Method for manufacturing an integrated electro-optical device, and electro-optical device thus obtained
Patent Information
- Application Number
- EP2024715195
- Authority / Receiving Office
- EP · EP
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2023-03-29
- Filing Date
- 2024-03-28
- Publication Date
- 2026-02-11
AI Technical Summary
Integrated electro-optical devices manufactured by annealed proton exchange (APE) exhibit degraded behavior under specific conditions such as vacuum and temperature, leading to fluctuations in half-wave voltage and slow drifts in low-frequency responses, resulting in unstable electro-optical performance.
A method involving surface machining to remove the residue of the layer modified by proton exchange, creating a machined zone between the electrodes, which reduces or eliminates temporal and frequency drifts, and stabilizes the electro-optical response by preserving the diffusion zone's optical guiding properties.
The method significantly reduces frequency and time-related drifts, maintaining stable electro-optical performance across varying environmental conditions and modulation frequencies, ensuring instantaneous and consistent response in electro-optical devices like phase modulators and Mach-Zehnder interferometers.
Smart Images

Figure EP2024058619_03102024_PF_FP_ABST
Abstract
Description
Method for manufacturing an integrated electro-optical device and the electro-optical device thus obtained Technical field of the invention
[0001] The present invention relates to the technical field of electro-optical devices and their manufacturing process.
[0002] It relates in particular to a manufacturing process for an electro-optical device integrated on an electro-optical substrate.
[0003] The invention also relates to an integrated electro-optical device having a time- and frequency-stabilized electro-optical response. The invention finds applications in an optical phase modulator and an optical intensity modulator of the Mach-Zehnder interferometer type. The invention also finds applications in an active directional coupler. State of the art
[0004] In the above field, it is known to fabricate and use an electro-optical device integrated on a planar electro-optical substrate, the electro-optical device comprising an optical waveguide and an electrode system adapted to modulate a light beam propagating in the optical waveguide by applying an electrical voltage across the terminals of the electrode system.
[0005] Various materials can be used to fabricate integrated electro-optical devices. Lithium niobate (LiNbOs), lithium tantalate (LiTaOs), and lithium antimoniate (LiSbOs) are particularly interesting because they exhibit the Pockels electro-optical effect. By positioning electrodes close to the integrated waveguide, it is possible to modulate the refractive index of the optical waveguide and thus modulate the phase of an optical signal propagating through it. In an integrated electro-optical device where the electrodes are separated by a few microns, applying a voltage of only a few volts is sufficient to generate an electric field and induce the desired phase modulation.
[0006] Various fabrication technologies for integrated electro-optical devices on lithium niobate substrates have been developed, such as diffusion of titanium (Ti) and annealed proton exchange (APE for Annealed Proton Exchange).
[0007] The proton exchange technology involves a first step in which a birefringent lithium niobate crystal is placed in a bath of benzoic acid heated to a temperature below 250°C (below the boiling point of benzoic acid), to replace lithium ions (Li + ) by protons (or H+ ions) +) from the surface exposed to the acid bath. This proton exchange (PE) process results in the formation of a proton exchange-modified layer 3 extending over the acid-exposed area from the substrate surface 2 to a shallow depth P3 within the substrate 1 (see Figure 1). Layer 3 exhibits a refractive index difference of approximately 0.1 compared to the substrate. The refractive index profile of layer 3 is step-shaped in the X and Z directions. Layer 3 is laterally limited to a width W3, typically between 1 pm and 10 pm. The depth P3 is less than or equal to 500 nm. Indeed, because the temperature is low, the proton exchange process cannot diffuse very far into the substrate.
[0008] This first step is followed by a thermal annealing step at a temperature below 500°C. During the annealing step, the higher temperature allows protons, or hydrogen ions, from the proton-exchange modified layer 3 to exchange with lithium ions in a scattering zone 4 around the initial location of layer 3. Annealing creates a refractive index gradient in the scattering zone 4 beneath the crystal surface, thus forming an optical waveguide 5. The scattering zone 4 extends into the substrate to a considerable depth P4, potentially reaching several microns, depending on the temperature and duration of the annealing (see Figure 2). The scattering zone 4 has a width W4 greater than the width W3 of the proton-exchange modified layer 3. The width W4 is typically between 1 pm and 10 pm.The width W4 and respectively the depth P4 are here measured at the maximum of the refractive index gradient divided by the number e in the base of natural logarithms (or exponential (-1 )), along the Z and X directions respectively.
[0009] After the thermal annealing step, the fabrication of the integrated electro-optical device includes a step of depositing a pair of electrodes. This pair electrodes allow an electric field to be applied in the optical waveguide 5 in order to control the electro-optical device, for example for the electro-optical modulation of an optical wave propagating in the optical waveguide 5.
[0010] Proton exchange on a birefringent lithium niobate crystal has the effect of both increasing the extraordinary index of the crystal, which creates guidance for a wave polarized along the extraordinary axis of the crystal, and reducing the ordinary index of the crystal, so that a wave polarized along the ordinary axis is not guided.
[0011] In electro-optical devices incorporating a proton-exchange optical waveguide on lithium niobate, the typical configuration is an X-section, with the X-axis of the uniaxial birefringent lithium niobate crystal perpendicular to the substrate surface, while the Y and Z axes of the crystal are parallel to the surface. The waveguide propagation direction is parallel to the Y-axis, and the TE mode (electric transverse, i.e., electric field parallel to the substrate surface) is parallel to the Z-direction. In this case, the proton-exchange optical waveguide guides only the TE polarization state; the TM cross-polarization state (magnetic transverse, i.e., magnetic field parallel to the substrate surface and therefore electric field perpendicular to the substrate surface) propagates freely within the substrate.The lithium niobate proton exchange technique thus makes it possible to fabricate a polarizing waveguide on an integrated optical circuit. Electrodes arranged on both sides of the optical waveguide 5 allow the application of an electric field parallel to the Z-axis within the optical waveguide 5.
[0012] There are also electro-optical devices comprising an optical waveguide formed by proton exchange on lithium niobate (or lithium tantalate or lithium antimoniate), following a Z-section. The Z-axis of the uniaxial birefringent lithium niobate crystal is perpendicular to the substrate surface, while the X and Y axes of the crystal are parallel to the surface. The waveguide propagation direction is parallel to the Y-axis, and the TE mode (electric transverse, i.e., electric field parallel to the substrate surface) is parallel to the Z-direction. One electrode covers the waveguide, while the other electrode is positioned next to the waveguide or on an opposite face of the substrate. These electrodes allow the application of an electric field parallel to the Z-axis within the optical waveguide. Moreover, in this case, the metallic electrode covering the waveguide absorbs the TM mode of the electromagnetic wave propagating in the optical waveguide 5 while it transmits the TE mode parallel to the Z axis.
[0013] However, the behavior of integrated electro-optical devices manufactured using APE is degraded under specific operating conditions such as vacuum and / or temperature. For example, an electro-optical phase modulator manufactured using APE has a transfer function defined by the ratio between an applied voltage V m across the terminals of the modulator electrodes and the optical phase shift <t> m induced, with V m >m = VK In where the signal VK is also called the half-wave voltage (VK). However, this signal VK fluctuates with the environment, particularly with temperature and / or vacuum.
[0014] Furthermore, the low frequency (between 100 Hz and 1 kHz) and very low frequency (between 0.0 Hz and 100 Hz) response of an electro-optical device made of APE is likened in some applications to a slow drift of the operating points or to a measurable hysteresis in the transfer function of the electro-optical device.
[0015] Several solutions to the time drift of an integrated optical phase modulator have already been proposed. For example, US patent 8,463,081 proposes combining in series a portion of an optical waveguide formed by titanium scattering and another portion of an optical waveguide formed by APE, the two portions having time drifts of opposite signs. However, the compensation is not perfect but depends on the temperature and the frequency of the signal applied at low and very low frequencies.
[0016] One of the aims of the invention is to provide a means of stabilizing the electro-optical response of an integrated electro-optical device as a function of time and / or the frequency of the modulation signal applied to the electrodes of such an integrated electro-optical device. It is desirable to reduce the slow drift of the operating points at low and very low frequencies. It is also desirable to eliminate or limit the hysteresis of the transfer function of such an integrated electro-optical device. In particular, in an electro-optical phase modulator based on an integrated electro-optical device, it is desirable to reduce fluctuations in the VK signal with the environment, particularly with temperature and / or vacuum.
[0017] Another objective of the invention is to propose a simple manufacturing method for such an integrated electro-optical device having an instantaneous electro-optical response, stable over time and / or as a function of the electrical modulation frequency, and independent of ambient variations in temperature and / or pressure. Presentation of the invention
[0018] In this context, the present invention proposes a method for manufacturing an electro-optical device.
[0019] More particularly, the invention proposes a method for manufacturing an electro-optical device comprising the following steps: manufacturing an optical waveguide in an electro-optical substrate by annealed proton exchange comprising the steps of: a) exposing an area of a surface of the substrate to a proton source so as to create in the substrate a proton exchange modified layer, said layer extending over said area from the surface of the substrate to a shallow depth (P3) in the substrate, and b) thermally annealing the substrate so as to diffuse hydrogen ions from said layer into a diffusion zone extending in the substrate around said layer to a deep depth (P4) so as to create an optical waveguide in the diffusion zone; and deposition of a pair of electrodes, the optical waveguide extending longitudinally between said pair of electrodes.
[0020] According to the invention, the process comprises, after the thermal annealing step, a surface machining step from the surface of at least a part of the substrate in an area of the optical waveguide corresponding to a residue of said layer to form a machined area located at least partly between said pair of electrodes, over a depth P of the order of the small depth of the layer modified by proton exchange.
[0021] Thus, the machining step allows for the removal, at least partially, of a residue from the proton-exchange modified layer, whose electro-optical properties are damaged by the proton exchange. Removing this residue reduces or even eliminates temporal and / or frequency drifts of the electro-optical device, when an electrical voltage is applied between the terminals of the electrodes.
[0022] Other non-limiting and advantageous features of the system according to the invention, taken individually or in all technically possible combinations, are as follows: - the machined zone extends into the substrate to a depth P, the depth P being greater than or equal to 0.02*P3 and less than or equal to 1.10*P3, where P3 is the shallow depth of the proton exchange modified layer before thermal annealing; - the machined area extends over at least part of a deposition surface of said pair of electrodes; - the machined area extends over the entire width of the substrate taken transversely to a longitudinal direction of the optical waveguide; - the machined area is limited to an area of the waveguide between said pair of electrodes; - the machining step is adapted to form several machined zones arranged in series along a longitudinal direction of the optical waveguide, said several machined zones being located at least partly between said pair of electrodes; - the machined area extends over the entire surface of the substrate; - the machining stage is carried out by wet engraving, plasma engraving, laser, lapping or mechanical polishing; - the process includes a step of depositing a dielectric layer, the dielectric layer being disposed on the machined area on the surface of the substrate and / or between the substrate and at least one electrode of the electrode pair.
[0023] The invention also relates to an electro-optical device obtained according to any one of the embodiments described.
[0024] In particular, the invention relates to such an electro-optical device, used in an electro-optical phase modulator, an optical intensity modulator, a Mach-Zehnder interferometer and / or an active directional evanescent wave coupler.
[0025] Of course, the different features, variants and embodiments of the invention can be combined with each other according to various combinations insofar as they are not incompatible or mutually exclusive. Brief description of the drawings
[0026] Furthermore, various other features of the invention become apparent from the attached description made with reference to the drawings which illustrate non-limiting embodiments of the invention and where:
[0027] Figure 1 is a cross-sectional view of an electro-optical substrate within which a proton-exchange modified layer is formed before annealing;
[0028] Figure 2 is a cross-sectional view of the electro-optical substrate after annealing to form a diffusion zone;
[0029] Figure 3 is a cross-sectional view of an integrated electro-optical device comprising an optical waveguide formed by APE and electrodes;
[0030] Figure 4 is a graph schematically illustrating the response of an electro-optical modulator manufactured by APE as a function of frequency, according to the prior art (solid line) and according to the present disclosure (dashed lines);
[0031] Figure 5 is a graph schematically illustrating the half-wave voltage, characteristic of the transfer function of an electro-optical modulator manufactured by APE as a function of frequency, according to the prior art (solid line) and according to the present disclosure (dashed lines);
[0032] Figure 6 is a graph schematically illustrating, on the left, the voltage applied across the terminals of an electro-optical modulator manufactured by APE as a function of time, and on the right, an example of phase shift produced respectively according to the prior art (solid line) and according to the present disclosure (dashed lines);
[0033] Figure 7 is a top view of an integrated electro-optical device comprising an optical waveguide formed by APE, electrodes and a machined area according to a first example;
[0034] Figure 8 is a top view of an integrated electro-optical device comprising an optical waveguide formed by APE, electrodes and a machined area according to a second example;
[0035] Figure 9 is a cross-sectional view of an electro-optical device integrated according to the first or second example;
[0036] Figure 10 is a top view of an integrated electro-optical device comprising an optical waveguide formed by APE, electrodes and a machined area according to a third example;
[0037] Figure 11 is a top view of an integrated electro-optical device comprising an optical waveguide formed by APE, electrodes and a machined area according to a fourth example;
[0038] Figure 12 is a top view of an integrated electro-optical device comprising an optical waveguide formed by APE, electrodes and a machined area according to a fifth example;
[0039] Figure 13 is a top view of an integrated electro-optical device comprising an optical waveguide formed by APE, electrodes and a machined area according to a sixth example;
[0040] Figure 14 is a cross-sectional view of an electro-optical device integrated according to any one of the third to sixth examples;
[0041] Figure 15 is a cross-sectional view of an electro-optical device integrated according to a variant of any one of the third to sixth examples;
[0042] Figure 16 is a cross-sectional view of a variant of an integrated electro-optical device comprising an optical waveguide formed by APE and electrodes;
[0043] Figure 17 is a cross-sectional view of another variant of an integrated electro-optical device comprising an optical waveguide formed by APE and electrodes.
[0044] It should be noted that in these figures the structural and / or functional elements common to the different variants may have the same references. Detailed description
[0045] Figure 3 schematically represents an integrated electro-optical device in cross-sectional view, i.e., perpendicular to the surface 2 of the electro-optical substrate 1 and transverse to a longitudinal direction of the guide optical waveguide 5. The integrated electro-optical device includes an optical waveguide 5 formed by APE and electrodes 6,7.
[0046] The electro-optical substrate 1 is, for example, a lithium niobate (LiNbOs) substrate. Alternatively, substrate 1 is a lithium tantalate (LiTaOs) or lithium antimoniate (LiSbOs) substrate. As a non-limiting example, the substrate is in an X-section configuration, with the X-axis of the crystal perpendicular to the substrate surface, while the Y and Z axes of the crystal are parallel to the surface, the Y-axis being parallel to the propagation direction of the optical waveguide. Alternatively, the substrate is in a Z-section configuration, with the Z-axis of the crystal perpendicular to the substrate surface, while the X and Y axes of the crystal are parallel to the surface, the Y-axis being parallel to the propagation direction of the optical waveguide (see Figures 16-17).
[0047] As described in relation to Figures 1 and 2, a proton-exchange modified layer 3 is formed in substrate 1. During thermal annealing, a diffusion zone 4 extends into substrate 1 around the initial location of layer 3. The diffusion zone 4 exhibits a refractive index gradient that allows the formation of an optical waveguide 5. The optical waveguide 5 is bounded by the refractive index gradient and by the surface 2 of substrate 1. The width W4 of the optical waveguide 5 is typically between 1 micrometer (pm) and 10 micrometers (pm). The optical waveguide 5 extends, for example, longitudinally along the Y direction. As is known, the optical waveguide 5 can be straight or have a curvature and / or a multi-branch junction, for example, a two-branch Y junction.The optical waveguide 5 is adapted to guide a single-mode optical wave so that it propagates along the longitudinal direction of the optical waveguide 5, for example here along the Y axis. Figure 3 schematically represents the spatial extent of the fundamental mode 15 of an optical wave propagating in the optical waveguide 5.
[0048] The electro-optical device illustrated in Figure 3 is in an X-section and Y-propagation configuration. The integrated electro-optical device comprises an electrode system including at least one pair of conductive electrodes 6, 7. Electrode 6 is deposited on one side of the optical waveguide 5 and electrode 7 is deposited on the opposite side of the optical waveguide 5. The electrode pair 6, 7 is generally deposited on the surface 2 of the substrate 1.
[0049] Alternatively, a dielectric layer 24, for example of silicon dioxide (SiCh) or silicon nitride (SiN), is placed between the surface 2 of the substrate 1 and each electrode 6, 7. The thickness of the dielectric layer 24 is, for example, between 100 nm and 2 pm. The dielectric layer 24 serves as protection for the surface of the optical waveguide 5. The dielectric layer 24 also isolates the optical wave from the metallic electrodes, for example, if the optical waveguide is very close to or may be partially in contact with an electrode, particularly for very broadband electrodes (operating in the range from 0 to several tens of GHz).
[0050] For clarity, we consider a straight optical waveguide 5 and a pair of straight electrodes 6, 7. The optical waveguide 5 extends longitudinally between said pair of electrodes 6, 7. The distance between the two electrodes 6, 7 is generally greater than or equal to the width W4 of the optical waveguide 5. For example, the distance D between the two electrodes 6, 7 is between 1 micrometer (pm) and 10 pm. The length L6 of the electrodes in the Y direction is generally between a few millimeters and several centimeters. The electrodes may have the same width or different widths. The optical waveguide 5 is centered or not between the two electrodes 6, 7. Optionally, another electrode is disposed on the rear face 9 of the substrate 1, that is to say the face opposite to the front face 2 where the optical waveguide is formed by APE.
[0051] In connection with Figure 3, we will now explain the electro-optical operation of the integrated electro-optical device. An electrical voltage is applied across the electrode pair 6, 7. For example, a positive voltage is applied to electrode 6 and a negative voltage to electrode 7. Alternatively, one of the two electrodes is grounded. Electric field lines Ί appear between the electrode pair 6, 7. In particular, electric field lines Ί oriented parallel to the Z-axis pass through the optical waveguide 5. As is known, modulating the electrical voltage applied between electrodes 6, 7 allows the optical phase shift of an optical wave propagating in the optical waveguide 5 to be modulated.
[0052] However, the electro-optical operation of the electro-optical device may exhibit certain drifts, illustrated in figures 4 to 6.
[0053] Figure 4 schematically represents, in solid lines, a response curve 16 of an electro-optical phase modulator manufactured by APE as a function of frequency according to the prior art. More precisely, curve 16 represents the ratio between the optical phase shift A <t>induced in the optical wave propagating in the optical waveguide 5 and the applied voltage V as a function of the modulation frequency f. At high frequencies, for example above 1 kHz, the ratio ± A <t>A / is constant. However, at low and very low frequencies, the ratio ± A <t>A / decreases drastically. The response of the electro-optical phase modulator manufactured by APE according to the prior art depends on the modulation frequency, which is undesirable.
[0054] Figure 5 schematically represents, in solid lines, a half-wave voltage curve (or VK signal) of an electro-optical phase or current modulator manufactured by APE as a function of frequency, according to the prior art. More precisely, curve 18 represents the VK signal as a function of the modulation frequency f of the voltage applied across electrodes 6 and 7. At high frequencies, the VK signal is constant. Conversely, at low and very low frequencies, the VK signal increases sharply. This curve 18 also indicates that the response of the electro-optical phase or current modulator manufactured by APE according to the prior art is adversely dependent on the modulation frequency f.
[0055] Figure 6 shows on the left a curve 20 of the voltage V applied between electrodes 6 and 7 as a function of time. The voltage V increases gradually in steps and then decreases gradually, also in steps. Figure 6 shows on the right a curve 21 of the optical phase shift A <t>The induced optical phase shift in the optical wave propagating in the optical waveguide 5 is a function of time, that is, a function of the applied voltage V. On curve 21 of the induced optical phase shift, we observe, for example, a step with a negative exponential shape when the voltage increases and a step with a positive exponential shape when the voltage decreases. In all cases, at each step of increasing or decreasing voltage, a stable value of the induced optical phase shift is only reached after a time that can range from one millisecond to several hours or even longer.
[0056] It follows from this disclosure that these slow drifts of the operating point of an electro-optical modulator as a function of time and / or modulation frequency can be greatly reduced by machining a portion of the substrate in a region of the optical waveguide 5 to a shallow depth corresponding to a residue of the proton-exchange modified layer 3 located between the electrode pair 6, 7. As schematically illustrated in Figure 3, the residue of the proton-exchange modified layer 3 extends from the surface 2 of the substrate to a shallow depth in the optical waveguide 5, and perpendicular to the longitudinal axis of the optical waveguide.
[0057] Without being bound by a theory, a physical interpretation is that during the proton exchange step, some of the lithium ions of substrate 1, for example lithium niobate, are substituted by protons or H+ ions + to form the proton-exchange modified layer 3. In cross-section along the X-axis, the proton-exchange modified layer 3 exhibits a step-shaped refractive index profile down to depth P3. The substitution rate in layer 3 can reach 60% or more of H ions + As a result, the crystalline structure of the proton-exchange modified layer 3 is degraded compared to the crystalline structure of substrate 1 around the initial location of the proton-exchange modified layer 3. In the layer 3 region, the high ion exchange rate locally induces a loss of the electro-optical properties of lithium niobate, which are necessary for any electro-optical modulation. During the thermal annealing step below 500°C, the scattering region 4, which forms the optical waveguide 5, extends into substrate 1. During thermal annealing, H+ ions + ions originating from layer 3 are exchanged with lithium ions at greater depths, for example, down to a depth of several microns, depending on the temperature and duration of the annealing. Annealing creates a refractive index gradient in the substrate, thus forming the optical waveguide 5. In cross-section along the X-axis, the scattering zone 4 generally exhibits a refractive index profile that can be approximated by a semi-Gaussian curve. The depth P4 corresponds to the maximum of this semi-Gaussian profile along the X-axis divided by the number e in the base of the natural logarithms (or exponential (-1)). Similarly, the width W4 of the optical waveguide 5 corresponds to the maximum of this profile along the Z-axis divided by the number e. During annealing In thermal annealing, the ion substitution rate in the diffusion zone is lower, generally less than 15%. This low lithium ion substitution rate helps preserve the electro-optical properties of the crystal. Following thermal annealing, the lithium ion substitution rate in the zone corresponding to the location of the proton-exchange modified layer 3 has decreased. Nevertheless, the crystal structure at the location of the proton-exchange modified layer 3 remains degraded. The electro-optical device conventionally fabricated by APE exhibits different crystalline properties in substrate 1, in the diffusion zone 4, and in the zone corresponding to the proton-exchange modified layer 3.
[0058] As illustrated in Figure 3, the electric field lines 14 pass not only through scattering zone 4 but also through a region beneath the substrate surface corresponding to a remnant of the proton-exchange-modified layer 3. The electro-optical response of the electro-optical device depends not only on the electro-optical properties of scattering zone 4 but also on the region corresponding to the location of the proton-exchange-modified layer 3. The crystalline degradation of the layer 3 region leads to an increase in its electrical conductivity when the fabricated electro-optical device is placed at a warm temperature, for example, around 50°C. This effect is accelerated and amplified as the temperature increases and when the electro-optical device is placed under a hot vacuum.
[0059] This disclosure proposes an additional step in the process of manufacturing an electro-optical device by proton exchange (PE). This additional step, following the thermal annealing step, comprises machining at least a portion of the optical waveguide 5 to a shallow depth corresponding to the residue of layer 3 in an area located or intended to be placed between the electrode pair 6, 7, so as to reduce or eliminate the residue of layer 3 in the machined area. The depth P of the machined area is on the order of the shallow depth P3 of the proton exchange-modified layer 3. In practice, the depth P may be less than, equal to, or slightly greater than the shallow depth P3. For example, the depth P of the machined zone is such that: 0.02*P3 < P < 1.10*P3, and preferably 0.50*P3 < P < 1.10*P3 and even more preferably 0.90*P3 < P < 1.10*P3.For example, for a depth P3 of 500 nm, the depth P of the machined area is approximately 500 nm ± 50 nm. The machining step is performed on. a shallow depth P to retain most of the scattering zone 4. The depth P of the machined zone is always less than the depth P4 of the scattering zone 4 so as not to degrade the optical guidance properties of the optical waveguide 5. Preferably, the depth P of the machined zone is less than 20-25% of the depth P4 of the scattering zone 4.
[0060] This machining step of the residue of said layer 3 can be carried out by different techniques: for example mechanical machining such as honing and / or polishing of the surface of the substrate or a part of the surface of the substrate, surface ablation by laser pulses, wet etching or dry etching by plasma.
[0061] This machining step of the residue from layer 3 can be performed on a wafer incorporating several batch-manufactured electro-optical devices or on an individual chip obtained, for example, after cutting such a wafer, each chip containing a single electro-optical device. The machining step of the residue from layer 3 can be performed before or after the deposition of electrodes 6 and 7 on a complete wafer or on an individual chip. The machining step of the residue from layer 3 can also be performed before or after transferring a chip onto a base or into an electronic package.
[0062] Figures 7 to 15 illustrate various examples of the embodiment of an electro-optical device according to this disclosure. In Figures 7 to 15, one or more machined areas 10, 11, 12 are shown on each electro-optical device.
[0063] In Figure 7, the top view of the electro-optical device 50 comprises an optical waveguide 5 formed by proton exchange annealed in an electro-optical substrate 1, and a pair of electrodes 6, 7 arranged parallel to the two longitudinal sides of the optical waveguide 5. The electro-optical device includes a machined area 10 extending between the two electrodes 6, 7 over a width W and a length L. The width W of the machined area 10 is less than or equal to the distance D between the two electrodes 6, 7. The length L of the machined area 10 is equal to the length L6 of the electrodes 6, 7. The width W of the machined area 10 extends along the optical waveguide. The width W is measured perpendicular to the longitudinal axis of the optical waveguide 5 and parallel to the plane of the surface 2. Advantageously, the width W is greater than or equal to the initial width W3 of the proton-exchange modified layer 3. Machining the residue of said layer 3 over this width W and along the entire length L6 of the electrodes 6, 7 completely eliminates the frequency and time drifts of the electro-optical device, as detailed below. In this configuration, machining of the machined area 10 can be performed before or after electrode deposition. However, when this machining is performed after electrode deposition, precautions must be taken to avoid damaging the electrodes 6, 7.
[0064] Figure 8 shows a second example of an electro-optical device, based on a variant of Figure 7. The machined area 10 here has a length L less than the length L6 of the electrodes 6 and 7, for example, such that 0.5*L6 < L < L6. The width W of the machined area 10 is less than or equal to the distance D between the two electrodes 6 and 7. This example illustrates certain configurations where it is more difficult to machine the area 10 over the entire length L6 of the electrodes. In this example, the machining of the machined area 10 can be performed before or after electrode deposition. In some applications, machining the area 10 over a portion of the electrode length is sufficient to limit the frequency and time drifts of the electro-optical device. In this case, the time and frequency drift phenomena are reduced proportionally to the machined area.
[0065] Figure 9 shows a cross-sectional view of the electro-optical device 50 of Figure 7 or 8. The machined area 10 corresponds here to a recessed area in the substrate, below the initial substrate surface 2 and perpendicular to the longitudinal axis of the optical waveguide. The machined area 10 has a depth P. For example, the depth P is between 10 nm and 550 nm for a depth P3 of 500 nm. The depth P of the machined area is less than, equal to, or slightly greater than the shallow depth P3 of the proton-exchange modified layer 3. In all cases, the depth P of the machined area is less than the initial depth P4 of the scattering area. Therefore, below the machined area between electrodes 6 and 7, the scattering area 5 has a depth equal to P4 - P. In general, the remaining depth, P4-P, of the optical waveguide 5 is much greater than the depth P of the machined area. For example, the remaining depth, P4-P, of optical waveguide 5 is greater than 75% of the initial depth of optical waveguide 5 before machining.
[0066] The machined area 10 of the electro-optical device 50 of figure 7, 8 or 9 can be machined before or after the deposition of the electrodes 6, 7.
[0067] Figure 10 shows a third example of an electro-optical device 50. The machined area 10 has a length L that is less than the length L6 of the electrodes 6 and 7. The width W of the machined area 10 is greater than the total width T of the electrodes 6 and 7. The width W of the machined area 10 is, for example, equal to the width of the substrate 1, which facilitates the use of mechanical machining. The machined area is then shaped before the electrodes are deposited.
[0068] Figure 11 shows a fourth example of an electro-optical device 50. The machined area 10 here has a length L equal to the length L6 of the electrodes 6, 7. The width W of the machined area 10 is here greater than the total width T of the electrodes 6, 7. The width W of the machined area 10 is for example equal to the width of the substrate 1 which facilitates the use of mechanical machining.
[0069] Figure 12 shows a fifth example of an electro-optical device 50. The machined area 10 has a length L greater than the length L6 of the electrodes 6 and 7. The width W of the machined area 10 is greater than the total width of the electrodes 6 and 7. The machined area 10 extends, for example, over the entire surface of the substrate 1. The machined area 10 can be shaped by etching, lapping, and / or polishing the surface so that the electro-optical device has a flat surface outside the electrodes. Machining the entire surface ensures the removal of all residual parts of the proton-exchange-modified layer 3. Furthermore, machining the entire surface 2 is easy to implement because it is a full-plate machining operation without masking.
[0070] Alternatively, the electro-optical device 50 illustrated in any one of Figures 10 to 12 includes one or more disjoint machined areas located outside the electrodes 6, 7. For example, in Figure 10 or 11, the device can also be shaped by forming a machined area located between the two electrodes 6, 7 and two machined areas outside the electrodes 6, 7. In this case, the disjoint machined area or areas can be shaped before or after the deposition of the electrodes 6, 7.
[0071] Figure 13 shows a sixth example of an electro-optical device 50 comprising several machined zones 10, 11, 12 arranged in series along the longitudinal direction Y of the optical waveguide 5. More precisely, a machined zone 10 extends over a length L less than the length L6 of the electrodes 6, 7 and over a width W greater than the total width of the electrodes 6, 7, for example, over the width of the substrate 1. Another machined zone 11 extends over a length L1 less than the length L6 of the electrodes 6, 7 and over a width W1 less than or equal to the distance D between the two electrodes 6, 7. Yet another machined zone 12 extends longitudinally over a length L2 partly between the electrodes 6, 7 and partly beyond the electrodes 6, 7. The machined area 12 extends transversely over a width W2 greater than the distance D between the two electrodes 6, 7 and less than or equal to the total width T of the electrodes 6, 7.The machined areas 10, 11, and 12 can be machined in a single operation using the same machining technique, or they can be manufactured sequentially using different machining techniques as described above. For example, the machined area 11 can be shaped after the electrodes have been depositioned, in cases where temporal or frequency drifts remain, in order to further limit these drifts.
[0072] Figure 14 shows a cross-sectional view of an electro-optical device 50 such as, for example, illustrated in one of Figures 10 to 13. Advantageously, the machined area 10, 11, and / or 12 is shaped before the deposition of the electrodes 6, 7. In this example, the machined area 10, 11, and / or 12 is machined over a width W greater than the total width T of the electrodes, and preferably over the entire width of the substrate 1. For example, the machined area(s) are formed here by removing a layer of thickness P over the entire width of the substrate after the thermal annealing step and before the electrode deposition step. In the cross-section, it can be seen that the surface machining has removed a layer of thickness P from the optical waveguide 5, where the residue of the proton-exchange modified layer 3 was located.Here too, below the machined area between electrodes 6,7, the diffusion area 5 has a depth equal to P4-P, which remains greater than 75% of the initial depth of the optical waveguide 5 before machining.
[0073] Figure 15 shows a cross-sectional view of an example of an electro-optical device 50 as illustrated in one of Figures 10 to 13. Here, the machined area 10, 11 and / or 12 is shaped before or after the deposition of the electrodes 6, 7. The area Machined areas 10, 11, and / or 12 are, in this example, shaped to a width W less than or equal to the distance D between the electrodes. Optionally, lateral areas are also machined outside the electrodes, between the total width T and the width of the substrate. For example, the machined area(s) are formed here by removing a central strip of thickness P along the entire length or part of the length of the electrodes, and possibly two lateral strips after the thermal annealing step and before or after the electrode deposition step. In the cross-section, it can be seen that the surface machining has removed a layer of thickness P from the optical waveguide 5, where the residue of the proton-exchange modified layer 3 was located. After deposition of electrodes 6, 7, the electro-optical device has a hollow between electrodes 6, 7 in the machined area 10. In this example, electrodes 6, 7 are located at a height greater than P relative to the optical waveguide.
[0074] Optionally, after the machined area has been manufactured, the process includes an additional step of depositing a dielectric layer 24, for example of silicon dioxide (SiCh) or silicon nitride (SiN). The dielectric layer 24 is deposited on the surface of the substrate after machining, between each electrode 6, 7 and the substrate 1. Alternatively, the dielectric layer 24 extends not only under the electrodes but over the entire surface of the substrate.
[0075] Alternatively or complementarily, the process includes another step of depositing a dielectric layer 24, for example of silicon dioxide (SiCh) or silicon nitride (SiN). The two electrodes 6, 7 are deposited on this dielectric layer 24. The dielectric layer 24 protects the surface of the optical waveguide 5. The dielectric layer 24 also isolates the optical wave traveling in the optical waveguide from the metallic electrodes.
[0076] The presence of a residue from the layer corresponding to layer 3 modified by proton exchange after thermal annealing can be detected and measured by various characterization techniques such as: optical or contact profilometry, Raman spectrometry, secondary-ion mass spectrometry (SIMS), X-ray diffraction (X-ray rocking curves), transverse electron microscopy (TEM), the Fourier transform infrared (FTIR) spectroscopy or UV ellipsometry are used to determine the required machining depth (P). This depth is generally less than 500 nm. The machining depth can be verified by checking the machined depth during manufacturing or, if necessary, by measuring the surface conductivity before and after etching. Furthermore, the electro-optical functionality of the device is checked after manufacturing.
[0077] This yields an integrated electro-optical device 50 comprising at least one machined zone 10, 11, 12 in the optical waveguide 5 over a shallow depth P, the machined zone 10, 11, 12 being located at least partially between the electrode pair 6, 7. The machined zone is adapted to remove at least partially a residue of the proton-exchange modified layer 3 in a region located at least partially between said electrode pair 6, 7. However, the depth of the machined zone in the optical waveguide is limited so as to preserve the optical guiding properties of the optical waveguide 5. Preferably, the depth of the machined zone in the optical waveguide is less than 25% of the initial depth of the optical waveguide 5 before machining. In some embodiments, the machined zone remains hollow and may form a depression relative to the substrate surface.In other embodiments, the machined area is at least partially filled by a dielectric layer, for example. Finally, in other embodiments, the machined area extends over the entire surface of the substrate, which then has the same surface relief as an unmachined electro-optical device.
[0078] The process also applies to an electro-optical device comprising a proton-exchange optical waveguide on lithium niobate (or lithium tantalate or lithium antimoniate), following a Z-cut, as illustrated in Figures 16-17. In this case, one electrode 6 is in contact with the optical waveguide 5. The other electrode 7 is disposed next to the optical waveguide 5 on the same face as electrode 6 (see Figure 16) or on the opposite face 9 of the substrate 1 (see Figure 17). In this Z-cut configuration, an applied electrical potential difference between the two electrodes generates an electric field E oriented along the Z-axis in the optical waveguide 5. According to the present disclosure, in the Z-cut configuration, analogously to the configuration In section X, after the thermal annealing step, at least a portion of the residue from layer 3 is machined in a region of the optical waveguide 5 intended to be placed in contact with electrode 6, and between the electrode pair 6, 7, so as to reduce or eliminate the residue from layer 3 in the machined area. The depth P of the machined area is on the order of the small depth P3 of the proton-exchange modified layer 3. In practice, the depth P can be less than, equal to, or slightly greater than the small depth P3. For example, the depth P of the machined area is such that: 0.02*P3 < P < 1.10*P3, and preferably 0.50*P3 < P < 1.10*P3, and even more preferably 0.90*P3 < P < 1.10*P3. For example, for a depth P3 of 500 nm, the depth P of the machined zone is approximately 500 nm ± 50 nm. The machining step is performed at a shallow depth P to preserve most of the diffusion zone 4.
[0079] In the Z-cutting configuration, the machining step is necessarily performed before the electrode 6 deposition step. The machining step can be carried out over the entire surface 2 of the substrate, after the thermal annealing step, so as to obtain a flat surface before the electrode 6 deposition, as illustrated for example in Figure 17. Alternatively, as illustrated in Figure 16, the machining is applied to a limited area in width and / or length of the face 2 of the substrate, above the optical waveguide 5. Partial or extensive machining over the entire surface can be applied whether the electrode 7 is located on the same face 2 as the electrode 6 or on the opposite face 9.
[0080] This gives us an electro-optical device in X-section or Z-section configuration, in which at least part of the residue of the proton-exchange modified layer 3 is removed in an area located or intended to be located between the electrode pair 6, 7.
[0081] In a non-intuitive way, this superficial machining, which can be limited locally to a small portion of the substrate between the pair of electrodes, allows for a great improvement in the response of the electro-optical device, whether as a function of time, temperature, pressure (for example under vacuum) or at low and very low modulation frequencies.
[0082] Figures 4 to 6 schematically represent the technical effects obtained on an integrated electro-optical device obtained by annealed proton exchange according to the process of the present disclosure with a machining step.
[0083] Figure 4 schematically represents, in dashed lines, a frequency response curve 17 of an electro-optical phase modulator manufactured by APE and machined according to this disclosure. It can be seen that the optical phase shift ratio A <t>The induced voltage (V) remains constant at low and very low frequencies. The ratio ± A <t> / V is independent of the modulation frequency f, especially at low and very low frequencies.
[0084] Figure 5 schematically represents, in dashed form, a half-wave voltage curve (or VK signal) of an electro-optical phase modulator manufactured by APE and machined according to this disclosure. It can be seen that the VK signal remains constant at low and very low frequencies. The VK signal curve is independent of the modulation frequency f, particularly at low and very low frequencies.
[0085] Figure 6 shows on the right a curve 22 of the optical phase shift A <t>The induced optical phase shift in the optical wave propagating in the optical waveguide 5 is a function of time, i.e., a function of the applied voltage V, following curve 20. On curve 22 of the induced optical phase shift, instantaneous steps are observed when the voltage increases or decreases. The desired optical phase shift value is thus obtained instantaneously with each change in the voltage applied to electrodes 6 and 7.
[0086] One possible interpretation of these improvements in the electro-optical response of the electro-optical device with a machined zone as disclosed herein is as follows. The electrical conductivity of the degraded crystal structure at the location of the proton-exchange modified layer 3 can vary depending on the temperature or under a hot vacuum, for example. As illustrated in Figure 3, the fundamental mode 15 of the wave guided by the optical waveguide 5 extends to the substrate surface and partially covers the location of the proton-exchange modified layer 3 remnant. When a voltage is applied between electrodes 6, 7, the electric field lines 14 pass through the different zones that have different electrical conductivities: the substrate 1, The diffusion zone 4 and the zone containing a proton-exchange-modified layer 3 remnant are distinct. The physical properties of these zones differ. In particular, their conductivity varies differently depending on environmental conditions, notably temperature and pressure. The increase in conductivity of the proton-exchange-modified layer 3 remnant differs from that of the optical waveguide 5 and the substrate. This difference in conductivity variation alters the electro-optical overlap and thus the response as a function of the modulation frequency applied to the electrodes. Machining the remnant in a specific, shallow area of the substrate 1 between electrodes 6 and 7 allows for at least partial removal of the layer 3 remnant. This shallow machining, to a depth P in the proportions detailed above, does not disrupt the guidance of the single-mode wave in the optical waveguide 5.
[0087] The resulting electro-optical device finds advantageous applications in electro-optical phase modulators, optical intensity modulators, and / or Mach-Zehnder interferometers. It also finds applications in active directional evanescent-wave couplers, such as a 2x2 active coupler on a lithium niobate substrate with electrodes arranged in the evanescent-wave coupling region.< / t> < / t> < / t> < / t> < / t> < / t> < / t> < / t>
Claims
Claims 1. A method of manufacturing an electro-optical device (50) comprising the following steps: manufacturing an optical waveguide (5) in an electro-optical substrate (1) by proton exchange annealing comprising the steps of: a) exposing an area of a surface (2) of the substrate (1) to a proton source so as to create in the substrate a layer (3) modified by proton exchange, said layer (3) extending over said area from the surface (2) of the substrate to a small depth (P3) in the substrate, and b) thermal annealing of the substrate so as to diffuse hydrogen ions from said layer (3) into a diffusion zone (4) extending in the substrate around said layer (3) to a large depth (P4) so as to create an optical waveguide (5) in the diffusion zone (4); depositing a pair of electrodes (6, 7), the optical waveguide (5) extending longitudinally between said pair of electrodes (6, 7);characterized in that the method comprises, after the thermal annealing step, a step of machining from the surface (2) at least a part of the substrate in a zone of the optical waveguide (5) corresponding to a residue of said layer (3) to form a machined zone (10, 11, 12) located at least partly between said pair of electrodes (6, 7), over a depth P of the order of the small depth (P3) of the layer (3) modified by proton exchange.; 2. Method according to claim 1 in which the machined zone (10, 11, 12) extends in the substrate (1) to a depth P, the depth P being greater than or equal to 0.02*P3 and less than or equal to 1.10*P3.
3. Method according to claim 1 or 2 in which the machined zone (10, 12) extends over at least a portion of a deposition surface of said pair of electrodes.
4. Method according to one of claims 1 to 3 in which the machined zone (10, 11, 12) extends over an entire width of the substrate (1) taken transversely to a longitudinal direction of the optical waveguide (5).
5. Method according to claim 1 or 2 in which the machined area (10) is limited to an area of the waveguide (5) between said pair of electrodes (6, 7).
6. Method according to one of claims 1 to 5 in which the machining step is adapted to form several machined zones (10, 11, 12) arranged in series in a longitudinal direction of the optical waveguide (5), said several machined zones (10, 11, 12) being located at least partly between said pair of electrodes (6, 7).
7. Method according to one of claims 1 to 4 in which the machined zone (10) extends over the entire surface of the substrate.
8. Method according to one of claims 1 to 7 in which the machining step is carried out by wet etching, plasma etching, laser etching, lapping or mechanical polishing.
9. Method according to one of claims 1 to 8 comprising a step of depositing a dielectric layer (24), the dielectric layer (24) being arranged on the machined area (10) on the surface of the substrate (1) and / or between the substrate and at least one electrode of the pair of electrodes (6, 7).
10. Electro-optical device (50) obtained according to the method of one of claims 1 to 9.
11. An electro-optical device according to claim 10, used in an electro-optical phase modulator, an optical intensity modulator, a Mach-Zehnder interferometer and / or an active directional coupler with evanescent waves.