WO2001014919A1

WO2001014919A1 - Alignment device

Info

Publication number: WO2001014919A1
Application number: PCT/GB2000/003062
Authority: WO
Inventors: Arbinder Singh Pabla; Colin William Ford
Original assignee: British Telecommunications Public Limited Company
Priority date: 1999-08-20
Filing date: 2000-08-08
Publication date: 2001-03-01
Also published as: GB9919847D0

Abstract

A device for aligning an optical fibre with respect to an optical component including an optical waveguide, such as a planar silica waveguide. The device has polymeric portions which cooperate. The portions (6, 7) include reference surfaces (13, 19) which determine the position of the optical fibre with respect to the waveguide. The portions are moulded from a polymer material from a mould made by exposing a photoresist to radiation. The parts are made to such a degree of accuracy that passive alignment of optical fibres to a waveguide is achieved.

Description

ALIGNMENT DEVICE

The present invention relates to an alignment device, typically for the alignment of an optical fibre with respect to an optical component. In particular, but not exclusively, the invention relates to the alignment of an optical fibre with respect to a planar optical component, such as a planar waveguide.

Increasingly, there is a trend towards so-called planar-waveguide devices in optical communications, and the waveguides of these devices typically need to be aligned with respective optical fibres. An example of such a planar-waveguide device is an arrayed-waveguide grating, which is used to separate the individual wavelengths of a wavelength division multiplexed system, so that the separated wavelengths can each be coupled into an appropriate one of several optical fibres coupled to the device.

Planar waveguide devices typically include a substrate, such as silicon or quartz, one of more waveguide regions (equivalent to the core of an optical fibre), and, surrounding the waveguide regions, one or more cladding regions or layers. In addition, one or more buffer layers may lie between the substrate and the waveguide regions. To achieve waveguiding, it is useful for the waveguide regions to be made of a material or materials having a higher refractive index or indices than the refractive index (or indices) of the cladding regιon(s) . While plastic materials can be used, currently silica and doped silica are preferred .

The waveguides are typically of square or rectangular cross-section with typical side lengths of 5 to 8 micrometres Typically the cladding and buffer layers each have a thickness approximately three times that of the waveguides.

In many planar waveguide devices, there are multiple waveguides and these need to be aligned with and coupled to multiple optical fibres In such cases, it is common for the optical fibres to be formed into an array by fixing them to a carrier, the carrier than being aligned with and secured to the waveguide device to effect alignment and coupling between the waveguides and the fibre. One common carrier type for forming the fibre array is what is known as micromachined silicon. Selective etching is used to form V-grooves, defined by the crystallographic planes in a single- crystal silicon substrate, the dimensions and the spacing of the V-grooves being chosen to suit the fibres and device which are to be aligned .

Even with the use of such fibre arrays, it is still typically necessary to use active alignment.

The importance of accurate alignment of an optical fibre with other optical components is well known. In aligning an optical fibre with an optical component, not only must the fibre be aligned angularly and in three co-ordinates for optimum coupling, but this alignment must also be fixed securely and must not be altered during fixation or subsequent processing. For a single-mode fibre, for example, final alignment must generally be held to within plus or minus one-half micrometer in dimensions radial to the fibre and to within one to two micrometers axially.

One prior art technique is, for example, disclosed in "Glass integrated optical devices on silicon for optical communications, M.F. Grant, Critical Review of Science and Technology Volume CR53, pp55-80 1 994. According to this conventional technique, a separate substrate to carry an array of fibres is used. An array of fibres is bonded into an array of well defined grooves, often V-grooves formed in silicon. The end face of the integrated optical chip and the fibre array are both polished . The silicon substrate on which the array of fibres is placed is then moved into contact with the optical component. Once in contact, the final alignment is achieved using active fibre alignment techniques (i.e. techniques requiring active monitoring of laser light launched into the fibre during the alignment process) to align the optical fibre with the planar component. Once the fibre is accurately aligned using the active technique it is securely glued in position with UV curing glue.

The problems with the conventional technique discussed above, are that the accuracy with which the fibre can be aligned is limited by the accuracy with which the grooves in the silicon substrate can be formed . In particular, in order to guarantee accurate alignment the depth of the groove is critical. Typically the depth must be of the order of 0.5 micrometres, and accurately achieving such a depth is difficult. Further, active alignment is expensive, time consuming and requires expertise to ensure that the fibres rest correctly in the grooves. Further, there is a limit to the number of optical fibres that can be aligned using a grooved block due to the tolerances in forming the grooves in the block. At present the maximum number of fibres alignable in this way is 1 6.

For ease of explanation only, the following text refers to the specific case of aligning an optical fibre with a planar silica waveguide. It will nevertheless be appreciated by those skilled in the art, that the invention has broader application, such as in integrated circuits fabricated on other materials, such as silicon or group lll-V semiconductors such as InP.

In accordance with a first aspect, the invention provides a device for aligning an optical fibre with an optical waveguide of an optical component, the component having first and second opposed surfaces joined by a face, said optical waveguide lying at a predetermined distance from said first opposed surface, an end of the waveguide being exposed at said face, the device including first and second polymeric portions, said first portion for attachment to said face and including a first reference surface located at a predetermined position, which, in use, determines the position of the fibre with respect to said first opposed surface, and a second reference surface, which in use, determines the position of said first and second portions with respect to each other and to said face, said second portion including a third reference surface for abutment with said first opposed surface, and a cooperating part for cooperation with said second reference surface of said first portion, so that the first and second portions cooperate, wherein, in use, the position of said first reference surface with respect to said face is such that the position of said optical fibre as determined by said first reference surface, aligns with the exposed waveguide at said face.

With embodiments in accordance with the first aspect of the invention, the dimensions of the alignment device can be moulded with sufficient accuracy to permit passive alignment of the fibre with respect to the optical component. According to a second aspect, the invention provides a device according to claim 1 for aligning a plurality of optical fibres with respect to a plurality of optical waveguides of an optical component including said plurality of optical waveguides, wherein for each optical fibre to be aligned, said first portion includes a respective first reference surface.

With embodiments in accordance with the second aspect, the dimensions of the alignment device can be moulded with sufficient accuracy to permit passive alignment of up to and in excess of 40 optical fibres with respect to the optical component.

According to third and fourth aspects, respectively, the invention provides a system for alignment of an optical fibre with respect to optical waveguide disposed within an optical component, including a device according to claims 1 or 6, whereby on said first opposed surface of said optical component, on which said second portion is engaged, a polymeric abutment means is provided, whereby said second portion abuts with said abutment means at a predetermined location on said first opposed surface of said optical component.

With embodiments in accordance with the third or fourth aspects, the alignment device for aligning a single or a plurality of optical fibres may be accurately positioned on the optical component.

According to a fifth aspect, the invention provides a device according to claim 1 , wherein said second reference surface defines a resilient portion arranged to act on said second portion, in use, to hold it in place. With embodiments in accordance with the fifth aspect, the first and second engaged portions are releasably assembled . Once assembled, they are held firmly in place.

According to a sixth aspect, the invention provides a device according to claim 1 , wherein said second portion includes at least two bars which engage with one of said opposed surfaces, wherein said cooperating part includes a part of each of said bars overhanging said end of the optical component. With embodiments in accordance with the sixth aspect, the dimensions of the first and second portions may be determined with sufficient accuracy that parts simply disposed with respect to one another provide accurate fibre alignment.

In accordance with the present invention there is further provided a method of manufacturing a device according to claim 1 or claim 6, said method including the steps of forming a mould with predetermined dimensions by exposing a photoresist to radiation, and thereafter using said mould to manufacture a moulded component. With embodiments in accordance with the method of the present invention, moulded pieces may be made whose dimensions are accurate to + /- 0.5 micrometres.

In accordance with the present invention, there is further provided a method of manufacturing an abutment means for use in the system according to claim 7, including the steps of disposing a photoresist layer on a surface of the optical component on which said abutment means is to be provided, exposing said photoresist to radiation through a mask defining the abutment means to be provided, removing those areas that have not been exposed to radiation, so that those exposed areas remaining form said abutment means.

In order that the invention may be more fully understood embodiments thereof will now be described by way of example only, and by way of contrast with a prior art device as previously described, reference being made to the accompanying drawings in which: Figure 1 shows a planar component in which waveguides are formed;

Figure 2 shows an end elevation of a face of the component of Figure 1 along x-x, depicting the structure of the planar waveguide structure;

Figure 3 shows an alignment device according to a first embodiment of the present invention; Figure 4 shows a side elevation of an alignment device according to the present invention in place on a planar silica waveguide component;

Figure 5 shows an alignment device according to a second embodiment; Figure 6 shows a system for alignment of an optical fibre according to a third embodiment of the present invention;

Figure 7 shows a system for alignment of an optical fibre according to a fourth embodiment of the present invention; Figure 8 shows a cross section taken along z-z of the system of Figures 6 and 7;

Figure 9 shows a cross section of a detail of the system of Figures 6 and 7 taken along line z-z;

Figures 1 0 and 1 1 show a system according to the third and fourth embodiments, respectively; Figures 1 2 and 1 3 show a plan view of the assembled system according to the third and fourth embodiments, respectively;

Figure 1 4 shows a first part of an alignment device according to a fifth embodiment;

Figure 1 5 shows a second part of an alignment device according of Figure 14;

Figure 1 6 shows a face on view of the alignment device of Figure 1 4, showing the first and second parts fitted together in use;

Figure 1 7 shows a plan view of an alignment device according to a sixth embodiment; and Figure 1 8 shows a face on view of the alignment device of Figure

1 7.

Figures 1 and 2 show an example of an optical component with which the alignment device of the present invention aligns an optical fibre. The alignment device of the present invention may be used to align any type of optical component. In particular, the alignment device may be used to align optical fibres to optical components including silica waveguides disposed on silicon substrates (known as "silica on silicon" technology) . Further applications are in lithium niobate components, polymer components, "silicon on silicon" components and lll-V semiconductor components.

Figure 1 shows an optical component, in this case a structure in which waveguides are disposed . The component comprises first and second opposed surfaces joined by a face. The optical waveguide lies at a predetermined distance from the first opposed surface, the top surface, (the position with respect to the opposed bottom surface may or may not be known accurately) . An end of the waveguide is exposed at the face. The optical component may comprise a single waveguide. Alternatively, it may comprise a plurality of waveguides. In this second case, the spacing of adjacent waveguides depends on the particular waveguides, but in the example shown in Figures 1 and 2, the spacing is of the order of 250 micrometres. Figure 2 shows, in more detail, but not to scale an end elevation of a face of an optical component, in particular, a planar waveguide structure along the line x-x. The component comprises a silicon substrate on which is disposed a glass material having a low refractive index relative to a "core" waveguidmg region. The lower refractive index glass region forms the cladding region of the waveguide. Patterned within cladding region are waveguides of a glass material having a higher refractive index. The higher refractive index waveguidmg regions are referred to as the "core" regions. By virtue of the difference in refractive index between the cladding and core regions, light is confined to the core due to total internal reflection, as in conventional optical waveguides. Disposed on the substrate 24, which has a thickness of approximately 1 mm, there is a buffer layer 26. On top of the buffer layer 26 there is a cladding layer 22. The core 20 of the waveguide structure is within the cladding layer 22. The refractive indices of the materials of the core 20 and the silicon substrate 24 are higher than the buffer 26 and the cladding layers 22. The buffer 26 is provided in order to separate the higher index core from the high index substrate and has a thickness of approximately 1 0-20μm. The dimensions of the core in which the light is confined, depend on each particular application, but may, for example, be in the range of 5x5 to 8x8 micrometres². The thickness of the cladding is chosen to be sufficient to confine the mode of the light in the waveguide, and may, for example, be in the region of 1 0-20 micrometres. The core 20 generally has a square or rectangular cross section, as such sections are easier to fabricate than the more desirable circular section. As a general rule of thumb, the thickness of the buffer and cladding layers are approximately three times that of the core.

Some waveguide structures have a flat upper surface, wherein the waveguidmg regions are embedded within the cladding layer 22. In other structures, however, such as that shown in Figure 2, the core is disposed within the cladding layer 22 in such a way that there is a slightly raised region 28 on the upper surface of the cladding which forms the upper surface of the waveguide structure. This resulting lack of flatness in the upper surface of the cladding layer may give rise to alignment problems depending on the alignment technique used .

As discussed above, the x-x direction in Figure 2 represents a polished plane, normal to the waveguides. In order to align an optical fibre with respect to a waveguide disposed in an optical component, such as the one shown in Figures 1 and 2, accuracy is required, as the cross sectional area of the optical fibre core is of the order of 25-64 micrometres². Where a plurality of fibres are to be aligned with a plurality of waveguides disposed in an optical component, the spacing between adjacent waveguides must be faithfully reproduced . This distance is approximately 250 micrometres. The waveguides are disposed in the optical component a distance e from the top surface of the optical component. The distance e may vary depending on the planar silica waveguide structure, but may be of the order of 1 0 micrometres. As described hereinbelow, it has been found that an alignment device which faithfully reproduces this dimension and which can be placed over the face of the optical component, where the waveguide is exposed, permits passive alignment of an optical fibre to that waveguide.

Figure 3 shows an alignment device according to the present invention which is used to align an optical fibre with an optical component, such as that shown in Figures 1 and 2.

In the description below, it will be apparent to the skilled person that configurations of the portions other than that shown in Figures 3 and 4 are possible. It is to be noted that in this specification, that unless the context clearly requires otherwise, expressions such as "top", "vertical", "horizontal", etc, are used for ease of description only and are not intended to convey a particular orientation of the device in use.

The device of the present invention is includes polymeric portions. The portions have a high aspect ratio, that is to say that the portions comprise opposed surfaces whose dimensions are predetermined to a high degree of accuracy, the opposed surfaces are joined by parallel walls. The extremely high precision of the polymeric parts moulded is due to the use of synchrotron radiation in the preparation of the mould. The intense highly collimated beam offers optimum feature resolution, aspect ratio, structural height and parallelism of the mould, from which portions are moulded . The moulds used in the preparation of the portions is designed using CAD, discussed in more detail hereinbelow.

In particular, first and second portions 6, 7 cooperate with each other. They are moulded so that they engage to be assembled simply by fitting the parts together with no machining being necessary. The alignment device shown in Figures 3 and 4 includes at least two polymeric portions. The portions are dimensioned so that they cooperate with one another. In cooperation they, preferably have substantially a T- shaped cross-section, where preferably, one arm 1 7 of the cross of the "T" 1 6, 1 7, is longer than the other 1 6, this is so that the second portion rests stably on one of the opposing surfaces of the optical component. This increases the stability of the device. The first portion 6 for attachment to the end face rests against the end face

1 1 of the optical component. In this embodiment, the first portion is essentially rectangular in shape. It is dimensioned so that its height extends from a lower edge at one of the opposing surfaces of the optical component to an upper edge which extends beyond the other opposing surface. The first portion includes a first reference surface 1 3 located at a predetermined position, which determines the position of the an optical fibre to be aligned with respect to the first opposed surface. The optical fibre is supported by the first reference surface, in use at the face of the optical component, with which the fibre is to be aligned. The first reference surface extends through the first portion 6 from the front face to the back face. The back face of the first portion is for attachment to the optical component at face 1 1 . The first reference surface 1 3 may define an opening including an aperture, a through hole or a recess, and extends through the first portion . In addition, the first portion may also include a gripping means 1 4 which may comprise an inverted V-groove extending from the lower edge of first portion to the first reference surface 1 3. By virtue of the combination of the groove and the first reference surface, a degree of resilience is introduced into the first portion. The dimensions of the groove and first reference surface are chosen such that the first portion possesses a degree of resilience, so that the first portion deforms elastically to permits the optical fibre to be inserted along the first reference surface and received. Once received, the first portion 6, 66 regains its original shape and the optical fibre is gripped in position. The first portion further includes a second reference surface, which determines the position of the first and a second portion with respect to each other and the face 1 1 of the optical component. In particular, the second reference surface may define either a recess or an aperture and is adapted to receive a cooperating part 9 located on the second portion 7.

The second portion 7 includes a third reference surface 1 9 to abut with one of said opposed faces 3, and includes a substantially rectangular piece having a first part which has a substantially constant width and thickness, a second tapered part having a constant thickness, and which tapers to a cooperating part 9 having a constant width and thickness, the width being smaller than that of the first part. The second portion 7 is to be positioned in use on the top surface 10 of the optical component 1 . The cooperating part may include a protrusion 9, which extends beyond the end face 1 1 of the optical component. The second portion further preferably is provided with a through hole 30. This facilitates the attachment of the second portion 7 to the integrated optical circuit. The portions 7, and 77 in Figure 5, are glued to the integrated optical circuit. This may be achieved by filling the through hole 30 with UV curable epoxy glue, which is subsequently cured . Once cured the glue fixes the portions 6, 66 permanently with respect to the integrated optical component. As the through hole 30 does nothing to facilitate accurate alignment, its dimensions are not critical. The through hole 30, is incorporated into the CAD design, however, in order to simplify the alignment device. The invention is applicable to all fibre types, but it is of particular application to single mode fibres having tight dimensional tolerances with respect to concentricity and diameter, of for example, 1 25 micrometres + /-1 micrometre or less.

The manufacture of portions 6 and 7 will now be discussed. Portions 6, 66 and 7, 77 are moulded from moulds which have been formed using the LIGA technique. The portions 6 and 7 can be moulded using a variety of polymer materials, such as PEEK (polyether ethyl ketone), PMMA (polymethyl methacrylate), POM (polyoxymethylene), PEI (polyetherimide), LCP (liquid crystalline copolyester) . These polymers have been selected as they have low thermal expansion coefficients and suffer minimal shrinkage on cooling. Preferably, the parts are moulded from PEEK. PEEK is a very stable, inert and insoluble, high performance polymer. It has been chosen as a preferable polymer material as it has been found to faithfully reproduce the mould . There is minimal shrinkage on cooling. Further, edge definition of PEEK moulded parts has been found to be good. Portions manufactured using moulds formed using the LIGA technique and PEEK as the material to be moulded have dimensions whose tolerances are within + /- 0.5 micrometres. Thus, the portions moulded from the LIGA technique manufactured moulds, are moulded to such a high degree of accuracy, that the dimension e, which is the distance from the centre of the waveguide in the optical component to the upper surface 1 0 of the optical component, can be defined to such accuracy in all 3 dimensions, that once accurately positioned in the x direction, the optical fibre 5 can be passively aligned with the waveguide in the optical component.

The alignment device shown in Figures 3 and 4 are moulded using a mould which has been micromachined using the LIGA technique. In the field of micromachining several techniques are known. LIGA is one of those techniques. The LIGA process is well known and is discussed in IMM, "The LIGA Technique", Commercial Brochure, IMM, 1 995. LIGA is an acronym for lithography (Lithographie), electroplating (Galvanoformung), moulding (Abformung) and has been applied to different technological fields demanding precision definition of high aspect features. Another technique uses UV radiation to expose a polymer, SU-8. Both techniques allow moulds to be produced from which high precision components are made. However, the nature of the processes involved in LIGA and UV exposure of SU-8, mean that these techniques have been limited in their applications. For example, to date LIGA has been used to manufacture motherboards for optical asymmetric demultiplexers, such as TOADs. Whereby the motherboard is machined to such accuracy that the components that constitute the TOAD can simply be dropped into place, as described in British Telecommunications Engineering, Vol. 1 7, Oct. 1 998. It is further known to use UV radiation to expose SU-8 polymer to fabricate moulds for watch components, such as gears, as described in Microsystem Technologies 4 ( 1 998) 1 43-1 46. In particular, the moulds from which the portions are moulded according to the present invention are manufactured according to the following technique. A CAD design of the mould is made. In designing the CAD mould, systematic error corrections are needed for oversizing or undersizing of LIGA parts due, for example, to temperature rises on exposure to x-rays.

An x-ray sensitive resist such as PMMA is spun on to an electrically conducting substrate to a certain thickness. The thickness will vary depending on the particular application, but may typically be of the order of 1000 micrometres. This is achieved simply by depositing a sample of the polymer on to the substrate and spinning the substrate at a speed of approximately 4000 rpm until the desired thickness of PMMA resist is achieved on the substrate. The thickness of the PMMA resist is chosen to correspond approximately to the thickness f of the moulded parts. The x ray sensitive resist is patterned by x-ray exposure though a photomask. The photomask bears the CAD design, which represents the "negative" of the mould. The photomask comprises, by way of example, a beryllium backing plate, which is transparent to x-rays, and a pattern made from a metal having a high atomic weight (a "heavy metal"), patterned according to the CAD design of the mould, which blocks x-rays. The heavy metal is preferably gold, but may be any other x-ray blocking heavy metal. The x-ray exposure of the PMMA causes some of the polymer chains in those areas that have been exposed to the x-rays to cross link. The result of increased cross linking in the exposed areas is that they harden. Subsequent resist development, achieved by dissolving the "unhardened" PMMA areas in a solvent, removes those areas of the PMMA that have not been exposed, leaving the exposed resist areas standing. Thus, the development yields a three-dimensional resist structure, having a depth f, where the thickness f may be of the order of 1 000 micrometres. The thickness is determined by the extent of the x-ray penetration and will depend, amongst other factors, upon those such as the x-ray power and exposure time. The 3D resist structure is subsequently electroplated using, for example, nickel. The depth f is chosen, and the exposure parameters are controlled so as to achieve the desired depth. The depth is chosen so as to allow a certain degree of slack between the first and second portions to be fitted together. The amount of slack is sufficient to allow the parts to be fitted together without requiring great force or further machining. The remaining structure is electroplated to form a metal structure on the substrate. This 3D metallic structure serves as a mould for precision plastic injection moulding or hot embossing. The plastic, as mentioned above, may be varied, and may be, for example, PEEK, PMMA, POM, PEI or LCP. However, preferably, the plastic used is PEEK.

As mentioned above, the extremely high precision of the plastic parts moulded from the nickel mould is due to the use of synchrotron radiation as the x-ray source. The intense highly collimated beam offers optimum feature resolution, aspect ratio, structural height and parallelism of the resist walls. The use of moulds manufactured according to the LIGA technique allows batch processable piece portions 6, 7 to be manufactured. The portions manufactured possess high aspect ratio definition. This term is used to describe parts having sheer gradient walls. The aspect ratio of a portion is determined by the ratio of the height of the part divided by the offset error of the part. For example, the aspect ratio for the first and second portions is approximately 25. Thus, for a first portion whose thickness is 25 micrometres, the offset error is 1 micrometre. A part manufactured using the LIGA technique will has a high aspect ratio as the highly collimated x-ray beam produces moulds having very sheer sides.

Figure 5 shows an alignment device according to a second embodiment. The alignment device according to a second embodiment aligns a plurality of optical fibres with respect to waveguides disposed in an optical component. The alignment device shown in Figure 5 is manufactured in the same way using the LIGA technique as the alignment device shown in Figures 3 and 4. The device shown in Figure 5 allows a plurality of optical fibres, whose number may vary depending on the application, but may be up to and in excess of 40, to be aligned to an optical component, such as an integrated optical circuit. As mentioned previously, waveguides disposed within a planar silica waveguide component are usually regularly spaced . This spacing d is constant for a particular type of component but may vary depending on the type of component. The spacing d is of the order of 250 micrometres. The spacing of the waveguides in the planar silica component shown in Figure 5 is also 250 micrometres. The alignment device comprises two polymeric portions 66, 77. As with the first embodiment, both portions cooperate with one another, and are moulded to such a degree of accuracy that they can be assembled simply by fitting the portions together. No further machining is necessary. Once assembled, the portions cooperate, having substantially a T-shaped cross-section, where preferably, one arm of the cross of the T is longer than the other, to give the assembled device stability.

The first portion 66 is attachable to the end face 1 1 of the optical component and, in use, rests against it. The second portion 77 includes a third reference surface 1 90 for abutment with one of the opposed surfaces of the optical component. Both portions include those features including the first, second and third reference surfaces 1 20, 1 30, 1 90 and gripping means 1 40 of the first and second portions 6, 7, respectively, discussed with reference to Figures 3 and 4. In particular, each portion 66, 77 for every optical fibre to be aligned, includes respective features for each optical fibre to be aligned with each waveguide disposed in the optical component.

The first portion 66 for attachment to the end face rests against the end face 1 1 of the optical component. The first portion 66 is essentially rectangular in shape. It is dimensioned so that its height extends from a lower edge at one of the opposing surfaces of the optical component to an upper edge beyond the other opposing surface. It is dimensioned in the x-x direction to extend along all waveguides disposed in the optical component, with which optical fibres are to be aligned. The first portion includes a plurality of first reference surfaces 1 30, each of which determines the position of each respective fibre with respect to the face 1 1 . Each optical fibre to be aligned is introduced along respective first reference surfaces. Each first reference surface 1 30 extends through the first portion 66 from the front face to the back face, which is for attachment to the optical component. Each first reference surface 1 30 may define an opening including an aperture, a through hole, or a recess that extends through the first portion. In addition, the first portion 66 may also include a plurality of gripping means 1 40 which may comprise an inverted V-groove extending from the lower edge of first portion to the first reference surface 1 30. The number of gripping means will depend on the number of waveguides to be aligned. By virtue of the combination of the groove and the first reference surface, and their dimensions, a degree of resilience is introduced into the first portion 66, which permits the optical fibre to be inserted along the first reference surface. Once received, the first portion 6, 66 regains its original shape and the optical fibre is gripped in position. The first 5 portion 66 further includes a plurality of second reference surface 1 20, which determines the position of the first and a second portion 66, 77 with respect to each other and the face 1 1 of the optical component. In particular, the second reference 1 20 surface may define either a recess or an aperture and is adapted to receive a cooperating part 1 90 located on the second portion 77. The second portion 77

10 includes a third reference surface 1 90 to abut with one of said opposed faces 3, and includes a substantially rectangular piece having a first part which has a substantially constant width and thickness, a second tapered part having a constant thickness, and which tapers to a plurality of cooperating parts 9 having a constant width and thickness, the width being smaller than that of the first portion. The second portion

1 5 77 is to be positioned in use on the top surface 1 0 of the optical component 1 . Each of the plurality of cooperating parts may include a protrusion 9, which extends beyond the end face 1 1 of the optical component. Each cooperating part cooperates with a respective second reference surface. It is noted that the number of second reference surfaces does not necessarily have to correspond to the number of optical

20 fibres to be aligned, as long as the position of the first and second portions can be maintained with respect to one another accurately. The number of protrusions and means 1 20 should be equal. However, this number does not necessarily have to be the same as the number of optical fibres to be aligned. The function of the cooperating parts and the second reference surfaces 1 20 is to ensure that the

25 position of the first and second portions with respect to one another are accurately defined .

The precision moulded parts discussed above allow the optical fibre to be aligned within a degree of accuracy of + /- 0.5 micrometres in the y direction shown in 30 Figures 3-7, that is in the direction perpendicular to both the longitudinal axis of the waveguides in the optical component and perpendicular to the direction in which adjacent waveguides are disposed within the optical component. The embodiment shown in Figure 6 facilitates accurate alignment of an optical fibre with respect to the x direction shown in Figures 3-7 that is in the direction in which adjacent waveguides are disposed within the optical component (and perpendicular to the longitudinal axis of the waveguides in the optical component) . Alignment in this direction is facilitated by the provision of an abutment means 8, 88. In the examples shown the abutment means is made from a polymer material.

Figure 6 shows a system for alignment of an optical fibre according to a third embodiment of the present invention. The second portion 7 is produced in the same way as described above. The abutment means 8 is made from a polymer such as SU-8 using a micromachining technique. In contrast to the LIGA technique, instead of exposing a photoresist to x-ray radiation, a photoresist is exposed to near ultraviolet radiation at around 400 nanometers. This technique also permits the production of high aspect ratio features. The aspect ratio of the features produced using this method is approximately 20, that is a stop having a height of 1 0 micrometres will have an error of 0.5 micrometres. The stop is micromachined in the following way. In the following example, the polymer SU-8 is used, however any polymer which is photosensitive to ultraviolet may be used, provided that its polymer structure is such that the same aspect ratios can be achieved. SU-8 is a negative- tone photoresist comprising of EPON SU-8 resin and is photosensitised with triaryl sulfonium salt, for example, Cyracure UVI from Union Carbide) . The photoresist has two important properties suited for micromachining the stop 8. First its low molecular weight allows dissolution in a variety of organic solvents. Second, the spun layer has a very low optical absorption in the near ultraviolet spectrum. SU-8 is spun onto the optical component, such as a planar silica optical component 1 , using conventional techniques. The viscosity of the SU-8 polymer is chosen so that when it is spun at 4000rpm, a desired thickness of polymer coverage over the planar silica optical component is achieved. For example, if SU-8- 1 0 is spun at 4000rpm, the resulting polymer thickness will be 1 0 micrometres. Similarly, if SU-8-500 is spun onto the planar silica at 4000 rpm, the polymer covering will have a thickness of 500 micrometres. Thus, the thickness can vary from approximately 1 0 to 500 micrometres. According to the embodiments shown in Figures 6 and 7 the viscosity of the SU-8 is selected so that when it is spun on to the planar silica optical component, the thickness of the covering achieved is preferably approximately 20 micrometres. Prior to spinning the planar silica is pre baked at 200 degrees Celcius for approximately 1 5 minutes. The SU-8 is then applied on to the planar silica. The planar silica is mounted for spinning and is then spun at 1 500 rpm for 1 5 seconds. The resist coated planar silica optical component is subsequently soft baked at 70 degrees Celcius before being further baked at 90 degrees Celcius for forty minutes. A standard chromium mask, which blocks ultraviolet radiation, is prepared. The mask which is designed using CAD and by standard photolithography techniques includes a window corresponding to the desired shape of the stop to be patterned onto the photoresist, and whose dimensions can be determined extremely accurately. The mask is then placed onto the coated component.

The resist is then subject to near ultraviolet exposure of around 400 nanometers through the mask. The exposure device may be a mask aligner in contact mode (for example, a SUSS M4A and MJB21 ) . Those areas of the photoresist polymer disposed beneath the window are exposed to the UV radiation and become hardened . The exposure dose is between 300 and 1 200 mJ/cm2, depending on the thickness of the photoresist layer and is of the duration of approximately 90 seconds. The exposed resist is subjected to postexposure baking on a hot plate at 50 degrees Celcius for approximately 3 minutes, followed by 90 degrees Celcius for a further three minutes. The exposed, post baked resist coated component is then developed in an organic solvent, such as propylene glycol methyl ether acetate

(PGMEA) for 1 minute, plus a further minute at room temperature. The development removes those areas of the photoresist that have not become hardened by the UV exposure.

After development in organic solvent, the area patterned in the shape of an abutment means 8, which has been hardened by exposure to the ultraviolet radiation remains on the planar component, whilst those areas disposed behind the mask, and which have not been hardened, as they were not exposed to UV radiation are dissolved and washed away by the solvent. The abutment means 8 thus comprises hardened SU-8 polymer and has a height h equal to the thickness of the spun SU-8 layer and a high aspect ratio of approximately 20. The positioning of the abutment means 8 on the planar silica component 1 can be accurately determined using CAD. This position is critical as it determines the position of the alignment device in the x direction, defined above as being the direction in which adjacent waveguides are disposed within the optical component (and perpendicular to the longitudinal axis of the waveguides in the optical component), as shown in Figure 6. The abutment means 8 is positioned so that when the second portion 6 of the alignment device abuts with it, the engaged first and second portions 6,7 are positioned with respect to the waveguides in the optical component so that the optical fibre can be passively aligned with it. Using the techniques described above it is possible to micromachine each of the portions 6 and 7 and abutment means 8, and define the position of abutment means 8 on the optical component with sufficient accuracy, that they are alignable with respect to each other to such a degree of accuracy that optical fibres can be passively aligned with respect to the waveguides disposed in the optical component. No further machining is necessary. Further, if it is found that optical loss occurs due to a misalignment in the y direction due to an error in the position of the stop, this is easily remedied simply by removing the polymer stop and forming it again using the same process, in the correct position. This represents an advantage over conventional plasma deposited oxide abutment means, where once the abutment means has been formed on the substrate, it is difficult to remove, and expensive to reform.

Figure 7 shows a system for alignment of a plurality of optical fibres according to a fourth embodiment of the present invention. According to a fourth embodiment, the abutment means 88, made according to the method described above, is abutted with a second portion 77 micromachine moulded to align a plurality of optical fibres with respect to a plurality of waveguides disposed within an optical component 1 .

Figures 8 to 1 3 show how the parts and the stop manufactured as described above, are assembled according to the third and fourth embodiments. Figure 8 shows a cross section taken along z-z shown in Figures 6 and 7 showing the abutment means 8, 88 and the waveguide 2 disposed within the optical component 1 . Once the abutment means 8, 88 has been patterned on the optical component, the second portion 7, 77 is abutted against it. Figure 8 shows the abutted second portion 7, 77, according to a cross section taken along line z-z shown in Figures 6 and 7. The first portion 6, 66 is fibre fed. This is achieved simply by introducing the optical fibres 5, 55, to be aligned with the waveguides disposed in the optical component, along the first reference surface 1 3, 1 30. The gripping means 1 4, 140 gives the structure of the first portion 6, 66 a degree of resilience, thus allowing the fibres to be fed more easily along the first reference surface 1 3, 1 30. However, once the fibre 5 , 55 is introduced along the first reference surface 1 3, 1 30, it is laterally gripped in position. Figures 1 0 and 1 1 depict the fibres inserted as discussed above, according to the third and fourth embodiments, respectively. According to third and fourth embodiments, an abutment means is provided on the optical component for positioning the alignment device on the optical component. The third embodiment relates to the case where a single optical fibre is aligned with an optical component. The fourth embodiment relates to the case where a plurality of optical fibres are aligned with an optical component. The fed fibres are aligned with the waveguides 2 disposed in the optical component 1 , by simply assembling the fibre fed first portion 6, 66 with the second portion 7, 77. The fibres 5, 55 are then glued in position to keep them permanently aligned with the waveguides 2. The parts 6, 66 are glued to the optical component. This is done by filling the through holes 30 formed in the portions with UV epoxy glue and fixed into place for permanent fixation. Further, in order to aid coupling from the waveguide to the fibre, the interface between the fibre and the end face 1 1 of the optical component in which the waveguide is disposed, is smeared with optical index matching material. Further, the fibre 5 is glued into the position defined by the first reference surface 1 3, 1 30 to permanently strengthen the join. Figures 1 2 and 1 3 show a plan view of the assembled system according to the third and fourth embodiments, respectively.

Figure 1 4 shows a first part of an alignment device for alignment of an optical fibre according to a fifth embodiment. According to the fifth embodiment the second reference surface defines a resilient means. As discussed above, it has been found that it is advantageous if a certain degree of slack is provided between the first and second portions 6, 66, 7, 77. Further, it has been found that in order to improve the fit between the first and second portions 6, 66, 7, 77, the CAD design for the portions may provide that the second reference surface 1 20 defines a resilient means 60. The resilient means may take many forms, however, one example is shown in Figures 1 4 and 1 6. It may be desired to improve the fit between the first and the second portions 6, 66, 7, 77, as the thickness f of the first and second portions is determined, as discussed above, by amongst other factors, the x-ray exposure time and power. Thus, the dimension f cannot be so accurately determined as, for example, the distance e, which is defined as being the distance between the centre of the core of the fibre to be aligned and the lower edge of the opening 1 2 in the first portion 6. This is because the accuracy of distance e is determined by the xray beam, which as discussed above, is highly collimated, and is determined within + /- 0.5 micrometres.

The first portion 6 made, as discussed above, using the LIGA technique from a polymer such as PEEK, is resilient to the degree that the resilient means 60 defined by the second reference surface will hold both parts firmly in place.

The resilient means shown Figure 1 4 may have many different configurations, provided that when the second portion 7 cooperates along the second reference surface 1 2, force is exerted between the first and second portions, thus, holding them in place. In Figure 1 4 the resilient means 60 is disposed on the upper inner surface of the second reference surface 1 2. Thus, when the first and second portions cooperate, the resilient means 60 exerts force downwards onto the first portion to force it towards the lower right hand corner of the second reference surface 1 2, and thus, hold it in place. It is not critical how far along the upper inner surface of the second reference surface 1 2, the resilient means 60 extends. However, in the example shown, the resilient means 60 extends half way along the upper surface of the second reference surface 1 2. The width of the second reference surface minus the length of the resilient means is equal to the distance H, shown in Figure 1 4. Further the resilient means 60 is provided with a rounded lower edge. This is provided to improve the ease with which the second portions and first portions cooperate.

Figure 1 5 shows a second part of an alignment device for alignment of an optical fibre according to the fifth embodiment. As shown in Figure 1 5, the second portion 7 comprises a main body and a cooperating part 9. The cooperating part 9 is tapered along one side 1 6. The tapered portion 1 6 is provided so that the first portion can be introduced along the second reference surface without being impeded by the resilient means. Both the tapered portion and the rounded under side of the resilient means improve the ease with which both portions can be fitted together. Further, it is preferable that the distance H is approximately the same as the distance J, shown in Figure 1 6, where the distance J is equal to the width of the tapered end section 1 7 of the second portion 7. It has been found that a better fit is achieved if the distance J is slightly less than the distance H.

Figure 1 6 shows a face on view of an alignment device according to the fifth embodiment, showing the first and second portions cooperating in use. In order to optimise the fit between the first and second portions, it has been found that the distance A1 , which is the width of the opening 1 2 in the first portion 6, should be slightly longer that the distance A2, which is the width of the cooperating part 9 on the second portion. This difference may be as small as a few micrometres, if necessary. Further, it has been found that the distance k, which is the distance between the bottom of the lower rounded side of the resilient means 60, should be slightly less that the distance f, which is the thickness of the second portion 6.

It is further envisaged that the resilient means 60 does not have to comprise a discrete feature. Alternatively, any of the walls defined by the second reference surface, however, preferably, the two vertical walls, of the opening 1 2 could simply be slightly bowed inwards.

Figures 1 7 and 1 8 shows an alternative alignment device according to a sixth embodiment. Using the techniques discussed above, abutment means 72 are patterned onto an optical component. First and second portions 74, 70 are designed using CAD are moulded out of a polymer material such as PEEK, using the LIGA technique discussed above. The rectangular shaped second portions 70 abut the abutment means 72. The use of the LIGA technique to create moulds from which portions 70 are made, ensures that the width 76 of the second portions and the dimensions of the first portions 74, including engaging means comprising fibre guiding grooves 78 can be set to within a very high accuracy of + /- 0.5 micrometres. The second portions 70 are affixed to the optical component using UV curing epoxy glue. The first portion 74 is brought into alignment with the second portion, using those parts of the second portion overhanging the optical component as alignment means. The dimensions of the fibre guiding grooves p and the distance from the lower edge of the fibre guiding grooves to the lower edge of the second part 74 is also determinable to + /- 0.5 micrometres by virtue of the LIGA technique.

Claims

1 . A device for aligning an optical fibre with an optical waveguide of an optical component, the component having first and second opposed surfaces joined by a face, said optical waveguide lying at a predetermined distance from said first opposed surface, an end of the waveguide being exposed at said face, the device including first and second polymeric portions, said first portion for attachment to said face and including a first reference surface located at a predetermined position, which, in use, determines the position of the fibre with respect to said first opposed surface, and a second reference surface, which in use, determines the position of said first and second portions with respect to each other and to said face, said second portion including a third reference surface for abutment with said first opposed surface, and a cooperating part for cooperation with said second reference surface of said first portion, so that the first and second portions cooperate, wherein, in use, the position of said first reference surface with respect to said face is such that the position of said optical fibre as determined by said first reference surface, aligns with the exposed waveguide at said face.

2. A device according to claim 1 , wherein said first reference surface defines either a groove or a first aperture.

3. A device according to claim 1 , wherein said second reference surface defines a second aperture, wherein said cooperating part is received by said second aperture.

4. A device according to claim 1 , wherein first and second portions include a moulded material.

5. A device according to claim 1 , wherein said first reference surface is disposed such as to allow said optical fibre to be introduced along said first reference surface in a direction along a longitudinal axis of the waveguide disposed in the optical component.

6. A device according to claim 1 for aligning a plurality of optical fibres with respect to a plurality of optical waveguides of an optical component including said plurality of optical waveguides, wherein for each optical fibre to be aligned, said first portion includes a respective first reference surface.

7. A system for alignment of an optical fibre with respect to optical waveguide disposed within an optical component, including a device according to claims 1 or 6, whereby on said first opposed surface of said optical component, on which said second portion is engaged, a polymeric abutment means is provided, whereby said second portion abuts with said abutment means at a predetermined location on said first opposed surface of said optical component.

8. A device according to claim 1 , wherein said second reference surface defines a resilient portion arranged to act on said second portion, in use, to hold it in place.

9. A device according to claim 1 , wherein said second portion includes at least two bars which engage with one of said opposed surfaces, wherein said cooperating part includes a part of each of said bars overhanging said end of the optical component.

1 0. A method of manufacturing a device according to claim 1 or claim 6, said method including the steps of forming a mould with predetermined dimensions by exposing a photoresist to radiation, and thereafter using said mould to manufacture a moulded component.

1 1 . A method according to claim 1 0, wherein said radiation includes a collimated x-ray beam.

1 2. A method of manufacturing an abutment means for use in the system according to claim 7, including the steps of disposing a photoresist layer on a surface of the optical component on which said abutment means is to be provided, exposing said photoresist to radiation through a mask defining the abutment means to be provided, removing those areas that have not been exposed to radiation, so that those exposed areas remaining form said abutment means.