Integratable non-reciprocal optical components, optical isolators, optical circulators, and integrated circuits

Abstract

The present invention relates to an integratable non-reciprocal optical component, in particular an optical isolator or circulator, having a guideway, in particular a sub-wavelength guideway, between two magneto-plasmonic interfaces, said interfaces being formed between a dielectric and a metal, respectively. An input signal (OS0) passes through a selection region (R1) which results in a selection signal (OS1) from an optical port (P1). The energy of the selection signal is concentrated in a single plasmon mode, LRSPP or SRSPP, through a selection aperture (R11) with a specific selection width (We). The width of the selection aperture for these modes determines whether they have significantly different optical impedances from each other, and whether one optical impedance (z1 eff ) is the input optical impedance (z0 eff ) is substantially equal to the input signal (OS0). The selection signal (OS1) passes through a differentiation region (R2) that creates or increases the asymmetry that exists between the two magneto-plasmonic interfaces, concentrating the energy of the selection signal onto a single magneto-plasmonic interface. The differentiation signal (OS2) passes through a non-reciprocal processing region (R3) formed by the two magneto-plasmonic interfaces (31, 32) of unequal shape, in particular through an absorbing cavity (311) on only one side, or through a detour that leads to different optical output ports (P2, P3). The input signal (OS0) thus undergoes different processing than the inverse signal (OS0B).

Description

Technical Field

[0001] The present invention relates to an integrable non-reciprocal optical component having a waveguide, in particular a sub-wavelength waveguide, between two magnetic plasmon interfaces. Specifically, the two magnetic plasmon interfaces are each formed between a dielectric and a metal.

[0002] The input signal passes through a selection region that provides a selection signal from an optical port, and the energy of the selection signal is concentrated in a single plasmon mode, namely, LRSPP (abbreviation for "long-range surface polariton plasmon") or SRSPP (abbreviation for "short-range surface polariton plasmon"), particularly via a selection aperture. The widths of the selection apertures for these modes have significantly different optical impedances from each other, and one of the optical impedances is close to or equal to the input optical impedance so that a selected optical signal that can be used in the selected plasmon mode is obtained.

[0003] The selection signal passes through a differentiation region that generates or increases the asymmetry existing between the two magnetic plasmon interfaces, and concentrates the energy of the selection signal on a single magnetic plasmon interface.

[0004] The differentiation signal passes through a non-reciprocal processing region formed by two magnetic plasmon interfaces of unequal shapes, in particular, an absorption cavity on only one side, or a detour that each leads to a different optical output port. The input signal thus undergoes a different process from the reverse signal.

[0005] The present invention also relates to an optical isolator, an optical circulator, and an integrated circuit that integrally include such components.

Background Art

[0006] The induction of optical signals is used in many fields, such as telecommunications, but also in many other fields such as sensors or data processing circuits. In optical or photonic circuits, various types of components are used to process or distribute such signals within different wavelengths, such as about 1.55 μm in telecommunications for example. Miniaturizing such circuits is actively sought to improve compactness and processing capabilities.

[0007] For about 20 years, photonics has been widely developed on the nanometer scale, enabling an increase in light-material interactions in many applications in photonics and optoelectronics in the fields of light emission, detection, imaging or induction.

[0008] To implement some types of functions within an optical circuit, it is necessary to make available components with non-reciprocity, i.e., components that process optical signals variably according to the propagation direction of this signal.

[0009] It is known to generate non-reciprocal optical components by combining the magneto-optical properties of the transverse magneto-optical Kerr effect (TMOKE, "transverse magneto-optical Kerr effect") with a metal layer that brings about a plasmon-induced effect, i.e., a metal layer that brings about resonance between the optical signal in the waveguide and the surface electrons of the metal wall.

[0010] Such non-reciprocal optical components can be, for example, optical isolators, which use absorption non-reciprocity to allow a signal to pass in one direction while blocking or significantly attenuating the signal in the other direction, as described in the publication, by Van Parys et al., "Transverse magnetic mode nonreciprocal propagation in an amplifying AlGaInAs / InP optical waveguide isolator", Applied Physics Letters 88, 071115 (2006).

[0011] These non-reciprocal optical components can also be circulator devices with three or more optical ports, where the three or more optical ports are permanently and optically connected to each other, and the circulation of signals between two given ports occurs in only one direction using non-reciprocal refractive indices that generate non-reciprocal phase shifts, as described in the publication, by Takei and Mizumoto, "Design and Simulation of Silicon Waveguide Optical Circulator employing Non Reciprocal Phase Shift", Jpn.J. Appl.Phys. 49 (2010) 052203.

[0012] Such non-reciprocal components are based, for example, on an optical signal confined within an optical waveguide, as in document US8849072, where one of the walls of the optical waveguide generates such a magnetic plasmon effect. This is different from a reciprocal optical circuit that uses plasmon induction where the induction occurs between two plasmon walls.

[0013] However, these non-reciprocal components are still bulky and difficult to integrate into integrated circuits because this non-reciprocal effect requires specific amplification to be sufficiently effective. This non-reciprocal effect is generally enhanced by resonators, such as the ring resonators described in document FR2981761, but this ring resonator is not only bulky but also complex and involves further constraints.

[0014] Therefore, manufacturing a device that is integrated within one and the same component and is small enough to efficiently integrate, for example, a laser source and a distribution circuit within it, and even smaller devices, is difficult and / or complex and thus costly.

[0015] To date, the proposed solutions have not been fully implemented to integrate within photonic circuits. The only isolator or circulator components available on the market are large garnet-based ones assembled with polarizers within non-inductive optical elements, which are several centimeters in length and are coupled to optical fibers.

Prior Art Documents

Patent Documents

[0016]

Patent Document 1

Patent Document 2

Non-Patent Documents

[0017]

Non-Patent Document 1

Non-Patent Document 2

Summary of the Invention

Problems to be Solved by the Invention

[0018] One object of the present invention is to overcome all or part of the drawbacks of the prior art.

[0019] In particular, while separating or forming a part of an optical circuit, it is required to obtain non-reciprocal processing for guiding an optical signal by a component that minimizes the bulkiness of the optical circuit and / or facilitates integration.

Means for Solving the Problem

[0020] The present invention provides an optical component, the optical component comprising at least one first port or optical waveguide and a second port or optical waveguide, the first port or optical waveguide and the second port or optical waveguide transmitting light to each other in a non-reciprocal manner, connecting the component into the induced optical circuit or enabling connection of the component, and the at least one first port or optical waveguide, or the second port or optical waveguide, being capable of exchanging an optical signal with the optical circuit in one first direction and a second direction opposite to the one first direction through at least one magnetic plasmon induction region.

[0021] The term magnetic plasmon as used herein refers to a combination of plasmon induction and a magneto-optical effect, and the magneto-optical effect is typically of the TMOKE type.

[0022] The term "plasmon induction" as used herein means performing a surface interaction with a metal of a type that is not always considered "plasmonic" in itself, for example iron-cobalt.

[0023] More specifically, the component according to the present invention is designed to process an optical signal having a wavelength of less than 50 μm, particularly less than 20 μm or 10 μm, more specifically less than 3 μm, for example about 1.55 μm, or another frequency used in telecommunication, and simulations have been performed and it is configured and determined as such.

[0024] According to the present invention, the component includes a plurality of regions through which an optical signal traveling from the first optical port in the first direction towards the second optical port can continuously pass.

[0025] These multiple regions at least continuously include, in this order, a selection region, a differentiation region, and a non-contradictory processing region, for example, all or some of them.

[0026] The selection region is configured as follows: - At the input, it receives an input optical signal that propagates according to the input optical impedance z0 eff , that is, the impedance of the guided mode. The impedance of the guided mode is typically proportional to the reciprocal of the input effective refractive index. This input optical signal is typically received inductively, but not compulsorily. In this selection region and possibly upstream of the selection region, the induction may or may not be plasmonic induction. Typically, it appears in a polarization form, preferably TE polarization. - At the output, it emits an optical signal called a selected optical signal by plasmonic induction. The selected optical signal includes a first plasmon mode and a second plasmon mode. The second plasmon mode has an amplitude of zero or significantly lower than that of the first plasmon mode, typically at least 1 / 50, preferably 1 / 100 of the intensity amplitude of the first plasmon mode. These first plasmon mode and second plasmon mode are 〇 One preferred first plasmon mode, that is, the LR mode (or LRSPP, abbreviation of "long-range surface polariton plasmon"), and 〇 The other preferred corresponding second plasmon mode, that is, the SR mode (or SRSPP, abbreviation of "short-range surface polariton plasmon") defined as.

[0027] The differentiation region is configured, at the input part, to receive at the input the selected optical signal arriving from the selection region and guide the selected optical signal to the output part. At the output part, the optical signal forms an optical signal called a differentiated optical signal by a guiding core called a differentiation core. The differentiation core extends between at least two interfaces called differentiation interfaces. This configuration includes, for example, the presence of magnetization.

[0028] These differentiated interfaces - are formed between a core material and at least one outer material, where one of the core material and the at least one outer material is a dielectric and the other is a metal, - are located on two opposite sides of this inductive core, - are determined to bring about magnetic plasmon induction between the core material and the at least one outer material.

[0029] In this differentiated region, the output part is configured to have a differentiation width Wd between the differentiated interfaces, and the differentiation width Wd is selected such that the first plasmon mode signal has an asymmetric amplitude greater than a determined threshold between the differentiated interfaces.

[0030] Accordingly, this differentiated region is configured to supply, at the output, an asymmetric optical signal called a differentiated signal, and this amplitude - asymmetric optical signal is concentrated on the first differentiated interface. That is, the differentiated signal is typically stronger on the first differentiated interface at a ratio greater than 50, especially at a ratio of 100 or more, and weaker or even negligible on the second differentiated interface.

[0031] The non - reciprocal processing region is configured to receive, at the input, the differentiated signal coming from the differentiated region and guide the differentiated signal to the output region via a processing part. In the output region, the differentiated signal forms an optical signal called a processed optical signal. This guidance occurs in a guiding core called a processing core that extends between at least one first processing interface and a second processing interface. This configuration includes, for example, the presence of magnetization.

[0032] These processing interfaces - are respectively formed between a processing core material and an outer material, where one of the processing core material and the at least one outer material is a dielectric material and the other is a metal material, - are located on the same two opposite sides as each of the first differentiated interface and the second differentiated interface, - are determined to bring about magnetic plasmon induction between the processing core material and the at least one outer material, - Between the input and output of the processing region, for example, by introducing a spatially symmetric interruption therebetween, it is configured such that, in particular, the effects on the signals over the entire length of the processing region have a shape where they are not equal to each other.

[0033] Therefore, this non-reciprocal processing region is configured to supply the processed optical signal in a first direction at the output, and this amplitude of the processed optical signal undergoes the processing brought about solely or mainly by the shape of the first processing interface alone. This is because the energy of this signal is concentrated on the first processing interface.

[0034] In this regard, a signal called the reverse signal passing through the non-reciprocal processing region in a second direction undergoes the processing brought about by the shape of the second processing interface alone, or by both of the two processing interfaces (when this reverse signal itself is also differentiated), and thus supplies the processed reverse signal. This processed reverse signal undergoes a processing different from the processing received by the processed signal traveling in the first direction.

[0035] Typically, each port includes these regions: a selection region, a differentiation region, and then the entire set of processing regions. Typically, two ports that communicate with each other share one and the same processing region, or a part of one and the same processing region. Therefore, one or more processing regions connecting two ports to each other always receive differentiated signals in both the first direction and the second direction. Each of the differentiated signals undergoes only the processing generated by the shape of the first processing interface, that is, the interfaces and shapes of the processing interfaces that are different according to the propagation direction.

[0036] For example, when the differentiation region is the same at each port and is configured to weaken or cancel the amplitude of the signal on the left interface considered in the propagation direction, the differentiated signal is concentrated on the right interface, and the processing signal undergoes the processing determined solely by this right interface.

[0037] In the path connecting two ports, it is possible to consider that on one side, plasmons of a signal propagating in the first direction pass through, and on the other side, plasmons of a signal propagating in the second direction pass through.

[0038] As can be understood, this continuous differentiation process and processing process are, unlike the components of the prior art, carried out by magnetic plasmon induction existing on two opposite sides of the induction core.

[0039] In the prior art, the induction between two adjacent magnetic plasmon interfaces has conventionally been considered to have no counter-effect.

[0040] In a different manner, the inventors have identified and implemented the asymmetry that can be obtained within the signal intensity distribution between these two interfaces by selecting one or more specific widths of the spacing separating these interfaces. This enables the present invention to process an input signal, concentrating an input signal of this intensity on only one of the two plasmon interfaces, and thus causing only one of these interfaces to undergo processing. In this case, the counter-effect is obtained by combining this lateral concentration with different processing for each interface.

[0041] According to a certain feature, the selection region is - an input part that forms optical induction, and possibly plasmon induction as well, but preferably forms only optical induction - an output part that forms magneto-optical plasmon induction between two magnetic plasmon interfaces and has.

[0042] These input and output parts communicate with each other through a selection aperture having a selection width We, and the selection width We is measured between two plasmon interfaces and is typically narrower than the input width of the selection part.

[0043] Therefore, the input optical signal received in the input part excites an optical signal in the output part, and after exiting the output part, forms a selection optical signal including a first plasmon mode and a second plasmon mode.

[0044] The optical guidance of this selection input unit is based on the input optical impedance z0 with respect to the signal received by the input. eff , that is, for example, it has the optical impedance of the guided mode in the form of the effective refractive index.

[0045] According to this feature, the selection width We is combined with the input optical impedance value z0 eff and 〇 The first plasmon mode has a first optical impedance z1 that is substantially equal to the input optical impedance z0 eff , that is, at least close enough to obtain a selectable optical signal that can be used in this first plasmon mode. eff has 〇 The second plasmon mode has a second optical impedance z2 that is significantly different from the first optical impedance z1 eff , that is, in the second plasmon mode, for example, different enough to obtain a selectable optical signal weakened to 1 / 50 or even 1 / 100. eff has and is determined as follows.

[0046] The selection width, and possibly other characteristics that affect the optical impedance, are adjusted to obtain an input signal. The input signal excites an optical mode according to the first plasmon mode through the selection aperture within the output section. The first plasmon mode excludes the second plasmon mode (or has an amplitude significantly larger than the amplitude of the second plasmon mode). Therefore, the optical signal exiting the output section (of the selection region) forms a selectable optical signal having only the first plasmon mode.

[0047] Note that the optical impedance is defined by the formula:

[0048]

Equation

[0049] as defined by.

[0050] Therefore, the optical impedance is proportional to the reciprocal of the effective refractive index n where n 2 ~= ε, and in the formula, ε = ε r ε0, ε is the dielectric constant, and ε r is the relative dielectric constant, ε0 is the permittivity of free space, μ r is the relative permeability, and μ0 is the permeability of free space.

[0051] In many cases, the optical impedance of the guided mode is directly related to the effective refractive index. Therefore, in many embodiments of the present invention, the term "optical impedance" may be replaced by the term "effective refractive index".

[0052] Other types of configurations are those that enable the selection of the first plasmon mode by excluding the second plasmon mode within the signal, and are used in the present invention to perform such a selection and can also generate a selected optical signal.

[0053] According to another feature, preferably but not compulsorily, in combination with the preceding feature, the differentiation region has a width that increases from the input width to the differentiation width Wd (which can be, for example, the same as the selection width We), and has a shape that is, for example, symmetric around the differentiation core and / or increases continuously or even regularly increases.

[0054] When selecting through an aperture having a selection width We, the differentiation region receives a selection signal that includes only a single plasmon mode. However, this selection width typically corresponds to a configuration in which the selected plasmon mode, i.e., the first plasmon mode, does not have a significant asymmetry.

[0055] Such a width variation in the differentiation region is selected such that at the output, this first plasmon mode reaches a differentiation width that has a significant asymmetry between the right and left interfaces defined in the signal propagation direction, i.e., between the first interface and the second interface, respectively.

[0056] Therefore, in the output from this differentiation region, the first plasmon mode accumulates on the first interface. In this regard, if the second plasmon mode still exists, it accumulates on the second interface. Since this second plasmon mode is pre-suppressed or minimized by the selection region, the differentiated optical signal accumulates its intensity only on the first interface.

[0057] Typically, as a possibility, in all embodiments, the non-reciprocal processing region also has a shape where the input width and / or the average width of the input width is substantially equal to the differentiation width Wd, and at least partially, the two processing interfaces have different shapes.

[0058] Other types of configurations are those that enable the differentiation of signals of the first plasmon mode and the second plasmon mode with respect to each other, and can also be used within the present invention to perform such differentiation and generate a differentiated optical signal.

[0059] The present invention enables the integration of a small and high-performance non-reciprocal optical component into a photonic circuit, such as an isolator or a circulator, particularly in a planar and simple technical manner in a broad sense. The photonic circuit can be, for example, a simple laser or even a circuit including multiple optical guiding functions.

[0060] The proposed solution has major advantages, particularly, - Such components can be made quite small (from a few microns to dozens of microns when operating in the near-infrared), and the loss of the guided light can also be limited. - The optical function is broadband (a spectral range of dozens of nanometers), while the currently proposed high-performance isolator resonates over a fraction of a nanometer. The spectral band is adjustable in the isolator and quite broad in the circulator. - In some versions (e.g., FeCo-silica-FeCo), the structure can be integrated into any photonic platform by standard manufacturing methods and a few steps, and can operate with residual magnetization. - The typical isolation rate of a structure with a length of several microns reaches about 10 dB in a wide band and is "cascadable" (addable). Even before any specific optimization, the isolation / loss trade-off is quite high compared to ordinary plasmonic structures. The present invention enables a total loss level of about 5 dB, for example, in a 10 dB isolator. This loss level can be easily compensated by amplification.

[0061] Preferably, although not mandatory, the selected region is configured such that the first plasmon mode corresponds to the LRSPP-type mode of the input signal.

[0062] According to a preferred feature, the differentiating core and / or the processing core (preferably both) are made of a dielectric material, and these outer materials are made of one or more metallic materials. Thus, a structure of a type that can be called a "slot", also known as MIM (abbreviation for "metal-insulator-metal") or MDM (abbreviation for "metal-dielectric-metal"), is obtained.

[0063] Alternatively, the differentiating core and / or the processing core (preferably both) are made of a metallic material, and these outer materials are made of one or more dielectric materials. Thus, a structure of a type that can be called a "rib" or "wire", also known as IMI (abbreviation for "insulator-metal-insulator") or DMD (abbreviation for "dielectric-metal-dielectric"), is obtained.

[0064] In either case, these materials are selected to create a plasmon-induced interface that implements a magneto-optical effect between the materials.

[0065] Preferably, although not mandatory, the differentiating core is fully or partially formed within a planar or two-dimensional layer included between a lower layer and an upper layer, and the interface that provides a conductive path within the core forms a planar or two-dimensional surface orthogonal to the layer. Thus, the conductive interface, particularly the magnetic plasmon interface, forms a "vertical" plane inside one and the same manufacturing layer of a component. Therefore, it is easier to manufacture the desired shape for these interfaces than when it is necessary to manufacture a shape configured between two stacked layers.

[0066] It should be noted that the upper layer and / or the lower layer is / are constituted by air or vacuum, and thus can form uncovered and / or suspended conductive paths.

[0067] Treatment by absorption According to a first group of embodiments, the first processing interface has a shape including one or more absorption cavities of, for example, the Helmholtz and / or Fabry - Perot type.

[0068] This cavity or these cavities have a shape configured to completely or partially absorb an optical signal of a determined frequency. In particular, there may be several cavities having the same dimensions and determined to absorb most of the signal.

[0069] The cavity may also be configured to completely or partially absorb an optical signal including a plurality of different frequencies, especially a plurality of frequencies shifted relative to each other, such that the processing interface absorbs all signals within a determined frequency range. In this case, there may be several cavities or groups of cavities having different dimensions, absorbing a plurality of wavelengths and thus determined to act over a wider frequency band.

[0070] In this case, the second processing interface has a shape configured to allow the passage of an optical signal of one or more of the said determined frequencies.

[0071] Thus, a component is obtained in which one or more frequencies are blocked in one direction but not in the other.

[0072] Preferably, the first optical waveguide or port and the second optical waveguide or port each include a selection region and a differentiation region, are connected to each other by at least one first processing region, and the first processing interface of the at least one first processing region is configured to completely or partially absorb an optical signal of one or more determined frequencies, and the second processing interface of the at least one first processing region is configured to allow the passage of the optical signal. In this case, the processing region of the first optical port can be regarded as forming the processing region of the second optical port.

[0073] Therefore, the selection region and the differentiation region of the second optical port act in the same way on the optical signal arriving in the reverse direction. The selection region and the differentiation region each select a first plasmon mode and then focus the first plasmon mode on the first interface as seen from the second direction, i.e., on the second interface as seen from the first direction.

[0074] Therefore, in the case of a given frequency or frequency range, when one of the processing interfaces is an absorption interface while the other processing interface is a passage interface, an optical isolator is obtained that allows passage in the second direction (forming the outer direction) and blocks in the first direction (forming the return direction of the isolator).

[0075] Such an isolator can be incorporated, for example, into an integrated circuit including a laser source, allowing the signal emitted by the laser to pass in the outer direction while blocking reflections that may return in the return direction, for example, enabling a power function without the risk of coherence loss due to the return light. Application examples are, for example, telecommunication, and in particular, the monolithic integration of lasers or semiconductor amplifiers into photonic circuits.

[0076] Processing by detour According to a second group of embodiments, the first processing interface of the first optical port has a shape that forms a detour with respect to the second processing interface. Such a detour is configured, for example, such that the first processing interface ends at a different optical port, or travels a different distance, or incorporates further or different processing.

[0077] According to a certain feature, the first processing interface of the first optical port has a shape that connects the first optical port to the second optical port, guides the differentiated optical signal arriving from the first port, and emits the output optical signal through the second port. In this regard, the second processing interface of the first optical port has a shape that connects the first optical port to the third optical port, receives the differentiated optical signal arriving from the third port, and emits the differentiated optical signal as an output optical signal through the first port, thus generating an optical circulator.

[0078] By applying the same configuration to each of the component optical ports, for example, when the last optical port loops back to the first optical port, a circulator can be obtained that causes each input signal to loop back to subsequent optical ports, potentially in a circular manner.

[0079] Therefore, it is possible to generate an integrated optical circulator that combines functions on a photonic circuit, such as with a microfluidic circuit, and functions such as continuous sensing or biological detection functions. Therefore, the integrated circulator function enables the circuit function to spread without adding bulk.

[0080] It is also possible to obtain a phase shift by generating two processing interfaces with different shapes configured to provide different interface lengths.

[0081] Integrated circuit According to another aspect, the present invention proposes a device comprising one or more non-reciprocal optical components as disclosed herein, generated to be integrated within an integrated optical circuit.

[0082] According to features that can be combined with each other, - This integrated optical circuit includes at least one integrated laser emitter and comprises at least one non-reciprocal optical component with non-reciprocal absorption disclosed herein. The non-reciprocal optical component is connected to the laser and allows an optical signal emitted by the laser to pass through in one direction while preventing or reducing an optical signal returning to the laser in the other direction. - This integrated optical circuit comprises a plurality of sensors generated or connected to be integrated, and at least one non-reciprocal optical component of the circulator type having a detour disclosed herein and connected to at least two of the sensors. - This integrated optical circuit comprises at least one non-reciprocal optical component of the circulator type having a detour disclosed herein. The non-reciprocal optical component is connected or configured to completely or partially integrally generate a Michelson interferometer, for example, in the field of signal processing.

[0083] According to another aspect, the present invention proposes a method for processing an optical signal. The method is characterized by including the processing of the optical signal by passing it through at least one non-reciprocal optical component disclosed herein or within a device disclosed herein.

[0084] It is contemplated that various aspects of the present invention incorporate all possible combinations of the various optional features disclosed herein.

[0085] Other features and advantages of the present invention will become apparent from the detailed description of non-limiting embodiments and the accompanying drawings.

Brief Description of the Drawings

[0086]

Figure 1

Figure 2

Figure 3a

Figure 3b

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Mode for Carrying Out the Invention

[0087] The present invention relates to a new structure of an inductive magneto-optical plasmon component involving non-reciprocal transmission.

[0088] In this example, the basic structure is composed of a metal-dielectric-metal type sub-wavelength waveguide, also called a metal-insulator-metal (MIM) type, and either the dielectric or the metal has magneto-optical properties within the relevant frequency range.

[0089] This structure enables the generation of a planar optical isolator or circulator that is independent of a photonic circuit or integrated within a photonic circuit and operates over a wide spectral range.

[0090] In this specification, the phenomenon that causes non-reciprocal action is a usage method specially adapted to the asymmetric spatial distribution phenomenon of the magnetic plasmon mode of MIM, and is obtained when the magnetization of material MO and the polarization of the propagating wave are selected to generate the TMOKE effect (magnetization parallel to the plane, and polarization of the wave orthogonal to this plane and the magnetization). However, (in this example) the two modes propagating in MIM, namely the even mode and the odd mode, have opposite asymmetries. For this reason, when both modes are excited simultaneously, they compensate each other in the conventional signal.

[0091] To utilize this physical phenomenon to obtain non-reciprocal transmission, the signal is processed so that only one of the two modes is excited. This selection process is here executed at the input of the MIM waveguide by passing through a narrow or narrowing waveguide to select only one of the plasmon modes, namely only the even mode or only the odd mode. In fact, the effective refractive indices of the even mode and the odd mode are significantly different in a very narrow waveguide. In this case, the selected mode is selected by using an aperture that generates an effective refractive index corresponding to the expected mode.

[0092] Figure 1 shows, in a comprehensive form, an exemplary non-reciprocal component 1 that details only a single optical port P1.

[0093] This optical component 1 includes a first port P1 and a second port P2, and the first port P1 and the second port P2 transmit light between them. Therefore, these optical ports form the connection of the component in the waveguide optical circuit. This circuit can be formed by an external optical fiber connected to the optical port. This circuit can also form a monolithic integrated circuit that enables a high degree of miniaturization. Therefore, in this optical circuit, this component 1 can pass an input optical signal OS0 in a first direction on one hand, and a reverse input signal OS0B in a second direction opposite to the first direction on the other hand.

[0094] In this comprehensive example, it is assumed that two optical ports P1 and P2 communicate with each other such that each input signal OS0 and OS0B exits in the form of output signals OS9 and OS9B via opposite poles. However, this component can definitely have three or more ports that communicate with each other differently, for example, as shown in FIG. 4.

[0095] The optical transmission between the ports is carried out by guiding materials 10, 20, 30 that form optical waveguides between cross-sections 110, 210, 310 with outer materials on one side and cross-sections 120, 220, 320 with outer materials on the other side.

[0096] As shown in the figure, the input optical signal OS0 entering through the first port P1 of this component 1 continuously passes through a plurality of regions R1, R2, R3, and the plurality of regions R1, R2, R3 include the following regions in the following order: selection region R1, differentiation region R2, processing region R3.

[0097] When the input optical signal OS0 arrives via the first optical port P1, it enters the selection region (R1). This selection region is here formed by a selection core 10, and the selection core 10 guides the signal received between the left cross-section 11 and the right cross-section 12.

[0098] This selection region (R1) has an input part (R10) that receives the input signal OS0. The selection region (R1) forms a guiding path for the input optical impedance of the guided mode for the signal received at the input, for example, the input effective refractive index z0 eff and forms a guiding path. This input part here includes a guiding path that operates only optically, but can also operate in a magnetoplasmon mode. Next, the optical signal propagates to the output part R12, and the output part R12 forms a magneto-optical plasmon guiding path generated between two magnetoplasmon interfaces.

[0099] The input part R10 and the output part R12 communicate with each other via a selection aperture (R11), and the selection aperture (R11) typically has an input width W (of this selection region R1) R0It has a determined selection width We that is narrower. Therefore, the input optical signal OS0 excites an optical signal called the selection optical signal OS1 through the selection aperture R11, and the selection optical signal OS1 propagates within the output section R12 toward the differentiation region R2. Here, this output section R12 can be regarded as merging with the starting section R20 of the differentiation region R2.

[0100] This selection optical signal OS1 propagating in the magnetic plasmon-induced mode has energy distributed between the first plasmon mode and the second plasmon mode.

[0101] One of these first plasmon mode and second plasmon mode is the LR mode (or LRSPP, the abbreviation of "long-range surface polariton plasmon"), preferably the first plasmon mode. The other of these first plasmon mode and second plasmon mode is the SR mode (or SRSPP, the abbreviation of "short-range surface polariton plasmon"), preferably the second plasmon mode.

[0102] The width We of this selection aperture R11, in combination with other characteristics of the output section R12 of the selection region R1, determines the selection optical impedance z1 for the selection optical signal OS1, for example, in the form of a selection effective refractive index. eff to determine.

[0103] These characteristics, especially the width We, are selected to generate the selection optical impedance z1. eff to generate, 〇 The first plasmon mode has a first effective refractive index that is substantially equal to the first optical impedance z1 of the induced mode, eff for example, the input optical impedance z0. eff 〇 The second plasmon mode has a second optical impedance z2 of the induced mode in the form of a second effective refractive index that is significantly different from, for example, the first optical impedance z1. eff eff to have.

[0104] ​​Therefore, the input signal OS0 excites the optical mode according to only the first plasmon mode through the selection aperture R11 within the output section R12. That is, the first plasmon mode excludes the second plasmon mode or has an amplitude significantly larger than the amplitude of the second plasmon mode, typically an amplitude with a 100-fold intensity.

[0105] Therefore, a selected optical signal OS1 that substantially contains only the first plasmon mode is obtained.

[0106] As indicated by the two thick black arrows, this signal OS1 has energy that is not necessarily uniform but is distributed across the two selection interfaces 11 and 12. In fact, for this frequency, the selected optical impedance z1 (selected such that the energy is distributed quite unevenly between the two plasmon modes in order to enable selection of the first mode) eff generally has only a very slight spatial asymmetry between the two magnetic plasmon interfaces.

[0107] The differentiation region R2 receives the selected optical signal OS1 arriving from the selection region R1 at the input R20 and guides the selected optical signal OS1 to the output section R32. In the output section R32, the selected optical signal OS1 forms an optical signal called the differentiated optical signal OS2.

[0108] This guiding occurs within a guiding core called the differentiation core 20, which extends between two lateral interfaces called the differentiation interfaces 21 and 22 located on two opposite sides.

[0109] These differentiated interfaces 21, 22 are each formed between a core material and at least one outer material (typically the same material on both sides), one being the dielectric 20 and the other being the metal materials 210, 220, which are selected and configured to generate magnetic plasmon induction therebetween. In this example, the differentiated interfaces 21, 22 are related to the TMOKE effect, which can be obtained, for example, by causing a magnetization M by exposing the differentiated interfaces 21, 22 to a magnetic field or under a residual magnetic effect.

[0110] As shown in the figure, the differentiated region R2 has a shape with a width increasing from the input width We to the differentiated width Wd. This spread is, for example, symmetric around the differentiated core and / or increases continuously, preferably regularly.

[0111] This differentiated width Wd, in combination with other characteristics of this differentiated region R2, is selected such that the selected optical signal OS2 of the first plasmon mode has a large spatial amplitude asymmetry, such as 100 times, between the differentiated interfaces, for example, from one interface to the other.

[0112] In this example, this output portion R22 is configured such that the energy of the differentiated optical signal OS2 is concentrated on the right interface, i.e., the differentiated interface 21. This asymmetry is shown by the arrows representing the distribution of the differentiated signal OS2 energy on the left and right interfaces, i.e., the thick arrows on the right interface 21 - 31 and the thin arrows on the left interface 22 - 32.

[0113] At this stage, as understood, the selection region R11 first concentrates the energy of the input optical signal OS0 in a single plasmon mode, and then the energy is concentrated in a single cross-section within the differentiated region R2, here on the first differentiated region 21 located on the right side of the figure.

[0114] In the non-inverting processing region R3, at the input R30, the differential signal OS2 arriving from the differentiation region R2 is received and guided to the output region R32 via the processing unit 31. In the output region R32, the differential signal OS2 forms an optical signal called the processed optical signal OS3 and passes through a guiding core called the processing core 30 that extends between at least one first processing interface 31 and a second processing interface 32.

[0115] In this example, the differentiation width Wd is substantially equal to the input width of the non-inverting processing region R3 and / or substantially equal to the average width of this non-inverting processing region R3, for example, for a part of the non-inverting processing region R3 where the two processing interfaces 31, 32 have different shapes from each other.

[0116] Typically, the input R30 of the processing region R3 can be considered to merge with the output R22 of the differentiation region R2.

[0117] The first processing interface 31 extends the first differentiation interface, and the second processing interface 32 extends the second differentiation interface 22.

[0118] These two non-inverting processing interfaces 31, 32 are respectively formed between the processing core material 30 and the outer materials 310, 320, where one of them is a dielectric and the other is a metal, which is the same as the case of the differentiation region R2 here.

[0119] The non-inverting processing interfaces 31, 32 are configured and determined to bring about magnetic plasmon induction between them. In this example, the configuration of the materials and / or magnetization is the same as that of the differentiation region except for the horizontal shape.

[0120] Furthermore, the non-inverting processing interfaces 31, 32 are configured to have shapes 319, 329 that are not equivalent to each other between the input R30 and the output R32 of the processing region R3.

[0121] Therefore, in this figure, the first processing interface 31 passes through the region 319 having the first processing shape, which is indicated by the hatched ellipse. Similarly, the second processing interface 32 passes through the region 329 having the second processing shape, which is indicated by the hatched ellipse, and the second processing shape is determined to bring about a process different from the process brought about by the first processing shape.

[0122] Generally, these two processing shapes 319, 329 are selected and determined to produce different effects on the signal over the entire length of the processing shapes 319, 329. This difference is preferably obtained by different shapes in the same configuration of material and / or magnetization. Alternatively, or in combination, this difference can also be obtained by different materials and / or different magnetizations.

[0123] Therefore, as can be understood from the figure, this non-inverting processing unit R3 supplies the processing optical signal OS3 in the first direction at the output, and the amplitude of this processing optical signal OS3 has undergone the first processing brought about by the shape of the first processing interface 21 alone. The result of this first processing is here indicated by the thick arrow with a square dashed line on the right interface 31 at the output of the first processing region 319.

[0124] On the second processing interface 32 on the left of the figure, the initial signal OS0 contains no energy passing through the second processing region 329 or only a very weak part of the energy, as indicated by the thin dashed arrow.

[0125] In the second direction or the reverse direction, the figure generally shows the passage of the reverse input signal OS0B, which is, for example, the same as the input signal OS0. The reverse input signal OS0B enters as the differentiated reverse signal OS2B and then passes through these same regions R3, R2, R1 in the reverse direction indicated by the thick gray arrow, i.e., here from the top to the bottom of the figure.

[0126] Therefore, on the condition that the differentiated inverse signal OS2B has a non-zero component on the second processing interface 32, this differentiated inverse signal OS2B undergoes a second process and emerges here in the form of a processed inverse signal OS3B indicated by a thick arrow with a round dot.

[0127] This inverse signal OS2B itself, when aggregated on the interface located on the right, then undergoes only the second process in the second area 329 and emerges in the form of an inverse output signal OS9B that has undergone only the second process via the first optical port P1.

[0128] Therefore, as can be understood, this component 1 supplies, at the output, an inverse output signal OS9B, and the inverse output signal OS9B has undergone a process different from the process received by the processing signal OS9 traveling in the first direction. The input signal OS0 and the inverse input signal OS0B passing through the first optical port P1 of this component 1 in two opposite directions emerge in two forms from different outputs OS9 and OS9B even if they are identical at the input.

[0129] Note that if this inverse signal OS2B is not pre-aggregated only on the right interface, the inverse signal OS2B can also undergo the first process within the first processing area 319. However, the output signal OS9B is composed of the energy that has undergone the first process 319 and the second process 329. Therefore, the output signal OS9B has undergone a process different from the input signal OS0 and emerges as OS9B in a different form.

[0130] Exemplary isolator FIG. 2 shows an exemplary embodiment of a first group in which two ports are provided for component 2, and the two ports communicate only with each other.

[0131] In this example, the second optical port P2 includes a selection area R1B and a differentiation area R2B, and the selection area R1B and the differentiation area R2B have functions similar to the functions of the areas R1 and R2 of the first optical port P1 and are, for example, identical.

[0132] The differentiated region of the second optical port P2 is interconnected by a common region that forms the non-inverting processing region R3 of the first port P1 and the non-inverting processing region R3B of the second port P2.

[0133] In embodiments including two selection ports and a differentiation port coupled to each other, this figure shows a specific example where the non-inverting processing regions R3, R3B are configured to result in absorption in direction 319 and transmission in the reverse direction 329.

[0134] In this schematic example, the first processing interface 31 has a cavity 311, and the cavity 311 is configured to completely or partially absorb the signal OS2 reaching the cavity 311 on this magnetic plasmon interface 31. The processed signal OS3 emerging from the cavity 311, and thus the output signal OS9, is thus weakened or even non-existent, resulting in a blocking direction with respect to the input via the first port P1.

[0135] In this schematic example, the first processing interface 31 has a shape including an absorption cavity 311, and the absorption cavity 311 has a shape configured to completely or partially absorb the optical signal OS2 of a determined frequency reaching the cavity 311 on this magnetic plasmon interface 31. The processed signal OS3 emerging from the cavity 311, and thus the output signal OS9, is thus weakened or even non-existent, resulting in a blocking direction with respect to the input via the first optical port P1.

[0136] In this regard, the second processing interface 32 has a straight equilateral and equiangular shape and is configured to allow the passage of the reverse input optical signal OS0B. The reverse input optical signal OS0B is OS2B that converges on the second interface 32 since it is on the right when viewed from the propagation direction of the optical signal. Thus, the processed reverse signal OS3B is transmitted as is and emerges in the form of the reverse output signal OS9B. The reverse output signal OS9B is substantially the same or at least represents the reverse input signal OS0, and thus results in a passing direction for the input via the second optical port P2.

[0137] Figure 3a shows a similar example where the shape of the first processing interface 311’ has several cavities. The example of Figure 3a, in particular, has several cavities of the same dimensions for absorbing most of the signal at the determined frequency. In this example, the first processing interface 31 also includes a plurality of absorption cavities, and the plurality of absorption cavities are sized to completely or partially absorb an optical signal including different frequencies, particularly a plurality of frequencies shifted relative to each other, so that the processing interface absorbs all signals within the determined frequency range.

[0138] Thus, each dimension corresponding to the determined frequency (or sub-range of frequencies) within this range results in a cut-off process for a wider band.

[0139] Such absorption cavities 311, 311’ are, for example, but not limited to, of the homogeneous or different types of Helmholtz or Fabry - Pérot type.

[0140] Furthermore, many combinations of cavities are possible, for example, several identical cavities for thin and deep absorption, and / or several individual absorption ranges spaced apart from each other.

[0141] In the examples of Figures 3a and 3b, the second processing interface 32 has a shape configured to allow an optical signal having one or more frequencies absorbed by the first processing interface 31 to pass through. Thus, as can be understood, this component forms an optical isolator for any optical signal located within the range absorbed by the first interface 31.

[0142] Such an isolator can, for example, isolate a laser source S0 connected to the second port P2 and inject the laser source S0 into an optical circuit via the first port P1. In this case, this laser source emits in the second direction that constitutes the passing direction or “outward” direction of this isolator, while the return light reaching from the circuit via the first port P1 is blocked in the first direction or “return” direction.

[0143] In other exemplary embodiments not detailed herein, the second processing interface itself may also have absorption cavities with different characteristics. For example, it blocks specific frequencies in one direction and other frequencies in the other direction.

[0144] Circulator FIG. 4 shows an exemplary embodiment of a second group in which the components are provided with three or more ports, and the three ports communicate with each other.

[0145] In this second group, the difference in processing between two processing interfaces 31 and 32 of the same processing region R3 is formed by these two interfaces bringing about processing in the form of a detour. That is, specifically, the processing interfaces 31 and 32 do not communicate with the same location, particularly the same optical port.

[0146] In such a component, the first processing interface 31 of the first optical port P1 has a shape that forms a detour with respect to the second processing interface 32. In particular, the first processing interface 31 ends at an optical port P2 different from the port P3 reached by this second processing interface 32, or travels a different distance, or incorporates further or different processing.

[0147] In the example of FIG. 4, the three optical ports P1, P2, and P3 all communicate with each other. Each optical port P1, P2, P3 includes its own selection region R1, R1B, R1C, then its own differentiation region R2, R2B, R2C, and the processing regions R3, R3B, R3C after leaving the differentiation regions R2, R2B, R2C, and the processing regions R3, R3B, R3C communicate with the regions of adjacent ports.

[0148] The guiding core of each of the three ports extends into the cores of adjacent ports and communicates with the cores of adjacent ports in the form of a continuous material that here forms a star creating a kind of branch for each port. Each of these branches forms a non-inverting processing core and is surrounded by two processing interfaces. For each optical port, e.g., P1, the first processing interface 31 extends on the outer periphery of the core and becomes the second processing interface 32B of the adjacent optical port on the first side, while the second processing interface 32 extends on the outer periphery of the core and becomes the first processing interface 31C of the adjacent optical port on the other side, i.e., the second side.

[0149] Accordingly, each optical port that receives an input optical signal transmits the input optical signal along the first processing interface towards the adjacent optical port on the first side, while it can receive an output optical signal arriving from the optical port on the second side along the second processing interface. As can be understood, an optical signal circulator is thus obtained, which here has a circulation among three ports.

[0150] Accordingly, the first signal OS0 entering the first optical port P1 is transmitted via the first processing interface 31 of the first port P1 towards the second interface 32B of the second port P2 and leaves the second port P2 as OS9.

[0151] In this regard, the second signal OS0B entering the second optical port P2 is transmitted via the first processing interface 31B of the second port P2 towards the second interface 32C of the third port P3 and leaves the third port P3 as OS9B.

[0152] The third signal OS0C entering the third optical port P3 is transmitted via the first processing interface 31C of the third port P3 towards the second interface 32 of the first port P1 and leaves the first port P1 as OS9C.

[0153] This example has three ports, but it should be understood as an illustration of any number of ports including 4, 5, 6, 7, 8 or any integer value made possible by an implementation of a high-precision technical method.

[0154] In other embodiments not described in detail herein, all or some of the processing interfaces combine and / or alternate some types of processing, such as bypass and absorption.

[0155] For example, it is possible to generate components where signals entering through a specific port are bypassed while other signals are absorbed, and / or components where specific frequencies are absorbed while other frequencies are bypassed.

[0156] The examples shown herein relate to "complete" circulations that can be classified as such as long as each port communicates with an emission port and another reception port. The execution of an incomplete circulation is assumed, for example, where a specific port forms a dead end at the input, together with, for example, a first absorption processing interface and a second passing processing interface.

[0157] The transmission of light inside components 1, 2, 3 is carried out by induction within optical cores 10, 20, 30, which are formed by core materials where the optical signal is induced between the core and the environment made of one or more materials at the interfaces.

[0158] In this example, this optical core is generated by the core material within the planar propagation layer C1 included between the lower layer C0 and the upper layer C2.

[0159] The optical signal is confined within this propagation layer C1 by the lower and upper interfaces formed by the lower and upper layers. Typically, these interfaces are planar or at least two-dimensional, and thus the lower and upper interfaces are uniform and / or form regular planar waveguides.

[0160] In these examples, the lateral confinement inside the propagation layer C1 is embodied by cross-sections that form a plane or two-dimensional surface crossing different layers C0, C1, C2. Inside the propagation layer C1, these cross-sections have a shape that varies along the signal propagation direction.

[0161] The configuration of the planar waveguide having such a cross-section is particularly advantageous. This is because, in particular, it is possible to easily provide any shape, and potentially quite complex shapes, simply by performing photolithography, for example, during the manufacture of the propagation layer.

[0162] In the examples shown in FIGS. 5 to 7, the lower layer and the upper layer are solid layers, one of C0 is the substrate, and the other is the cover layer. In the example shown in FIG. 8, the propagation layer C1 is suspended between two layers formed by the ambient air, that is, between the lower layer and the upper layer.

[0163] Magnetic plasmon structure FIGS. 5 to 8 show examples of magnetic plasmon structures that can be used to generate a magnetic plasmon interface for components according to the present invention, for example, the components of FIGS. 1, 2, and 4.

[0164] In the current preferred embodiment, particularly the embodiment shown here, the magnetic plasmon induction core is made of a dielectric material for the differentiation core 20 and the non-opposite processing core 30 here. In this regard, the outer material is made of one or more metallic materials. Thus, a structure of a type that can be called a "slot", also referred to as MIM (abbreviation for "metal-insulator-metal") or MDM (abbreviation for "metal-dielectric-metal"), is obtained.

[0165] Alternatively, in a manner not shown in cross-section herein, the components can be produced in a ribbed configuration or by DMD with the same type of material.

[0166] FIG. 5 shows a first exemplary configuration of materials for the interfaces 21, 22 of the differentiation region R2 and the interfaces 31, 32 of the non-opposite processing region R3.

[0167] In this configuration, the outer materials 220, 320 and 210, 310 are plasmonic metals or plasmonic metal alloys, referred to here as "MP", and are particularly selected from gold, silver, copper, aluminum, and combinations of two, three, or four of these materials.

[0168] The materials of the induction cores 20 and 30 are dielectric materials of the magneto-optical material type, herein referred to as DMO, and in particular, garnet oxides, or dielectric materials doped with ferrite particles such as, for example, mineral silica doped with cobalt ferrite nanoparticles.

[0169] Figures 3a and 3b show the non-reciprocal processing section R3 for an exemplary isolator according to this structure. The induction core 30 is made of a magneto-optical dielectric material of the magneto-optical garnet oxide type, for example, in a propagation direction parallel to the axis x. This magneto-optical dielectric material is herein bismuth iron garnet ("BIG"), but known materials such as Bi:YIG or Ce:YIG can be used. Since the outer materials 310 and 320 are made of gold (Au) herein, magnetic plasmon interfaces 31 and 32 are brought about under the effect of the transverse magnetization "M" along the axis y. The first processing interface 31 has one or more loss cavities, and the one or more loss cavities extend laterally into the outer material 310 along the axis z and are formed on one of the walls of the wall of the MIM waveguide. These cavities 311' have different dimensions and several examples of each dimension. The cavities can have different shapes, for example, particularly rectangular of the Helmholtz type.

[0170] Figure 6 shows a second exemplary configuration of the materials for the interfaces 21 and 22 of the differentiation region R2 and the interfaces 31 and 32 of the non-reciprocal processing region R3.

[0171] In this configuration, the outer material is a metal structure formed by the superposition of two layers of two different metals MMO and MP.

[0172] In this case, these two metals are - On the one hand, a plasmonic metal or a plasmonic metal alloy, herein referred to as MP, and in particular, selected from gold, silver, copper, aluminum, and combinations of two, three, or four of these materials, - On the other hand, a magneto-optical metal, herein referred to as MMO, and in particular, an iron-cobalt alloy.

[0173] In this case, the materials of the cores 20, 30 are typically dielectrics that are not of the magneto-optical type, and are herein referred to as "D", and are selected in particular from silica, silicon nitride, germanium, silicon, and some combinations of these materials.

[0174] Figure 7 shows a third exemplary configuration of materials for the interfaces 21, 22 of the differentiation region R2 and the interfaces 31, 32 of the non-reciprocal processing region R3.

[0175] In this configuration, the outer material is a magneto-optical type metal, herein referred to as MMO, and in particular an iron-cobalt alloy.

[0176] In this case, the materials of the cores 20, 30 are typically dielectrics that are not of the magneto-optical type, and are herein referred to as "D", and are selected in particular from silica, silicon nitride, germanium, silicon, and some combinations of these materials.

[0177] Figure 8 shows a fourth exemplary configuration of materials of a modified form of the configuration of Figure 7 here. In this example, the propagation layer C1 is suspended and not covered. That is, the lower layer C0 and the upper layer C2 are each formed by air or any other ambient atmosphere. Note that the characteristics of the suspended and / or uncovered propagation layer C1 are also applicable to other magnetic plasmon structures.

[0178] Naturally, the present invention is not limited to the described examples, and numerous modifications can be made to the examples as long as the scope of the present invention is not exceeded.

Explanation of reference numerals

[0179] 1 Non-reciprocal component 2 Isolator 3 Circulator 10 Selective core 11 First interface 12 Second interface 110 Outer material of the first selective interface 120 Outer material of the second selective interface 20 Differentiation core 21 First Differentiated Interface 22 Second Differentiated Interface 210 Outer Material of the First Differentiated Interface 220 Outer Material of the Second Differentiated Interface 30 Non - Opposite Processing Core 31 First Non - Opposite Processing Interface 32 Second Non - Opposite Processing Interface 310 Outer Material of the First Non - Opposite Processing Interface 320 Outer Material of the Second Non - Opposite Processing Interface 311 Absorption Cavity 311’ Group of Absorption Cavities 319 Shape of the First Non - Opposite Processing 329 Shape of the Second Non - Opposite Processing 31B First Non - Opposite Processing Interface 32B Second Non - Opposite Processing Interface (Second Optical Port) 31C First Non - Opposite Processing Interface 32C Second Non - Opposite Processing Interface (Third Optical Port) C0 Lower Layer C1 Propagation Layer C2 Cover Layer OS0 First Input Optical Signal (Return Signal) OS0B Second Input Optical Signal (Inverse Signal, Outward Signal) OS0C Third Input Optical Signal OS1 First Selected Optical Signal (Return Signal) OS2 First Differentiated Optical Signal (Return Signal) OS3 First Processed Optical Signal (Return Signal) OS9 First Output Optical Signal (Return Signal) OS1B Second Selected Optical Signal (Inverse Signal, Outward Signal) OS2B Second Differentiated Optical Signal OS3B Second Processed Optical Signal OS9B Second Output Optical Signal OS1C Third Selected Optical Signal OS9C Third Output Optical Signal P1 First Optical Port P2 Second Optical Port P3 Third Optical Port R1 Selection Region Input section of R10 (selection area) Selection aperture of R11 (selection area) Selection section of R12 (selection area) Differentiation area R2 Input section of R20 (differentiation area) Output section of R22 (differentiation area) Non-inverse processing area R3 Input section of R30 (processing area) Output section of R32 (processing area) Selection area R1B (second optical port) Selection area R1C (third optical port) Differentiation area R2B (second optical port) Differentiation area R2C (third optical port) Non-inverse processing area R3B (second optical port) Non-inverse processing area R3C (third optical port) Laser source S0 Differentiation width Wd Selection width We W R0 Input width of selection area R1

Claims

1. An optical component (1, 2, 3), wherein the optical component (1, 2, 3) comprises at least one first optical port (P1) or an optical waveguide, and a second optical port (P2) or an optical waveguide, and the first optical port (P1) or the optical waveguide, and the second optical port (P2) or the optical waveguide transmit light to each other in a non-reciprocal manner, connecting or enabling the connection of the components within an induced optical circuit, and the at least one first optical port (P1) or the optical waveguide, or the second optical port (P2) or the optical waveguide can exchange optical signals with the optical circuit in one first direction and a second direction opposite to the one first direction through at least one magnetic plasmon induction region, The component includes a plurality of regions through which an optical signal traveling from the first optical port towards the second optical port in the first direction can continuously pass, and the regions at least include, in order, - a selection region (R1); - a differentiation region (R2); - a non-reciprocal processing region (R3) and includes, The selection region (R1) 〇 At the input (R10), receives an input optical signal (OS0) propagating according to an input optical impedance z0 ｅｆｆ and, 〇 At the output (R12), emits an optical signal called a selected optical signal (OS1) by plasmon induction. The optical signal (OS1) includes a first plasmon mode defined as one LR mode and a second plasmon mode defined as the other SR mode. The second plasmon mode has an amplitude of zero or significantly smaller than the first plasmon mode. The differential region (R2) is configured to receive, at an input (R20), the selected optical signal (OS1) arriving from the selection region (R1) and guide the optical signal (OS1) to an output section (R32). In the output section (R32), the optical signal (OS1) forms an optical signal called a differential optical signal (OS2) by an induction core called a differential core (20). The differential core (20) extends between at least two interfaces called differential interfaces (21, 22). The differential interfaces include a first differential interface (21) and a second differential interface (22). The differential interface is 〇 formed between a core material and at least one outer material, and of the core material and the at least one outer material, one is a dielectric material and the other is a metal material. 〇 located on two opposite sides. 〇 determined to bring about magnetic plasmon induction with a magneto-optical effect between the materials. The output section (R22) is configured to have a differential width Wd between the differential interfaces. The differential width Wd is selected such that the signal of the first plasmon mode has an asymmetric amplitude greater than a determined threshold between the differential interfaces. The differential region supplies, at the output, an asymmetric optical signal called a differential signal (OS2). The amplitude of the differential signal (OS2) is concentrated on the first differential interface (21). The non-inverting processing region (R3) is configured to receive, at an input (R30), the differential signal (OS2) arriving from the differential region (R2) and guide the differential signal (OS2) to an output region (R32) via a processing section (31). In the output region (R32), the differential signal (OS2) forms an optical signal called a processed optical signal (OS3). via an induction core called a processing core (30) that extends between at least one first processing interface (31) and a second processing interface (32). The processing interfaces (31, 32) are 〇 Each is formed between a processing core material (30) and an outer material (310, 320), one of the processing core material and the outer material being a dielectric and the other being a metal. 〇 Located on two opposite sides, each the same as that of the first differentiation interface (21) and the second differentiation interface (22). 〇 Determined to bring about magnetic plasmon induction between the materials. 〇 Configured to have shapes (319, 329) that are not equivalent to each other between the input (R30) and the output (R32) of the processing region (R3). The non-inverting processing region (R3) supplies a processed optical signal (OS3) in a first direction at the output, and the processed optical signal (OS3) of the amplitude is undergoing processing brought about by the shape of the first processing interface (21) alone, while A signal (OS2B) passing through the processing region (R3) in the second direction, called an inverse signal, undergoes processing brought about by the shape of the second processing interface (32) alone or by both of the two processing interfaces, and supplies a processed inverse signal (OS3BB), and the processed inverse signal (OS3BB) undergoes processing different from the processing received by the processed optical signal (OS3) traveling in the first direction, characterized component (1, 2, 3).

2. The selection region (R1) - An input portion (R10) forming a conductive path having an input optical impedance z0 for a signal received at the input. ｅｆｆ And an output portion (R12) forming a magneto-optical plasmon induction between two magnetic plasmon interfaces. And having, the input portion (R10) and the output portion (R12) communicating through a selection aperture (R11) having a selection width (We), the value of the input optical impedance z0 And the value of the selection width (We) are ｅｆｆ 〇 The first plasmon mode has a first optical impedance z1 substantially equal to the input optical impedance z0 And, ｅｆｆ And having, ｅｆｆ And, the second plasmon mode is the first optical impedance z1 ｅｆｆ A second optical impedance z2 that is significantly different from ｅｆｆ have 2. The component according to claim 1, characterized in that the input optical signal (OS0) excites, through the selection aperture (R11), at the output (R12) an optical mode according to the first plasmon mode to form a selection optical signal (OS1), the first plasmon mode being to the exclusion of the second plasmon mode or being of an amplitude significantly larger than the amplitude of the second plasmon mode.

3. The differentiation region (R2) has a shape whose width increases from its input width to the differentiation width (Wd), 3. A component according to claim 1 or 2, characterized in that the non-reciprocal processing region (R3) has a shape in which the input width of R3 and / or the average width of the input widths of R3 is substantially equal to the differentiation width Wd.

4. the differentiation core (20) and / or the processing core are made of a dielectric material and the outer material (210, 220, 310, 320) is made of one or more metallic materials; Component according to claim 1 , characterized in that the materials are selected to provide a plasmon-guiding interface between them implementing a magneto-optical effect.

5. the differentiation core (20) and / or the processing core are made of a metallic material and the outer material (210, 220, 310, 320) is made of one or more dielectric materials; Component according to claim 1 , characterized in that the materials are selected to provide a plasmon-guiding interface between them implementing a magneto-optical effect.

6. The various cores (10, 20, 30) are completely or partially generated within a planar or two-dimensional layer (C1) included between a lower layer (C0) and an upper layer (C2), and the interfaces (21, 22, 31, 32) that bring about induction within the core form a plane or two-dimensional surface orthogonal to the layers (C0, C1, C2). The component according to any one of claims 1 to 5, characterized in that.

7. The selection region (R1) is configured such that the first plasmon mode corresponds to the LRSPP type mode of the input optical signal (OS0). The component according to any one of claims 1 to 6, characterized in that.

8. The first processing interface (31) has a shape including one or more absorption cavities (311), and the one or more absorption cavities (311) have a shape configured to completely or partially absorb an optical signal of a determined frequency, while The second processing interface (32) has a shape configured to allow the passage of an optical signal (OS0B) of the determined frequency. The component (1, 2) according to any one of claims 1 to 7, characterized in that.

9. The first processing interface (31) includes a plurality of absorption cavities, and the plurality of absorption cavities are dimensioned to completely or partially absorb optical signals including different frequencies, while the processing interface absorbs all signals within a determined frequency range, The second processing interface (32) has a shape configured to allow the passage of an optical signal of one or more frequencies within the determined frequency range. The component according to any one of claims 1 to 8, characterized in that.

10. The first optical waveguide or port (P1) and the second optical waveguide or port (P2) each include a selection region (R1, R1B) and a differentiation region (R2, R2B), and are connected to each other by at least one first processing region (R3). The first processing interface (31) of the at least one first processing region (R3) is configured to completely or partially absorb an optical signal of one or more determined frequencies. The second processing interface (32) of the at least one first processing region (R3) is configured to allow the passage of the optical signal, and thus, to generate an optical isolator for the optical signal. The component (2) according to claim 8 or 9, characterized in that.

11. The first processing interface (31) of the first optical port (P1) has a shape that forms a detour with respect to the second processing interface (32). In particular, the first processing interface (31) ends at a different optical port (P3), or travels different distances, or incorporates further or different processes. The component (3) according to any one of claims 1 to 10, characterized in that.

12. The first processing interface (31) of the first optical port (P1) has a shape that connects the first optical port (P1) to the second optical port (P2), guides a differentiated optical signal (OS0) arriving from the first optical port (P1), and emits an output optical signal (OS9) through the second optical port (P2). The second processing interface (32) of the first optical port (P1) has a shape that connects the first optical port (P1) to a third optical port (P3) different from the second optical port (P2), receives a differentiated optical signal (OS9C) arriving from the third optical port (P3), and emits the differentiated optical signal (OS9C) as an output optical signal (OS9C) through the first optical port (P1). Thus, the component (3) according to any one of claims 1 to 11, characterized in that it generates an optical circulator.

13. The component is configured and determined to process an optical signal having a wavelength of less than 50 μm, and is characterized in that it is the component (1, 2, 3) according to any one of claims 1 to 12.

14. Regarding the interfaces (21, 22, 31, 32) of the differentiation region (R2) and / or the processing region (R3), - The metal material is a plasmonic metal or a plasmonic metal alloy, - The dielectric material is a magneto-optical material, and is characterized in that it is the component (1, 2, 3) according to any one of claims 1 to 13.

15. Regarding the interfaces (21, 22, 31, 32) of the differentiation region (R2) and / or the processing region (R3), the metal material is at least - On the one hand, a plasmonic metal or a plasmonic metal alloy, and - On the other hand, it is a stack or combination of layers of a magneto-optical type metal, and is characterized in that it is the component (1, 2, 3) according to any one of claims 1 to 13.

16. Regarding the interfaces (21, 22, 31, 32) of the differentiation region (R2) and / or the processing region (R3), - The metal material is a magneto-optical type metal, - The dielectric material is a dielectric that is not of the magneto-optical type, and is characterized in that it is the component (1, 2, 3) according to any one of claims 1 to 13.

17. A device comprising one or more non-reciprocal optical components (1, 2, 3) according to any one of claims 1 to 16, which are generated integrally within an integrated optical circuit.

18. The integrated optical circuit includes at least one integrated laser emitter, and the device comprises at least one non-reciprocal optical component (1, 2) according to any one of claims 8 to 10 connected to the laser, the non-reciprocal optical component (1, 2) being capable of passing an optical signal emitted by the laser in one direction while preventing or reducing the optical signal from returning towards the laser in the opposite direction, the device according to claim 17.

19. The integrated optical circuit comprises a plurality of sensors generated or connected in an integrated manner, and the device comprises at least one non-reciprocal optical component (3) according to claim 12 connected to at least two of the sensors, the device according to claim 17 or 18.

20. The integrated optical circuit comprises at least one non-reciprocal optical component (3) according to claim 12, the non-reciprocal optical component (3) being connected so as to generate an integrated Michelson interferometer, either completely or partially, the device according to claim 17 or 18.

21. A method for processing an optical signal, the method comprising processing the optical signal by passing it through at least one non-reciprocal optical component (1, 2, 3) according to any one of claims 1 to 16 or a device according to any one of claims 17 to 20, a method for processing an optical signal.

Images