WO2003007034A1

WO2003007034A1 - An integrated optical device

Info

Publication number: WO2003007034A1
Application number: PCT/GB2002/003150
Authority: WO
Inventors: Ian Edward Day; Ralf-Dieter Pechstedt; Daniel Kitcher; Andrew Alan House
Original assignee: Bookham Technology PLC
Current assignee: Lumentum Technology UK Ltd
Priority date: 2001-07-10
Filing date: 2002-07-09
Publication date: 2003-01-23
Anticipated expiration: 2004-01-10
Also published as: GB2377503A; WO2003007034A8; GB0116859D0

Abstract

An integrated optical device comprising at least one optical waveguide (2, 3) formed on a substrate, the waveguide (2, 3) being of elongate form with an optical axis extending along its length, at least one doped region (1) being provided in the substrate adjacent at least one side of the waveguide (2, 3), an edge of the doped region (1) adjacent the waveguide (2, 3) having one or more projections (1E, 1F) shaped to intercept stray light travelling substantially parallel to the waveguide (2, 3) in a region (4) of the optical substrate between the waveguide (2, 3) and the remainder of the doped region (1) and to absorb said light and/or deflect said light away from the waveguide (2, 3).

Description

AN INTEGRATED OPTICAL DEVICE

FIELD OF THE INVENTION

This invention relates to an integrated optical device comprising at least one waveguide or other optical device formed on a substrate and, in particular, to an arrangement for reducing problems caused by stray light within the substrate.

BACKGROUND ART

A common problem with waveguides of an integrated optical device is the presence of stray light in the substrate on which the waveguides are formed. Although most of the light is guided by the waveguides, some light inevitably escapes to the substrate, e.g. where light is input into an end of a waveguide or where light leaves the end of a waveguide or due to leakage of light from the waveguide, e.g. around bends in the waveguide or at junctions between waveguides. Such stray light can cause cross-talk between waveguides or may reach light detectors provided on the device. In either case, it reduces the signal/noise ratio for the device.

It is known to use doped areas to absorb stray light as described in WO-A- 99/28772. Doping is preferred in some case to trenches for optically isolating two areas, as doped areas absorb the stray light rather than merely redirecting it. However, in many cases, it is desired to minimise the area of doped regions provided on a device as they can give rise to heating of the chip during processing, which, in turn, can lead to distortion of the chip. It is also desired to minimise the area of doped regions positioned close to devices such as waveguides as they attenuate a portion of the optical signal extending beyond the confines of the waveguide due to the free carrier dispersion effect, particularly as, in many cases, the optical mode may extend 2 microns or more beyond the edges of the waveguide.

The present invention seeks to reduce the problem caused by stray light by improving the effectiveness of doped regions used to absorb the light whilst minimising the above-mentioned problems associated with the use of doped regions.

SUMMARY OF THE INVENTION

According to a first aspect of the invention, there is provided an integrated optical device comprising at least one optical waveguide formed on a substrate, the waveguide being of elongate form with an optical axis extending along its length, at least one doped region being provided in the substrate adjacent at least one side of the waveguide, an edge of the doped region adjacent the waveguide having one or more projections shaped to intercept stray light travelling substantially parallel to the waveguide in a region of the optical substrate between the waveguide and the remainder of the doped region to absorb said light and/or deflect said light away from the waveguide.

According to a second aspect of the invention, there is provided an integrated optical device formed on an optically conductive substrate having light absorbing means in one or more selected areas of the substrate, the light absorbing means comprising one or more doped areas where the doping concentration is greater than that of areas of the substrate forming the optical device so as to absorb stray light in the substrate, the doped area having diffuse boundaries such that the concentration of dopant falls from that of the doped area to that of the adjacent substrate over a distance greater than 0.1 microns and preferably over a distance of at least 0.5 microns and most preferably over a distance of at least 1.0 microns. Preferred and optional features of the invention will be apparent from the following description and from the subsidiary claims of the specification.

BRIEF DESCRIPTION OF THE DRAWINGS

The invention will now be further described, merely by way of example, with reference to the accompanying drawings, in which:

Figure 1 is a plan view of a doped region between a pair of waveguides, the doped region having side projections at one end thereof according to a first embodiment of the invention;

Figure 2 is a plan view of a doped region between a pair of waveguides, the doped region having a plurality of side projections along each side thereof according to a second embodiment of the invention;

Figure 3 is a plan view of doped regions either side of a waveguide, the doped regions having a plurality of saw-tooth shaped projections along the sides thereof according to a third embodiment of the invention;

Figure 4 is a plan view of doped regions either side of a waveguide, the doped regions having a single saw-tooth shaped projection on the sides thereof according to a fourth embodiment of the invention;

Figure 5 is a plan view of doped regions between waveguides within a waveguide array, each doped region having saw-tooth shaped projections on each side thereof at one end of the region so the end of the region flares outwardly according to a fifth embodiment of the invention; Figure 6 is a plan view of doped regions between pairs of waveguides, each doped region having side projections at one end thereof according to a sixth embodiment of the invention;

Figure 7 is a plan view of a doped region between a pairs of waveguide, the doped region having side projections at one end thereof according to a seventh embodiment of the invention;

Figures 8A to 8C are plan views of other shaped doped regions according to further embodiments of the invention; and

Figure 9 is a plan view of yet another embodiment of the invention.

DETAILED DESCRIPTION OF THE EMBODIMENTS

Figure 1 shows a schematic plan view of a doped region 1 between a pair of substantially parallel waveguides 2, 3. Stray light (as indicated by arrows S) travelling substantially parallel to the waveguides 2, 3 (typically within 10 degrees of the optical axes of the waveguides) in the substrate between the waveguides 2, 3 is intercepted by the doped region 1 and absorbed therein. The sides 1A and 1B of the doped region are preferably spaced from the waveguides 2, 3, respectively, by a distance of at least 8 microns so as not to adversely affect the light transmission within the waveguides. However, this leaves a channel 4 of this width on each side of the doped region 1 between the doped region and the adjacent waveguide through which stray light may still be transmitted. If the sides 1A, 1 B of the doped region are flat, it is found that they tend to guide the stray light through this channel 4 or reflect light incident thereon towards the waveguide. This is because the stray light S is incident on the sides 1A, 1B at a glancing angle (often less than 1 degree) and as the dopant in the doped region 1 tends to suppress the refractive index thereof, the light undergoes total internal reflection at the faces 1A and 1B (a refractive index step of 0.001 gives rise to a critical angle for total internal reflection of 1.6 degrees). To reduce this problem, the sides of the doped region 1 are provided with projections 1C, 1D at one end thereof (the end facing the source of stray light S) which project from the sides of the doped region towards the adjacent waveguides 2, 3. The width of the entrance to each of the channels 4 referred to above is thus substantially reduced. The projections 1C, 1D may terminate close to the waveguides 2, 3, e.g. at a distance in the range 3 to 5 microns. As the close proximity of the doped region to the waveguides at these points is localised and does not extend for any significant length along the waveguides 2, 3, the deleterious affect this has on the transmission of light signals in the waveguides is small.

As shown, the projections 1C, 1 D are in the form of arms which diverge from each other at distances further from the doped region. The inwardly facing sides 1 E, 1 F of the arms thus serve to direct stray light S incident thereon into other parts of the doped region 1.

Preferably, the distal ends of the projections 1C, 1D are rounded as it is found that if they have pointed ends facing in the direction from which the stray light is coming, light can be reflected from the flat faces either side of the point whereas rounded points alter the angle of incidence such that more light is absorbed rather than reflected.

Figure 1 shows projections 1C, 1D at one end of the doped region 1. In some cases this may be sufficient but in other cases, further projections 1G, 1H, 11, U may be provided at intervals along the sides 1A, 1B of the doped region 1 , as shown in Figure 2. These serve to intercept any light that does enter the channels 4 past the projections 1C, 1D and either absorb the light or direct it towards the bulk of the doped region. To this end, the projections 1G - 1 J are also in the form of arms which extend at an angle from the sides 1A and 1B of the doped region inclined towards the direction from which stray light may be received.

Figure 3 shows a plan view of doped areas 11 , 12 on each side of a waveguide 13. In this case, the sides of the doped areas adjacent the waveguide 13 are saw-tooth shaped, with each saw tooth shape 11 A, 12A pointing in the direction from which stray light S may be received. The saw tooth shape is useful as the point of the tooth is located close to the waveguide and so extends into the channel 14 in the substrate between the waveguide 13 and the bulk of the doped region 11 , 12. Stray light intercepted by this point is incident on a surface which is substantially perpendicular to its direction of travel. The stray light thus tends to pass into the doped region (rather than being reflected from its surface) and the remainder of the tooth shape serves to absorb the light and/or direct it into the bulk of the doped region for absorption.

As with the projections described above, the close proximity of the points of the saw tooth shapes to the waveguide occurs only in localised regions and the bulk of the doped region is spaced from the waveguide by a greater distance. The deleterious effects caused by close proximity of the doped regions to the waveguide are thus minimised.

In a further embodiment, rather than providing a series of saw tooth shapes along the sides of the doped regions 11 , 12, a single pair of saw tooth shapes 11B, 11C may be provided on opposite sides of the waveguide 13 as shown in Figure 4. These provide a localised narrowing of the channels 14 to prevent transmission of stray light along these channels. As only a single pair of tooth shapes are used, they can be formed so that their points extend into even closer proximity, e.g. 2 to 4 microns, to the waveguide 13 than would be desirable if a series of such shapes is used as in the arrangement shown in Figure 3. Figure 5 shows an array 20 of waveguides such as may be used to receive light from the output of an arrayed waveguide grating (AWG), not shown. In this Figure, each waveguide 21 is indicated by a single thick line (rather than two thin lines as in previous Figures). The waveguides 21 diverge from each other and, once the spacing between adjacent waveguides is large enough, it is desirable to form doped regions 22 between the waveguides 21 to absorb stray light. To increase the effectiveness of these doped regions 22 without suffering from the consequences of forming them very close to the waveguides 22 throughout their length, the ends of the doped regions 23 are broadened, as shown in the enlarged view of one such doped region 22 in the Figure. This broadened end 22A is, in effect, provided by a single saw tooth shape on each side of the doped region 22 at the end thereof facing the source of stray light S.

Figure 6 shows another embodiment somewhat similar to that of Figure 1 in which doped regions 30 between waveguides 31 , 32 (and between waveguides 32, 33) are provided with projecting arms 30A, 30B on each side thereof at the end facing the source of stray light S. The arms diverge from each other so as to form a forked or bifurcated arrangement, the arms terminating in close proximity to the adjacent waveguide and so reducing the width of the entrance to channels 34 between the doped regions 30 and the adjacent waveguides. The arms 30A, 30B are inclined to the optical axes of the adjacent waveguides by a small angle, e.g. less than 5 degrees and preferably less than 1 degree so the light incident on the inner faces 30C, 30D of the arms is reflected from these surfaces into the bulk of the doped region for absorption. Light received between the arms 30A, 30B thus becomes trapped and is eventually absorbed by the doped region. In a typical arrangement, the arms 30A, 30B have a length of around 250 microns (measured parallel to the waveguides 31 ) and a width of around 4 microns. Figure 7 shows another bifurcated doped region 40 provided between waveguides 41, 42, the end of the doped region 40 facing the source of stray light S having projections 40A, 40B on each side thereof which together form a curved or crescent shape for intercepting the stray light. The curved shape helps ensure that the most of the stray light is incident on a surface which is substantially perpendicular to its direction of travel (so as to maximise transmission of the light into the doped region 40 for absorption) or, if it is incident on the curved surface at a smaller angle, it is deflected towards another portion of the curved shape until it is absorbed; the thin ends of the curved area thus serving to reflect the light towards the thicker parts thereof.

Figures 8A - 8C are plan views of edges of doped regions adjacent waveguides showing projections of other shapes which may be used. Figure 8A shows a doped region 50 adjacent a waveguide 51 , the edge 50A of the doped region having a substantially sinusoidal shape (although other curved wave shapes may also be used). Figure 8B shows a doped region 60 adjacent a waveguide 61, the edge 60A of the doped region having projections in the form of parallelograms inclined towards the source of stray light S. Figure 8C shows a doped region 70 adjacent a waveguide 71 , the edge 70A of the doped region have triangular projections inclined towards the source of tray light S.

Projections of a wide variety of other shapes may be used but the preferred shapes are such as to minimise glancing angle reflections of stray light incident thereon back towards the adjacent waveguide. It is also preferred if the surfaces of the projections which intercept the stray light are shaped and/or angled so as to maximise transmission of the stray light into the doped region and/or deflect it towards the bulk of the doped region, i.e. away from the adjacent waveguide. The critical angle (for total internal reflection) is typically around 2 degrees so the projections are preferably arranged to increase the angle of incidence to above 2 degrees for substantially all the stray light received. It is also preferred if the projections are shaped so that light intercepted thereby and transmitted through the surface thereof travels through a sufficient length (typically 10 or 20 microns or more) of the doped region so as to be substantially absorbed therein before reaching another external surface thereof, particularly where that surface may deflect the light back towards the adjacent waveguide. Preferably, the surface of the projections facing away from the source of stray light S are positioned and/or angled so that if some light does pass through the projection without being absorbed, the light exiting the said face is deflected towards another portion of the doped region so as to be absorbed thereby rather than being deflected back towards the adjacent waveguide. For these reasons the dimensions of the projections in a direction substantially parallel to the adjacent waveguide is preferably at least 10 microns and most preferably, at least 20 microns.

It will be appreciated that the above arrangements involve providing some form of discontinuity in the edges of the doped regions adjacent the waveguide to satisfy the requirements discussed above. The form or nature of the doped region between the edges thereof is of less importance. Whilst in most cases the remainder of the doped region, or the bulk thereof, will comprise a doped region extending either to the edge of the chip or to a further edge thereof provided with discontinuities (e.g. when the doped region is provided between 2 waveguides), the remainder of the doped region may take other forms. It may not necessarily comprise a uniformly doped area so long as it provides sufficient absorption of stray light received by the doped region or otherwise prevents the stray light being directed back towards the waveguides. Other forms of absorption means or light traps may, for instance, be provided adjacent the doped regions for receiving stray light directed by the doped regions away from the adjacent waveguide.

The use of doped regions to absorb the stray light is also advantageous as they are integrated with the device and there is only a small refractive index difference between the doped and undoped regions so preventing back reflection of the stray light for a wide range of angles of incidence on the doped region. The boundary between the doped and un-doped region may also be diffuse so further inhibiting back reflections therefrom. The doped regions both suppress back-reflection of stray light and suppress propagation of stray light in the channel(s) between the doped region and the adjacent waveguide(s).

Figure 9 shows another form of device having doped regions adjacent a diode formed across a waveguide. The type of diode shown in Figure 9 is similar to that described in GB0019771.5 the disclosure of which is incorporated herein. The device comprises a waveguide 80, with tapered portions 80A at each end thereof, with doped regions (not shown) of opposite type formed on each side thereof to form one or more p-i-n diodes across the waveguide 80. Preferably an alternate series of p-i-n diodes are formed across the waveguide as described further in GB0019771.5. These can then be electrically connected in series by electrical conductor 81 from an anode 81 A at one end of the series to a cathode 81 B at the other end of the series, as shown in figure 9.

Figure 9 shows doped regions 82 formed adjacent the series of diodes, the doped regions comprising longitudinal portions 82A extending substantially parallel to the waveguide 80 and arm portions 82B extending substantially perpendicular from the longitudinal portions 82A towards (or away from) the waveguide 80. As in the embodiments described above, the arm portions 82B extend close to the waveguide 80, typically terminating about 4.5 -8 microns from the waveguide 80 and, as shown, the arm portions 82B are preferably located between portions of the electrical conductors 81. Such doped regions may be provided on both sides of the waveguide (as shown) or on only one side thereof. In a particularly preferred embodiment, the doped regions 82A and 82B comprise a p-doped region extending lengthwise along the doped region with an n-doped region on each side thereof extending lengthwise along the doped region. The doped regions thus comprise an npn structure across their width. In other arrangements, this structure may instead be by n-i-p-i-n or pnp or p-i- n-i-p. This type of structure has the advantage that as well as providing optical isolation (by absorbing stray light incident thereon), it also provides electrical isolation as it provides two pn (or np) junctions back-to-back which prevent the flow of electrical current across the doped region.

The embodiments described in relation to Figures 1-8 may also use doped areas having this type of structure.

Further details of this type of doped region are given in GB0023133.2 and a US application filed 8 May 2001 claiming priority therefrom, the disclosures of which are incorporated herein.

Figures 3 and 8 show a series of projections along the sides of the doped regions. Such projections may, however, be spaced apart from each other in the manner shown in Figure 2. The spacing between projections (measured between corresponding points on each projection) is preferably at least 10 or 20 microns but may typically be greater, e.g. 100 to 200 microns. The projections themselves preferably extend at least 1 to 20 microns, and preferably at least 5 microns, away from the sides of the doped regions (measured as the perpendicular distance between the distal end of the projection and the position where it joins the bulk of the doped region). In a typical case, the channel between the side of the doped region and the waveguide has a width of at least 8 microns. The distal end of a projection extending 5 microns into this gap is thus spaced from the side of the waveguide by a distance of at least 3 microns. Preferably, the spacing between the distal end of the projection and the side of the adjacent waveguide is at least 4.5 microns and no greater than 8 microns. This distance is determined to some extent to that which can reliably be fabricated at this level of spacing as there is an exponential increase in attenuation of the optical mode in the waveguide the closer the doped area is to the waveguide.

The arrangements described above may be used with a wide variety of substrates and waveguide types but are particularly suited to use with optical devices fabricated on silicon, e.g. on silicon-on-insulator chips comprising a silicon layer separated from a supporting substrate (usually also of silicon) by an optical confinement layer (usually a layer of silicon dioxide). In such a device, the waveguides are typically rib waveguides and the doped regions preferably extend through the upper silicon layer to the confinement layer. Figures 3 and 4 show the rib 13 of a rib waveguide and trenches 13A, 13B on either side thereof defining the rib 13. As shown, the doped regions 11 and 12 extend into the trenches 13A, 13B to a position close to, but spaced from, the rib 13.

The doped regions described above may be provided adjacent both straight waveguides and curved waveguides.

The doped areas preferably comprise phosphorus or boron as the dopant, e.g. to a level in the range of 10¹⁷ - 10¹⁹ and typically around 10¹⁸ atoms/cm³. However, other dopants, e.g. arsenic, may be used.

Doping may be carried out, for example, by ion implantation. This produces a relatively narrow boundary between doped and undoped areas, with the level of dopant falling from the figures given above to the background level in the surrounding intrinsic material (typically in the range of 10¹³ - 10¹⁵ atoms/cm³) over a distance of about 0.1 microns. However as mentioned above, an advantage of used doped areas to absorb the stray light is that they can have diffuse boundaries to reduce further the likelihood of the stray light being back reflected from the boundary rather than entering the doped area and being absorbed. A diffuse boundary can be formed by heating the area, e.g. to 900- 1200 degrees C for about an hour, whereby the concentration of dopant at the boundaries of the doped area falls off, following an approximately Gaussian curve, over a distance of 1 micron or more.

This feature can also be applied to other forms of doped region for absorbing stray light, i.e. such doped regions should preferably have diffuse boundaries with the dopant level falling by a factor of 100 or more over a distance greater than 0.1 microns, preferably over a instance of 0.5 microns or more and most preferably over a distance of 1.0 microns or more. For example, the concentration may fall from the range 10¹⁷-10¹⁹ atoms/cm³ to the range of 10¹³-10¹⁵ atoms/cm³ over such distances.

Claims

1. An integrated optical device comprising at least one optical waveguide formed on a substrate, the waveguide being of elongate form with an optical axis extending along its length, at least one doped region being provided in the substrate adjacent at least one side of the waveguide, an edge of the doped region adjacent the waveguide having one or more projections shaped to intercept stray light travelling substantially parallel to the waveguide in a region of the optical substrate between the waveguide and the remainder of the doped region and to absorb said light and/or deflect said light away from the waveguide.

2. An integrated optical device as claimed in claim 1 in which a series of said projections are provided along the side of the doped region adjacent the waveguide.

3. An integrated optical device as claimed in claim 1 or 2 in which the or each projection comprises a first surface which is positioned and/or angled so as to absorb light intercepted thereby and/or deflect said light away from the waveguide.

4. An integrated optical device as claimed in claim 3 in which the first surface is substantially perpendicular to the optical axis of the waveguide or angled towards the source of stray light.

5. An integrated optical device as claimed in claim 3 or 4 in which the or each projection has a second surface facing away from the source of stray light, the second surface being angled such as to deflect light which has been transmitted through the projection from the first surface to the second surface away from the waveguide.

6. An integrated optical device as claimed in any preceding claim in which the or each projection has a length (measured in the direction parallel to the optical axis of the waveguide) of at least 10 microns, and preferably at least 20 microns.

7. An integrated optical device as claimed in any preceding claim in which the or each projection projects from the remainder of the doped region towards the waveguide by a distance of at least 1 micron, and preferably at least 5 microns (measured in a direction perpendicular to the optical axis of the waveguide).

8. An integrated optical device as claimed in any preceding claim in which the region of the substrate between the doped region (ignoring the projections) and the waveguide has a width of at least 8 microns (measured in a direction perpendicular to the optical axis of the waveguide).

9. An integrated optical device as claimed in any preceding claim in which the spacing between the distal end of the or each projection and the waveguide is at least 3 microns and preferably at least 4.5 microns (measured in a direction perpendicular to the optical axis of the waveguide).

10. An integrated optical device as claimed in any preceding claim in which the spacing between the distal end of the or each projection and the waveguide is 8 microns or less (measured in a direction perpendicular to the optical axis of the waveguide).

11. An integrated optical device as claimed in any preceding claim in which the doped region is provided between two waveguides the optical axes of which are substantially parallel to each other, one or more projections being provided on each side of the doped regions adjacent the respective waveguide.

12. An integrated optical device as claimed in claim 11 in which a projection is provided on each side of the doped region at the end of the doped region towards the source of said stray light.

13. An integrated optical device as claimed in claim 12 in which the projections at the end of the doped region have the effect of broadening said end of the doped region, an end surface of the doped region facing the source of stray light being substantially perpendicular to the optical axes of the waveguides.

14. An integrated optical device as claimed in claim 12 in which the projections at said end of the doped region diverge from each other and together form a bifurcated arrangement such that stray light received between the projections is trapped therebetween and directed into the doped region.

15. An integrated optical device as claimed in any of claims 11 to 15 in which the two waveguides are part of an array of waveguides, a doped region being provided between each adjacent pair of waveguides in the array.

16. An integrated optical device as claimed in claim 15 in which the array of waveguides is arranged to receive the output of an arrayed waveguide grating.

17. An integrated optical device as claimed in any preceding claim in which the or each doped region comprise phosphorus or boron as the dopant.

18. An integrated optical device as claimed in claim 17 in which the dopant is provided at a level in the range 10¹⁷-10¹⁹ atoms/cm³.

19. An integrated optical device as claimed in any preceding claim in which the doped region has diffuse boundaries with the level of dopant falling from that of the bulk of the doped area to that of the substrate adjacent the doped region over a distance of at least 1 micron.

20. An integrated optical device as claimed in any preceeding claim in which the doped region comprises an elongate region of a first type of dopant with elongate regions of a second type of dopant on each side thereof, the first and second types of dopant being of opposite polarity.

21. An integrated optical device as claimed in any preceding claim in which the or each waveguide is formed in a substrate comprising a silicon layer.

22. An integrated optical device as claimed in claim 21 in which the silicon layer is separated from a supporting layer by an optical confinement layer.

23. An integrated optical device as claimed in claim 22 in which the optical confinement layer comprises silicon dioxide.

24. An integrated optical device as claimed in claim 22 or 23 in which the or each doped region extends through the silicon layer to the optical confinement layer.

25. An integrated optical device as claimed in any preceding claim in which the or each waveguide comprises a rib waveguide.

26. An integrated optical device formed on an optically conductive substrate having light absorbing means in one or more selected areas of the substrate, the light absorbing means comprising one or more doped areas where the doping concentration is greater than that of areas of the substrate forming the optical device so as to absorb stray light in the substrate, the doped area having diffuse boundaries such that the concentration of dopant falls from that of the doped area to that of the adjacent substrate over a distance greater than 0.1 microns and preferably over a distance of at least 0.5 microns and most preferably over a distance of at least 1.0 microns.

27. An integrated optical device as claimed in claim 25 in which the doped region comprises dopant at a concentration in the range of 10¹⁷-10¹⁹ atoms/cm³ and the adjacent substrate comprises dopant at a concentration of 10¹⁵ atoms/cm³ or less.