CA2292806A1 - Optical coupling system - Google Patents

Abstract

An optical coupling system for providing an output beam of light which is collimated, is disclosed wherein a length of single mode optical first fibre having a first outer diameter is optically coupled with an end face of a second optical fibre lens having a graded index region about a longitudinal optical axis of the lens. The second optical fibre lens has a same outer diameter as the first optical fibre. The ratio of the diameter of the graded index region of second optical fibre lens to the diameter of the core of the first optical fibre is 6.5:1 or greater. Conveniently, as the diameters of the fibre and the fibre lens are the same, they can be placed inside a sleeve such as a connector ferrule to provide a collimated beam at an output end. The lens and fibre can be joined by fusion, adhesive or within a sleeve can be spaced apart.

Description

Doc. No. 10-134 CA Patent Optical Coupling System Field of the Invention The invention relates to the use and fabrication of optical fibre lenses for use in optical components such as optical switches, optical filters, optical isolators, optical circulators and optical connectors and optical amplifiers to name just some. As well this invention relates to optical systems and devices utilizing microlensed optical fiber and methods for producing such.
Description of Related Art In the development of optical communications, the miniaturization of the optical devices and parts used therein has become a desired objective. In particular, optical isolators, circulators, and multiplexors are in demand which achieve both miniaturization and simplification when they are coupled with optical fibers.
Furthermore there is a need to improve upon conventional collimating systems which, heretofore have had inherent requirements for micron precision adjustment of the optical fiber and the optical axis position of the lens it is to be coupled with;
obviating this requirement would substantially reduce the cost of producing optical products that require collimated beams of light passing through their components.
Furthermore, from the standpoint of miniaturization, fiber collimator beams of sufficiently narrow (e.g. less than 100 Vim) light flux are required. However, conventional optical fiber collimators are not able to obtain a light flux narrower than 300 pm.
In optical components such as optical circulators or isolators, costly birefringent crystals and rotators in most instances have heretofore been provided with collimated beams of light. Typical diameters of these collimated beams has been about 350 Vim, and accordingly, costly components such as birefringent crystals have been sized to 3o accommodate one or more of these large beams.

Doc. No. l0-134 CA Patent By substantially reducing the size of the collimated beam passing through these components, their physical size and hence their cost can be substantially reduced.
Until now, and in most instances, coupling a graded index (GRIN) collimating lens with an optical fibre has required precision active alignment. It is an object of this invention to obviate the requirement for actively aligning an optical fibre and a collimating GRIN
lens it is to be coupled with.
Various patents are known which disclose the provision of a spherical lens at the tip of an to optical fibre. Although the lenses disclosed in these patents may perform their intended function, some better than others, it is an object of this invention, to provide a lens that collimates a beam passing through and exiting the lens. It is also an object of this invention to provide a lens which is relatively easy to manufacture and which has a planar end face, suitable for directly coupling with planar end faces of other optical 15 components. And yet still further, it is an object of this invention to provide collimating system wherein a fibre lens has an outer diameter that is substantially the same as the diameter of the optical fibre it is being coupled with, thereby obviating active alignment between the lens and the optical fibre.
2o It is possible to provide a collimating lens on the end of an optical fibre having a uniform core diameter, by first increasing the core diameter of the fibre near the fibre end tip, and then by providing a ball or spherical lens that attempts to focus light exiting the lens tip.
Notwithstanding, producing an accurate ball lens that will collimate light is a formidable and costly task. Yet still further, this ball lens is not a most convenient surface to couple 25 with planar optical components.
Although a crude spherical lens is easy to produce and is generally sufficient to the meet the needs of some systems in which it is placed, a spherical lens is subject to spherical aberration. Such spherical aberration lowers coupling efficiency and thus renders such a 3o fiber less preferred for low loss, high gain uses, such as laser optical amplifiers.

Doc. No. 10-134 CA Patent U.S. Pat. No. 4,565,558 (Keil et al.) and U.S. Pat. No. 4,589,897 (Mathyssek et al.) both relate to the formation of a spherical or aspherical lensed end on an optical fiber. The apparatus disclosed in these patents utilizes two clamps, at least one of which moves relatively to the other while an electric arc heats a portion of the fiber between the clamps. As a constriction appears as the result of constant tension and heat, the tension is dropped and a further constriction occurs leading to a separation which solidifies when the heat is cut off to form a lens on a tapered fiber.
Blaudau and Rossberg, Journal of Lightwave Technology, Vol. LT-3, No. 3, April teach making an aspherical lens by first forming a bulbous spherical lens on a fiber and then welding a cylinder of pure quartz at the center of the bulb. Upon remelting the pure quartz flows out to form an aspherical surface.
U.S. Pat. Nos. 4,243,349 and 4,370,021 (both to Khoe et al.) teach flattening the end of an optical fiber to produce a semi-elipsoidal lens.
United States Patent 5,551,968 in the name of Pan issued September 3, 1996 describes a method of forming a microlens at the tip of a fiber by jerking apart two fused fibers. One drawback to this invention is that the lens formed at the end of the optical fibre is a ball lens and is not precisely formed with ease. Furthermore, it is prefered to provide a lens which is a true graded index collimating lens, wherein centering of the lens and the optical fibre to which it is coupled is automatic. Yet still further, it is preferred to have a lens having a planar end face for coupling directly to optical components having a complementary planar end face. The lens of Pan in the '968 patent cannot provide a collimated beam.
A preferred embodiment of this invention provides a single mode optical fibre, and a graded index fibre lens to be coupled to one another preferably by fusion splicing; a first single mode optical fibre having a known outer diameter is jointed to a second graded 3o index fibre lens having a same outer diameter and having a length suitable for collimating light exiting its end face; the core diameter of the second fibre is substantially larger than Doc. No. 10-134 CA Patent the average diameter of the core of the single mode optical fibre. The core of the second fibre may or may not having a cladding region about the graded index region radiating outward from the centre of the fibre.
GRADED INDEX optical fibres are known in which the refractive index of the core decreases from the axis to the core-cladding interface in accordance with a pseudo-parabolic law. A refractive index profile such as this one minimizes the temporal broadening of a light pulse travelling along the fibre and therefore maximizes the band pass of the fibre. Such graded index fibres are well known for use in high-speed optical l0 transmission. The index of refraction of the core material is graded in the radial direction so as to yield an optical fiber with very low total dispersion and therefore high bandwidth. Since the work of Maxwell in the late nineteenth century, elucidating the electromagnetic wave nature of light, it has been known that electromagnetic radiation in the optical region of the spectrum is inherently capable of transmitting more information than electromagnetic radiation in lower frequency (higher wavelength) regions of the spectrum.
In the mid-1960's, it was suggested by Kao and Hockham (Proceedings of IEE, Vol.
113, No. 7, July 1966, page 1151 ) that pure silica would be able to transmit light in the 2o optical region of the electromagnetic spectrum with losses less than 20 dB/km, a value viewed by many as representing the onset of commercial, economic viability.
Hence silica would constitute the long searched-for low loss optical transmission medium.
In the early 1970's, a number of manufacturing techniques were developed to form silica based fibers of loss less than 20 dB/km. Such processes depended, by and large, upon the formation of silica-based glass from appropriate glass precursor vapors.
Processes include the Bell System MCVD process (U.S. Pat. No. 4,217,027), the Corning soot process (U.S. Pat. Nos. 3,711,262, Re. 28,029), and the VAD process (U.S.
Pat. Nos.
3,966,446, 4,135,901, 4,224,046) all incorporated herein by reference, pursued, among others, by a number of Japanese companies. As a result of these developments, optical Doc. No. 10-134 CA Patent fibers are now routinely fabricated in commercial processes with losses less than 2 dB/km in the optical region of the spectrum.
While losses associated with optical fibers have been reduced, such reduction is due primarily to the selection of appropriate material systems and to the use of appropriate fabrication processes. Added losses due to cabling, or induced by curvature, microbending, or splicing, continue to be of some concern, especially when ultra-low loss fibers (less than 1 dBlkm) are considered.
Despite the continuing concern with fiber loss, fiber characteristics have reached the point where repeater spacing is in many instances limited, not by the loss characteristics of the fiber, but rather by the bandwidth characteristics of the fiber.
Repeaters are required where the optical signal is well above detectable levels, since the individual pulses have reached a point where they overlap sufficiently to result in bandwidth degradation.
To understand the bandwidth limitation of optical fibers it should be remembered that, originally, fibers were constructed primarily with parameters appropriate to the support of numerous propagating modes within a single fiber, hence the term "multimode fibers."
2o Different propagation velocities, associated with each of these modes, result in dispersion, commonly referred to as mode dispersion, which affects the bandwidth of the fiber deleteriously. Early in the development of optical fibers, it was appreciated radial gradations in the index of refraction resulted in lowered mode dispersion and improved bandwidth of multi-mode fibers.
United States Patent 4,242,117 issued December 30, 1980 entitled Method of producing optical fibers for telecommunications describes a method of producing a glass fiber, having a graded refractive index profile, by means of the double crucible method, the initial materials being core and cladding glass compositions having mutually different 3o alkali ions, a profile is usually obtained which greatly deviates from the desired parabolic form. The invention furnishes the possibility of approximating this parabolic form very Doc. No. 10-134 CA Patent closely. This is attained by a partial substitution of the core alkali ion by the cladding alkali ion.
United States Patent 4,249,924 incorporated herein by reference issued February 10, 1981 entitled Process for the production of graded index optical fibres teaches a process for the manufacture of a glass blank in the form of a rod from which the optical fibre is drawn, said blank being produced starting from a molten bath of glass. The method teaches the creation of a radial refractive index gradient and an optical cladding.
to Heretofore graded index optical fibre has been primarily used in the field of high-speed optical communications.
In the past decade, un-clad graded index (GRIN) rod lenses, have become nearly ubiquitous in the field of optical component manufacture.
United States Patent 4,686,195 incorporated herein by reference relates to the manufacture of gradient index glass and a composition therefor. More particularly, it relates to a novel sol-gel diffusion process whereby large diameter, high gradient index glass can be manufactured 2o relatively inexpensively, and relates to a composition useful therefor in which the ions that form the metal oxides are of unequal valence.
Gradient index glass is a structure wherein the index of refraction changes across the structure. Typically, such structures are cylinder-shaped and the index changes in a radial fashion with the highest index being centrally located and decreasing as the perimeter is approached.
Such structures recently have been in great demand for use as lenses for coupling with optical fibre. Several methods are known for the making of such structures.
These 3o methods include ion-exchange in glass, "molecular stuffing" of porous glass and a phase separation and partial leaching of glass. These methods have been reviewed by Doc. No. 10-134 CA Patent Mukherjee in Gradient Index Lens Fabrication Processes: A Review, in Topical Meeting on Gradient-Index Optical Imaging Systems, Optical Society of America ( 1981 ).
Each of those methods has several drawbacks. The ion exchange process is slow because of low diffusion rates and has not been practical to make lenses of greater than 10 mm aperture. The molecular stuffing process can make larger lenses with low index gradients;
moreover, the gradients are not always uniform. The phase separation/leaching method may produce lenses of 10 mm aperture but they will have low index gradients which are not always uniform.
to In his 1981 review, Mukherjee also described a method of diffusion of inorganic oxides in a gel monolith. He proposed a sol-gel system whereby a gel monolith is preformed with modifying ion specie which subsequently diffuse out and are replaced with another modifying ion specie of equal valence. The resulting structure would have a large size, with a high index gradient which is uniform. Mukherjee previously applied this method to the making of non-gradient index glass. See Mukherjee, Sol-Gel Processes in Glass Science and Technology, J. Non-Cryst. Solids, 42:477 (1980); and Brinker and Mukherjee, Conversion of Monolithic Gels to Glasses in a Multicomponent Silicate Glass System, J. Mat. Sci., 16:1980 ( 1981 ).
Kurosaki (see U.S. Pat. No. 4,436,542) has attempted to practice a method for the manufacture of gradient index glass. Kurosaki employs silicic acid in a thallium, rubidium or cesium nitrate solution which is neutralized with acid. This forms a colloidal matrix which may be poured into a mold to produce a gel. The gel then is immersed in a leachate solution of water, acid, alkali metal salt, ketone or alcohol. This results in a gradient of Rb, Cs or Tl which decreases radially outwardly while the leachate ion gradient decreases radially inwardly. The gel then is dryed and sintered to form transparent glass. The structures produced are disclosed to have a diameter of 13 mm.
The process taught is essentially a leaching process and not one of ion inter-diffusion.

Doc. No. 10-134 CA Patent Commercially available GRIN lenses lenses are produced under the trade name "SELFOC"; the mark is registered in Japan and owned by the Nippon Sheet and Glass Co. Ltd. At a location along the lens, indicated as 0.25 pitch, the input beam becomes collimated. At the 0.5 pitch location midway between the endfaces of the lens 10, the input beam becomes inverted. This phenomenon is further demonstrated in FIG. 2 .
Although commercially available GRIN lenses adequately perform their intended function, such lenses used in the optoelectronic component industry are generally have a diameter of approximately 350 Vim. These lenses are crude and don't provide adequate coupling for use with single mode fibre. Some GRIN lenses have a smaller diameter of 250 ~m and are used as imaging lenses in photocopiers. However, there are several drawbacks to utilizing a lens having such a large diameter. Standard GRIN
lenses for use with optical fibre have an outer diameter of approximately 1.8mm and provide a collimated beam having a diameter of about 300 Vim. Consequently components coupled with the lens, for receiving the collimated beam must be of a dimension, i.e.
greater than the beam diameter of 300 ~m to support or carry the beam propagating therethrough. Yet another drawback of coupling an optical fibre to a commercially available GRIN
lens, is that the fibre must be precisely positioned at a particular location at the end face of the GRIN lens. Misalignment of even a few ~m is generally unacceptable, since the core of the fibre to which the light must be coupled efficiently through the GRIN lens has a 2o diameter of approximately 10 pm.
It is an object of this invention to provide a fibre GRIN collimating lens having approximately the same outer diameter as the optical fibre to which it is being coupled.
It is an object of the invention, to utilize a graded index optical fibre that can receive and couple light efficiently and provide collimation of the received light.
Summary of the Invention 3o In accordance with the invention, there is provided, an optical coupling system for providing a collimated beam of light, comprising:

Doc. No. 10-134 CA Patent a length of single mode optical fibre having an end face for directly coupling in contact with an end of an optical fibre having a graded index core, the ratio of the diameter of core of the optical fibre having the graded index core to the diameter of the core of single mode fibre the being substantially about 6.5:1 or more.
In accordance with the invention, an optical connector is provided comprising two sleeves and means for containing and holding the two sleeves in a secured relationship so that light can be coupled from one sleeve to the other, at least one of the sleeves having contained within a first length of single mode optical fibre having a first outer diameter and an end face in direct contact with an end face of a second length of optical fibre having a graded index core, the second length of optical fibre having a same outer diameter as the first length of optical fibre, the ratio of the diameter of the core of second length of optical fibre to the diameter of the core of the first optical fibre the being substantially about 6.5:1 or greater, wherein the second optical fibre is of a length that will provide collimation of light exiting the end face of the second optical fibre when launched into the second optical fibre via the first optical fibre.
In accordance with the invention, there is provided an optical coupling system comprising a block of material having a plurality of bores extending therethrough each 2o bore being sized to accommodate and tightly accept a length of single mode optical first fibre having a first outer diameter and having an end optically coupled with an end face of a second optical fibre lens having a graded index region about a longitudinal optical axis of the lens, the second optical fibre lens having a same outer diameter as the first optical fibre, the ratio of the diameter of the graded index region of second optical fibre lens to the diameter of the core of the first optical fibre being substantially about 6.5:1 or greater, wherein the second optical fibre lens is of a length that will provide collimation of light exiting an output end face of the second optical fibre when launched into the second optical fibre via the first optical fibre.

Doc. No. 10-134 CA Patent Brief Description of the Drawings Exemplary embodiments of the invention will now be described in conjunction with the drawings in which:
Fig. 1 is a cross section of a length of single mode optical fibre in accordance with the prior art.
Fig. 2 is a prior art cross section of the single mode optical fibre shown in Fig. 1 coupled to a GRIN lens;
Fig. 3 is a prior art cross sectional view of a length of single mode fibre having a ball lens 1 o at an end face;
Fig. 4 is a cross section of a length of single mode optical fibre coupled with a length of graded index optical fibre having a core diameter of approximately 50 Vim;
Fig. 5 is a cross section of a length of single mode optical fibre coupled with a length of graded index optical fibre having a core diameter of greater than 50 Vim;
Fig. 6 is a cross section of a length of single mode optical fibre having a thermally expanded core at an end face thereof coupled with a length of graded index optical fibre having a core diameter of greater than 50 Vim;
Figs. 7a to 7c are side cross sectional views of a sleeve for using in an optical connector having fibre and fibre lenses contained within a bore;
2o Fig. 7d is a side view of an optical connector arrangement having an optical element disposed between connector ends;
Figs. 8a and 8b are cross sectional views of a ribbon connector block having fibres and lenses contained therein; and, Figs. 9a and 9b illustrate yet other embodiments wherein bore are not completely separated, and wherein a v-grooved surface accommodates the fibres and fibre lenses.

Doc. No. 10-134 CA Patent Detailed Description Turning now to prior art Fig. 1, a length of single mode optical fibre is shown having a core l Ob having a diameter of approximately 10 ~tm and a cladding l0a surrounding the core having a diameter of approximately 125 Vim. When light launched into an end of a single mode fibre exits the fibre, it non-collimated and in fact its rays are highly divergent as shown in the figure. It is not possible to polish the end of the single mode fibre to obtain a collimated beam. Commonly, in the field of optoelectronics, a GRIN
lens 16 is to utilized and is directly coupled to the end of a single mode fibre so that the beam exiting the lens 16 as is shown in Fig. 2 is collimated. It is preferred to have a collimated beam passing through optical components in optical isolators, circulators, dichroic filters, etalons, etc. The term directly coupled used within this specification is to be taken as meaning that the fibre and fibre lens are either in direct contact or are directly coupled via an adhesive which may be present between the fibre and fibre lens. Hence for direct coupling to take place, the fibre end and the fibre lens end, need not be touching one another. In a preferred embodiment, however, they do make contact with one another.
Fig. 3 illustrates a less than preferred solution to having a highly divergent beam exiting an end face of a single mode optical fibre being optically coupled to other components. A
ball lens 18 is provided at an end face of the single mode optical fibre for focusing a beam that would otherwise be divergent. There are several drawbacks to this solution.
Firstly the beam is not truly collimated, second the end of the ball must be polished accurately and control of this process is difficult, and third, the round end of the ball lens in not preferred for coupling with planar end faces of other optical components.
Referring now to Fig. 4, a first length of single mode optical fibre 40 is coupled with a second length of multi-mode graded index optical fibre 42. The single mode optical fibre has a core diameter of approximately 10 ~m and the core diameter of the second length of standard graded index fibre typically used for high speed data transmission is approximately 50 ~m or less.

Doc. No. 10-134 CA Patent Providing a collimated beam of light by coupling the single mode fibre 40 with a preferred length of graded index high speed transmission fibre 42, and ensuring that the length of the fibre 42 is substantially about a quarter pitch, provides a collimated output beam that is somewhat lossy, i.e. wherein the beam does not completely conform to a desired shape and wherein some light leakage occurs through the cladding.
Since the beam exiting the single mode optical fibre 40 is highly divergent upon exiting the fibre 40 a lens, in this instance a fibre lens is required which has an acceptance angle that will shape the incoming light into a collimated beam without substantial loss.
Hence, it is 1o preferred to have a gradient core diameter of 65 ~m of more as shown in Fig. 5. The region surrounding the 65 ~m of more diameter may constitute a cladding region however, this may not be necessary or preferred.
Another embodiment in accordance with the invention which provides a beam of a 15 suitable shape and which provides a preferred coupling of light is shown in Fig. 6. Here a first optical fibre having a thermally expanded core (TEC) end is coupled with a second fibre having a core having a gradient index of refraction wherein the core diameter of the second fibre is approximately 60 ~m or more. However, the necessity of using a diameter of 60 ~m or more may be obviated when the end of the single mode fibre has an 2o expanded core, expanded by thermal or other means. By providing an expanded core single mode fibre, having a smaller numerical aperture the refractive index gradient change within fibre lens may be more gradual, and as well the tolerance on the length of the lens can be relaxed. Coupling a standard single mode fibre that does not have an expanded core end, to a standard SOpm core multimode fibre will result in too much loss 25 and poor coupling, and therefore the core diameter of the graded index regions should be at least 65p.m or alternatively, the mode field of the end of the single mode fibre should be expanded. Of course other methods of expanding the core of the first fibre can be employed. One of most advantageous aspects of this invention is that by coupling an optical fibre with an optical fibre lens either by using fusion or by affixing the fibres with 3o an adhesive, the beams propagating from one fibre to the other are essentially automatically centred, as opposed to affixing an optical fibre to a standard GRIN lens Doc. No. 10-134 CA Patent having a diameter of approximately 1800 ~m where aligning the fibre with the lens is not without difficulties.
Turning now to Figs. 7a through 7c, three sleeves are shown, for example as for use as an optical connector ferrule, wherein in the sleeve in Fig. 7a a first single mode optical fibre 73 is directly coupled via an epoxy adhesive coupling 75 with an optical fibre lens 71 in accordance with the teachings of this invention, for collimating light at an output end of the sleeve 70. In Fig. 7b single mode optical fibre 73 and the optical fibre lens 71 are shown fused together at 76 prior to be placed in the sleeve 71; after placement in the to sleeve 71, the collimating end of the optical fibre lens 71 and the sleeve can be polished to a desired angle and length. By providing a pair of mating connector sleeves or ferrules as described, coupling is enhanced most notably where an otpical component is to be placed between the collimating connector ends, i.e. at the end face of 71.
Each length of graded index fibre lens 71 is of an index and length to provide a quarter pitch lens for collimating a beam of light exiting its end. Providing a coupler or connector that collimates light exiting its end has numerous advantages. For example in Fig.
7d an optical connector is shown having a pair of sleeves as described in one of Figs. 7a, 7b, or 7d, and wherein an optical component 79, for example an optical isolator, is disposed within the connector housing. This is highly advantageous; here a small collimated beam 2o is provided to a miniaturized isolator. Alignment is passive and demanding, time consuming centering of fibre with lens with component is obviated.
Active alignment can also be obviated in the optical ribbon connector shown in Fig. 8a, wherein a plurality of ribbon fibres 73 are optically coupled with optical fibre collimating lenses 71 having a graded index profile as described heretofore. In Fig. 8a, for example, eight optical fibres 73 forming a ribbon are disposed within a block of material 82, having precisely defined bores therein for accommodating 125 pm single mode fibres. In one embodiment, the fibres are fused with optical fibre lenses prior to being placed within openings within the block 82. After the fibres and lenses are inserted the end is polished to a desired length. In yet another embodiment the fibres and lenses are separated inserted from opposite respective ends. Since the block is precisely Doc. No. 10-134 CA Patent manufactured having a precise gap 86 between the fibre lens and single mode fibre, and having a predetermined length for accommodating the fibre lens, collimation of light exiting the lens that has been launched into the single mode fibre is achieved.
of the block without active alignment.
Referring now to Fig. 8b, a block is shown having a bores extending therethrough, each bore has a portion for accommodating an optical fibre 73 and portion having a same diameter for tightly receiving an optical fibre lwns 71 having an outer diameter of 125~m; and a small inner bore 86 joins the two larger bores to allow light to pass to therethrough. Bores 88a, 88b, and 88c are shown in the figure, absent fibre 73 or fibre lens 71.
Figs. 9a another embodiment wherein adjacent bores are not completely separated. In Fig. 9a three optical fibres 71 are shown inserted into block 92. Not shown in the figure are fibre lenses 71 coupled with the fibres 73 inside the block 92.
Turning now to Fig. 9b, a V-grooved slab is shown accommodating an array of ribbon fibres each coupled to an optical fibre lens of a same outer diameter.
Depending upon the depth of the grooves, the slab can conveniently be covered by a flat or complementary 2o slab to contain the fibres and lenses.
Of course numerous other embodiments may be envisioned without departing from the spirit and scope of the invention.

Claims (20)

1. An optical coupling system for providing an output beam of light which is collimated, comprising:
a length of single mode optical first fibre having a first outer diameter and having an end optically coupled with an end face of a second optical fibre lens having a graded index region about a longitudinal optical axis of the lens, the second optical fibre lens having a same outer diameter as the first optical fibre, the ratio of the diameter of the graded index region of second optical fibre lens to the diameter of the core of the first optical fibre being substantially about 6.5:1 or greater, wherein the second optical fibre lens is of a length that will provide collimation of light exiting an output end face of the second optical fibre lens when launched into the second optical fibre lens via the first optical fibre.

2. An optical coupling system as defined in claim 1 wherein the graded index region is a core of the second optical fibre lens.

3. An optical coupling system as defined in claim 2, wherein the graded index core of the second optical fibre lens has a gradient which varies radially from the optical axis of the optical fibre lens, and wherein the first optical fibre end is directly coupled with the end of the second optical fibre lens.

4. An optical coupling system as defined in claim 3 wherein the end of the single mode optical fibre is fused with the end of the second optical fibre lens.

5. An optical coupling system as defined in claim 4, wherein an end face of the second optical fibre lens is polished to provide a length suitable for collimating light exiting said end face thereof.

6. An optical coupling system as defined in claim 4, wherein the diameter of the graded index core is more than 65 µm.

7. An optical coupling system as defined in claim 6 wherein the length of the second optical fibre lens is less than 1 mm.

8. An optical coupling system as defined in claim 7, wherein the second optical fibre lens is absent a cladding layer.

9. An optical coupling system as defined in claim 6 wherein an end face of the first optical fibre has an expanded core having a diameter substantially larger than the average core diameter of the first optical fibre.

10. An optical fibre coupling system for as defined in claim 1, wherein the first optical fibre and the second optical fibre lens are contained within a sleeve or a grooved holder.

11. An optical fibre coupling system as defined in claim 10, wherein the end face of the second optical fibre lens is at an end face of the sleeve or grooved holder, and the second optical fibre lens and end face of the sleeve or grooved holderare polished to a desired length.

12. An optical fibre coupling system as defined in claim 11 wherein the optical fibre and the optical fibre lens are optically coupled inside the sleeve form a connector end, for coupling with another optical connector end.

13. An optical fibre coupling system as defined in claim 12, wherein light exiting one end of the optical connector end is substantially collimated.

14. An optical fibre coupling system as defined in claim 13, wherein the connector end is spaced apart from a similar connector end when coupled together, the space for accommodating another optical component.

15. An optical connector comprising two sleeves and means for containing and holding the two sleeves in a secured relationship so that light can be coupled from one sleeve to the other, at each of the sleeves having contained within a first length of single mode optical fibre having a first outer diameter and an end face in direct contact with an end face of a second length of optical fibre having a graded index core, the second length of optical fibre having a same outer diameter as the first length of optical fibre, the ratio of the diameter of the core of second length of optical fibre to the diameter of the core of the first optical fibre the being substantially about 5:1 or greater, wherein the second optical fibre is of a length that will provide collimation of light exiting the end face of the second optical fibre when launched into the second optical fibre via the first optical fibre.

16. An optical connector as defined in claim 15, wherein the second optical fibre contained within each sleeve is a substantially quarter pitch collimating fibre lens.

17. An optical connector as defined in claim 16 wherein the core at an end of the first length of optical fibre in at least one of the sleeves is expanded and has a greater diameter than the average diameter of the core of the first length of optical fibre.

18. An optical connector as defined in claim 16 wherein the core of the second optical fibre has a graded index region having a diameter of at least 65 µm.

19. An optical connector as defined in claim 15 wherein an end faces in direct contact are fused together.

20. An optical coupling system comprising a block of material having a plurality of bores extending therethrough or grooved extending along a surface thereof, each bore being sized to accommodate and tightly accept a length of single mode optical first fibre having a first outer diameter and having an end optically coupled with an end face of a second optical fibre lens having a graded index region about a longitudinal optical axis of the lens, the second optical fibre lens having a same outer diameter as the first optical fibre, the ratio of the diameter of the graded index region of second optical fibre lens to the diameter of the core of the first optical fibre being substantially about 6.5:1 or greater, wherein the second optical fibre lens is of a length that will provide collimation of light exiting an output end face of the second optical fibre when launched into the second optical fibre via the first optical fibre.