Mode spot converter based on gradient silica optical waveguide
Technical Field
The invention belongs to the technical field of integrated optoelectronics, and particularly relates to a mode spot converter based on a gradient silica optical waveguide, which is used for edge coupling of the silica optical waveguide and a single-mode optical fiber in optical path integration.
Background
The photon chip can be used in the fields of optical communication, big data calculation, artificial intelligent systems and the like. Functional integration in optoelectronics refers to the fabrication of multifunctional, high performance devices, such as photonic chips, by integrating elements of different functions together, where optical components (waveguides, modulators, switches, detectors, etc.) and electronic components (field effect transistors, etc.) are fabricated on the same substrate, thereby reducing area, cost, and power consumption. In optoelectronic integration, low-loss connection of an optical fiber and a waveguide is an important precondition for a high-performance chip, and is one of key problems to be solved.
Spot-size converters are commonly used for the connectorized coupling of optical fibers to waveguide devices on photonic chips. Since waveguides and optical fibers differ greatly in structure and size, coupling loss due to the difference in mode field size and shape becomes a major factor affecting insertion loss.
In order to realize the matching of the optical fiber mode field and the waveguide mode field, the structure and the preparation method of the mode spot converter are required to be optimized, and the effective refractive index matching and the mode field matching of the optical fiber and the waveguide are enhanced, so that the coupling loss is reduced, and meanwhile, the process complexity is reduced.
The integrated optical device based on the silicon dioxide material has the advantages of small optical loss, large process tolerance, compatibility with a CMOS process, good matching with a single-mode optical fiber mode field and the like, and has wide application in optical communication, optical interconnection and integrated optics.
Disclosure of Invention
The invention aims to provide a gradient silica optical waveguide-based mode spot converter which has high coupling efficiency, compact structure, low process complexity and insensitive wavelength.
The invention relates to a mode spot converter based on a gradient silicon dioxide optical waveguide, which is characterized in that:
1. The device consists of a basal layer, a lower cladding layer, a core layer waveguide and an upper cladding layer in sequence from bottom to top. As shown in fig. 1, the lower cladding layer is located on the base layer, the core waveguide and the upper cladding layer are located on the lower cladding layer together, and the core waveguide is covered by the upper cladding layer;
2. The lower cladding and the upper cladding are made of silicon dioxide, and the refractive indexes are 1.445; the core layer waveguide material is germanium-doped silicon dioxide, and the refractive index is 1.481; the basal layer is a silicon wafer, and the refractive index is 3.455;
3. As shown in fig. 2 (a), the core waveguide is located above the lower cladding layer and is covered by the upper cladding layer; as shown in fig. 2 (b) and 2 (c), the Core waveguide is constituted by an input waveguide Core 1, a tapered transition region waveguide Core 2, and an output waveguide Core 3 in the light transmission direction; The input waveguide Core 1 and the output waveguide Core 3 are straight waveguide structures, the Core 1 is rectangular in cross section perpendicular to the input light, the width W 1 =8 μm, Height H 1 = 6 μm, length L 1 = 25 μm; The width of the tapered transition region waveguide Core 2 is gradually narrowed in the light transmission direction, the input end width W 1 =8 μm, the output end width W 2 =3.5 μm, the waveguide height does not become H 1 =6 μm, Length L 2 = 75 μm; Core 3 has a rectangular structure in a cross section perpendicular to the input light, a width W 2 =3.5 μm, a height H 2 =3.5 μm, and a length L 3 =18 μm; The bottom surfaces of the input waveguide Core 1, the tapered transition region waveguide Core 2, and the output waveguide Core 3 are located in the same plane and are co-located above the upper surface of the lower cladding;
4. The effective optical mode field area of the input waveguide Core 1 of the Core layer waveguide is matched with the optical mode spot size of the optical fiber transmission optical field, the height of the input waveguide Core 1 of the coupling input end of the Core layer waveguide is the same as that of the waveguide Core 2 of the conical transition zone, and the height of the input waveguide Core 1 of the coupling input end of the Core layer waveguide is larger than that of the output waveguide Core 3, so that the optical mode spot size in the waveguide is changed from large to small.
The working principle of the spot-size converter is as follows:
The coupling efficiency of a waveguide and an optical fiber refers to the ratio of the energy of signal light coupled into the waveguide by the optical fiber to the total energy of the output light of the optical fiber (or the ratio of the energy of signal light coupled out of the waveguide into the optical fiber to the total energy of the output light of the waveguide). The coupling loss of a single-mode fiber and a silica optical waveguide mainly refers to the mode mismatch loss of the waveguide and the optical fiber caused by the difference of structure, size and effective refractive index. According to the optical mode principle, as the width of a silica optical waveguide gradually decreases, the size of the optical mode field supported by the silica optical waveguide also decreases. When the signal light from the optical fiber enters the tapered transition region waveguide Core 2 from the input waveguide Core 1 of the Core waveguide, the signal light mode field in the waveguide is gradually compressed and gradually changed from a circular optical fiber mode field distributed in a Gaussian manner (shown in figure 3) to an elliptic waveguide mode field distributed in an hermite manner (shown in figure 4) along with the width change of the tapered transition region waveguide Core 2; the mode field size supported by the input waveguide Core 1 of the Core layer waveguide is matched with the optical mode field size supported by the optical fiber, so that the optical loss caused by mode field mismatch can be effectively reduced; the height of the input waveguide Core 1 of the Core layer waveguide is the same as the height of the tapered transition region waveguide Core 2, and H 1,H1 is greater than the height H 2 of the output waveguide Core 3, so that when an optical signal enters the output waveguide Core 3 from the tapered transition region waveguide Core 2, the mode spot size is reduced, and low-loss mode spot conversion from single-mode optical fiber signal light to single-mode silica optical fiber signal light can be realized (as shown in fig. 5).
Compared with the prior device, the invention has the beneficial effects that: compared with the traditional optical fiber waveguide of the tapered mode spot converter, the optical fiber waveguide of the tapered mode spot converter has the same height (V.Vusirikala,S.S.Saini,R.E.Bartolo,S.Agarwala,R.D.Whaley.1.55-μm InGaAsP-InP laser arrays with integrated-mode expanders fabricated using a single epitaxial growth.IEEE Journal of Selected Topics in Quantum Electronics,1997,3(6):1332~1343),, the heights of the silicon dioxide input waveguide and the tapered transition region waveguide of the tapered mode spot converter are larger than the height of the single-mode silicon dioxide waveguide, and the mode field size of the silicon dioxide input waveguide is increased, so that the mode field matching of the silicon dioxide waveguide and an optical fiber output optical signal is enhanced. Through the gradient waveguide structure, the low-loss conversion from the Gaussian optical mode field transmitted by the optical fiber to the Hermite-Gaussian optical mode field can be realized, and the integration and encapsulation are facilitated. In different optical fiber-waveguide coupling structures, the structure has the characteristics of high coupling efficiency, compact structure, easy packaging, low process complexity and insensitivity to wavelength variation, and has wide application prospect.
Drawings
FIG. 1 is a schematic diagram of a mode spot-size converter based on a graded silica optical waveguide according to the present invention; wherein, each part name is: the base layer 4, the lower cladding layer 3, the core waveguide 1 and the upper cladding layer 2 are located together on the lower cladding layer 3 and the core waveguide 1 is coated in the upper cladding layer 2.
FIG. 2 is a cross-sectional view (a), top view (b) and side view (c) of an input end of a graded silica optical waveguide based spot-size converter according to the present invention;
FIG. 3 is a graph of the mode field of a circular fiber of Gaussian distribution transmitted in a single mode fiber and input waveguide Core 1;
FIG. 4 is a diagram of the mode field of an elliptical waveguide of hermite distribution transmitted in the output waveguide Core 3;
Fig. 5 is a diagram of a low loss optical mode field transmitted by the input waveguide Core 1, the tapered transition region Core 2, and the output waveguide Core 3 as a whole (mode spot-size converter);
FIG. 6 is a graph showing the coupling efficiency of a single mode fiber with a spot-size converter made in accordance with the present invention as a function of wavelength of light;
FIG. 7 is a flow chart of a process for preparing a graded silica optical waveguide-based spot-size converter according to the present invention; comprises 1, cleaning a silicon substrate; 2. thermally oxidizing and growing a low refractive index silicon dioxide lower cladding; 3. depositing a high refractive index silica core waveguide; 4. growing a polysilicon mask layer; 5. photoetching and polysilicon etching to form a waveguide mask pattern; 6. etching to form a silicon dioxide waveguide; 7. growing a polysilicon mask layer again; 8. photoetching and polysilicon etching to form a waveguide mask pattern; 9. etching the SiO 2 core layer; 10. removing the polysilicon mask layer; 11. and depositing an SiO 2 upper cladding layer.
Detailed Description
The invention will be described in further detail below with reference to the drawings by means of specific embodiments.
Example 1
1. The widths W 1 of the input ends of the Core waveguide's input waveguide Core 1 and the tapered transition region waveguide Core 2 are first determined, The output end width W 2 of the tapered transition region waveguide Core 2 and the width W 2 of the output waveguide Core 3, And a height H 2 of the output waveguide Core 3. The supported optical mode field is shown in FIG. 3, since the single mode fiber core diameter is 8.3 μm; the optical mode field distribution supported by SiO 2 single-mode waveguide Core 3 is shown in fig. 4. The input end widths W 1 = 8 μm of the Core waveguide input waveguide Core 1 and the tapered transition region waveguide Core 2 are determined from the beam propagation method and finite time domain difference method (Huei-Min Yang,Sun-Yuan Huang,Chao-Wei Lee,et al.High-Coupling Tapered Hyperbolic Fiber Microlens and Taper Asymmetry Effect.IEEE Journal of Lightwave Technology,2004,22(5):1395~1401) calculations, The output end width of the tapered transition region waveguide Core 2 and the output waveguide Core 3 have a width W 2 =3.5 μm, and the height of the output waveguide Core 3 is selected to be H 2 =3.5 μm.
2. The height H 1 of the Core input waveguides Core 1 and Core 2 is determined. The finite time domain difference method determines that when the heights of the input waveguide Core 1 and the conical transition region waveguide Core 2 are H 1 =6μm, the mode matching between the waveguide mode field of the mode spot converter and the optical fiber mode field can be effectively enhanced, and the conversion from the circular optical fiber mode field with Gaussian distribution to the oval waveguide mode field with hermite-Gaussian distribution can be realized.
3. The Core waveguide input waveguide Core 1 length L 1, the tapered transition region waveguide Core 2 length L 2, and the output waveguide Core 3 length L 3 are determined. Considering that the oversized device is not beneficial to integration and packaging, the length L 1 = 25 μm of the Core input waveguide Core 1, the length L 2 = 75 μm of the tapered transition region waveguide Core 2, and the length L 3 = 18 μm of the output waveguide Core 3 are determined by a finite time domain difference method.
4. Fig. 5 shows a diagram of the light field transmitted along the gradient-structured silica waveguide mode spot-size converter with signal light input by the Core 1, entering the cone-shaped transition zone Core 2 and the output waveguide Core 3. Therefore, the mode field size at the input end of the Core 1 of the Core waveguide is matched with the size of the signal optical spot in the optical fiber, so that the optical coupling efficiency can be effectively improved; the stepped waveguide can realize low-loss conversion from a large-size optical mode field in the Core 2 to a single-mode waveguide mode field in the Core 3, no obvious light leakage exists, and the structure can realize low-loss conversion from a Gaussian distributed optical fiber mode field to an hermite distributed waveguide mode field, so that coupling loss caused by mismatch of the optical mode field is effectively reduced.
5. FIG. 6 shows the variation of coupling efficiency of the silica waveguide mode spot-size converter with a single mode fiber according to the present invention with respect to the wavelength of light. The result shows that when the wavelength of the signal light is 1550nm, the coupling efficiency is 95.6%; the coupling efficiency of the spot-size converter is varied by 95-95.9% within the wavelength range of 1500-1620 nm.
Example 2
The following details the preparation mode of the invention with reference to fig. 7, the specific steps are as follows:
1. cleaning the silicon substrate (4): selecting a monocrystalline silicon wafer as a substrate, sequentially ultrasonically cleaning the silicon substrate by using acetone, ethanol and deionized water, and removing impurities on the surface of the silicon substrate;
2. Thermal oxidation growth of low refractive index silica lower cladding (3): growing a layer of silicon dioxide film as a lower cladding on a cleaned silicon substrate by a wet thermal oxidation method at 1000 ℃, and controlling the flow rate of water vapor, the temperature of the substrate and the reaction time to ensure that the thickness of the lower cladding of the low-refractive-index silicon dioxide is kept at 10 mu m;
3. Depositing a high refractive index silica core waveguide: depositing germanium (Ge) -doped high-refractive-index silicon dioxide by using a plasma enhanced chemical Vapor Deposition (PLASMA ENHANCED CHEMICAL Vapor Deposition, PECVD), controlling the flows of the reaction gases GeCl 4、SiH4 and N 2 O to be respectively 32sccm, 20sccm and 40sccm, controlling the radio frequency power to be 50W, and controlling the substrate temperature to be 200 ℃ and the reaction time to form a high-refractive-index silicon dioxide core waveguide with H 1 =6mu m;
4. And (3) growing a polysilicon mask layer: siH 4 and H 2 are used as reaction gases, a hot filament CVD method is adopted, and under the conditions that the growth air pressure is 1Pa, the substrate temperature is 200 ℃ and the dilution ratio V (H 2)/(V(SiH4)+V(H2))=98.4%, a layer of polycrystalline silicon layer with the thickness of 1 mu m is grown on the surface of the high refractive index silicon dioxide core layer waveguide by controlling the reaction time to serve as an etching mask layer of the high refractive index silicon dioxide core layer waveguide;
5. Photoetching and polysilicon etching to form a waveguide mask pattern: spin-coating photoresist AZ1500 on the surface of the polysilicon layer formed in the step 4, carrying out ultraviolet lithography and development on an i line (365 nm), transferring a waveguide pattern on the mask plate, which is the same as a silicon dioxide core layer waveguide structure to be prepared, to the surface of the photoresist, removing the polysilicon layer without a photoresist protection part by utilizing a Reactive Ion Etching (RIE) method, exposing and developing to remove the photoresist, and obtaining the polysilicon mask layer which is the same as the silicon dioxide core layer waveguide structure to be prepared; the polysilicon mask layer is composed of three parts, wherein one part is of a structure with the same size as the output waveguide Core 3, one part is of a structure with the same size as the conical transition area waveguide Core 2, and the third part is of a structure with the same size as the input waveguide Core 1;
6. Etching to form a silicon dioxide core layer waveguide: controlling the gas flow ratio of SF 6、CHF3 and O 2 by using the Reactive Ion Etching (RIE) method in the step 5, and removing the high refractive index silicon dioxide core layer waveguide of the part without the polysilicon mask protection by using the chemical corrosion and physical bombardment effect of fluorine ions;
7. And (4) regrowing a polysilicon mask layer: using SiH 4 and H 2 as reaction gases and adopting a hot filament CVD method, and under the conditions of 1Pa of growth air pressure, 200 ℃ of substrate temperature and dilution ratio V (H 2)/(V(SiH4)+V(H2))=98.4%, growing a polysilicon layer with the thickness of 2 mu m on the surface of the high refractive index silicon dioxide Core layer waveguide as an etching mask layer of the output waveguide Core 3 by controlling the reaction time;
8. Photoetching and polysilicon etching to form a waveguide mask pattern: spin-coating a photoresist AZ1500 on the surface of the polysilicon layer formed in the step 7, carrying out ultraviolet lithography and development on an i line (365 nm), transferring a waveguide pattern on a mask plate to the surface of the photoresist, removing the polysilicon layer without a photoresist protection part by utilizing a Reactive Ion Etching (RIE) method, exposing and developing to remove the photoresist, and obtaining the polysilicon mask layer with the same waveguide structure; the polysilicon mask layer is composed of two structural areas, wherein the A structural area is a structure with the same size as the output waveguide Core 3, the length of the B structural area is the sum of the lengths of the taper transition area waveguide Core 2 and the input waveguide Core 1, and the width is the width of the whole spot-size converter;
9. Etching the SiO 2 core layer: the SF 6、CHF3 and O 2 gas flows are controlled to be 10:1:1, enabling the transverse etching rate of fluorine ions to the silicon dioxide Core layer waveguide along the +/-y direction to be larger than the longitudinal etching rate along the-z direction, so as to partially remove the polysilicon mask layer of the A structure region and the high refractive index silicon dioxide Core layer waveguide covered by the polysilicon mask layer, and form an output waveguide Core 3 with the length of L 3, the width of W 2 and the height of H 2, wherein the output waveguide Core 3, the input waveguide Core 1 and the bottom surface of the conical transition region waveguide Core 2 are all in the same horizontal plane, namely the upper surface of a lower cladding layer;
10. Removing the polysilicon mask layer: removing the residual polysilicon mask layer by using 15% KOH aqueous solution by mass fraction;
11. Deposition of SiO 2 upper cladding (2): and (3) depositing low-refractive-index silicon dioxide on the surface of the device structure obtained in the step (10) by utilizing a PECVD method to serve as an upper cladding, controlling the thickness of a silicon dioxide upper cladding film by adjusting the gas flow, the reactant proportion, the radio frequency power and the time, and adopting a chemical mechanical polishing method to enable the thickness of the SiO 2 upper cladding to be 20 mu m, namely the thickness of the whole silicon dioxide cladding to be 10+20=30 mu m.