GB2375830A

GB2375830A - Creating Bragg gratings in optical waveguide devices

Info

Publication number: GB2375830A
Application number: GB0115059A
Authority: GB
Inventors: Douglas Charles John Reid; Elizabeth May Ward; Neil William Forbes
Original assignee: Marconi Caswell Ltd; Marconi Optical Components Ltd; Bookham Technology PLC
Current assignee: Marconi Caswell Ltd; Marconi Optical Components Ltd; Lumentum Technology UK Ltd
Priority date: 2001-05-22
Filing date: 2001-06-20
Publication date: 2002-11-27
Also published as: AU2002302766A1; GB0115059D0; GB0112384D0

Abstract

The method comprises exposing a photoresist coated wafer (8) twice to a source of coherent light adapted to imprint a grid into the photoresist. In between the two exposures, the wafer is rotated through a predetermined angle d eg. about a central axis 23, whereby two intersecting grids are formed. Development of the result produces a plurality of points where the grid intersections occur. These points form a number of Bragg grids, the pitch of which is determined by the angle d . Hence the process may be used to produce a wide variety of Bragg gratings simply by rotating the substrate wafer through a predetermined angle, thereby enabling a much faster procedure.

Description

METHOD OF CREATING BRAGG GRATINGS IN OPTICAL WAVEGUIDE DEVICES The present invention relates to a method of creating Bragg gratings. More particularly, but not exclusively, it relates to a holographic process for creating a plurality of Bragg gratings within an optical waveguide device. The gratings thus formed may be part of a DFB laser and tuneable DBR laser for use with a wavelength division multiplex (WDM) optical communication system.

In order to increase the information carrying capability of optical fibre communication systems, wavelength division (WDM) systems are being developed in which a number of different wavelengths, often termed wavelength channels, may be carried over a single optical fibre. As. is known there are two principal communications wavergth bands, at 1300nm, and 1550nm, of which the latter band is receiving the greater commercial exploitation because of its suitability for a variety of different generic communication applications. Currently 1550nm WDM communication systems are evolving into

architectures comprising eighty 2.5Gb/s modulated wavelength channels on a single fibre. The same 1550nm band is also being proposed for forty lOGb/s modulated channels thereby doubling the fibre information capacity.

To provide forty or eighty individual wavelength circuits within the 1550nm band requires a light source that can be accurately set to specific wavelengths and maintained at these wavelengths, within specified limits, over the equipment operating life. Such light sources historically comprise a plurality of distributed Bragg reflector (DBR) lasers, each of which is operated to produce light of one of the wavelength channels and optical switching means for selecting a required wavelength channel. Whilst such an arrangement is effective it is impracticable for more than ten or so wavelength channels.

To overcome this problem wide tuning range lasers are being developed, such as sampled grating distributed Bragg reflector (SGDBR) lasers as described in A M Gulisano, D J Robbins, P J Williams, and P Verhoeve (1997)"Widely Tuneable Sampled Grating DBR Lasers to address 100 channels over 40nm for WDM applications", Proc. Euro. Conference on Integrated Optics. The tuning mechanism of the SGDBR laser is by differential current steering of the operating wavelength by means of the currents injected into the front and rear grating Bragg reflector sections, with fine tuning (trimming) being possible by means of current control in the phase section in combination with ganged current injection in the front and rear Bragg grating sections.

An explanation of the operation of a Bragg grating can be found in"Optical Fiber Communications", J M Senior, Prentice Hall International Series in Optoelectronics ISBN 0-13-635426-2.

Figure 1 illustrates the prior art and shows an optical waveguide 330 being part of a semiconductor laser 300. A grating 320 is illuminated with light of wavelength AI that propagates along the waveguide 330. Such a structure acts as a diffraction grating, and light that satisfies the Bragg condition may be reflected back along the original path. The

wavelength I of light that will be reflected back along its original path is given by :Al = (2xnXA)/m (1) where ne is the effective refractive index of the waveguide material 330, A is the pitch of the grating, and m is the number of wavelengths kl in 2 x A, i. e. 1, 2, 3, 4 etc. m may sometimes be referred to as the order of the grating and for many applications in is unity and thus equation (1) simplifies to ; = 2 x ne x A. A necessary condition for reliable lasing is that a single mode of light travels through the waveguide Thus the total reflection wavelength is directly proportional to the pitch or periodicity A of the grating. At this wavelength, for a given set of structure conditions, substantially all the incident light power is totally reflected. Bragg gratings are therefore widely used as

wavelength selective filters and in particular as mirrors in semiconductor DBR and wide bandwidth tuneable lasers.

An SGDBR laser has two grating structures: a front grating, and a rear grating. The fabrication of semiconductor lasers uses wafers of Group III-V semiconductor substrate, in particular Indium Phosphide (InP). The specific laser structures can be grown onto the wafers using Low Pressure Metal Organic Vapour Epitaxy or other processes. Various dopant materials may be grown into the substrate, and various materials may be grown onto the substrate. These materials and dopants typically comprise Gallium and Arsenic. For the purposes of this patent specification the term substrate means the base semiconductor material and such growth layers as pertain to the device in fabrication, depending upon the stage reached in its fabrication process. Typical laser devices are Indium Gallium Arsenide Phosphide structures. It is not necessary for an understanding of this patent specification to detail the precise architecture of various optical waveguide devices; these will be known to those skilled in the art. The monolithic structures are fabricated in many hundreds on circular semiconductor substrates typically of 50,100 or 155 mm diameter, commonly known as wafers. Typically laser devices are grown in bars across the wafer, which is then cleaved into individual devices.

The pitch of gratings for photonic applications are sub-micron, for example 200-300nm, and are therefore not readily amenable to being defined using standard photolithography techniques. Electron beam techniques can be used to write a grating directly onto a

photoresist coating covering the wafer, but this is very costly as a production process. Alternative production means are required for producing gratings which are to be manufactured in large quantities.

A process is known from a paper published in 1982 by The Aerospace CorporationElectronics Research Laboratory, (Optical Engineering, SPIE, May/June 1982, volume 21 No 3,"Sub-micron grating fabrication on GaAs by holographic exposure"), which involves cleaning the wafer substrate, applying a high definition photoresist coating (positive or negative), making a holographic exposure in a special precision aligned rig, developing the resist to leave the desired mask pattern and then transferring the grating to the wafer by means of dry or wet etching.

Figure 2 shows diagrammatically the principle, which is known in the art, that two intersecting coherent plane wavefront beams of optical energy 2 and 4 interfere to produce fringes. In the area of intersection the wavefronts interfere to produce a sinusoidal pattern of light and dark bands 6. The pitch Ao of the sinusoidal pattern is given by :- Ao = ko/ (2 x Sin 8/2) (2) where . o is the wavelength of the optical energy, and 0 is the angle between the two beams of optical energy. Ao is hereinafter referred to as the native pitch.

From equation (2) it can be seen that the native pitch Ao of the sinusoidal pattern of light and dark bands (fringes) is inversely proportional to the angle 6 and thus as the angle increases so the pitch decreases, and conversely as the angle decreases the pitch increases.

From equation (1) it can be seen that by changing the pitch A of the gratings, the Bragg wavelength of the fringes may be varied.

The interference fringes can be used to produce sub-micron gratings on a wafer that has been coated with photoresist. Figure 3 illustrates diagrammatically the prior art effect on a substrate wafer 8a prepared with a surface layer of photoresist which may be introduced into the interference fringe domain. The resultant interference grating 6 is shown on a plan view of the wafer 8b, after the photoresist has been developed and the grating etched. The wafer 8a and 8b may, for some applications, be a dielectric material. Clearly, when fabricating monolithic laser devices a grating over the whole area of the wafer is not usually required.

Those areas of the wafer not requiring a grating can be created either by secondary etching or spatial masking.

An existing practical arrangement for producing gratings on a substrate wafer is shown in Figure 4. This is a known holographic interferometer a ngement that will be explained by way of example to produce a 240nm pitch grating in Indium Gallium Arsenide Phosphide, for use with a target semiconductor laser operating in the 1550nm band.

With reference to Figure 4,10 is an Argon Ion Ultra Violet laser emitting laser light at 351. 1nm. The emitted beam 11 is reflected by a high efficiency mirror 12 to produce beam 13 which is half reflected by 50: 50 beam splitter 14 to form beam 15, while the balance of the light forms beam 19. Beam 15 illuminates a focusing lens 22 which passes the collimated light through a pin hole 24. Together 22 and 24 form a spatial filter. The resultant divergent beam 17 is reflected by a high efficiency mirror 18 to illuminate the photoresist coated target wafer 8a. Similarly, beam 19 enters an equivalent spatial filter comprising a focusing lens 26, and a pin hole 28, and the resultant divergent beam 21 illuminates the target wafer 8a after reflection by high efficiency mirror 16. The angle between the two light beams 17 and 21 is 8. The resultant sinusoidal interference grating 6 is shown for clarity of explanation.

The process relies upon just one time controlled exposure of the target wafer to give the required degree of photoresist hardness for the development and etch time. As a consequence of the beams 21 and 17 being divergent wavefronts the pitch of the interference grating progressively increases away from the centre of the target wafer. However, by appropriate design of the optical system the magnitude of this increase is very small compared to the pitch of the sinusoidal interference grating being recorded, and as such is within the tolerances required for all practical purposes.

Putting Ao as 240nm, and ko as 351. 1 run, into equation (2) and solving for 0 gives a result of 8 = 94 .

Semiconductor lasers are monolithic devices grown with epitaxial layer deposition. Laser wafer gratings need to be orthogonal to the direction of optical power propagation. Thus wafers for optical device manufacture must have a known orientation. Figure 5 shows a wafer 8b whose orientation is known by reference to quadrature major 7, and minor 9, flats on the circumference. As a rule the major flat 7 is substantially parallel to the crystallographic structure of the wafer. For optimum performance Bragg grating rulings (fringes) need to be substantially parallel to the crystallographic structure of the wafer. Therefore, in one convention, optical grating grooves 6 are fabricated to be parallel with the major flat 7 on the wafer, and the laser bars 23, each containing a plurality of lasers 27, are parallel to the major flat. For the purposes of clarity, the plurality of bars 23 across the wafer is not shown.

The holographic arrangement as described with reference to Figure 4 needs to be precision aligned on an optical bench. The process of getting the right interference gratings typically takes many days of set up and producing test grating samples for inspection using an electron microscope.

Once set up, the holographic interferometer process for producing optical wavelength Bragg gratings works well, but it is not possible to make a quick change over if another grating pitch is required. For example with the arrangement of Figure 4, to produce a grating for use in the 1300nm band requires 6 to be approximately 120 to give a grating pitch of

nominally 200nm for use with a typical Indium Gallium Arsenide Phosphide laser structure.

This complex alignment situation is exacerbated by the need normally for the mark space ratio of the grating to be one to one, which requires multiple trials of exposure time, photoresist development, etch times, and electron microscope measurement to achieve the desired result.

It is an object of the present invention to provide a process for creating holographic interference gratings in which changing the resultant Bragg grating wavelength is straightforward and rapid. The process should be ideal for large scale volume production of optical waveguide devices incorporating Bragg gratings.

According to a first aspect of the present invention, there is provided a method of defining gratings in an optical waveguide device which comprises the steps of providing a target substrate surface having a photosensitive coating, providing means to irradiate said coating with two coherent electromagnetic beams, exposing the coating to such irradiation to define a first grating therein, rotating said target substrate surface through a predetermined angular offset and exposing said coating a second time to define therein a second grating which intersects said first grating. The photosensitive coating is preferably photoresist.

The target substrate may be a. dielectric material or preferably may be a semiconductor material, such as layers of Indium Gallium Arsenide Phosphide.

The photosensitive coating may be developed to establish in the substrate a planar pattern of the intersecting grating structure, and the exposure times and/or development time of the coating may be varied to give a planar pattern of the resultant grating structure which is substantially an array of rectangular forms, circular forms, elliptical forms or square forms.

Between the first and second exposures, in some embodiments, measures may be taken to make the grating pitch created by the second exposure dissimilar to the pitch created by the first exposure, such measures comprising, for example: - displacement of the target substrate along the axis of the bisector of the two coherent beams ; and/or non-orthogonal rotation of the target substrate; and/or tilt of the target substrate, and/or

- shift of the target substrate in the plane of the target substrate surface.

The duration times of the first and second exposures may be dissimilar whilst keeping the overall exposure tme constant, whereby there is a blazing of the defined lattice grating.

According to a second aspect of the present invention, there is provided a Bragg grating created by a method as described above wherein the pitch of the grating depends upon the

wavelength of the two interfering electromagnetic beams, upon the angle between the two interfering electromagnetic beams and upon the angular offset, between the first and second exposure, of the target substrate surface.

In this case, the pitch AM ()-) o) of the grating M (l-1 0) may be given by the equation :-

where #0 is the native pitch of the holographic interference fringe and is the angular offset of the target substrate surface between the first and second exposures.

Alternatively, the pitch ^M (l1 0) of the grating M (I 1 0) may be given by the equation :-

where Ao is the native pitch of the holographic interference fringe and # is the angular offset of the target substrate surface between the first and second exposures.

Other possibilities include the pitches #M(-2 1 0) and AM (I -2 0) of the gratings M (-2 1 0) and M (I-2 0) being given by the equations :-

where Ao is the native pitch of the holographic interference fringe and is the angular offset \ I of the target substrate surface betwe,'the first and second exposures, and

the pitches AM (2 1 0) and AM (i 2 o) of the gratings M (2 1 0) and M (l 2 0) being given by the equations :-

where Ao is the native pitch of the holographic interference fringe and is the angular offset of the target substrate surface between the first and second exposures, The defined grating structure may be etched into the target substrate surface at such a depth as to strengthen the effect of the grating upon incident waveguided electromagnetic energy.

The grating rulings, or fringes, are preferably substantially parallel to the crystallographic structure of the target substrate.

Other Bragg gratings may be created in which: the pitch ^M (I -I 0) of the grating M (1 -1 0) is given by the equation :-

where Ao'is the native pitch of the first grating, Ao"is the native pitch of the second grating, and is the angular offset of the target substrate between the first and second exposures; the pitch/\M (l) o) of the grating M (1 1 0) is given by the equation :-

where #0' is the native pitch of the first grating, Ao"is the native pitch of the second grating, and # is the angular offset of the target substrate between the first and second exposures; the pitch AM (2 -I 0) of the grating M (2-1 0) is given by the equation :-

where Ao'is the native pitch of the first grating, Ao"is the native pitch of the second grating, and is the angular offset of the target substrate between the first and second exposures; the pitch AM (-12 0) of the grating M (-l 2 0) is given by the equation :-

where Ao'is the native pitch of the first grating, Ao"is the native pitch of the second grating, and is the angular offset of the target substrate between the first and second exposures; the pitch AM (2) 0) of the grating M (2 1 0) is given by the equation :-

where Ao'is the native pitch of the first grating, Ao"is the native pitch of the second grating, and is the angular offset of the target substrate between the first and second exposures; or the pitch AM (I 2 0) of the grating M (1 2 0) is given by the equation :-

where a s the native pitch of the first grating, Ao"is the native pitch of the second grating, and < I > is the angular offset of the target substrate between the first and second exposures.

Embodiments of the invention will now be more particularly described by way of example and with reference to Figures 6 to 15 of the accompanying drawings, in which: Figure 1 is a schematic representation of a prior art Bragg grating function in a laser context; Figure 2 is a schematic representation of holographic beam interference and the resultant grating, according to prior art; Figure 3 is a schematic representation of holographic beam interference applied to a semiconductor wafer for grating creation, according to the prior art; Figure 4 is a schematic representation of an arrangement for holographic Bragg grating creation, according to the prior art; Figure 5 is a schematic representation of an example of orientation of Bragg gratings in semiconductor laser wafer fabrication process, according to the prior art; Figure 6 is a schematic representation of an arrangement for creation by holographic means of a Bragg grating embodying the invention; Figure 7 i's a schematic representation of intersecting gratings creating a lattice structure; Figure 8 is a schematic representation to an enlarged scale of the lattice gratings shown in Figure 7, identifying new grating structures;

Figure 9 is a geometric representation of the lattice gratings showing the derivation of the pitch relationship of the new gratings; Figures lOa, lOb, 10e and 10d are schematic representations of the Moire Indices description of individual lattice gratings; Figures lla, lib, lie and lld are schematic representations of a variety of etched planar patterned lattice gratings; Figures 12a and 12b are schematic representations of one means of producing two dissimilar native pitch gratings ; Figures 13a and 13b are schematic representations of the geometry of first order lattice gratings produced with dissimilar native pitch gratings; Figure 14 is a schematic representation of the geometry of second order lattice gratings produced with dissimilar native pitch gratings; and Figure 15 is a schematic representation of a blazed second order Moire grating.

Figures 1 to 5 are acknowledged to show the prior art and are discussed in some detail above.

The invention relates to a new process for creating holographic gratings. The invention relies upon using'a fixed geometry holographic interferometer arrangement as shown in Figure 6 that upon exposure of the target wafer 8a creates a first native pitch Ao grating. The target wafer 8a is then rotated through an angle ()) about a central axis 23 substantially perpendicular to the wafer plane and a second exposure made which produces a second

grating with a native pitch of Ao. If desired, alternative orthogonal centres of rotation on the target surface may be used. As a consequence of the beams 21 and 17 being divergent wavefronts the pitch of both of the interference gratings progressively increases away from the centre of the target wafer. However, by appropriate design of the optical system the magnitude of this increase is very small compared to the pitch of each of the sinusoidal interference gratings being recorded, and as such is within the tolerances required for all practical purposes. The wafer 8a now has two sinusoidal gratings 30 and 32 at an angle ()) one to the other as shown in Figure 7 in its plan form 8b, where < j) is the angle of rotation between the first and second exposure of the target wafer 8a. With typical positive photoresist material, where the gratings 30 and 32 overlap the photoresist becomes more soluble in the developer and a lattice structure in photoresist is defined. By suitable developing of the photoresist the lattice intersections may be made dominant as shown in Figure 8, where each of the intersecting peaks of the sinusoidal gratings is identified with a black dot 33.

Thus the invention is a holographic process for the creation of Bragg gratings in optical semiconductor waveguide devices comprising a photosensitive recording medium attached to a target substrate surface, and an optical arrangement to create two interfering coherent electromagnets beams whereby the recording medium is irradiated using a first and second controlled exposure in which the first and second exposure are angularly offset by means of rotating the target surface about a substantially orthogonal axis to create two intersecting

gratings on the recording medium, wherein the resultant grating structure creates planar lattice Bragg gratings.

The target substrate surface material is a semiconductor, such as Indium Gallium Arsenide Phosphide, or a dielectric material.

Figure 8 shows the doubly exposed wafer 8b with a section 40 of the lattice intersection grating enlarged. The enlargement shows the original gratings 30 and 32 both with native pitch Ao at angle +. The intersections of the gratings 30 and 32 are marked with dots 33 showing the areas of photoresist corresponding to lattice beaks.

As may be seen from Figure 8 these dots 33 define not only points along the written gratings but also define alternative lattice gratings such as for example the orthogonal gratings of pitch Ah and Av Referring now to Figure 9 (which is an enlargement of part of Figure 8) the geometry of the relationship between the original cross gratings 30 and 32 of native pitch Ao is shown in relation to prime implicit lattice gratings 34 and 36 of pitches respectively Av and Ah In order to illustrate-this relationship with more clarity the angh i) in Figure 9 is shown greater than it was in Figure 8. From the geometry of the lattice the following relationships can be derived :- Av=Ao/ (2 Sin /2) (3)

Ah = Ao/ (2 Cos $/2) (4) The two prime implicit lattice gratings are orthogonal and the Bragg condition for reflecting light back along its path of origin is that the reflecting means is orthogonal to the light direction of travel. Thus only one of the identified prime implicit gratings can operate upon the light for a specific orientation.

In an optical waveguide a grating is created by the periodic variation produced in the refractive index of the composite material. The periodicity is determined by the lattice grating structure and is maximised by the choice of fill factor (the planar profile dimensions) selected at the time of lattice creation. For example, first order gratings require circular or elliptical dots of fill factors of substantially 25% or 75%; and with second order gratings square or rectangular profiles of fill factors of substantially 50% are required.

There are more lattice gratings than the prime implicit gratings identified in Figures 8 and 9.

The act of producing intersecting native gratings immediately produces a finite but extremely large number of interference or lattice gratings. Whichever Bragg lattice grating is chosen for use, for optimum performance, the grating rulings, or fringes, need to be substantially parallel to the crystallographic structure of the target wafer as indicated by the major flat discussed in relation to the prior art Figure 5. A lattice grating can be seen to exist wherever a straight line can be drawn between any two intersection points of the superposed original Ao holographic gratings. The closer together are adjacent intersection

points the stronger will be the-lattice grating if illuminated with orthogonally directed light at an appropriate wavelength. In practice the dominance of a lattice grating is also influenced by the exposure, development and etching processes used.

In order to describe which lattice grating is intended for optical use, there may be used the Miller Indices used for crystallography in which the three mutually perpendicular axes may

be defined as x, y and z. Using Miller notation a specific direction can be defined by [ Xl Yl z,] where xi, y, and zu are whole number values relative to an arbitrarily defined origin, corresponding to their corresponding locations on the x, y and z axes. Using Miller notation a plane is defined as (X Y Z) where X, Y and Z are the reciprocal of the intercept points on the corresponding x, y and z axes, scaled to be the lowest whole number. The whole number values X, Y and Z are known as the Miller Indices.

Unlike three dimensional crystallography, holographic lattice gratings are constrained principally to two axes, and in the general case the holographic axes are not mutually perpendicular. Using Miller notation rules a terminology proposed for holographic lattice gratings is called Moire Indices, in which the third axis is always perpendicular to the first and second axes and its corresponding Moire Index is always 0 i. e. an intercept at infinity.

The first and seçbnd axis can be at any angle relative one to, the other but both are perpendicular to the third axis. As with the third axis the Moire Indices for the first and second axis are the reciprocal of the intercept points on the respective axis.

The Moire Indices description of a holographic lattice grating will be explained by reference to Figures lOa, b, c and d in which two native pitch Ao holographic gratings 30 and 32 are at

an included angle . Grating 30 is arbitrarily selected as the a axis, and grating 32 is selected as the b axis. The c axis is perpendicular to the plane of Figures lOa, b, c and d wherein Figure lOa, b, c and d are drawn intersecting the virtual c axis at its origin. The Moire position of each intersection, according to the co-ordinate system a, b, c is marked, for example position 35 corresponds to a, b, c position 4 3 0 indicating that it is 4 units along a, 3 units along b, and at the 0 c plane. In Figure lOa, a lattice, or Moire, grating clearly exists shown as 37a, 37b, 37c. The gratings 37 continue throughout the lattice, or Moire, holographic grating but for clarity of the drawing, just three are shown. Applying the Moire Indices rules, analogous to the Miller Indices rules, the Moire Indices for 37a are calculated as :- (l/a lib lIc) where a is 1 b is I and c is oo (infinity) Thus the Moire Indices are 1 1 0, and the Moire grating is defined as (1 1 0) the first order, or prime, lattice grating.

Applying the same analysis process to 37b, 37c and so on produces the same Moire grating definition.

In Figure lOb, applying the same rules to Moire grating 39a gives the Moire grating definition (2 1 0), which is also the result for 39b, 39c and so on. This is a second order Moire grating.

In Figure 10c, applying the same rules to Moire grating 41a gives the Moire grating definition (3 1 0), which is also the result for 41 b, 41c and so on. This is a third order Moire grating.

In Figure 10d, applying the same rules to Moire grating 42a gives the Moire grating definition (1-1 0), which is also the result for 42b and so on. This is again a first order Moire grating, but orthogonal to the (1 1 0) Moire grating. As with Miller indices so Moire indices with negative terms may also be written with a negative sign printed above the corresponding index. For convenience they are written herein as negative quantities.

The nomenclature for such grating pitches is AM (p) where (p) is the Moire Indices of the grating, and M is indicative that it is a Moire pitch. Similarly the corresponding Bragg wavelength associated with a (p) grating is referred to as M (p) wherein, using equation (1) the Bragg wavelength is given by :- km (p) = (2 He AM (p))/m (5)

where ne is the effective refractive index of the substrate or waveguide, and m is the number of wavelengths km (p)/ne inside the material in terms of 2 AM (p) ; often m is unity and for this particular case km (p) = 2 l1e AM (p).

Hereinafter, the description refers to deriving the Moire grating pitch AM (p). However, by substituting for AM (p) in equation (5), the Bragg wavelength km (p). may be derived.

In deriving the Moire Indices of a grating, the intercept point on a chosen axis may be negative i. e. a Moire Index of-1 or-2 etc- First order Moire gratings are always in pairs, e. g. M (I 1 0) and M (l-1 0) (M (1 -1 0) is equivalent to M (-1 1 0) ), whereas there are always four gratings of a chosen higher order e. g. M (l 2 0), M (2 1 0), M (-1 2 0) and M (2-1 0).

The diminution of the strength of the higher order Moire gratings can be in part compensated by invoking the third axial dimension c, which in practical terms means that the grating is etched deeper into the structure semiconductor material.

By this means the Bragg grating pitch : in dependence upon the wavelength of the two interfering electromagnetic beams, the angle between the two interfering electromagnetic beams and the angular offset of the target substrate surface between the first and second exposure, wherein the depth of etching of the defined grating structure into the target

substrate surface is used to strengthen the effect of the grating upon the incident waveguided electromagnetic energy.

By comparing Figure lOa and Figure 9, and using equations (3) and (4) the pitch of the first order Moire gratings AM (I -I 0) and AM (ll 0) are given by :-

AM (I-1 0) =Ao/ (2 Sin+/2) and AM (t1 0) =Ao/ (2 Cos /2).

The first order grating pitch AM (I I 0) is shown on Figure lOa, and the first order grating pitch AM (1 -1 0) is shown on Figure 10d. Only first order Moire gratings are orthogonal, higher order gratings are not.

Substituting for Ao from equation (2) gives :-

AM (t -I 0) = (ÀO) I ( (2 x SinO/2) x (2sin+/2)) (6) and

It can also be shown using trigonometry and equation (2) that pitch AM (-i 2 0) or AM (-2 o) and

AM (I 2 0) or AM (2 I 0) of the second order Moire gratings are given by :-

AM (-2) o) = AM (-I 20) = AO/ (I + 8 x Sin ( {)/2)' (8) and AM (2] 0) = AM () 20) = Ao/ (1 + 8 X Cos/2)' (9).

Above a second order Moire grating the strength of the gratings become progressively weaker and use of deeper etching into the semiconductor substrate is used as a means of strengthening their effect. There is a finite, but very large, series of higher order Moire gratings with decreasing pitch. The native pitch used in fabricating a Moire grating would be suitably selected to accommodate a high order grating if this was required to be used.

Equations (6) and (7) show that the Bragg grating under consideration has a pitch which depends upon the wavelength of the two interfering electromagnetic beams, the angle between the two interfering electromagnetic beams and the angular offset, between the first and second exposure, of the target substrate surface.

Hence, for a chosen angle 9, it is easy to produce a Bragg grating of a desired pitch simply by selecting an appropriate angle , through which the wafer is rotated between exposures.

The gratings have been portrayed diagrammatically as simple lines. In reality the native gratings are sinusoidal producing graded density of hardness across the pitch. Thus where the native lattice gratings intersect there is a further graded hardening of the photo-resist. By choice of photoresist, angle of holographic interference, and etch time, the profile of the Moire gratings can be manipulated to produce a variety of cross section profiles as shown in Figures 11 la, b, c and d. Thus by varying the exposure times and or development time of the photosensitive recording medium the planar pattern of the intersecting grating structure may be determined. The developed photoresist planar pattern has the form of peaks and wells. For a given exposure and development time a unique mark space ratio of peaks and wells will be produced in the fabricated grating.

With a first order Moire grating M (1 1 0) produced with equal native gratings in the first and second exposure, and 0 (the angle between interfering common wavelength coherent energy beams) of 90 , and a 25% or 75% fill factor i. e. etching to 25% or 75% of the sinusoidal degree of hardness, or softness, depending upon whether positive or negative photoresist has been used, substantially circular planar pattern grating structures 50 are formed as shown in Figure 11 a.

In general with a. first order Moire grating M (1 1 0) produced with equal native gratings in the first and second exposure, and 0 (the angle between interfering common wavelength coherent energy beams) not 90 and a 25% or 75% fill factor i. e. etching to 25% or 75% of the sinusoidal degree of hardness, or softness, depending upon whether positive or negative

photoresist has been used, substantially elliptical planar pattern grating structures 52 are formed, as shown in Figure 11 b.

With a second order Moire grating M (2 1 0) produced with equal native gratings in the first and second exposure, and 8 (the angle between interfering common wavelength coherent energy beams) of 90 , and a 50% fill factor i. e. percentage of the sinusoidal degree of hardness, or softness, depending upon whether positive or negative photoresist has been

used, substantially square planar pattern grating structures 54 are formed, as shown in Figure lie.

With a second order Moire grating M (2 1 0) produced with equal native gratings in the first and second exposure, and 0 (the angle between interfering common wavelength coherent energy beams) not 90 , and a 50% fill factor i. e. percentage of the sinusoidal degree of hardness, or softness, depending upon whether positive or negative photoresist has been used, substantially rectangular planar pattern grating structures 56 are formed, as shown in Figure 11 Id.

A variation of the invention is to make the native pitch of the first and second exposure ', dissimilar, as weil as making an angular offset of the target substrate surface between the first and second exposure. There exist a number of measures that can be taken to achieve this :-

1. displacement of the target substrate along the axis of the bisector of the two coherent beams; and/or 2. non-orthogonal rotation of the target substrate; and/or 3. tilt of the target substrate; and/or 4. shift of the target substrate in the plane of the target substrate surface.

The magnitude of any of the above variables represent small changes that are minor adjustments to the holographic equipment wafer holder.

Figure 12a schematically shows the first of these measures. The two coherent beams 17 and 21 illuminate the target substrate 8a. As a consequence of the beams 21 and 17 being divergent, the pitch of the interference grating progressively increases away from the centre of the target wafer. Normally, by appropriate design of the optical system the magnitude of this increase is very small compared to the pitch of the sinusoidal interference grating being recorded, and as such is within the tolerances required for all practical purposes. However, by displacing the target substrate 8a along the axis of the bisector of the two beams 400, a small defined distance bu to either of positions 8d, the target substrate surface can be exposed to holographic interference fringes of different native pitches as diagrammatically represented in Figure 12b, with gratings 20 a, b or c.

Tilt of the target substrate can be achieved by mounting the wafer on a wedge. Nonorthogonal rotation of the substrate surface can be achieved by rotation of the wafer on a wedge.

The purpose of creating different native pitch gratings is to enhance the effectiveness i. e. a stronger reflective or transmissive effect, of the selected Moire grating by means of modifying the planar cross section profile of the structure making up the grating.

Irrespective of the method used to achieve dissimilar native pitches the magnitude of the required differences are small. As a consequence only minor adjustments to the holographic equipment wafer holder are required. Hence exposure calibrations are unaffected.

Figures 13a and b. show dissimilar native pitch Ao'and Ao"grating geometries but for clarity with grossly exaggerated pitch differences so that the first order Moire gratings M (I 1 0) and M (1 -1 0) may be more clearly identified and their respective geometries appreciated.

Figure 13a shows the Moire grating M (I 1 0) pitch AM (I I 0). Figure 13b shows the Moire grating M (l -1 0) pitch AM (1 -10) The pitch of the M (I 1 0) Moire grating AM (1 10) is given by

where Ao'is the native pitch of the first holographic exposure, Ac"is the native pitch of the second holographic exposure, and # is the angular offset of the target substrate between the first and second exposures.

The pitch of the M (1 -1 0) Moire grating AM (l-I 0) is given by :-

where Ao'is the native pitch of the first holographic exposure, Ao"is the native pitch of the second holographic exposure, and is the angular offset of the target substrate between the first and second exposures.

Figure 14 shows the geometry of the second order Moire gratings M (1 2 0), M (2 1 0), M (-l 2 0) and M (2-1 0) produced with dissimilar native pitch gratings Ao'and Ao". The pitch of the gratings is defined by their corresponding shortest distance to the point O.

The pitch of the M (2-1 0) Moire grating AM (2 -I 0) is given by :-

where #0' is the native pitch of the first holographic exposure, Ao"is the native pitch of the second holographic exposure, and is the angular offset of the target substrate between the first and second exposures.

The pitch of the M (-1 2 0) Moire grating AM (- 20) is given by :-

where Ao'is the native pitch of the first holographic exposure, Ao"is the native pitch of the second holographic exposure, and < j) is the angular offset of the target substrate between the first and second exposures.

The pitch of the M (1 2 0) Moire grating AM (l2 0) is given by :-

where #0' is the native pitch of the first holographic exposure, is the native pitch of the

second holographic exposure, and ( is the angular offset of the target substrate between the I first and second exposures.

The pitch of the M (2 1 0) Moire grating AM (21 0) is given by :-

where Ao'is the native pitch of the first holographic exposure, Ao"is the native pitch of the second holographic exposure, and c is the angular offset of the target substrate between the first and second exposures.

In equations (10) to (15), the parameters Ao'and Ao"may be restated in terms of the wavelength of the laser used to produce the holographic system and the angle 6 between the two interfering beams of laser energy, viz :-

/\o'=o/ (2xSine72) and ^o" = o/ (2 x Sin 8"/2).

9'and 0"will not be the same when the target substrate is displaced along the axis of the bisector of the two coherent beams.

A further method of strengthening Moire gratings is to use the technique of blazing. With Moire gratings this is achieved by having dissimilar exposure duration times for the first and second controlled exposures whilst keeping the overall exposure time constant. The effect of

such a blazing process is that the grating laid down in the photoresist has a shorter (or longer) first exposure interference pattern relative to that of the second exposure. The choice of longer or shorter first exposure is governed by the desired blaze direction. During photoresist development the first interference grating can consequently be made to have greater or lesser dominance in the definition of the Moire Grating structure. An example of a blazed second order lattice grating is shown at 130 in Figure 15.

By making the first exposure take 35-45% of the overall exposure time, for example, the resultant blazing of the second order lattice grating transforms what would have been rectangular or square profiled forms into linked parallelogram shapes, and what would have been circular profiled forms become elliptical in character. By suitable manipulation of exposure time ratios, and etch times (fill factor) the modified second order lattice grating forms are made to blend in the direction of the chosen grating.

The benefit of blazing is that the chosen grating becomes strengthened. Blazing also has the advantage that the depth of etching to form the Moire grating in the waveguide structure is reduced and a corresponding reduction in the growth layer thickness. Both these benefits result in an improved production yield of the device.

Claims

CLAIMS 1. A method of defining gratings in an optical waveguide device comprising the steps of providing a target substrate surface having a photosensitive coating, providing means to irradiate said coating with two coherent electromagnetic beams, exposing the coating to such irradiation to define a first grating therein, rotating said target substrate surface through a predetermined angular offset and exposing said coating a second time to define therein a second grating which intersects said first grating.
2. A method as claimed in claim 1, wherein the photosensitive coating is photoresist.
3. A method as claimed in either claim 1 or claim 2, wherein the target substrate is a dielectric material.
4. A method as claimed in either claim 1 or claim 2, wherein the target substrate is a semiconductor material.
5. A method as claimed in claim 4 wherein the target substrate comprises layers of Indium Gallium Arsenide Phosphide.

<Desc/Clms Page number 34>
6. A method as claimed in, any one of the preceding claims, wherein the photosensitive coating is developed to establish in the substrate a planar pattern of the intersecting grating structure.
7. A method as claimed in claim 6, wherein the exposure times and/or development time of the coating may be varied to give a planar pattern of the resultant grating structure which is substantially an array of rectangular forms.
8. A method as claimed in claim 6, wherein the exposure times and/or development time of the coating may be varied to give a planar pattern of the resultant grating structure which is substantially an array of circular forms.
9. A method as claimed in claim 6, wherein the exposure times and/or development time of the coating may be varied to give a planar pattern of the resultant grating structure which is substantially an array of elliptical forms.
10. A method as claimed in claim 6, wherein the exposure times and/or development time of the coating may be varied to give a planar pattern of the resultant grating structure wHich is substantially an array of square forms.
11. A method as claimed in any one of the preceding claims, wherein, between the first and second exposures measures are taken to make the grating pitch created by the

<Desc/Clms Page number 35>

second exposure dissimilar to the pitch created by the first exposure, said measures comprising : displacement of the target substrate along the axis of the bisector of the two coherent beams ; and/or non-orthogonal rotation of the target substrate ; and/or tilt of the target substrate; and/or shift of the target substrate in the plane of the target substrate surface.
12. A method as claimed in any one of the preceding claims, wherein the duration times of the first and second exposures are dissimilar whilst keeping the overall exposure time constant, whereby there is a blazing of the defined lattice grating.
13. A method of creating gratings in an optical waveguide device substantially as described herein with reference to any one of Figures 6 to 15 of the accompanying drawings.
14. A Bragg grating created by a method as claimed in any one of the preceding claims wherein the pitch of the grating depends upon the wavelength of the two interfering electromagnetic beams, upon the angle between the two interfering electromagnetic beams and upon the angular offset, between the first and second exposure, of the target substrate surface.

<Desc/Clms Page number 36>
15. A Bragg grating as claimed in claim 14, wherein the pitch #M(1 -1 0) of the grating M (1 -1 0) is given by the equation :-

where Ao is the native pitch of the holographic interference fringe and < )) is the angular offset of the target substrate surface between the first and second exposures.
16. A Bragg grating as claimed in claim 14, wherein the pitch ^M (I I 0) of the grating M (l 1 0) is given by the equation :-

where Ao is the native pitch of the holographic interference fringe and is the angular offset of the target substrate surface between the first and second exposures.
17. A Bragg grating as claimed in claim 14, wherein the pitches #M(-2 1 0) and #M(1 -2 0) of the gratings M (-2 1 0) and M (1 -2 0) are given by the equations :-

where Ao is the native pitch of the holographic interference fringe and is the angular offset of the target substrate surface between the first and second exposures.

I
18. A Bragg grating as claimed in claim 14, wherein the pitches AM (2 I o) and AM (1 2 0) ouf the gratings M (2 1 0) and M (1 2 0) are given by the equations :-

<Desc/Clms Page number 37>

where Ao is the native pitch of the holographic interference fringe and ()) is the angular offset of the target substrate surface between the first and second exposures.
19. A Bragg grating as claimed in any one of claims 14 to 18, wherein the defined grating structure is etched into the target substrate surface at such a depth as to strengthen the effect of the grating upon incident waveguided electromagnetic energy.
20. A Bragg grating as claimed in any one of claims 14 to 19, wherein the grating rulings, or fringes, are substantially parallel to the crystallographic structure of the target substrate.
21. A Bragg grating created by a method as claimed in Claim 11, wherein the pitch AM (t- 10) of the grating M (1 -1 0) is given by the equation :-

where Ao'is the native pitch of the first grating, Ao"is the native pitch of the second grating, and is the angular offset of the target substrate between the first and

second exposures.

\
22. A Bragg grating created by a method as claimed in Claim 11, wherein the pitch AM (I 10) of the grating M (l 1 0) is given by the equation :-

<Desc/Clms Page number 38>

where Ao'is the native pitch of the first grating, Ao"is the native pitch of the second grating, and # is the angular offset of the target substrate between the first and second exposures.
23. A Bragg grating created by a method as claimed in Claim 11, wherein the pitch AM (2- 10) of the grating M (2-1 0) is given by the equation :-

where #0' is the native pitch of the first grating,/\o"is the native pitch of the second grating, and is the angular offset of the target substrate between the first and second exposures.
24. A Bragg grating created by a method as claimed in Claim 11, wherein the pitch ^M (-l 20) of the grating M (-l 2 0) is given by the equation :-

where Ao'is the native pitch of the first grating, Ao"is the native pitch of the second grating, and is the angular offset of the target substrate between the first and second exposures.

I 1
25. A Bragg grating cheated by a method as claimed in Claim 11, wherein the pitch AM (2 1 0) of the grating M (2 1 0) is given by the equation :-

<Desc/Clms Page number 39>

where Ao'is the native pitch of the first grating, Ao"is the native pitch of the second grating, and 4 > is the angular offset of the target substrate between the first and second exposures.
26. A Bragg grating created by a method as claimed in Claim 11, wherein the pitch AM (l 2 o) of the grating M (1 2 0) is given by the equation :-

where Ao'is the native pitch of the first grating, Ao"is the native pitch of the second grating, and # is the angular offset of the target substrate between the first and second exposures.
27. A Bragg grating substantially as described herein with reference to any one of Figures 6 to 15 of the accompanying drawings.