AU649621B2

AU649621B2 - Opto-electronic component having a positioned optical fiber associated therewith

Info

Publication number: AU649621B2
Application number: AU61449/90A
Authority: AU
Inventors: Melvin Harry Johnson; Jerald Dana Lee; Frank Marsden Willis
Original assignee: EI Du Pont de Nemours and Co
Current assignee: EIDP Inc
Priority date: 1989-08-02
Filing date: 1990-07-17
Publication date: 1994-06-02
Anticipated expiration: 2010-07-17
Also published as: IE902781A1; JPH04507304A; KR920704162A; EP0485426A4; WO1991002271A1; AU6144990A; CA2022475A1; EP0485426A1

Description

WO 91/02271 1 OPTO-ELECTRONIC COMPONEN OPTICAL FIBER ASSOCI/ CROSS REFERENCE TO REL Subject matter disclosed herei A 0 in copending pacontemporaneously herewith, titled The Center Of An Optical Fiber Al Reference Axis".

pCr/US9/039 3

TITLE

T HAVING A POSITIONED ATED THEREWITH ATED APPLICATION in is disclosed and claimed Zf filed "Apparatus For Positioning ong a Predetermined BACKGROUND OF THE INVENTION ~1I Field Of The Invention The present invention relates to an opto-electronic component having a positioning appa' 'us for positioning the center of an optical fiber along a predetermined reference axis of an opto-electronic device independently of variations in the outside diameter of the fiber.

Description Of The Prior Art Devices are known for positioning an optical fiber so that the axis of the fiber is positioned with respect to a reference axis. A typical expedient used in such devices is a generally V-shaped groove that is formed in a substrate material and which serves as a cradle to accept the fiber being positioned. Representative of such devices is thai shown in United States Patent 4,756,591 (Fischer et wherein a V-groove is formed in a silicon substrate and an elastomeric member is biased against the fiber to hold it in the groove. The groove may be s..pped to k' o 1 WO 91/02271 PCT/US90/03903 2 provide a deeper groove segment to hold the jacket of the fiber within the device.

United States Patent 4,756,591 (Sheem) discloses a grooved silicon substrate having a pair of intersecting Vgrooves therein. A fiber to be positioned is disposed in one of the grooves while a shim is disposed in the other of the grooves. The shim may take the form of a tapered or an eccentric fiber, which when respectively slid or rotated under the first fiber serves to lift the same to bring the axis thereof into alignment with a reference axis. A cover may be positioned above the substrate to assist in clamping the first fiber into position.

United States Patent 4,802,727 (Stanley) also discloses a i positioning arrangement for optical components and Swaveguides which utilizes a V-grooved structure. United i States Patent 4,826,272 (Pimpinella et al.) and United States Patent 4,830,450 (Connell et al.) discloses arrangements for S 20 positioning an optical fiber that utilize members having frustoconical apertures therethrough.

It is believed that single crystalline silicon is the material of choice of the devices above mentioned because of the proclivity of crystalline silicon to be etched along precise crystallographic planes, thus forming precise grooves or structural features by photolithographic microfabrication techniques. Etclants exist that act upon a selected crystallographic plane to a differential degree than upon an adjacent plane, permitting the needed precise control. Vgrooves, in particular, can be etched to a controlled width and truncated depth. Under some conditions V-grooves may be etched in a self-limiting operation. The photolithographic microfabrication process is generally described by Brodie and i c 1 i "1 WO 91/02271 PCT/US90/03903 3 Muray, "The Physics of Microfabrication", Plenum Publishing, New York (1982).

Optical fibers include an inner core having a predetermined index of refraction surrounded by a cladding layer of a lower index. The inner core is the medium in which the optical energy is guided, while the cladding layer defines the index boundary with the core. The outer diameter of the fiber may vary in dimension about a predetermined nominal dimension. It has been seen, for example, that two nominally identical fibers from the same manufacturer may vary in ioutside diametrical dimension by as much as plus or minus four micrometers. This fiber to fiber variation in outer diameter makes difficult the accurate positioning of the axis of the core of a fiber with respect to a predetermined reference axis.

In view of the foregoing it is believed advantageous to make use of the ability of microfabrication techniques to form accurate structures, channels and/or surfaces in a crystalline material to construct a positioning apparatus that will fI accurately position the center of the fiber, or of any other elongated generally cylindrical member having small jj dimensions (such as capillary tubing), with respect to a predetermined reference axis. Moreover, it is believed advantageous to provide a positioning apparatus that cjnsistently aligns the predetermined point on the fiber or other cylindrical member with the reference axis without requiring great technical skill, expensive apparatus, and extensive alignment procedures.

L L_ WO91/02271 PCT/US90/03903 4 SUMMARY OF THE INVENTION The present invention relates to a relates to an optoi 5 electronic component having a positioning apparatus for positioning the center of an optical fiber along a predetermined reference axis of an opto-electronic device independently of N variations in the outside diameter of tile fiber. The optoelectronic device can take the form of a solid state laser or a solid state light responsive diode such as a photodiode. The opto-electronic device can be an edge active or a surface active device. The optical fiber can be a single-mode or a multi-mode Sfiber.

The positioning apparatus includes a first and a second arm, each of which has at least a first and a second sidewall that cooperate to define a groove therein. The groove in each arm is preferably a converging groove so that when the arms are arranged in superimposed relationship the converging grooves cooperate to define a funnel-like channel over at least a predetermined portion of its length. The channel has an inlet V end and an outlet end and a reference axis extending therethrough. A fiber introduced into the inlet end of the V channel with its axis spaced from the reference axis is displacable by contact with at least one of the sidewalls on one of the arms to place a predetermined point on an end face of the member into alignment with the reference axis where it is there held by contact with the first and second arms. To guide the fiber toward the inlet end of the channel each of the first and the second arms includes a trough therein, each trough Sbeing disposed on an arm a predetermined distance behind the groove in that arm, so that in the closed position the troughs cooperate to define a guideway.

I

wo 15 S 20 i{ 91/02271 PC/US9/03903 1 The arms having the converging grooves therein may, as is preferred, be movable from a first, closed, position to a second, centering, position. The superimposed arms are, in this instance, moujnted cantilevered fashion, to a foundation. Means is provided foi biasing each of the arms with a substantially equal and oppositely directed biasing force toward the first position. In the preferred implementation the biasing means comprises a reduced thickness portion in each of the first and the second arms, the reduced thickness portion defining a flexure in each arm which, when each arm is deflected by contact with the cylindrical member, generates a force on each arm to bias each arm toward the closed position.

It should be understood that so long as tile arms are movable and biased toward the closed position, it is not required that the grooves formed therein are converging grooves. Accordingly, other positioning apparatus in which the arms are movable but in which the grooves in each of the arms have a form other than a converging groove are to be construed as lying within the contemplation of the invention.

Succinctly stated, the present invention encompasses any positioning apparatus having arms that are movable whether the groove in each arm take the formn of a converging groove or a groove of an alternate form. Alternately, the present invention also encompasses any positioning apparatus in which the groove in each arm is converging in form, whether the arms are movable or fixed with respect to each other.

In whatever embodiment realized, it is preferred that the positioning apparatus be fabricated from a crystalline material, L~ l: such as single crystal silicon, using microfabrication techniques.

Each structural element of the positioning apparatus (viz., each of the arms and each foundation) is fabricated in mass on a wafer of silicon. The finished wafer are aligned, superimposed, and bonded, and each of the resulting positioning apparatus 1|

I

S91/02271 PC/US90/03903 WO 91/02271 7 BRIEF DESCRIPTION OF THE DRAWINGS The invention will be more fully understood from the following detailed description thereof, taken in connection with the accompanying drawings, which form a part of this application, and in which: Figure 1 is a perspective, exploded view of a positioning apparatus in accordance with the preferredbedIrrbi4Tinent of the present invention for positioning the center point on the end face of an optical fiber with respect to a predetermined reference axis; Figure 2 is a perspective view of the positioning apparatus of Figure 1 in the fully assembled condition; Figure 3 is a front elevation view of the assembled i? positioning apparatus of Figures 1 and 2, taken along view lines 3-3 in Figure 2; Figure 4 is a sectional view, in elevation, of the assembled positioning apparatus of Figure 2, taken along section lines 4-4 in that Figure illustrating the truncated V-groove therein; Figure 4A is a view generally similar to Figure 4 in which a .full V-groove is formed in the positioning apparatus while Figure 4B is a view generally similar to Figure 4 in which both a full V-groove and a truncated V-groove are formed; Figure 5 is a plan view one of the arms of the positioning apparatus of Figure 1 illustrating the relationships of the axes of the groove and tile guideway therein; 2 J~ PCT/US90/03903 WO 91/02271 8 Figures 6A and 6B, 7A and 7B, and 8A and 8B are diagrammatic elevational and end views of the action of the clips disposed on the arms of the positioning apparatus shown.

in Figures 1 and 2 in response to the introduction of a fiber thereinto; Figures 9 and 10 are exploded and assembled perspective views, generally similar to Figures 1 and 2, of another alternate embodiment of a positioning device in accordance with the present invention in which the arms have nonconverging grooves therein and in which the arms are articulably movable with respect to each other along one axis only; Figures 11 are 12 are sectional views taken along section lines 11-11 and 12-12 in Figure Figures 13 and 14 are exploded and assembled perspective views, generally similar to Figures 1 and 2, of another alternate embodiment of a positioning device in accordance with the present invention in which only one of the arms has a nonconverging groove therein and in which both of the arms are articulably movable with respect to each other; Figures 15 are 16 are sectional views taken along section lines 15-15 and 16-16 in Figure 14; Figures 17 and 18 are exploded and assembled perspective views, generally similar to Figures 1 and 2, of an alternate embodiment of a positioning device in accordance 30 with the present invention in which the arms have converging grooves tlerein and in which the arms are fixed with respect to each other; Figure 19 is an end view taken along view lines 19-19 in Figure 18; 1

I

i 1 i.

":I

i pC/US90/03903 WO 91/02271 Figure is a side sectional view, taken along view lines 20-20 in Fig,.,e 18, illustrating the position of the fiber within the channel of the a positioning apparatus in accordance with the alternate embodiment of the invention shown therein; Figure 21 is an exploded isometric view of a pair of positioning apparatus as shown in Figure 1 used to form a fiber-to-fiber connector in accordance with the present 10 invention while Figure 22 is an isometric view of the fully assembled connector of Figure 21; Figures 23 and 24 are, respectively, a top view in section and a side elevation section view of a pair of positioning apparatus in accordance the embodiment of the invention as shown in Figure 17 used to form a fiber-to-fiber connector in accordance with the present invention; Figures 25 and 26 are isometric views of a housing used for the fiber-to-fiber connector shown in Figures 21 and 22 in the open and in the partially closed positions, respectively, while Figure 27 is a section view of the housing of Figure 25 in the fully closed position taken along section lines 27-27 of Figure 26; Figure 28 is a section view generally similar to Figure 27 of a housing used for the fiber-to-fiber connector shown in Figure 24; Figure 29 is a isometric view of an alternate housing for a fiber-to-fiber connector formed of a pair of positioning apparatus; Figures 30 and 31 are isometric exploded and assembled views, respectively, illustrating the use of a positioning

I!

CALLINAN LAWRIE

.^W

i

J

I

re pCU/US90/03903 WO 91/02271 apparatus in accordance with the present invention to position an optical fiber with respect to the axis of an edge emitting active device, in which the device is surface mounted; Figure 31A is a side elevational view generally similar to ,Figure 31 showing a positioning apparatus in accordance with the present invention positioning a lens with respect to an opto-electronic component; Figures 32 and 33 are isometric exploded and assembled views, respectively, generally similar to Figures 30 and 31, illustrating (lie use of a positioning apparatus in in accordance with the present invention to position an optical fiber with respect to the axis of a device having active surface device, in which the device is edge mounted; b 7 30 Figure 34 is a perspective view of a wafer used used to fabricate a plurality of arms or foundations used in a positioning apparatus in accordance with the present invention; Figure 35 is a perspective view of a mask used in the photolithographic process forming a plurality of arms or foundations for a positioning apparatus in accordance with the present invention; Figure 36 is an enlarged view of a portion of the mask used for creating a plurality of arms on the wafer 34; Figures 37A through 37E are schematic representations of the process steps effected during fabrication of the wafer; Figure 38 is is an enlarged view of a portion of the mask used for creating solder masks on the wafer;

I

lI 1 This form may be completed and filed after the filing of a patent application but the form must not be :igned until after it has been comlelY filled in as indicated bythe marinal notes, The place and date of si3.nin must be filled in. ompan taps or seals should not be sed.

I-,

I

30 I ^I PCT/US90/0 3 9 0 3 WO 91/02271 11 Figure 39 is an enlarged view of a portion of the mask used for creating foundations on the wafer; Figures 40A through 40D are schematic representations of the steps used to form a plurality of fiber-to-fiber connectors fronm superimposed wafers having the arms and foundations thereon; Figure 41 is a definitional drawing illustrating the characteristics of a converging groove as that term is used in this application; and Figures 42A through 42F are end views showing alternate arrangements of movable arms each holding a cylindrical member along at least three contact points in accordance with the present invention.

DETAILED DESCRIPTION OF THE INVENTION

I

Throughout the following detailed description similar reference numerals refer to similar elements in all Figures of the drawings.

With reference to Figures 1 and 2 a positioning apparatus generally by reference character 20 in accordance with the present invention is shown in an exploded and in a fully assembled condition. As will be developed herein, in the preferred instance the positioning apparatus 20 is microfabricated from single crystal silicon or another differentially etchable single crystal material. These materials are preferred because they permit the accurate formation of the structural features of the apparatus 20 using the process of differential etching.

i- Now pcT/US90/03 90 3 WO 91/02271 The positioning apparatus 20 is useful in accurately positioning a predetermined point on the end face of a cylindrical member, typically a point on the central axis of the member, along a reference axis. Typically this reference axis is itself collinearly aligned with respect to another axis, as an operative axis of a device operably associable with the cylindrical member. Throughout this application the description of the cylindrical member is cast in terms of an optical fiber, but it is to be understood that the present invention may be effectively utilized with any other member having the form of small diameter cylindrical object. By way of example and not limitation, the positioning apparatus invention may be used to position a point on the end face of a length of microtubing or capillary tubing. By small diameter it is generally meant less than 0.04 inch (one millimeter), but usually less than 0.020 inch. Moreover, it should be further understood that the term cylindrical is not to be strictly limited to an object having a completely circular outer configuration, but would apply to any object whose outer contour is symmetrical to its central axis. Thus, again by way of further example and not limitation, the positioning apparatus of the present invention may be used to position a point on the end face of a polygonal shaped member or an elliptical member.

t As noted, the cylindrical member preferably takes the form of an optical fiber. The positioning apparatus of the present invention is particularly adapted to place a predetermined point P on the end face E of an optical fiber F along a predetermined reference axis R. In practice the point is the geometric center and lies on the axis A of the core C (Figures 6A ard 6B) of the fiber F. The core C is itself surrounded by an outer cladding layer L. A jacket I is provided about the cladding layer L but is stripped from the fiber F prior to the insertion of the fiber into the positioning apparatus 20. The jacket may comprise more than one layer.

WOW-

W0 91/02271 13 pTU9/ 3

O

i As discussed previously, the dimension of the outer diameter D of the cladaing layer L of the fiber F may vary from fiber to fiber. Typically this diametrical variation from fiber to fiber is on the order of three micrometers. This situation makes difficult the positioning of the point P along the reference axis R using the positioning devices of the prior art. The fiber may be a single-mode or a multi-mode fiber. As will be seen from the following the positioning capability of the positioning apparatus 20 is especially adapted for positioning the point P of a single mode fiber within the precise tolerances required to effectively couple light emanating from the single mode fiber into another fiber or to position the fiber with respect to an

I

ii optical device.

With reference to Figures 1 and 2 it is seen that the positioning apparatus 20 includes a first and a second arm 22A, 221, respectively. Preferably, each of the arms 22A, 22B is identically formed in a manner to be discussed, so the structural details of only one of the arms, e. the arm 22A, will be discussed. It should be apparent, however, that each structural detail of the arm 22A finds a counterpart in the other arm 22B. Accordingly, corresponding reference numerals with the appropriate alphabet suffix will denote corresponding structural details on the arm 22B. If the arms are no, substantially identical (as, for example, in the embodiments of Figures 13 through 16 and Figure 42) adjustments must be made to provide the requisite biasing forces to maintain the point P on the reference axis R The arm 22A includes a base portion 24A having a first major surface 26A and a second, opposed, major surface 28A.

The base portion 24A extends along the full length of the arm 22A and the dimension of the central region 25A of the base portion 24A defines the basic dimension of the arm 22A. A clip generally indicated by the reference character 30A is lI (57) Abstract of either the edge or the surface active type mounted An opto-electronic component includes an opto-electron device (20) of either the edge or the surface active type mounted on a pedestal. A fiber is positioned by a positioning device 1 See back of page 1 1.

ri pCT/US90/0 3 9 0 3 WO 91/0227i defined at a first end of the arm 22A. The clip 30A is formed in a relatively thicker abutment portion 32A that lies on the first surface 26A of the arm 22A. The abutment 32A has a planar surface 34A thereon that preferably lies parallel to the first major surface 26A. To provide some feeling for the physical dimensions involved, the arm 22A has an overall length dimension on the order of twenty eight hundred (2800) micrometers and a width on the order of three hundred fifty (350) micrometers. In the central region 25A the arm 22A has a thickness dimension on the order of fifty (50) micrometers, while the remaining portion of the arm 22A has a thickness dimension on the order of one hundred twenty five (125) micrometers.

i I 15 As may be better seen with reference to Figures 3 and 4 a generally converging V-shaped groove 36A is defined in the abutment 32A of the clip 30A by generally planar first and second sidewalls 38A, 40A, respectively, and the forward end region of the first surface 26A of the base 24A. The sidewall 38A has an upper edge 39A (Figure 1) thereon while the sidewall 40A has an upper edge 41A thereon. It should be understood that the term "planar" is meant to encompass a surface formed in a single crystal material by etching in which microscopic steps are of necessity produced owing to the lattice structure of the crystal.

With reference now to the definitional drawing of Figure 41, the meaning of the term "converging" when applied to a groove (using the reference characters of Figures 1 and 2) may be made more clear. As used herein, a "converging" groove is a groove 36 defined from at least two planar sidewalls 38, and has an enlarged inlet end 42A and a narrower outlet end 43A. The respective upper edges 39, 41 of the sidewalls 38, of the groove 36 lie in a reference plane RP having a reference axis R lying therein. The planar surfaces 34 also lie in the i -i (tisclier et wherein a V-groove is formed in a silicon substrate and an elastomeric member is biased against the fiber to hold it in the groove. The groove may be s..,pped to W 1 pCT/US90/0390 3 WO 91/02271 reference plane RP. The reference axis R extends in the reference plane RP from the inlet end 4: to the the outlet end 43 of the groove 36. Each point on the reference axis R is spaced in the reference plane RP an equal distance from the respective upper edges 39, 41 of the sidewalls 38, 40. The distance between the upper edges of the sidewalls decreases from the inlet end 42 to the outlet end 43 of the groove 36.

The surfaces of the sidewalls 38, 40 are equally and oppositely inclined with respect to the reference plane at an angle A greater than zero and less than ninety degrees. The angle of inclination A is determined by the lattice structure of the crystal, and in the case of (100) silicon, is 54.74 degrees.

The projections of the sidewalls 38, 40 intersect in a line L that itself intersects the reference axis R forwardly past the outlet end 43 of the groove 36. The line L is inclined with respect to the reference plane RP at an angle b that is greater than zero degrees but less than ninety degrees. In the reference plane RP the upper edges 39, 41 of the sidewalls 38, 40 each converge toward the reference axis R at an angle C that is on the order of two and one-half to five degrees (2.5 to 5) degrees, and most preferably at about three degrees. The angle B is dependent upon the values of the angles A and C and typically the angle B lies in the range from about four to five degrees. As used herein a "fully funnel-like" channel is a channel that is defined by the cooperative association of at least two converging grooves. A "partially funnel-like" channel is a channel that is defined by one converging groove and a surface.

i From the foregoing it may be readily understood that a "uniform width" groove is one in which each point on the reference axis R is spaced in the reference plane RP a uniform distance from the edges 39, 41 of the sidewalls 38, 40 as one progresses from the inlet end 42 to the outlet end 43 of the i i~ grooves, in particular, can be etched to a controlled width and truncated depth. Under some conditions V-grooves may be etched in a self-limiting operation. The photolithographic microfabrication process is generally described by Brodie and PCT/US90/0396., WO 91/02271 16 groove 36. The sidewalls of a uniform width groove may be inclined with respect reference plane RP, or they may extend perpendicularly to it, as desired. A channel formed from one or two uniform width groove(s) is termed a "uniform width" channel. Such a channel may have a rectangular cross section in a plane perpendicular both to the reference plane and to the reference axis, assuming no inclination of the sidewalls of the groove.

A tapering groove is one in which the planar sidewalls are perpendicular to the reference plane but the distance in the reference plane between the reference axis and the edges of the sidewalls decreases as one progresses from the inlet to the outlet of the groove such that the extensions of the planar sidewalls intersect in a line that itself intersects perpendicularly with the reference axis.

In the preferred embodiment seen in Figures 3 and 4 the groove 36 is a converging groove, and more preferably, is a Vgroove truncated by the presence of a third sidewall defined by a portion of the major surface 26 of the arm 22 in which it is disposed. The truncated V-groove has the same depth throughout its length, when measured along a dimension line erected perpendicular to the surface 34A of the abutment 32A in a direction extending toward the major surface 26A.

It should be understood that the V-shape of the groove 36A may take alternate forms and remain within the contemplation of the invention. For example, as seen in Figure 30 4A, the groove 36A may be defined by only the first and second sidewalls 38A, 40A, respectively, in which event the groove 36A appears as a full V-shape throughout its length.

The apex 42A of the groove 36A thus appears throughout the full length of the groove 36A.

i'

I

2 5

S

3 0 r pCT/US9o/03 9 0 3 WO 91/02271 Figure 4B shows another alternative arrangement in which a truncated V-groove (defined by the first and second sidewalls 38A, 40A, respectively, and the portion of the major surface 26A) extends for some predetermined axial distance while a full V-groove (defined by the first and second sidewalls 38A, 40A, respectively) extends for some second predetermined distance. Thus, as seen in Figure 4B, when measured along a dimension line erected perpendicular to the surface 34A of the abutment 32A in a direction extending toward the major surface 26A the depth that the groove 36A extends into tle abutment 32A is greater at its inlet end 42 (as indicated by the dimension arrow 44A) than it is at its outlet end 43 (as indicated by the dimension arrow 46A).

The fully truncated V-groove shown in Figure 4 is preferred. For purposes of ease of manufacturability, as will be made clear herein, it is also preferred that the groove 36A does not converge throughout the full axial distance through the abutment 32A. Owing to the provision of tabs 48A, 48B (Figures 1 and 5) formed near the ends of the abutments 32A, 32B, the sidewalls 38A, 40A defining the groove 36A do not converge throughout the full length of the groove, but define a short uniform width portion just past the converging portion of tihe groove 36A. The overall axial length of the groove 36 (including both the converging and the uniform width portions) is on the order of three tenths of a millimeter, while the uniform width portion of the groove occupies an axial length of one tenth of a millimeter. As is believed best seen in Figure 5 the converging and nonconverging portions of the groove 36A have a common axis 50A associated therewith.

i Again with reference to Figure 2, an extended enlargement region 54A having a planar surface 56A lies on the base portion 24A of the arm 22A spaced a predetermined axial distance 58A behind the abutment 32A. The distance I c:i

I

I PCr/US90/03903 WO 91/02271 58A is on the order of one millimeter. The surface 56A is coplanar with the surface 34A. The enlargement 54A is provided with a nonconverging, uniform width, truncated Vshaped trough 60A defined by inclined planar sidewalls 62A, 64A, respectively, and by a portion of the major surface 26A of the base portion 24A near the second end thereof. In the embodiment shown in Figures 1 and 2 the trough 60A is uniform in depth along its axial length, as measured with respect to a dimension line erected perpendicular to the surface 56A toward the major surface 26A. The trough communicates with a converging lead-in 68A. If desired, the walls 62A, 64A may be inclined with respect to each other so that the trough 60A may be a full V-shape or a partial Vshape, similar to the situation illustrated in connection with Figures 4A and 4B for the groove 36A. Alternatively, the walls S62A, 64A defining the troughs 60A, 60B may be parallel or Sotherwise conveniently oriented with respect to each other. As is believed best seen in Figure 5 the trough 60A and the leadin 68A have a common axis 70A. Tile length of the trough and associated lead-in 68A is on the order of 1.59 millimeter.

i i Figure 5 is a plan view of one of the arms 22A. In the preferred implementation the axes 50A, 70A (respectively through the groove 36A and the trough/lead-in 60A/68A) are offset a predetermined distance 72 in the reference plane RP (the plane of Figure Preferably, the offset 72 is at least one-half the difference between the diameters of the anticipated largest and smallest fibers to be positioned. As will become clearer herein offsetting the axes 50A, 70A of the Sv 30 structures 36A, 60A/68A facilitates the centering action of the i:

I~

I:

positioning apparatus 20 by insuring that a fiber, as it is introduced into the apparatus 20, is biased to strike one of the sidewalls 38A, 40A forming the groove 36A (and analogously, the sidewalls 381, 40B forming the groove 361). This insures wall contact with the fiber at at least two spaced locations.

i- WO 1 PCT/US90!03903 WO 91/02271 19 However, the presence of the offset 72 necessitates additional manufacturing considerations, as will be discussed. It should be noted that the force resulting from biasing the fiber in the manner just discussed (or the force on the fiber due to gravity) is much i;maller in magnitude than the biasing force of the arms which serves to center the fiber on the reference axis.

In the assembled condition the arms 22A, 22B are disposed in superimposed relationship one above the other, 1u with the groove 36A, the trough 60A and the lead-in 68A on the one arm 22A registering with the corresponding groove 36B, trough 60B and lead-in 681 on the other arm 22B. The registered converging grooves 36A, 36B in the abutments 32A, 32B cooperate to define a generally fully funnel-shaped channel 92 having an input end 94 (Figure 4) and an output end 96 (Figures 4 and (Note that if the tabs 48 -are provided, the channel 92 so defined has a uniform width portion just preceding the outlet end 96 thereof.) The reference axis R extends centrally and axially through the channel 92. Preferably, the reference axis R lies on the intersection of the reference plane RP (which contains the conjoined surfaces 34A, 34B) with the plane containing the axes 50A, 50B of the converging grooves 36A, 36B.

The registered troughs 60 and lead-ins 68 cooperate to define a guideway 98 (Figure Similarly, the axis R' through the guideway 98 lies on the intersection of the plane containing the conjoined surfaces 56A, 56B of the enlargements 54A, 54B (which is the reference plane in the preferred case) with the plane containing the axes 70A, 70B (Figure 5) of the trough/lead-in 60A/68A, 601/681. The axes R and R' both lie in the reference plane RP (the plane of the surface" 34A, 34B, 56A, 561) although the axes R and R' are laterall offset with respect to each other in this reference plane by a pCT/US90/03903 WO 91/02271 predetermined offset distance 100. For a fiber the offset distance 100 is typically on the order of five micrometers.

The inlet end 94 of the fully funnel-like channel 92 (best seen in Figures 4 and 5) is sized to circumscribe and thereby to accommodate a fiber F whose cladding layer L (or outside surface) has the largest expected outer diameter dimension.

The outlet end 96 of the channel 92 (best seen in Figure 3) is sized to circumscribe and thereby to accommodate a fiber F whose cladding layer L (or outside surface) has a dimension somewhat smaller than the minimum expected outer diameter dimension of the fiber F. In practice, to position an optical fiber having a nominal outer diameter dimension of one hundred twenty five (125) micrometers, the largest expected outer diameter dimension is on the order of one hundred twenty nine (129) micrometers while the smallest -expected outer diameter dimension is on the order of one hundred twenty one (121) micrometers.

The dimension of each of the troughs 60A, 60B is such that the guideway 98 so formed by the registered troughs is sized to accommodate a fiber F whose cladding layer L hIas the largest expected outer diameter dimension. Despite its dimension with respect to the fiber, the guideway 98 assists in the insertion of a fiber into the positioning apparatus 20 and is advantageous in this regard.

8i: In the embodiment shown in Figures 1 through 5 the surfaces 34A, 34B on the respective arms 22A, 22B, 30 respectively, are, when in a first, closed, position, either in contact with each other or, if desired, within a predetermined close distance to each other. For optical fibers the predetermined close distance is typically on the order of five to twenty-five (25) micrometers. In this embodiment the planar surfaces 34A, 34B on thie abutments 32A, 32B of the apparatus ol figure 1 illustrating the relationships of the axes of the groove and the guideway therein; SPCT/US90/03903 WO 91/02271 21 clips 30A, 30B are not secured to each other and may move to a second, centering, position, as will be described. The planar surfaces 56A, 56B on the respective arms 22A, 22B are secured to each other by any convenient means of attachment, as by fusing or soldering. It should be understood that any other mechanical securing expedient may be used to attach or otherwise hold together the surfaces 56A, 56B to each other.

The positioning apparatus 20 further includes, in the preferred instance, a mounting foundation 74 (Figures 1 and 2).

The mounting foundation 74 is provided with a planar attachment surface 76 thereon. A step 78 in the mounting foundation 74 serves to space the attachment surface 76 a predetermined clearance distance 80 from a second surface 82.

The opposite major surface, the surface 28A, of the arm 22A is secured, as by fusing or soldering, to the planar attachment surface 76 on the foundation 74. Of course, it should be again understood that any alternative mechanical attachment expedient may b. used to attach or otherwise hold together the second major surface of the arm to the foundation 74.

Although the second surface 82 of the foundation is shown in the Figures as being generally planar in the preferred case, it should be understood that this surface 82 may take any desired configuration. As will be more fully appreciated herein, so long as the opposite surface 28A of the arm 22A affixed to the foundation 74 is, at least in the region of the clips spaced at least a predetermined clearance distance 30 from the second surface 82 (assuming the surface 82 is parallel to the surface 76), the movement of the clip on the arm 22A attached to the foundation (in tile drawings, the clip 30A) to be described will not be impeded.

I?

I,

Figure 19 is an end view taken along view lines 19-19 in Figure 18; pcT/USg0/03903 WO 91/02271 22 When assembled, the clips 30A, 30B disposed at the ends of the arms 22A, 221, respectively, are supported in a cantilevered fashion from tile conjoined enlargements 54A, 54B at the opposite ends of the arms. The arms 22A, 22B are rigid in x-z plane, as defined by the coordinate axes shown in Figure 1. Moreover, the relatively thin dimension of the central region 25A, 25B of the base portion 24A, 24B of the arms 22A, 22B axially intermediate the respective abutments 32A, 32B and the enlargements 54A, 54B acts as a flexure and permits each arm 22 to flex, springboard fashion, in the directions of the arrows 88 in the y-z plane. As used herein it should thus be appreciated that a flexure is a spring member that is rigid in one plane and is constrained to flex in tihe orthogonal plane.

It should further be appreciated that when a clip 30B is deflected in its corresponding respective direction 88A, 881, the resiliency of the thinner central region 25A, 25B of the base 24A, 241, acting as a flexure, defines means for biasing the clips 30A, 30B toward the first, closed, position. The biasing force acts on the clip 30A, 30B in a direction shown by the arrows 90A, 90B, counter to the direction of motion 88A, 88B of the arms. The biasing forces must be substantially equal and in opposite directions. In general, whatever the number of arms used in the positioning apparatus, the force on each arm passes through the reference axis and the sum of forces when in the centering position position substantially equals zero. Biasing means employing the thinner central region of the base 24 as a flexure (as shown in the Figures 1 to 4) is preferred, because when implemented in a single crystal material using a microfabrication technique precise control of 30 the biasing forces is able to be attained. Typically the bias force on each arm is on the order of five grams.

It should be understood that any other convenient mechanism may be used to define the means for biasing the Figures 30 and 31 are isometric exploded and assembled views, respectively, illustrating the use of a positioning PCT/US90/03903 WO 91/02271 23 arm- and the clips 30 thereon toward the closed position so long as the force on each arm passes ;hroughl tle reference axis and the sum of forces oni the arms when they are in the centering position is substantially equal to zero. Whatever form of biasing means is selected the bias force must increase with deflection of the arm.

-o-0-o- Having defined the structure of the positioning apparatus the operation thereof in positioning a point P on the end face E of an optical fiber F along a predetermined reference axis R may be readily understood in connection with Figures through 7.

In operation the fiber F is inserted into the positioning apparatus 20 in the direction of the arrow 102 (Figure 6A).

The lead-in portions 68A, 68B (Figure 1) cooperate to guide the fiber F into the guideway 98 (Figure 2) defined by the registered troughs 60A, 60B in thie enlargements 54A, 54B (Figure Because the axis R' of the passage 98 is offset from the axis R of the fully funnel shaped channel 92 the guideway 98 serves to guide the face E of the fiber F toward the inlet end 94 of the channel 92 at a predetermined azimuth with respect to the axis R.

As a result the end face E of the fiber F enters the channel 92 and is initially displaced through contact with at least one of the sidewalls 38A or 38B, 40A or 40B (or portions of the major surface 26A, 26B, if these are used to define the grooves 36A, 36B, as in Figure 4) on one of the clins 30A, respectively, to the extent necessary to place a predetermined point P on an send face E of the F'iber I into alignment with the reference axis R.

anl caiiugea view ol a portion 01 the mask used for creating solder masks on the wafer; 30 3 9 0 3 WO 91/02271 At some point on the path of axial insertion of the fiber F into the channel 92, as the end E of the fiber F moves toward the outlet end 96, the outer diameter of the cladding layer L of the fiber F exceeds the dimension of the channel 92. The arms 22A, 22B respond to a force in the directions 88A, 88B imposed thereon by the fiber F by moving against the biasing force from the first, closed, position, shown in Figures 7A, 78, toward a second, centering, position showing in Figures 8A, 8B. In the centering position the clips 30A, 30B open against the bias force acting in the directions 90A, 90B generated by the flexing of the arms 22A, 228, to separate the surfaces 34A, 34B thereon. However, this movement of the arms 22A, 22B from the first toward the second position maintains the point P on the end face E of the fiber F on the reference axis R. The end face E of the fiber F thus exits through the outlet end 96 of the fully funnel shaped channel 92 with the point P precisely aligned with within one micrometer of) the reference axis R, as is shown in Figures 8A, 8B. The fiber F is held in this position by contact with the sidewalls 38A, 38B, 40A, and If the tabs 48A, 48B are formed on the abutments 32A, 32B these tabs cooperate to define a passage of uniform width along its axial length that communicates with the outlet of the funnel-like channel. The fiber F passes through and emerges from such a conduit with the point P on the end face of the fiber still along the reference axis R.

It should be noted that the movement of the arms could be other than the flexing thereof as described heretofore. It therefore lies within the contemplation of this invention to have the arms move in any other manner, as, for example, by any form of pinned or jointed (articulated) motion.

-o-0-oi differential etching.

ipCT/US90/03903 WO 91/02271 With reference now to Figures 9 through 12 an alternate embodiment of the positioni..g apparatus 20' in accordance with the present invention is shown. In this embodiment the arms 22' are, similar to ihe embodiment earlier discussed, articulably movable in caitilevered fashion with respect to each other against the bias of the flexure defined by the central portion 25' thereof. However, the grooves 36' formed in the arms 22' are not converging grooves, but are uniform width grooves. Accordingly the channel 92' formed by the cooperative association of the arms 22' when superimposed one on the other is a uniform width channel. The maximum dimension of such a channel 92' in the plane perpendicular to the reference R is less than the outside diameter of the smallest anticipated fiber F.

A further modification to the positioning apparatus may be seen from Figure 12. It is first noted that the planar walls 62', 64' of the troughs 60' are parallel, rather than inclined with respect to each other. Moreover, the offset 100' between the axes R and R' lies in the vertical plane, that is, in the plane containing the axes 70' of the troughs 60', as opposed to being offset laterally in the plane containing the surfaces The lead-in portions 68'A, 68'B are ommitted here but may be provided.

In operation, a fiber F is inserted into the positioning apparatus 20' and guided by the passage 98' defined by the registered troughs 60'A, 60'B. Because the axis R' of the passage 98' is vertically offset from the axis R of the channe' p 30 92' the surface 26'B of the arm 22'B bounding the passage 98' serves to guide the fiber F toward the inlet end 94' of the channel 92'. The fiber F enters the channel 92' and contacts with the edges of the sidewalls 38'A, 38'B, 40'A and 40'B. Due to the sizing of tile grooves 36'A, 36'B the fiber F does not touch the major surface 26'A, 26'B of the arms 22'A, 22'B, layui L. i- jaCKeC J IS provided about the cladding layer L but is stripped from the fiber F prior to the insertion of the fiber into the positioning apparatus 20. The jacket may comprise more than one layer.

1 WO 91/02271 26 respectively. The fiber may be chamfer mechanical device may be used to facili fiber into the channel 92'.

Since the fiber F exceeds the dirm the clips 30'A, 30'13 are displaced from toward a second, centering, position. TI clips 30'A, 30'B maintains the point P on fiber F on the reference axis R. The end thus exits through the outlet end 96' of the point I' precisely aligned on the refe F is held in this position by contact witi sidewalls 38'A, 38'B, 40'A, and 40'B, as character LC.

Thie embodiment of Figures 9 to modified, as seen in Figures 13 to 16.

arm 22"B differs from those shown earl provided therein. In this embodiment, i converging groove, a partially funnel-li The fiber F is guided by contact against and held in position on the reference ax major surface 26"B and the edges of th again as indicated by the character LC.

-o-0-o- Figures 17 and 18 are exploded perspective views, generally similar to 30 anotlher lllterniate I:Ihnlil nl n :i n n 3 9 0 3 ed or tapered or a itate insertion of the ension of the channel 92' the first, closed, position his movement of the the end face E of the face E of the fiber F the channel 92' with rence axis R. The fiber 1 the edges of the indicated by the 12 can be further In this modification, the ier in that no groove is f the groove is a ke channel is defined.

Sthe major surface 26"B is R by contact with the e sidewalls 38'A, 38'B, i and assembled Figures I and 2, of siinnin, ilnnarilius in accordance with the present invention while Figure 19 shows the end view thereof. In this embodimient, instead of the arms being articulably movable as described earlier, lthe arms are fixed relative to each other. Each of tile arms 22"'A 335 and 22"'13 has a co\nverging groove therein and the channel 92"' ua alunig ie lull length of the arm 22A and tlie dimension of tle central region 25A of the base portion 24A defines the basic dimension of the arm 22A. A clip generally indicated by the reference character 30A is i WO 91/02271 27 formed by tile cooperative association c superimposed one on tile other is fully The channel 92"' defines a minimum d perpendicular to the reference R that is than the outside diameter of the smalle In operation, a fiber F is inserted apparatus 20"' and guided through the inlet end 94"' of the channel Thel S 10 funnel-like channel 92"' and is guided more of the sidewalls 38'A, 38'B, 40'A surfaces 26"'A, 26"'B to place the point axis R. However, since the arms 22"' a Seach other, the fiber F can only advanc to the axial location where the outer di equals the local dimension of the chann location within the channel the fiber is minimum of four point contacts (indic PC) between the fiber F and each of th 40'A, and 40'B. The dimension of tile fiber is not able to contact the major s wheni it is held along tile reference axi the fiber as the same is held within the spacing 104 between the end face E of outlet end 96"' of the channel 92"' va outer diameter dimension of the fiber -o-O-o- The positioning apparatus in acci pCT/US90/03903 f the arms 22"' when funnel-like in form.

inension in the plane ,near its outlet end, less st anticipated fiber F.

into the positioning passage 98"' toward the fiber F enters tile by contact with one or Sand 40'B and/or major P of tile fiber F on the are fixed with respect to :e within tile channel 92"' ameter of the fiber F el At this axial held in position by a ated by the characters e sidewalls 38'A, 38'B, channel is such that the urfaces of the arms 22"' s R. Figure 20 illustra.es channel The axial the fiber F and the ries, dependent upon the

F.

ordance with any of the above-described embodiments of the invention may be used in a variety of applications which require the precise p itioning of a point P on the end face E of a fiber F along a reference axis 0, II I i ivw v I ULICL CIIU 43A. The respective upper edges 39, 41 of the sidewalls 38, of the groove 36 lie in a reference plane RP having a reference axis R lying therein. Thle planar surfaces 34 also lie in the i>T 1 -J 1 WO 91/02271 In Figures 21 and 22, 1, 20-2 (corresponding to the and 2) are arranged to defi generally indicated by the r arrangement tile apparatus disposed with respect to the the respective channels 92 distance 122 with the resp being collinear. To effect 74 is extended in an axial is provided with a planar a positioning apparatus 20-1, attachment surface 76.

a The fibers F-I and F the lead-ins 68 of the rest 20-2. Each positioning ap manner described above, sei face E of the respective fi disposed axes The fibei respective apparatus 20-1, 2 The ends E of the fibers Fdescribed holding action of desired an suitable index n ultraviolet curing adhesive Electro-Lite Corporation, D 82001ELC4480, may be us 30 The fihber-to-fiber co pC/US90/03903 a pair of positioning apparatus embodiment shown in Figures 1 ne a fiber-to-fiber connector eference character 120. In this 20-1, 20-2 are confrontationally other so that the outlet ends 96 of .therein are spaced a predetermined ective reference axes R therethrough such an arrangement the foundation direction and each axial end thereof ttachment surface 76. Each 20-2 is mounted to its respective F-2 to be connected are inserted into )ective positioning apparatus 20-1, ,aratus 20-1, 20-2, acting in tile rves to place the point P on the end ber F-I or F-2 along the collinearly rs F-1, F-2 are inserted in to the !0-2 until the end faces E, E' abut.

F-2 are secured due to the abovethe positioning apparatus. If hatching adhesive, such as an such that manufactured and sold by anbury, Connecticut as number ed.

nnector mayv he implemented using any of tile above-discussed alternative embodiments of the positioning apparatus. In the event a pair positioning apparatus as shown in Figure 17 is used (see Figures 23 and 24), the confronting ends of the positioning apparatus 20"'-1, 20"'-2 are preferably abutted and secured, or the pair of I- vIn; i1 wiIICII each point on the reference axis R is spaced in the reference plane RP a uniform distance from the edges 39, 41 of the sidewalls 38, 40 as one progresses from the inlet end 42 to the outlet end 43 of the I 2PCT/US90/03903 WO 91/02271 29 positioning apparatus formed integrally with each other. The spacing 122 between the end faces E of the fibers F-I, F-2 is, in this embodiment, defined by the sum of the distances 104-1, 104-2. The spacing 122 is filled with an index matching material, such as the adhesive defined above. To this end, an access port 124 is provided to permit the introduction of the index matching material into the region between the confronting end face of the fibers F-I, F-2.

Prior to insertion into the positioning apparatus (of whatever form) it should be understood that tile jacket J (Figure 29) of the fiber F is stripped in its entirety a predetermined distance from the free end thereof. The exposed portion of the fiber is cleaned with alcohol. The fiber is cleaved to form the end face E. If desired the end face E may be ground into a convex shape to yield a point or be lensed.

-o-0-o If desired the fiber-to-fiber connector 120 may be disposed in a suitable housing 130 (figure 25). The preferred form of the housing 130 is generally similar to that disclosed in United States Patent 4,784,456 (Smith), assigned to the assignee of the present invention. This patent is hereby incorporated by reference herein. The housing 130 includes a base 132 and a cover 134. The base 132 is, in all cases, provided with a recess 136 that is sized to closely receive the connector 120. If the connector 120 is realized using any form of the positioning apparatus that articulates, the cover 134 i 9 30 must be provided with a corresponding recess 138 located so as to permit the articulating motion of the arms of positioning apparatus used to form the connector. If the connector 120 is realized using the form of the positioning apparatus shown in Figures 23 and 24, the recess 138 need not be provided. Such a housing 130 is shown in Figure 28.

I W91/02271 CT/US90/03903 WO 91/02271 The cover 134 is segmented into three sections, 140A, 1401, 140C, each which is hinged to the base 132. The base 132 has, adjacent to each end of the recess 136, V-shaped grooved regions 142A, 142B. The top end sections 140A, 140B each contain respective generally tapered lands 143A, 143B.

Each of the lands has serrations 145A, 145B respectively thereon.

In use a connector is inserted in the recess 136 of the housing 130. It is there held in place by friction but may be otherwise secured if desired. The central section 140C of the cover may then be closed, if desired. An optical fiber having a predetermined length of its jacket J stripped and cleaned, is inserted through one of the V-shaped grooved regions 142A, 142B to dispose the stripped end of the fiber into the connector 120. The grooved region serves to properly orient and position the fiber with respect to the connector 120 in the recess 136.

The associated top end section 140A, 140B, as the case may be, is then closed and latched to the corresponding portion of the base 132 (Figure 25, with the fiber ommitted for clarity).

When the top is secured to the base the serrations 145 act against the jacket of the fiber to urge, or to bias, the fiber toward the connector. A second fiber is correspondingly introduced into the housing and connector in an analagous manner. If not already done so, the central section 140C of the cover is then closed. The housing 130 is preferably formed by injection molding.

30 As seen in Figure 29, in another form the housing 130 may be implemented using a mass 160 of index matching material, such as that identified above. The mass 160 extends over both the connector 120 (to embed the same therein) and some predetermined portion of tile jackets J of the fibers F-l, F-2.

L ii WO 91/02271 PCT/US90/0390 3 31 -o-0-o- The reference axis R on which the point P of the fiber F is positioned may itself extend collinearly with the axis X of any of a variety of devices. Accordingly, a positioning apparatus may be used to accurately position the point P on the end face E of the fiber F with respect to the axis X of a particular device 170. Figures 30, 31 and Figures 32, 33 illustrate several examples of the use of a positioning apparatus 20 to locate a fiber F along an axis X of a device 170. The device 170 may, for example, be realized by any active optical component, such as a solid state laser, a photodiode, a light emitting diode, whether these devices are edge active devices or surface active devices. Although in the discussion that follows the reference character 20 is used to indicate the positioning apparatus, it should be under, ood that any one of the embodiments of the positioning apparatus heretofore described may used.

When used in connection with an edge active device 170 the arrangement in Figures 30 and 31 is preferred. In this arrangement the foundation 74 is axially extended to define a pedestal 174 at the axial end thereof. The upper surface 176 of the pedestal 174 defines a planar attachment surface. The surface 176 is spaced a predetermined distance above the attachment surface 7( or otherwise located such that when the active optical component 170 is mounted the surface 176 the axis X of the device 170 and the reference axis R are collinear.

0-f. ,With the axes R and X collinear, the positioning of the point P on the fiber F along the axis R will 'utoiatically position that point P in the same relationship with ,he axis X. The device 170 musl be accurately mounted onl the surface 176 so that its axis X is collinearly aligned with the axis R.

L ;-1 l(u tu LU, is Dlase 10 striKe one of the sidewalls 38A, 40A forming tile groove 36A (and analogously, the sidewalls 38B, 403 forming the groove 36B). This insures wall contact with the fiber at at least two spaced locations.

PCT/US90/03903 WO 91/02271 32 To mount the device 170 the surface 176 may be provided with a layer of solder layer, such as a gold/tin solder.

The device 170 may have a corresponding layer of the same material. The device 170 is positioned on the surface 176 using a suitable micropositioning apparatus, such as a vacuum probe. The device is aligned to the edge, heated above the melting point of the solder and cooled, so that the solder forms a bond.

When used with a surface active device, as seen in Figures 32 and 33, the active surface of the device 170 is secured to the front surface 178 of the pedestal 174. Attaching the device to the front surface 178 is believed to provide sufficient bonding area to secure the device 170 to the positioning apparatus 20. The surface 176 of the pedestal 174 is relieved to avoid obstruction between the active region of the device 170 and the end face E of the fiber F.

It should also be appreciated, as is illustrated in Figure 31A, that the positioning apparatus in accordance with any one of the embodiments heretofore described may be configured to accurately position a lens, such as a ball or a rod lens L, with respect to the axis X of the device 170 (whether the same is an edge active or a surface active device). The positioning apparatus would be modified to provide a seat 31S in the clips thereof sized to accept the lens L.

-o-0-os, pCT/US90/03903 WO 91/02271 33 The photolithographic microfabrication technique used to manufacture a positioning apparatus 20 may .be understood from the following discussion taken in connection with Figures 34 to 40. Although the discussion is cast in terms of the manufacture of a fiber-to-fiber connector 120 using the preferred embo,;ment of the positioning device 20, as shown i "igure 22, the teachings are readily extendable to the manufacture of any of the embodiments of the positioning apparatus heretofore described, including their use in the various other applications previously set forth.

A silicon wafer 200 having an appropriate predetermined crystallographic orientation is tihe starting point for fabrication of the arms 22 of a positioning apparatus 20 in accordance with the present invention. It should be understood that other single crystalline substrate materials, such as germanium, may be used provided appropriate alternative etchants and materials compatible with the selected alternative substrate are used. The wafer 200 is polished on at least one surface.

Suitable silicon wafers are available from SEll America, Inc., a subsidiary of Shin-Etsu Ilandotai Co. Ltd., Tokyo, Japan, located at Sparta, New Jersey. It should be understood that the wafer 200 can be of the "p-type", "n-type" or intrinsic silicon.

The substrate material is preferably (100) surface silicon because this material can be etched by anisotropic etchants which readily act upon the (100) crystallographic plane but substantially do not etch tie (111) plane. As a result the preferred truncated V-shaped grooves 36A, 36B, the troughs S 30 60A, 60B, the lead-ins 68A, 68B and tie central region of the arms 22A, 22B between the abutments 32A, 32B and the enlargements 54A, 54B are easily formed. The width and depth of such features are dependent upon the preselected width of the opening in the photolithographic mask being used and the time during which tile etchants are permitted to act.

i" 1 f- I a I I J a EIM, predetermined close distance is typically on thie order of five to twenty-five (25) micrometers. In this embodiment the planar surfaces 34A, 34B on the abutments 32A, 32B of the PCT/US90/0390 3 WO 91/02271 34 Etchants operate on 100 surface silicon in an essentially selflimiting manner which property is useful in forming a full V-groove. One of skill in the art will recognize that if other cross-section configurations are required, other predetermined crystallographic orientations of the silicon may be used. For example, if square cross-section features are desired, (110) surfaces silicon wafers can be used. Otler cross sectional configurations for the features are, however, significantly more expensive and, as will be seen later, would require a more complicated configuration to obtain the fiber centering action equivalent to that inherent in a V-groove.

Figure 34 is a plan view of the wafer 200. The wafer 200 has peripheral flats 201 and 202, as specified by the SEMI Standard. The flats 201, 202 primarily indicate orientation of the crystallographic structure of the silicon and are also used for wafer identification and mask alignment. The longer flat 201 indicates the direction of crystallographic plane (110).

The shorter flat 202 is placed a predetermined angular amount on the periphery of the wafer with respect to the flat 201, the magnitude of the angle depending upon the doping of the crystal.

As will be developed, the peripheral regions 203 of the wafer 200, when prepared, carry alignment grooves, while the central region 204 of the wafer 200 has the structural features of the arm or foundation, as the case may be, of the positioning apparatus formed thereon.

Figure 35 shows a mask 210 with a patterns 212 of orthogonal alignment grooves thereon.. The grooves in each i pattern 212 are graduated in size to accommodate various sized (diameter) quartz alignment fibers. The grooves 212 have a V-shaped cross section to accept fibers ranging in width from about 0.004825 inches (0.123 mim) to 0.005000 inches Li described will not be impeded.

m' a pCT/US90/03903 WO 91/02271 (0.127 mm) in 0.039370 inch (0.1 mm) steps, five grooves 212 having been illustrated. The groove width (at the open top of the groove) is larger than the diameter of the fiber so that the center of the fiber is substantially coplanar with the surface of the wafer when the fiber is disposed in its associated groove.

Accordingly, for a 0.123 mm fiber, a groove 0.1506 mm is provided. Similarly, for a 0.124 mm fiber, the open top dimension of the groove is 0.1518 mm. For a 0.125 mm fiber, the open top dimension of the groove is 0.1531 mm; for a 0.126 mm fiber the open top dimension .f tihe groove is 0.1543 mm.; and for a 0.127 mm fiber, the open top dimension of the groove is 0.1555 mm.

A central area 214 of the mnask 210 has provided thereon a repetitive pattern 220 (one of which is shown in Figure 36) containing to a predetermined number of structural features arms or foundations) of the positioning apparatus 20 being formed. Since the typical wafer 200 is about 3.9381 inches (101.028 mm) in diameter and a typical connector 120 measures about three hundred fifty (350) micrometers at the widest location and is about two thousand eight hundred (2800) micrometers in length, the structural features for approximately one thousand (1000) connectors 120 may be formed from the central region 204 of the wafer 200.

Figure 36 is an enlarged view of a portion 220 of the pattern provided on the central region 214 of the mask 210.

In Figure 36, the pattern illustrated is that used to form a plurality of conjoined arms 22 used in a connector 120 (Figure P 30 22). The pattern 220 is formed on the surface of the central region 214 of the mask 210 using a well-known step and repeat process to cover the entire area.

The repetitive pattern 220 shown in Figure 36 is comprised of a plurality of columns 224 which are defined

I

It should be understood that any other convenient mechanism may be used to define the means for biasing the 1

I

91/02271 pCT/US90/03903 SWO 91/02271 36 between an array of adjacent parallel scribe lines 226 and a first and a second separation line 227A and 227B. Each column 224 contains ten (10) discrete zones 228A through 228E that are symmetrical within the column 224 about a cutting line 230.

Seen between two next adjacent scribe lines 226 is the configuration of two arms 22 joined front end to front end.

Seen between three next adjacent scribe lines 226 is the configuration of two arms 22 joined lengthwise side to side.

The zone 228A corresponds to features defining the region of the lead-in 68A of an arm 22A. The zone 228B corresponds to features defining the region of the trough 60A of the arm 22A.

Similarly, zone 228C corresponds to the central portion 25A of the arm 22A, while the zone 228D corresponds to features defining the region of the converging groove 36A on the arm 22A. The axis 50A of the converging groove 36A is offset from the axis 70A of the trough 60A by the offset distance 100.

Finally, if provided, the zone 228E corresponds to features defining the region of the tabs 48A of an arm 22A. Note that in the mask illustrated in Figure 36 the position of the offset 100 on one side of the cutting line 230 is reversed from the position of the offset 100 on the opposite side of the cutting line, although this arrangement is not necessarily required.

S 30 The repetitive pattern for a mask of the arm 22B will be similar to that shown in Figure 36 except that the direction of the offset distances 100 for the arm 22B will be the mirror image of the pattern for the arm 22A. As will be come clearer herein, this mirror image relationship between the offsets is necessary so that so that features on the resulting arms 22A, 22B will register with each other when one is inverted and superimposed on the other. Of course if the offset 100 is eliminated, masks for the arms 22A and 22B will be identical.

i i respectively, to tile extent necessary to place a predetermined point PI on in end face E of the fiber F into alignment with the reference axis R.

WO 91/02271 36 36 between an array or adjacent parallel scribe lines 226 and first and a second separation line 227A and 227B. Each c 224 contains ten (10) discrete zones 228A through 228E t are symmetrical within the column 224 about a cutting lir 230.

Seen between two next adjacent scribe lines 226 is configuration of two arms 22 joined front end to front end Seen between three next adjacent scribe lines 226 is the configuration of two arms 22 joined lengthwise side to sid The zone 228A corresponds to features defining the regior the lead-in 68A of an arm 22A. The zone 228B correspon features defining the region of the trough 60A of the arm Similarly, zone 228C corresponds to the central portion 25 the arm 22A, while the zone 228D corresponds to feature, defining the region of the converging groove 36A on tlhe 22A. The axis 50A of the converging groove 36A is offset the axis 70A of the trough 60A by the offset distance 100 Finally, if provided, the zone 228E corresponds to feature defining the region of the tabs 48A of an arm 22A. Note the mask illustrated in Figure 36 the position of the offse on one side of the cutting line 230 is reversed from the p of the offset 100 on the opposite side of the cutting line, although this arrangement is not necessarily required.

The repetitive pattern for a mask of the arm 22B w similar to that shown in Figure 36 except that the directio the offset distances 100 for the arm 22B will be the mirr image of the pattern for the arm 22A. As will be come c S 30 herein, this mirror image relationship between the offsets pCT/US90/03903 a olumn that ne the

I.

e.

Sof ds to 22A.

A of s arm from s that in t 100 osition ill be >n of or learer is necessary so that so ihat features on thle resulting arms 22A, 22B will register with each other when one is inverted and superimposed oni the other. Of course if the offset 100 is eliminated, masks for the arms 22A and 22B will be identical.

PCT/US90/03903 WO 91/02271 37 The cross-hatched areas shown in Figure 36 preferably correspond to those areas of the central region of the wafer 200 that will be protected by a layer of resist material (as will be described) while tie areas shown without hatching will be left unprotected during subsequent etching steps. A negative resist is employed but it should be apparent that the location of the hatched and clear areas of Figure 36 may be reversed if desired. This would alter somewhat subsequent steps, but in a manner known to those in the art.

Figures 37A through 37E illustrate the process steps whereby a wafer 200 of crystalline silicon may be formed into an array of arms 22A corresponding to the array shown on the mask of Figures 35 and 36. As seen in Figure 37A the wafer 200 is preliminarily covered with a layer 232 of a material that acts in a manner similar to a mask. Silicon nitride (Si 3 N4) is preferred, and is surfaced onto tile polished operative surface 200S of the silicon wafer 200 by thermally growing the silicon nitride layer in an oxygen atmosphere at elevated temperature (circa seven hundred fifty (750) degrees Celsius), as is known.

As indicated, silicon nitride is used because available etchants that attack silicon will also attack known photoresists but will not affect silicc nitride. A suitable nitride layer is grown onto the wafer by CVD Systems and Services, Incorporated, Quakertown, Pennsylvania.

The layer of silicon nitride 232 is then covered with a photoresist 234. Preferred is a positive resist, such as the mixture of 2-ethoxyethyl acetate, N-butyl acetate and xylene 1 30 sold by Shipley Company, Incorporated of Newton, Massachusetts, as "Microposit Photoresist" 1400-37. The resist is spun onto the surface of the silica nitride in accordance with instructions set forth in the Shipley Microelectronic Products Brochure (1984) using standard apparatus such as that Si z, t i

F-

e 1' IlC pCU/US90/0 3 9 0 3 WO 91/02271 available from Headway Research Incorporated of Garland, Texas under model number ECR485.

The mask 210 is mounted atop the wafer 200 and is aligned with respect to the flats 201, 202 of the wafer 200 using alignment bars 235. Thus, in a finished wafer the alignment grooves 212 are precisely positioned with respect to the flats on the wafer through the use of alignment bars 213 on the mask. The wafer 200 is exposed to ultraviolet light through the mask 210 and subsequently developed.

Since a positive resist is used the unexposed areas of the resist are washed away using de-ionized water, leaving the layered arrangement of exposed, hardened resist 234, silica nitride 232 and wafer 200, as shown in Figure 37B.

Next the pattern of the mask 210 is etched into the silica layer 232. Phosphoric acid (-1 3 P0 4 is preferred. This step results in the arrangement shown in Figure 37C. Those skilled in the art will recognize that process variables such as, for example, concentration, time and temperature are all adjusted appropriately to optimize results in all of the wet processing steps described.

Thereafter, a second, differential, etching step is performed to etch the silicon to form the features of the arms 22A. Preferably using an anisotropic etchant such as ethylene diamine pyrocatechol and water. A mix of 750 ml ED, 120 gm P and 240 ml water is preferred. A two-step etch using potassium hydroxide (KOII) may also be used if desired.

This etching produces the structural feature in the surface of the silicon illustrated schematically in Figure 37D by reference character 236. The depth of the feature 236 is controlled by controlling the etching time, as is well known. Of course, t I, f WO 91/02271 39 differential etching is self-limiting for the inside angles of structure, if left to proceed.

The silicon nitride layer 232 is then removed by et with phosphoric acid and a layer of silica, silicon dioxi grown on the surface. Next, resist is deposited on the sui of the wafer and is imaged through a mask, as shown Fig 38. This results in a layer 238 of hardened resist being on those predetermined portions of the wafer that are to bonded (corresponding to zones 228C through 228E and t troughs 60 (see Figure 36)).

The silicon layer is then etched from areas that are be bonded (See, Fgiure 37E) using hydrofluoric acid (HF).

resist layer 238 is stripped using acetone, leaving a finis wafer ready for bonding.

This completes the fabrication of the first wafer 20 having the array of arms 22A thereon.

-o-0-o- As noted earlier, since the axis of the guideway 98/ offset from the axis of tile groove 92A, the mask for the 22B is not identical with the mask used to form the arms Accordingly, a second wafer having an array of arms 22E thereon must be prepared in accordance with the metho illustrated in Figure 37. The finished second wafer (not specifically illustrated but hereinafter referred to by ch 30 200') is similar in all resnects excent in location of the ol PCT/US90/039 0 3 Sthe ching ide, is rface ure formed be o not to The hed 0 A is arms 22A.

d steps aracter ffset 100.

A third wafer 200" is prepared using a foundation mask, a portion of whicl, is shown in Figure 39. Figure 39 is an enlarged view of a portion of the pattern 220' provided on the i-~L i -1 t.PI .i L. u. E, of a point P on the end face E of a fiber F along a reference axis

R.

i PCT/US90/03903 WO 91/02271 central region of the foundation mask (analagous to the pattern of the arm mask shown in Figure 36). The repetitive pattern 220' is comprised of a plurality of columns 244 which are defined between an array of adjacent parallel scribe lines 246 and a first and a second separation line 248A and 248B. Each column 244 contains four discrete zones 250 that are symmetrical within the column 244 about a center line 252.

The zones 250A define mounting surface 76 on a foundation 74. The zones 250B correspond to the surfaces 82 provided on the foundation. The wafer 200" containing th(lie foundations 74 is exposed in a manner analagous to that shown in Figure 37, with the exception that the exception that the solder mask exposure is not carried out. However, the layer of silica is removed from the surface of the wafer 200".

Having prepared wafers for the arms 22A (the wafer 200), the arms 22B (the wafer 200') and the foundations 74 (the wafer 200"), the final assembly of the connector 120 may be made as is shown in Figure The wafer 200' is placed on top of the wafer 200. The registration of the features on the wafer 200' to those on the wafer 200 is effected using at least two and preferably four lengths of a stripped optical fiber and the corresponding appropriate one of the alignment grooves in each array 212 of grooves. The diameter of each length of the optical fiber is measured by micrometer, accurate to plus or minus micrometers. Each of the fibers is placed in groove in the groove array 212 that most closely corresponds to the measured diameter. Each alignment fiber thus sits in the selected alignment groove such that the axis of the alignment fiber lies in tile plane of the surface of thie wafer 200 with the remaining portion of each fiber protrudes above that surface.

i: i _11.

i j pubitluiong apparatus. In the event a pair positioning apparatus as shown in Figure 17 is used (see Figures 23 and 24), the confronting ends of the positioning apparatus 20"'-1, 20"'-2 are preferably abutted and secured, or the pair of PC/US90/03903 WO 91/02271 4 1 The wafer 200' is inverted and placed atop the wafer 200, with the corresponding grooves in the wafer 200' receiving the protruding portions of the alignment fibers thereby to precisely align the pattern of the two wafers. Since the alignment grooves on each wafer are formed simultaneously with the formation of the features on the wafer, and s;nce the mask for each wafer is formed optically one from the other, precise alignment between the wafers is achieved. It is noted in Figures 40A and 40B only one of the fibers 254 and grooves 121 is shown, for clarity of illustration.

The assembly of superimposed wafers 200, 200' shown in Figure 40A is bonded in a wet controlled atmosphere furnace according to methods described the paper by Shimbo et al., "Silicon-to-silicon direct bonding method" published 10/86 in the Journal of Applied Physics, and in the paper by'Lasky et al., "Silicon on Insulator (SO) By Bonding and Etchback", IEDM As seen in Figure 40B the exterior surface 256' of the wafer 200' is lapped to reduce its thickness from it original thickness (typically approximately seventeen (17) micrometers) to a final thickness of five micrometers.

The resulting bonded structure is inverted and the exterior surface 256' of the wafer 200' is mounted atop the wafer 200". The alignment of these wafers is effected using a fixture employing quartz blocks 260 abutting against the flats 201, 202 of the wafer 200. The wafer 200' is then bonded to the wafer 200". It is to be understood that other bonding techniques, such as those discussed in the paper by 30 Wallis and Pomerantz "Field Assisted Glass-Metal Sealing" published 9/69 in the Journal of Applied Physics may be used to bond the wafers. Still other alternate bonding techniques would include metallic or glass solder bonding.

~I

ulit cuor. II Hle COnrcluIVi Iv J realized using the form of the positioning apparatus shown in Figures 23 and 24, the recess 138 need not be provided. Such a housing 130 is shown in Figure 28.

PCT/US90/03903 WO 91/02271 4 2 The exterior surface 256 of the wafer 200 is then lapped until the dimension of the wafer 200 is that of the wafer 200'.

Thus, the substantial equality of thle biasing forces imposed by the flexure is provided.

The resultant three wafer bonded stack shown in Figure may then be cut. Only tile top two wafers 200, 200' of the bonded stack (containing the arms 22-11, 22-21 and the arms 22-1A, 22-2A, respectively, Figure 22) are first cut along the lines in thie wafers corresponding to the cutting lines 230, 230' on the arm masks (Figdre 36). This cut is made using a blade that is on the order of 0.003 inches to create the distance 122 in Figure 22. The bonded stack is thereafter cut, using a blade that is 0.015 inches thick, along the lines in the wafers corresponding to the separation lines 227A, 227B on the wafer 200, the separation lines 227A', 227B' on the wafer 200', and thle separation lines 248A, 248B on the wafer 200", as well along the scribe lines 226, 226' and 246 (on the respective wafers 200, 200' and 200") all of which are registered with each other, thereby to yield from the bonded stack about one thousand of the fiber-to-fiber connectors 120.

-o-0-o- Those skilled in thle art, having the benefits of the teachings of the present invention as hereinabove set forth, may impart numerous modifications thereto.

For example, as seen in Figure 42, in addition to the various embodiments of tile two-armed configurations for the positioning apparatus of the present invention previously disclosed, it lies within the contemplation of this invention for a positioning apparatus to exhibit more than two arms 22. In this regard Figures 42A to 42C illustrate i positioning apparatus having three arms 22A, 22B and 22C while Figures maitc ria, sucn as that identified above. The mass 160 extends over both the connector 120 (to embed the same therein) and some predetermined portion of the jackets J of the fibers F-l, F-2.

I,

pCT/US90/03903 WO 91/02271 43 42D through 42F illustrate a positioning apparatus haviig four arms 22A, 22B, 22C and 22D. The extension to even greater number of arms would be readily apparent to those skilled in tile art.

In Figure 42A each the arms are configured similar to the form of tile arms discussed above. The arms may, if desired carry a groove, although it should be understood that such is not required. In Figures 428 and 42E the arms are configured from rods. Although the rods shown aS round in cross section it should be understood that they can have any desired alternate cross section. In Figures 42C and 42F the arms are configured in a generally planar bar form. In Figure 42D the four arms may be formed by sawing tile upper and lower arms (indicated by the characters 22A, 22B in Figures I to 4) along a cut line extending perpendicular to the major suraces 26 and 28 of each of the arms.

However configured the arms are shown in Figures 42A through 42F as angularly juxtaposed in a surrounding relationship to the channel 92 defined their cooperative association. Similar to tile situation described heretofore the resiliency of the arms defines tile biasing means which urge the arms toward the closed position. However, it should be understood that the biasing means may be otherewise defined, so long as the force on each arm passes through the reference axis and the sum of forces oni the arms when they are in tile centering position is substantially equal to zero. Whatever form of biasing means is selected the bins force must increase 30 with deflection of tle arm. The arms act against the fiber F inserted into the channel along tihe various lines of contact LC illustrated in Figure 42 to maintain the predetermined point on thile fiber on thie reference axis R.

I

S/u musi be accurately mounted on the surface 176 so that its axis X is collinearly aligned with the axis R.

WO 91/02271 44 PCr/US90/03903 It should be understood that such modifications as herein preented and any others are to be construed as lying within the contemplation of the present invention, as defined by the appended claims.

WHAT IS CLAIMED IS: Ir

Claims

1. An opto-electronic component comprising, in combination: a foundation having a pedestal thereon; a positioning apparatus itself comprising a first and a second arm, the first arm being mounted to the foundation, at least the first arm having at least a first and a second sidewall cooperating to define a groove therein, the arms being arranged in superimposed relationship, each arm being movable from a first, closed, position to a second, centering, position, means for biasing each of the arms with a substantially equal and oppositely directed biasing force toward the first, closed position, in the closed position the arms cooperating to define a channel having a reference axis therethrough, the channel having an inlet end and an outlet end, each of the arms being arranged such that a cylindrical member introduced into the inlet end of the channel with the axis of the member spaced fro.n the reference axis is initially displaceable by contact with at least one of the arms to move a center point on an end face of the member toward alignment with the reference axis, the arms being responsive to further axial movement-of the member through the channel by moving against the bias force toward the centering position to position the point on the face of the member into alignment with the reference axis by contact between the member and both the first and the second arms; and an opto-electronic device mounted to the pedestal, the opto- electronic device having a reference axis therethrough, the reference and depth of width of the and the lime suci features are dependent upon the preselected opening in the pliotolilliographic mask being used during which the etchants are permitted to act. imq 1 ii A1 ij 46 axis of the opto-electronic device aligning collinearly with the reference axis of the channel.

2. The opto-electronic component of claim 1 wherein the first and the second sidewalls in the first arm cooperate to define a converging groove therein, the channel being partially funnel-like in shape over at least a predetermined portion of its axial length.

3. The opto-electronic component of claim 1 wherein the second arm has at least a first and a second sidewall disposed therein, the first and second sidewalls in the second arm cooperating to define a converging groove therein, the converging groove in the first arm and the converging groove in the second arm cooperating to define the channel, the channel being fully funnel-like in shape over at least a predetermined portion of its axial length.

4. The opto-electronic component of claim 1 wherein the first and second sidewalls in the first arm cooperate to define a groove having a uniform width dimension throughout its length, the channel being rectangular in cross sectional shape over at least a predetermined portion of its axial length.

The opto-electronic component of claim 4 wherein the second arm has at least a first and a second sidewall disposed therein, the first and the second sidewalls on the second arm cooperating to define therein a groove having a uniform-width dimension throughout its length, the uniform groove in the first arm and the uniform groove in the second arm cooperating to define the channel, the channel being rectangular in cross sectional shape over at least a predetermined portion of its axial length.

6. The opto-electronic component of claim 5 wherein each sidewall of the groove in the first arm and each sidewall of the *I i L -I i LLL aie giaouaiea in size to Iccomnmodate various sized (diameter) quartz alignment fibers. The grooves 212 have a V-shaped cross section to accept fibers ranging in width from about U.004825 inches (0.123 imm) to 0.005000 inches 1~ L 47 groove in the second arm has an edge thereon, the edges of the sidewalls contacting the member.

7. The opto-electronic component of claim 3 wherein each of the arms has a major surface thereon, a portion of the major surface connecting the first and the second sidewalls and cooperating to define the groove therein, wherein the groove so defined in each arm has a truncated V-shape.

8. The opto-electronic component of claim 3 wherein each of the arms has a major surface thereon, a portion of the major surface connecting the first and the second sidewalls and cooperating to define the groove therein, wherein the converging groove so defined in each arm has a truncated V-shape. I

9. The opto-electronic component of claim 1 wherein biasing means comprises a reduced thickness portion in each of the first and the second arms, the reduced thickness portion defining a flexure in each arm which, when each arm is deflected by contact with the member, generates a force on each arm to bias each arm toward the cosed position.

The opto-electronic component of claim 1 wherein the opto-electronic device is an edge active device.

11. The opto-electronic component of claim 1 wherein the opto-electronic device is a surface active device.

12. The opto-electronic component of claim 1 wherein the opto-electronic device is a solid state laser.

13. The opto-electronic component of any one of claims 1 to 4 wherein the opto-electronic device is a solid state light responsive diode. C The repetitive pattern 220 shown in Figure 36 is comprised of a plurality of columns 224 which are defined r 48

14. The opto-electronic component of any one of claims 1 to 4 or 9 wherein the first and the second arms are each fabricated from a crystalline material. D A T E D this 31st day of March E.I. DU PONT DE NEMOURS AND COMPANY By their Patent Attorneys: CALLINAN LAWRIE

1994. I 1 r :_ii L_ r 1