two waveguides as illustrated with the help of
Method 600 begins with operation 605 in which the pump laser is turned on and coupled to the frequency conversion element. Once the laser is powered and suitably coupled to the frequency conversion element, the optimization of operation is initialized with operations 610, 615 and 620. In operation 610, the common tuner (corresponding to, for example, 220, 420) is operated with the goal of improving the frequency conversion efficiency. In operation 615, the fundamental tuner (corresponding to, for example, 210, 470) is operated with the goal of improving the frequency conversion efficiency. Operations 610 and 615 are shown to be sequential with 610 preceding 615 in method 600, but the order can be inverted, or they can be simultaneously executed. Once both tuners are adjusted, method 600 determines at operation 620 if a predetermined metric of efficiency (conversion efficiency from the fundamental mode to the 2nd harmonic mode) has been met. If not, the method goes back to operation 610 and performs another combination of operating fundamental and common tuners until the efficiency metric is met. Once that efficiency metric is met, the method proceeds to operation 625 in which the frequency converted output is utilized. The ability to independently operate at least the fundamental resonance condition utilizing a fundamental tuner enables the resonant architecture to achieve efficient non-linear conversion at any frequency by matching both resonances.
It is to be understood that these illustrative embodiments teach just some examples of frequency conversion elements utilizing the present invention, and many other, similar arrangements can be envisioned where tunings of the resonator at two frequency ranges is decoupled. Furthermore, frequency conversion elements can be combined with multiple other components to provide additional functionality or better performance such as various filtering elements, amplifiers, monitor photodiodes, modulators and/or other photonic components.
Embodiments of the present invention offer many benefits. The ability to de-couple tuning at two frequency ranges enables realization of highly efficient non-linear generation using resonant elements at essentially any wavelength, and to account for any fabrication variation. This in turn enables scalable manufacturing of PICs utilizing non-linear conversion including high density devices.
Embodiments of the optical devices described herein may be incorporated into various other devices and systems including, but not limited to, various computing and/or consumer electronic devices/appliances, communication systems, medical devices, sensors, and sensing systems.
It is to be understood that the disclosure teaches just a few examples of the illustrative embodiment and that many variations of the invention can easily be devised by those skilled in the art after reading this disclosure and that the scope of the present invention is to be determined by the following claims.
