WO2021060981A1

WO2021060981A1 - Broadband optical coupler and switch

Info

Publication number: WO2021060981A1
Application number: PCT/NL2020/050588
Authority: WO
Inventors: Bakiye Imran AVCI
Original assignee: Stichting Vu
Priority date: 2019-09-25
Filing date: 2020-09-24
Publication date: 2021-04-01
Also published as: NL2023901B1

Abstract

A broadband optical coupler (50) comprises a set of waveguides (10,20). Each waveguide comprises a respective curved starting sections (11,21) where the waveguides curve towards each other, straight middle sections (12,22) where light can couple between the waveguides, and ending sections where the waveguides curve away from each other. To increase bandwidth while maintaining a compact design, the straight middle sections (12,22) are disposed at a tangential offset (Y1) with respect to each other along a direction (Y) of their respective length (L2,L6). An optical switch (100) comprises at least two of such couplers connected in series. An optical device (1000) employs the switch or coupler in broadband applications such as optical coherence tomography (OCT).

Description

Title: BROADBAND OPTICAL COUPLER AND SWITCH

TECHNICAL FIELD AND BACKGROUND

The present disclosure relates to the coupling or switching of optical signals, in particular broadband signals which may include an extended range of wavelengths.

Optical signals, e.g. continuous or pulsed light, can be delivered by guiding the light via waveguides such as optical fibers and/or chip-based optical lanes and other structures. An optical coupler can be used to couple and/or distribute light between waveguides. For example, a 3dB coupler can be used to split incoming light into equal parts distributed over different waveguides. An optical switch can be used to selectively switch the passage of light between different waveguides. For various applications, e.g. telecommunication, an interest exists to allow the coupling or switching of broadband optical signals. Since the optical structures may be dimensioned to efficiently operate for limited wavelengths, it can be a challenge to extend satisfactory operation to a larger wavelength range while managing optical losses.

As background, Tran et al. [DOI: 10.1109/IPCon.2016.7831302] describe the design and fabrication of a broadband thermo-optic Mach- Zehnder Interferometer switch using two 3-dB adiabatic couplers on the 500 nm Silicon-On-Insulator platform. The fabricated switch has shown 120 nm optical bandwidth with more than 15 dB extinction ratios. The coupler comprises three following sections: the first section (300 pm long) where the two waveguides (500 nm and 700 nm wide) slowly come close to each other through S-bends, the coupling section (500 pm) where the two asymmetric waveguides are linearly tapered to 600 nm wide waveguides at the end, and the third section (300 pm) where each 600 nm wide waveguide is guided to one of the output waveguides by two other S-bends.

There is yet a need for further improvement of the known designs of optical couplers and switches, e.g. with regards to bandwidth, extinction ratio, losses, and compactness. SUMMARY

Some aspects of the present disclosure relate to a broadband optical coupler comprising a set of waveguides. Typically, each waveguide of the coupler comprises a respective curved starting section, straight middle section, and curved ending section. At least one of the starting sections is configured to receive light at its respective entry. The curved starting sections follow a respective curved path towards each other starting at a first distance apart and getting closer together along their respective length. Each curved starting section transitions at its respective end into a respective straight middle section. The straight middle sections are parallel to each other over at least a subsection of their respective length. The parallel subsections are at a second distance apart, which is smaller than the first distance to couple at least part of the light between the waveguides. Most preferably widths of the middle sections taper inward and outward, respectively, in opposite directions between the waveguides along their respective length. Each straight middle section transitions at its respective end into a respective curved ending section. At least one of the ending sections is configured to output a respective part of the light. The ending sections follow a respective curved path curve away from each other getting further apart along their respective length to end at a third distance larger than the second distance.

As described herein, the straight middle sections are preferably disposed at a tangential offset with respect to each other along a direction of their respective length. In other words, the first middle section is parallel but shifted with respect to the second middle section. Specifically, the middle sections are shifted along a length coordinate of the sections, i.e. tangential or along a line parallel to the straight middle sections. The tapered straight middle sections may also be referred to as adiabatic sections. Without being bound by theory, it will be appreciated that the asymmetry between the two tapered waveguides (i.e. asynchronicity) has a critical role on the wavelength insensitive bandwidth of the coupler. Typically, increased asynchronicity improves the bandwidth at the expense of coupler length. By adding a tangential offset, as described herein, the adiabatic part can be considered as a combination of a some curved waveguides and tapered waveguides. Advantageously, the present configuration increases the asynchronicity between the two waveguides without changing the in and out waveguide widths of the tapered section. As will be further appreciated, adding a tangential offset can help to satisfy the asymmetry condition for broadband operation of the adiabatic couplers with minimal or no increase of the length of the device.

Other or further aspects of the present disclosure relate to a broadband optical switch, e.g. comprising an arrangement of at least two couplers as described herein. Yet further aspects relate to an optical device employing one or more switches, or couplers. For example, new applications for sensory devices used in optical coherence tomography may be enabled by the present teachings, as will be described herein. Also new methods of designing optical couplers, switches, and devices can be envisaged.

BRIEF DESCRIPTION OF DRAWINGS

These and other features, aspects, and advantages of the apparatus, systems and methods of the present disclosure will become better understood from the following description, appended claims, and accompanying drawing wherein:

FIGs 1A-1C illustrate respective views of an optical coupler;

FIG 2 illustrates a top view of an optical switch;

FIG 3A illustrates an optical switch with a switching mechanism to control a respective path of the light;

FIG 3B illustrates use of an optical switch in an optical device; FIG 4A illustrates a graph of a splitting ratio of two coupler designs;

FIG 4B illustrates a graph of relative transmission or extinction ratio of corresponding designs;

FIG 5A illustrates a similar graph as FIG 4B, but for a range of different tangential offsets;

FIG 5B illustrates a plot of the bandwidth where the extinction ratio is better than 30dB;

FIGs 6A and 6B illustrate propagation through the waveguides of a switch without and with tangential offset, respectively.

DESCRIPTION OF EMBODIMENTS

Terminology used for describing particular embodiments is not intended to be limiting of the invention. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. The term "and/or" includes any and all combinations of one or more of the associated listed items. It will be understood that the terms "comprises" and/or "comprising" specify the presence of stated features but do not preclude the presence or addition of one or more other features. It will be further understood that when a particular step of a method is referred to as subsequent to another step, it can directly follow said other step or one or more intermediate steps may be carried out before carrying out the particular step, unless specified otherwise. Likewise it will be understood that when a connection between structures or components is described, this connection may be established directly or through intermediate structures or components unless specified otherwise.

Embodiments described herein provide a broadband optical coupler comprising a set of at least two waveguides. Preferably, each waveguide comprises a respective curved starting sections where the waveguides curve towards each other, straight tapering middle sections where light can couple between the waveguides, and ending sections where the waveguides curve away from each other. Most preferably the straight middle sections are disposed at a tangential offset with respect to each other along a direction of their respective length. An optical switch can comprise at least two of such couplers, e.g. connected in series. An optical device can employ the switch or coupler in broadband applications such as optical coherence tomography.

The invention is described more fully hereinafter with reference to the accompanying drawings, in which embodiments of the invention are shown. In the drawings, the absolute and relative sizes of systems, components, layers, and regions may be exaggerated for clarity. Embodiments may be described with reference to schematic and/or cross- section illustrations of possibly idealized embodiments and intermediate structures of the invention. In the description and drawings, like numbers refer to like elements throughout. Relative terms as well as derivatives thereof should be construed to refer to the orientation as then described or as shown in the drawing under discussion. These relative terms are for convenience of description and do not require that the system be constructed or operated in a particular orientation unless stated otherwise.

FIGs 1A-1C illustrate respective views of an optical coupler 50. The broadband optical coupler 50 comprises a set of at least two adjacent waveguides 10,20.

FIGs 1A and 1C illustrates corresponding front and backside cross-sectional views of the coupler. In some embodiments, waveguides can be formed by sections of wave guiding material, e.g. typically material that is transmissive to the wavelength l used, surrounded by material having sufficiently different index of refraction to confine the waves. For example, waveguides of silicon can be formed onto or into a surrounding silicon dioxide matrix. Also other materials can be used depending on the wavelength. Typically, the waveguides have a height HI on the order of the wavelength used, e.g. similar to the widths as described herein. For example, HI is 0.1 — 10 pm. In some embodiments, the waveguides can e.g. be defined by an extra height H2 of the waveguide sections, e.g. H2>0.1 pm, compared to the surrounding non-guiding sections. For example, the thickness of the non-guiding surrounding sections (H1-H2) is less than the smallest width of the waveguides. In other or further embodiments, the waveguides can be completely surrounded by non-guiding material, i.e. H1=H2.

In a preferred embodiment, e.g. as shown, the optical coupler 50 is essentially planar. For example, the waveguides (forming the coupler) run adjacent each other with varying distance in the same layer. In this way the coupler can be relatively compact and easy to manufacture. Alternatively, a non-planar coupler can be envisaged, e.g. wherein the waveguides are disposed in different layers. This can provide similar function, with a varying distance between the waveguides being determined by the respective shapes of the waveguides in the respective layers, and a distance between the layers, e.g. as a result of an intermediate layer. For example, it can be envisaged that the waveguides partially overlap in a top view projection while being distanced by a layer there between. In one embodiment (not shown), the coupler comprises a set of waveguides disposed in different layers. For example, the curved starting and ending sections follow a respective curved path in a respective layer towards and away from each other, at least in a top projection view. For example, the straight middle sections are disposed in different layers at a tangential offset with respect to each other along a direction of their respective length, and with an optional layer there between, e.g. of the same material forming the surrounding matrix. FIG IB illustrates a top view of the coupler showing respective paths of the waveguides 10,20. It is noted that the relative scale in FIG IB (and FIG 2) is indicated as X:Y ~ 200:1 - meaning that the X dimension (transverse to the main path) is highly exaggerated compared to the Y dimension (along the main path) by about a factor two hundred. This is done for better illustrating the design, in particular with respect to the comparatively small distances (D1-D3) between the waveguides, and respective width W1-W4 - compared to the much larger lengths (L1-L5) of the sections (ll-13;21-23). As a result of the distorted scale, it may appear e.g. that a respective length of the curved sections has a considerable transverse component, while in reality the length may be predominantly determined by the extent of the section along the Y dimension. As described herein a respective length L1-L7 of a section may thus indicate a length along the Y direction, or path length along the section, which can be very similar, e.g. within one percent or even less than half a percent. Another apparent effect of the scale is that the curvature and change of width over the curved sections may seem relatively abrupt, e.g. for the curved section Si, while in reality the curvature / width change can be quite smoothly distributed over a much larger length than illustrated.

In some embodiments, each waveguide comprises a respective curved starting section 11,21. Typically, at least one of which is configured to receive light IA at its respective start. Preferably, the curved starting sections 11,21 follow a respective curved path Sl,S4, e.g. towards each other starting at a first distance D1 apart and getting closer together along their respective length L1,L7.

In some embodiments, each curved starting section 11,21 transitions at its respective end into a respective straight middle section 12,22. Preferably, the straight middle sections 12,22 are parallel to each other over at least a subsection L26 of their respective length L2,L6. Typically, the parallel subsections L26 are at a second distance D2 apart which is smaller than the first distance Dl. Accordingly, at least part, e.g. AO or OB. of the light IA can be coupled (C) between the waveguides 10,20. Most preferably, the widths W2^W3, W4^W3 of the middle sections 12,22 taper inward and outward, respectively, in opposite directions between the waveguides (indicated in the figure by the arrows Ti,To.

In some embodiments, each straight middle section 12,22 transitions at its respective end into a respective curved ending section 13,23. Typically each ending section is configured to output a respective part OA,OB of the light IA (e.g. to a further section, not shown here). Preferably, the ending sections follow a respective curved path S2,S3, e.g. curving away from each other getting further apart along their respective length L3,L5. Accordingly the waveguides may end at a third distance D3 larger than the second distance D2.

In a preferred embodiment, the straight middle sections 12,22 are disposed at a tangential offset Y1 with respect to each other along a direction Y of their respective length L2,L6. For example, the tangential offset Y1 is selected for a specific design with a length that is on the one hand large enough to create sufficient asymmetry between the waveguides, while on the other hand being small enough to let a fundamental mode in the combined system get excited properly. Too short offset may give smaller operational bandwidth and too long offset may make the splitting ratio away from what it was designed for.

In some embodiments, to have an appreciable effect with regards the asymmetry, the tangential offset Y1 is at least ten micrometers, preferably at least twenty or fifty micrometers, more preferably at least hundred micrometers, or more. In other or further embodiments, to let the fundamental mode in the combined system get excited properly, the tangential offset Y1 is less than five hundred micrometer, preferably less than four hundred or three hundred micrometer, more preferably less than two hundred micrometer.

In some embodiments, the selection of the tangential offset may also depend on the rest of the design. One way to scale the tangential offset Y1 can be to specify the offset in relative terms, e.g. with respect to the length L2 or L6 of the tapering straight middle sections 12,22. For example, this may provide a measure for a length of the parallel subsection L26 relative to the total length L2 or L6 and/or the relative path length over which the curved section 21 approaches the straight section 12. In a preferred embodiment, the tangential offset Y1 is a fraction F of the length L2 and/or L6 of the straight middle sections 12,22. For example, the fraction F is between five and sixty percent, preferably between ten and fifty percent, more preferably between twenty and forty percent, e.g. about thirty percent in the preferred embodiment shown.

In some embodiments, the respective curved starting sections 11,21 of the different waveguides 10,20 have different lengths, e.g. L1<L7 in the embodiment shown. Preferably, the second starting section 12 is substantially longer than the first starting section 11 (or vice versa), e.g. by at least ten percent, at least twenty percent, at least fifty percent (factor

I.5), twice as long (factor two), or more.

In some embodiments, e.g. as shown, the curved starting sections

II,21 start with the same width W1 to transition into a respective different width W2,W4 at the start of the respective middle sections 12,22. For example, the uniform starting width W1 is selected for optimal propagation of the light in a preceding optical circuit. In other embodiments (not shown), the curved starting sections 11,21 may start with respective different widths, e.g. the widths W2,W4 of the respective middle sections 12,22.

In some embodiments, the respective straight middle sections 12,22 of the different waveguides 10,20 have the same length L2=L6. In this case the tangential offset Y1 at the start of the middle sections 12,22 is the same at the tangential offset Y2 at the end of the middle section. In other words the middle section which starts at an earlier position (along the length coordinate Y), also ends at an earlier position. For example, by providing the middle sections 12,22 the same but offset lengths L2=L6, each section may taper inward or outward respectively, in a similar but opposite way.

In some embodiments, a width W2^W3 of one of the middle sections 12 decreases along its respective length L2, and a width W4^W3 of another of the middle sections 22 increases along its respective length L6. For example, the widths W2,W3,W4 of the middle sections vary over a range of values in accordance with a bandwidth of the optical coupler. In a preferred embodiment, a respective ending widths W3 of the middle sections 12,22 are equal. More preferably, the middle sections 12,22 start with a first width W2, and a different second width W4, respectively, while ending with a third width W3, which is the same for both middle sections 12,22, wherein the third width W3 is between the first width W2 and the second width W4, most preferably half-way there between, e.g. W3 = (W2+W4)/2 (plus-minus some tolerance, e.g. ±10%, preferably ±5%, or less). By ending the respective middle sections 12,22 with the same width W3, the coupler can be easily connected to a succeeding part optical circuit, e.g. another coupler as will be described later with reference to FIG 2. Of course, the endings can also have different widths, e.g. when the coupler is used in another (non- symmetric) configuration. For example, a single coupler can be used as a stand-alone splitter or combiner; and/or the widths of the endings can be different to match other waveguides connected to the endings.

In some embodiments, the widths W2,W3,W4 of the middle sections 12,22 decrease and increase, respectively, at a constant rate of change along their respective length L2,L6. In other words, the change of the width per unit length is preferably constant along the length. In other words, the edges within each middle sections 12,22 are preferably straight, along their respective length, and only at a (very small) tapering angle with respect to each other. It is noted that the angle between the edges of the middle sections is barely visible in the figures, yet already highly exaggerated due to the scale X:Y.

In some embodiments, the variation in width W2^W3 and W4^W3 per unit length is relatively small. Without being bound by theory, a relatively smooth variation in width along the respective sections may provide a more adiabatic behavior with fewer losses. For example, the smooth variation in width per unit length can be expressed by the angle between the edges of a respective middle section. The smaller the angle, the more gradual the transition. However, too small angle may lead to excessive lengths of the design.

In some embodiments, opposing side edges 12i,12j of a respective middle section 12,22 are at a (fixed) tapering angle of between one microradian (10^-6 rad) and ten milliradians (10^-2 rad), preferably between ten and thousand microradians (10^-5 rad to 10^-3 rad), most preferably between hundred and five hundred microradians, e.g. around two-hundred- fifty microradians (250 prad). For example, in the non-limiting specific embodiment of the coupler design, as shown, using typical widths W2=1.44pm, W3=1.32pm, and length L2=500pm, an angle between the inward tapering edges of the first middle section 12 can be approximated as tan^_1( 1.44-1.32 / 500 ) = 2.4 · 10^-4 radians (240 prad) or 0.013 degrees plane angle. The outward tapering edges of the second middle section 22 may have the same or similar angle there between, e.g. using typical widths W4=1.2pm, W3=1.32pm, and length L6=500pm.

In some embodiments, a distance D2 between the middle sections 12,22 is constant over the respective length of the parallel subsections L26, e.g. within margin of less than one percent of the distance, more preferably within one per mille, most preferably within manufacturing accuracy. The distance D2 is preferably relatively small to allow coupling between the waveguides, e.g. on the order of a wavelength of the light. For example, the distance D2 is between ten nanometers and ten micrometers, preferably between hundred nanometers and five micrometer, most preferably between five hundred nanometers and two micrometers, e.g. one micrometer depending on manufacturability. While the inner edge 12i of one the middle section 12 may have a slight tapering angle, the opposing edge of the other middle section 22 may have the same or similar angle keeping the distance still constant. Alternatively, the tapering angle can be exclusively on the outside edges (e.g. 12j), or a small deviation may be tolerated.

In some embodiments, the curved starting sections 11,21 and curved ending sections 13,23 are shaped as S-bends S1-S4 to bring the waveguides together and apart. For example, a (furthest) distance D1 between respective starts of the curved starting sections 11,21 is larger than the (closest) distance D2 between the middle sections 12,22 by at least a factor one and half, preferably at least a factor two, most preferably at least a factor three, or more, to prevent inadvertent coupling before the middle sections. The (furthest) distance D3 between respective ends of the curved ending sections 13,23 can be similar as the distance Dl, or higher as shown. For example, the distance D3 is larger than the distance D2 by at least a factor two, three, four, or more, to prevent inadvertent coupling after the middle sections (e.g. between the couplers when used in a switch).

In some embodiments, the width W3 is constant over the curved ending sections 13,23. In other or further embodiments, the respective curved ending sections 13,23 of the different waveguides 10,20 have different lengths L3>L5. For example, a difference in length (L7 - LI) of the curved starting sections 11,21 is the same as a difference in length (L3 - L5) of the curved ending sections 13,23.

FIG 2 illustrates a top view of an optical switch 100. In a preferred embodiment, the optical switch 100 comprise a first optical coupler 50a as described herein and a second optical coupler 50b as described herein. For example, respective curved sections 13a, 23a of the first optical coupler 50a are connected to respective curved sections 23b, 13b of the second optical coupler 50b.

In some embodiments, the second optical coupler 50b has identical features as the first optical coupler 50a. In one embodiment, the second optical coupler 50b is identically shaped as the first optical coupler 50a. In another or further embodiment, the second optical coupler 50b is shaped as a rotated copy of the first optical coupler 50a. For example, a respective waveguide 110,120 of the optical switch 100 is formed by a combination of a respective one of the set of waveguides of each optical coupler 50a, 50b. In one embodiment, e.g. as shown, the straight middle sections 12a, 22b of the interconnected optical couplers 50 along the same waveguide of the optical switch 100 either both taper inward (Ti) or both taper outward (To).

In some embodiments, the respective curved sections 13a-23b; 23a-13b of the optical switch are interconnected by respective intermediate straight sections 55 between the optical couplers 50a, 50b. Preferably, the intermediate straight sections 55 have constant width W3. For example, the width W3 is the same as the ending width of the preceding curved ending sections 13a, 23a and/or the starting width of the succeeding curved starting sections 23b, 13b. Preferably, the intermediate straight sections 55 are parallel to each other, e.g. having a constant distance D3 there between. For example, the distance D3 is the same as the ending distance of the preceding curved ending sections 13a, 23a and/or the starting distance of the succeeding curved starting sections 23b, 13b.

In some embodiments, the intermediate straight sections 55 have a length L4 sufficient to place a switching device (SW) to at least one of the arms (e.g. shown in the next FIG 3A). For example, the length L4 is at least ten micrometers, preferably at least twenty micrometer, more preferably at least fifty micrometer, e.g. hundred micrometer. To apply a switching effect such as heating exclusively to one of the arms with minimal effect in the other arm, preferably the distance D3 is sufficiently high, e.g. at least two micrometer, preferably at least three micrometer, e.g. between four and ten micrometer, or more.

FIG 3A illustrates an optical switch 100 with a switching mechanism SW to control a respective path of the light. In a preferred embodiment, the optical switch 100 comprises a switching mechanism SW to switch light output OA,OB from the optical switch 100.

In some embodiments, the switching mechanism SW is arranged to affect wave propagation in at least one of the waveguides of the optical coupler 50. In one embodiment, all light is coupled exclusively to one of the output arms, e.g. the opposite arm, when the switch is not active, i.e. wave propagation is the same in both intermediate straight sections 55. In another or further embodiment, by slightly delaying wave propagation in one of the arms compared to the other, the waves may constructively or destructively interfere to couple exclusively in the other of the two output arms.

In some embodiments, the switching mechanism SW affects wave propagation by changing a refractive index in a least part of a respective one of the waveguides of the optical coupler 50. Preferably, the switching mechanism SW is applied at one of the intermediate straight sections 55.

For example, the switching mechanism SW comprises a heating element configured to apply heat to at least one of the waveguides, which can affect the refractive index. Also other switching mechanisms can be envisaged, e.g. electrical and/or magnetic.

In some embodiments, the switch may act as a sensor device for heat, electric/magnetic interaction, or other stimuli such as the presence of chemicals, or relative forces e.g. stress/strain applied to one of the waveguides. In one embodiment, the switch is configured to detect a specific chemical composition e.g. binding to a receptor disposed on one of the waveguides and affecting wave propagation. For example, an output intensity of light from one of the arms can be used as a measure for a quantity to be detected by the sensor device.

FIG 3B illustrates use of an optical switch 100 in an optical device 1000. For example, the optical device 1000 comprises at least one optical switch and/or optical coupler as described herein, to enable broadband signals propagating through the device.

In some embodiments, the device comprises light guiding means to direct light from a light source (not shown) via the optical switch. In other or further embodiments, the device comprises a controller configured to selectively control a target of the light by controlling the optical switch. Preferably, the optical device is an integrated device, e.g. fabricated on a chip, including or excluding the controller logic.

In some embodiments, the optical device, switch, and/or coupler, is configured to act as part of an optical coherence tomography system. In one embodiment, the system comprises at least one light source for generating broadband light. In another or further embodiment, the optical device 1000 as described herein is configured for sequentially directing the light of the same at least one light source to illuminate respective different locations of a sample.

In some embodiments, the system or device comprises one or more optical couplers 50 according as described herein (or otherwise), each configured to split the light into different arms for illuminating in parallel the same location, or different locations, of the sample. In one embodiment, e.g. as shown, the different arms for illuminating the sample in parallel provide different lengths Ah for respective light paths. In another or further embodiment, e.g. as shown, a common light path II is split into a first set of light paths Ila-Ild having a first set of different lengths AL=0,3x,6x,9x. For example, each light path Ila of the first set of light paths is split into a second set of light paths having a second set of different lengths AL=0,x,2x. Accordingly, a total length of each light path is determined by a combination of the first and second sets of different lengths. For example, the second set of different lengths AL=0,x,2x further subdivide the first set of different lengths AL=0,3x,6x,9x, or vice versa, to provide each light path formed by a combination of light paths from the first and second sets with a unique length AL=0,x,2x,3x, etc.

The frame rate of an optical coherence tomography (OCT) system is typically limited by the speed of the sensor (camera) or the sweep rate of the light source. This can be alleviated by multiple-beam imaging, in which different locations on the sample are illuminated by an array of light simultaneously. This may allow parallel imaging from multiple sample locations and therefore improve e.g. OCT axial scan rate by a factor equal to the number of beams used simultaneously which can go up to very high frequency ranges e.g. MHz. The speed improvement can be especially beneficial in OCT-imaging because it has to potential to open up for new applications.

In a parallel-OCT system, increasing the number of the sample arms may thus contribute to increased imaging speed and/or larger field of view. However, as the number of sample arm increases, the optical power at each location gets smaller, which can limit the overall system signal to noise ratio SNR and degrade the imaging performance. Using a high power light source can be expensive or not available at all. As an improved solution, the present methods and systems of broadband light switch enable a sequentially- switched parallel OCT approach. Here, instead of using all sample arms simultaneously (which reduces the output power at each sample arm), a certain number of them, i.e. bundle, can be used at a time. In one embodiment, by using an optical switch between each bundle, all sample arms can be sequentially scanned. The optical switch can be based on e.g. on thermal or electro-optic effect. As a further benefit, the latter option would not degrade the imaging speed as the switching speed can be in the GHz range. With this method, degradation of the SNR of the OCT system can be alleviated. An example of such a sequentially- switched parallel OCT sample arm is shown FIG 3B. Only one switch and two bundles are shown used to prove the working principle of the idea. By using more switches and bundles, it is possible to obtain a much larger field of view. By integrating the system on a chip a miniaturized OCT system becomes available, e.g. without the need for a mechanical scanner.

FIG 4A illustrates a graph of a splitting ratio S in decibel [dB] of two coupler designs indicating relative intensity of light OA/IA; OB/IA in respective arms as a function of wavelength l. In the graph, a coupler design without tangential offset (Y1=0) is compared to the same design but with tangential offset (here Yl=140pm).

As will be appreciated, the coupler design with tangential offset is able to achieve a relatively flat splitting ratio over a broad band of wavelengths. In a preferred embodiment, the broadband coupler as described herein has a fifty-fifty percent splitting ratio of outgoing light intensity OA,OB coming out of each arm (waveguide) of the coupler with respect to the light IA coming into the coupler (e.g. in one of the arms). This is also referred to as a -3dB coupler since this is very close to a factor half. Most preferably, the broadband coupler as described herein provides close to fifty-fifty percent splitting ratio, e.g. within one percent, preferably within one per mille, or closer, over a wavelength bandwidth “A” of at least hundred nanometer, preferably at least two hundred nanometer, or even more than two hundred fifty nanometer. FIG 4B illustrates a graph of relative transmission T in decibel, dB or extinction ratio of corresponding switch designs indicating relative intensity of light Il/IA; I2/IA in respective arms as a function of wavelength l. In the graph, a switch design without tangential offset (Y1=0) is compared to the same design but with tangential offset (here Yl=140pm).

As will be appreciated, the switch design with tangential offset is able to achieve a very high extinction ratio (e.g. less -30dB, i.e. <0.1%) over a broad band of wavelengths “A”, e.g. more than 250 pm.

FIG 5A illustrates a similar graph as FIG 4B, but for a range of different tangential offsets Y1 = (100pm — 180pm). FIG 5B illustrates a plot of the bandwidth A where the extinction ratio is better than -30dB, i.e. less than one per mille of the light is left in the non-selected waveguide. As will be appreciated, the bandwidth of a design can be optimized by varying the tangential offsets Yl. For example, here an optimum bandwidth (A) of about 280 nm is found around a tangential offset Yl=130pm.

In a preferred embodiment, the tangential offset Yl is selected to provide the optical switch 100 with an extinction ratio of better than (at least) 20 dB, preferably better than 30 dB over a wavelength range (A) of more than two hundred nanometer, preferably more than two hundred fifty nanometer. For example, less than one percent, preferably less than one per mille (0.1%) of the light IA entering the switch is output in a non-selected output of the switch. In other words more than ninety-nine percent, or even >99.9%, of the light is switched to go into the selected output waveguide. For example, in the specific embodiment shown, selecting the tangential offset Yl in a range between about 120 and 140 pm provides an extinction ratio better than 30 dB over a wavelength range of more than 250nm

Some aspects can be embodied as a method of manufacturing an optical device with an optical coupler 50 or optical switch 100 as described herein. For example, some embodiments comprise determining an initial design including the set of waveguides 10,20 each comprising the respective curved starting sections 11,21, respective straight middle sections 12,22, and respective curved ending sections 13,23. Other or further embodiments comprise varying a tangential offset Y1 at which the straight middle sections 12,22 are disposed with respect to each other along a direction Y of their respective length L2,L6 to optimize a bandwidth of the light coupled or switched by the optical device. Other or further embodiments comprise manufacturing the optical device with the tangential offset Y1 corresponding to the optimized bandwidth.

FIGs 6A and 6B illustrate propagation through the waveguides of a switch without and with tangential offset, respectively. The left hand side of each figure illustrates a map of the relative intensity I/IA for various positions in the waveguides. The right hand side of each figure illustrates the relative intensity I/IA for each arm as a function of position Y along a length of the respective arm. For example, the switching performance of the waveguides may be inferred from the transfer of relative intensity between the arms. Where at the start 100% of the light is located in one of the arms, at the end of the switch the most of light is located in the other arm. While for the switch without tangential offset, there remains a small part of the light in the original arm, for the switch with tangential offset virtually all light is transferred. As will be appreciated, the transfer of the light is also much more smoothly in the latter case of the tangential offset, e.g. without the fluctuations shown in the case without the offset.

In interpreting the appended claims, it should be understood that the word "comprising" does not exclude the presence of other elements or acts than those listed in a given claim; the word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements; any reference signs in the claims do not limit their scope; several "means" may be represented by the same or different item(s) or implemented structure or function; any of the disclosed devices or portions thereof may be combined together or separated into further portions unless specifically stated otherwise. Where one claim refers to another claim, this may indicate synergetic advantage achieved by the combination of their respective features. But the mere fact that certain measures are recited in mutually different claims does not indicate that a combination of these measures cannot also be used to advantage. The present embodiments may thus include all working combinations of the claims wherein each claim can in principle refer to any preceding claim unless clearly excluded by context.

Claims

1. A broadband optical coupler (50) comprising a set of waveguides

(10.20), each waveguide comprising a respective curved starting section

(11.21), at least one of which curved starting sections is configured to receive light (IA) at its respective start, wherein the curved starting sections (11,21) follow a respective curved path (Sl,S4) towards each other starting at a first distance (Dl) apart and getting closer together along their respective length (L1,L7), wherein each curved starting section (11,21) transitions at its respective end into a respective straight middle section (12,22), wherein the straight middle sections (12,22) are parallel to each other over at least a subsection (L26) of their respective length (L2,L6), wherein the parallel subsections (L26) are at a second distance (D2) apart, wherein the second distance (D2) is smaller than the first distance (Dl) to couple (C) at least part (OB) of the light (IA) between the waveguides (10,20), wherein widths (W2^W3; W4^W3) of the middle sections (12,22) taper inward and outward (Ti,To), respectively, in opposite directions between the waveguides, wherein each straight middle section (12,22) transitions at its respective end into a respective curved ending section (13,23) each configured to output a respective part (OA,OB) of the light (IA), wherein the ending sections follow a respective curved path (S2,S3) away from each other getting further apart along their respective length (L3,L5) to end at a third distance (D3) larger than the second distance (D2), characterized in that the straight middle sections (12,22) are disposed at a tangential offset (Yl) with respect to each other along a direction (Y) of their respective length (L2,L6).

2. The optical coupler (50) according to claim 1, wherein the tangential offset (Yl) is a fraction (F) of the length (L2,L6) of the straight middle sections (12,22), wherein the fraction (F) is between ten and fifty percent.

3. The optical coupler (50) according to any of the preceding claims, wherein a distance (D2) between the middle sections (12,22) is constant over the respective length of the parallel subsections (L26), wherein the constant distance (D2) is between hundred nanometers and five micrometer, wherein the respective length of the parallel subsections (L26) is at least hundred micrometers.

4. The optical coupler (50) according to any of the preceding claims, wherein the middle sections (12,22) start with a first width (W2), and a different second width (W4), respectively, while ending with a third width (W3), which is the same for both middle sections (12,22), wherein the third width (W3) is half-way between the first width (W2) and the second width (W4).

5. The optical coupler (50) according to any of the preceding claims, wherein the widths (W2,W3,W4) of the middle sections (12,22) decrease and increase, respectively, at a constant rate of change along their respective length (L2,L6), wherein opposing side edges (12i,12j) of a respective middle section (12,22) are at a constant tapering angle of between ten and thousand microradians.

6. The optical coupler (50) according to any of the preceding claims, wherein the curved starting sections (11,21) and curved ending sections (13,23) are shaped as S-bends (S1-S4) to bring the waveguides together and apart, wherein a distance (Dl) between respective starts of the curved starting sections (11,21) an distance (D3) between respective ends of the curved ending sections (13,23) are both larger than the distance (D2) between the middle sections (12,22) by at least a factor two.

7. The optical coupler (50) according to any of the preceding claims, wherein the respective curved starting sections (11,21) of the different waveguides (10,20) have different lengths (L1<L7), wherein the respective straight middle sections (12,22) of the different waveguides (10,20) have the same length (L2=L6), wherein the respective curved ending sections (13,23) of the different waveguides (10,20) have different lengths (L3>L5), wherein a difference in length (L7-L1) of the curved starting sections (11,21) is the same as a difference in length (L3-L5) of the curved ending sections (13,23).

8. An optical switch (100) comprising a first optical coupler (50a) according to any of the preceding claims connected to a second optical coupler (50b) according to any of the preceding claims wherein respective curved sections (13a, 23a) of the first optical coupler (50a) are connected to respective curved sections (23b, 13b) of the second optical coupler (50b).

9. The optical switch (100) according to claim 8, wherein the second optical coupler (50b) is shaped as a rotated copy of the first optical coupler (50a).

10. The optical switch (100) according to claim 8 or 9, wherein the optical switch (100) comprises a switching mechanism (SW) to switch light output (OA,OB) from the optical switch (100), wherein the switching mechanism (SW) is arranged to affect wave propagation in at least one of the waveguides of the optical coupler (50).

11. An optical device (1000) comprising at least one optical switch according to any of the preceding claims; light guiding means to direct light from a light source via the optical switch; and a controller configured to control the selectively control a target of the light by controlling the optical switch.

12. An optical coherence tomography system comprising at least one light source for generating broadband light; and the optical device (1000) of claim 11 for sequentially directing the light of the same at least one light source to illuminate respective different locations of a sample.

13. The system according to claim 12, comprising one or more optical couplers (50) according to any of the preceding claims, each configured to split the light into different arms for illuminating in parallel the same location, or different locations, of the sample, wherein the different arms for illuminating the sample in parallel provide different lengths (AL) for respective light paths.

14. A method of manufacturing an optical device with the optical coupler (50) or optical switch (100) according to any of the preceding claims, the method comprising determining an initial design including the set of waveguides

(10.20) respectively comprising the curved starting sections

(11.21), straight middle sections (12,22), and curved ending sections (13,23); varying a tangential offset (Yl) at which the straight middle sections (12,22) are disposed with respect to each other along a direction (Y) of their respective length (L2,L6) to optimize a bandwidth of the light coupled or switched by the optical device; and manufacturing the optical device with the tangential offset (Yl) corresponding to the optimized bandwidth.

15. The method according to claim 14, wherein the tangential offset (Yl) is optimized to provide the optical switch (100) with an extinction ratio of better than 20 dB over a wavelength range (L) of more than two hundred nanometer.