GB2304929A - An optical waveguide comprising hydrogenated aluminosilicate doped with rare earth ions - Google Patents

Description

i" 2304929 1 Method of Dhotosensitisina an olDtical waveauide made from an

aluminosilicate material and wavecruide obtained sa d method The present invention concerns a method of photosensitising aluminosilicate optical waveguides.

It also concerns the optical waveguides obtained by this method.

The invention finds one particularly advantageous application in writing arrays such as Bragg arrays in optical fibres and planar waveguides.

The optical fibres most widely used in telecommunications are monomode germanosilicate fibres.

However, the typical order of magnitude of the index variations occurring on UV illumination of germanosilicate fibres is usually found to be limited to a few 10-5.

Many methods of increasing the photosensitivity of germanosilicate fibres have already been proposed.

Attempts have -been made to render them photosensitive by modifying their matrix and introducing photosensitising agents such as boron and/or germanium oxide.

It has also been found that hydrogenation of these fibres by highpressure (> 50 atmospheres) molecular hydrogen diffusion can produce particularly high photoinduced index variation amplitudes Q x 10-3 for a fibre doped with only 3.5% of Ge02)- The diffusion of hydrogen (or of molecular deuterium) is a technique now widely used by laboratories specialising in the production of components with arrays written in the fibres. It is also a research topic that is pertinent to both fundamental research and applied research. In particular, some Raman spectroscopic investigations have shown that the high sensitivity obtained in hydrogenated germanosilicate fibres is 2 related to the presence during exposure of elongation modes attributed to SiO-H, =- Ge-H and -Ge-H.

As yet, the details of the photosensitisation mechanisms are not understood, however.

Although germanosilicate fibres are the most widely used, other optical fibres or planar optical waveguides not doped with germanium oxide are also manufactured to meet specific requirements in the fields of laser optical communications or fibre sensors.

This applies, for example, to optical waveguides or optical fibres with a phosphosilicate, alUMinophosphosilicate or aluminosilicate matrix.

The article in ELECTRONICS LETTERS, vol. 30, no. 16, 4 August 1994, page 1311/1312 XPOO0468807, K. TAGAWA T ET AL: " SINGLE- FREQUENCY ER3±DOPED SILICABASED PLANAR WAVEGUIDE LASER WITH INTEGRATED PHOTO-IMPRINTED BRAGG REFLECTORS" discloses a method of photosensitising a phosphosilicate matrix by hydrogenation.

Aluminosilicate optical waveguides have the advantage of allowing use of all of the range of wavelengths from visible light to the near infra-red and therefore the three telecom windows: 0.85 gm, 1.3 gn and 1.55 gm.

Doping with aluminium is especially beneficial in the manufacture of 1.55 pm amplifying fibres, at which wavelength such fibres have a high and more uniform gain than germanosilicate fibres.

As early as 1991 it was shown that the incorporation of Ce3+ or Eu3+ ions into the core of aluminosilicate fibres made the fibres photosensitive. In this respect, reference may advantageously be had to the following publication:

(1] M.M. BROER, R.L. CONE and J.R.SIMPSON "Ultraviolet-induced distributed-feedback gratings in 3 Ce3l doped silica optical fibers", Opt Lett 16, 1991, p. 1391-1393, or, in the case of phosphosilicate optical waveguides or optical fibres, to the publication:

(2) K. 0 HILL, B. MALO, F. BILODEAU, D.C. JOHNSON, T.F. MORSE, A. KILIAN, L. REINHART and KYUNGHWAN on "Photosensitivity in Eu2+: A1203 - doped core fiber: preliminary results and application to mode converters" Digest of Conference on optical communication, 1991, paper PD3.

On the other hand, the person skilled in the art has traditionally believed that hydrogenation of an aluminosilicate matrix cannot improve its photosensitivity.

Unexpectedly, the inventors have found that hydrogenation significantly increases the effects of photosensitisation of doping an aluminosilicate matrix with rare earth ions.

At this time the inventors are not aware of any logical explanation of this phenomenon, the mode of action of the hydrogen probably being very different from that operative in the case of fibres containing germanium.

Accordingly, photosensitising the invention proposes a method of an optical waveguide made from an aluminosilicate material for the purpose of writing a Bragg array onto the waveguide, characterised in that the waveguide is hydrogenated by diffusion of molecular hydrogen at high pressure, the aluminosilicate material of the waveguide being doped with rare earth ions.

The invention also consists in optical waveguides obtained by this method and a method of making waveguides incorporating a Bragg array.

It further consists in optical fibres made by this method and optical components incorporating optical 4 is waveguides obtained in this way.

Other features and advantages of the invention will emerge further from the following description which should be read in conjunction with the accompanying drawings, in which:

- figure 1 shows the transmission spectra of Bragg arrays written in an aluminosilicate fibre doped with Tb3+ ions; - figure 2 shows the refractive index modulation as a function of the dose of ultraviolet radiation impinging on an aluminosilicate fibre doped with Tb3+ ions for various pump laser wavelengths; - figure 3 is a graph showing the spectrum characteristics of a Bragg array as it is written in a hydrogenated aluminosilicate fibre doped with Tb3+ ions; - figure 4 is a graph showing curves of the refractive index modulation as a function of the number of incident pulses for a hydrogenated fibre and for a non-hydrogenated fibre; - figure 5 is a graph showing the transmittance of a plurality of Bragg arrays during their writing, for a plurality of pump laser wavelengths; figure 6 is a graph showing the spectral transmittance of Bragg arrays subject to isochronal thermal destruction for a hydrogenated fibre doped with Ce3+ ions; - figure 7 shows the spectra of Bragg arrays before and after gamma irradiation, these arrays being written on a hydrogenated fibre doped with Ce3+ ions and a nonhydrogenated fibre doped with Ce3+ ions; - figure 8 shows the evolution of the refractive index modulation for Bragg arrays subject to isochronal thermal destruction, these arrays being written in a hydrogenated fibre doped with Tb3+ ions and corresponding to the spectra shown in figure 1; - figure 9 shows the absorption spectrum induced in the 1.4 gm spectral region by exposure of a hydrogenated aluminosilicate fibre doped with Ce3+ ions; - figure 10 shows the transmission spectra of an aluminosilicate preform plate doped with Ce3+ ions at different stages of writing; - figure 11 shows the absorption spectra induced in a hydrogenated aluminosilicate fibre during exposure and after exposure is stopped; - figure 12 shows the evolution of the loss excess spectra created by exposure of hydrogenated or nonhydrogenated aluminosilicate fibres doped with Ce3+ ions; - figure 13 shows the refractive index modulation as a function of the dose of ultraviolet radiation impinging on an aluminosilicate fibre doped with Tm3+ ions for various pump laser wavelengths.

The high-pressure (> 1 Bar, preferably between 100 bars and 1 000 bars) molecular hydrogen diffusion photosensitisation process was used on aluminosilicate fibres doped with Ce3+ or Tb3+ ions. The results of these experiments are explained hereinafter.

The inventors believe that these results are not limited to these specific dopants, but can be generalised to any aluminosilicate fibres doped or co-doped with rare earth ions having a transition in the band of ultraviolet radiation used (Er3+, Tm3+, Pr3+, Yb3+, for example).

FIBRE CHARACTERISTICS of 12 5 gm Table I summarises the optogeometric properties of three types of fibres (or preforms) Fl, F2, F3 used in the experiments. These fibres have an outside diameter Parallel sided plates with a typical thickness of 100 PM were cut from the preforms corresponding to these fibres and then polished to sufficient optical quality for use in spectroscopic 6 experiments.

Note that the photosensitising elements (cerium or other rare earths) are introduced as dopants (< 1%) whereas the constituent elements of the various matrices aluminium, germanium, phosphorus) are present in concentrations of a few percent (typically 5% to 30%).

Generally speaking, rare earth doping is optimum for concentrations between 1 000 ppm and 10 000 ppm.

Below 1 000 ppm the effects remain but are weak.

Above 10 000 ppm (i.e. 1%) it is difficult to retain the good spectral transmission properties of the glass. The rare earth oxide becomes a constituent element of the glass and it is no longer possible to speak of an aluminosilicate matrix.

EXPERIMENTAL PROTOCOL The photosensitivity of the fibres was estimated by measuring the reflectivity of Bragg arrays written in the fibres. It corresponds to the maximal value of the index modulation obtained for typical writing conditions. The arrays were written by exposing each fibre laterally to a system of ultraviolet fringes obtained with the aid of a "Lloyd" type mirror interferometer (or using a phase mask).

Various types of source were used:

- a pulsed laser that could be tuned in the ultraviolet band; - a KrF laser (X = 249 nm); - a continuous laser with an output optical power of 20 mW at 244 nm; - a continuous laser with an output optical power in the order of 100 mW at 244 nm or at 257 nm.

The principle of the experiment was to compare the photosensitivity of fibres loaded with hydrogen to that of untreated fibres. The arrays were written using various pump radiation wavelengths to determine the 7 spectral region in which to tune the pump laser to obtain the optimum photosensitivity.

HYDROGENATI The conditions for preliminary treatment of the fibres were as follows.

The fibres Fl doped with Tb3+ ions were placed for 70 days in an enclosure containing hydrogen at a pressure of 110 bars.

The fibres F2 doped with Ce3+ ions were kept for 120 days at a hydrogen pressure of 100 bars.

The fibres F3 were placed in a container of hydrogen at 180 bars for 14 days.

The hydrogenation techniques were those familiar to the person skilled in the art. As a general rule, hydrogenation is dependent on the molecular hydrogen pressure to which the fibres are subjected and the time and temperature of exposure.

Pressur Optical fibres are hydrogenated to improve their photosensitivity, usually at a pressure between a few bars and a few hundred bars, the percentage of molar hydrogen incorporated being proportional to the pressure.

For example, a pressure of 100 bars enables incorporation of approximately 1% molecular hydrogen at saturation. A pressure of 500 bars enables incorporation of approximately 5%.

The pressure to choose is therefore dependent on the degree of photosensitisation required.

For a standard germanosilicate fibre the appropriate pressures are generally between 50 bars and 100 bars.

For fibres that are highly photosensitive a hydrogen pressure of a few bars may suffice.

For fibres having a low photosensitivity very high pressures may be needed (up to 750 bars).

8 Above 200 bars (the standard packaging pressure), the method is difficult to implement.

Duration The duration of hydrogenation under pressure is determined by the time needed for the hydrogen to diffuse to the core (to the centre of the 125 jim diameter fibre).

At room temperature, saturation is obtained after 10 days to 15 days.

TemQez-at:ur The H2 saturation rate is highly dependent on the temperature. Heating speeds up the diffusion rate enormously.

On heating to 700C, for example, the time necessary to saturate the fibre with H2 is less than 24 hours. In practice, heating leads to a significant time saving.

To summarise, the hydrogenation pressures most commonly used are between 50 bars and 200 bars.

The duration (order of magnitude to obtain saturation) is in the order of 12 days.

Heating accelerates the process and a temperature of 700C enables the duration to be reduced to less than one day.

MAIN RESULTS Untreated fibres The array detection threshold corresponded to a reflectivity of 3/100.

The untreated fibres Fl and F3 were not found to be photosensitive.

The untreated fibre F2 was photosensitive provided that a pulsed laser was used for the exposure.

The range of pump wavelengths explored ranged between 240 nm and 300 nm. The maximum photosensitivity was obtained when the pump wavelength was around 265 nm. Using this pump wavelength, the inventors were able to inscribe an array 6.3 mm long with a reflectivity of It 1 9 96/100. These characteristics correspond to a maximal index modulation equal to 3.7 x 10-4 at 1.3 gm.

The growth of the array was a monotonic function of the dose of radiation. In the example given, inscription was stopped after 155 000 pulses (F + 255 mj/cm2).

When the pump wavelength was less than 250 nm, the photosensitivity dropped off sharply (8nmax < 3 x 10-5 at 1P = 243 nm).

Treated fibreg 1 Diffusing hydrogen into the fibre F, increased its photosensitivity by at least an order of magnitude.

Using a pulsed laser emitting at an appropriate wavelength and a fibre loaded with hydrogen, it is possible to write arrays with a reflectivity close to Rmax - Figure 1 shows the spectra of three arrays G1, 'G2 and G3, 8 mm long written in the same F1 type fibre using a pulsed laser and 15 000 pulses for G1, 1.2 x 105 pulses for G2 and 1.2 x 105 pulses for G3 at a pump wavelength of Xp = 240 nm, the dose being 150 mj/cm2 each time.

Figure 2 shows the evolution of the refractive index modulation at the wavelength of 1 335 nm as a function of the dose to which the fibre was exposed, for a wavelength of 240 nm (measured values represented by triangles) and 244 nm (measured values represented by lozenges)., In the spectral area explored, the photosensitivity appeared to increase as the pump laser wavelength was reduced.

It was found that the refractive index modulation An is conveniently represented as a function of the dose D impinging on the f ibre by a law of the type An=Da (with a < 1 and D in j/cm2) (curves Cl and C2).

Figure 3 shows the evolution of the spectral characteristics of an array as a function of the number of incident pulses impinging on the fibre F1 (measured values of transmittance (triangles) and Bragg wavelengths (lozenges) as a function of the number of pulses).

The Bragg wavelength of the array was shifted towards the red during exposure, representing an increase in the mean effective index of the fibre.

Array writing experiments were repeated for 7 days after removal of the fibre from the hydrogen enclosure. The results obtained were similar to those shown in figure 3 without any drop in photosensitivity being apparent. However, two weeks after removing the fibre from the enclosure the photosensitivity was sharply reduced (5nmax = 2 x 10-5).

No attempts were made to write networks using continuous lasers.

Treated fibres E2 Loading them with hydrogen increased the photosensitivity of the fibres F2 doped with Ce3+ ions. However, as when using the untreated fibres, the experiments to write arrays using continuous lasers failed.

Figure 4 compares the refractive index modulation evolution created by exposure of a hydrogenated fibre (solid circles in the figure) and a nonhydrogenated fibre (hollow circles). As with the aluminosilicate fibre doped with Tb3+ ions, the index modulation evolution could be represented as a function of the incident radiation D by a power of a law (a < 1) (curve C3) and the mean effective index increased on exposure.

The dependency of the photosensitivity on the pump laser wavelength was investigated in a narrow spectral region (250 nm - 240 nm}.

relativelv weak The dependency seemed in this spectral domain since the maximal amplitude of the index variation increased from 8.8 x 10-4 with Xp = 250 r1M to 11 x 10-4 with Xp = 245 = 11 or 240 nm.

Figure 5 illustrates this observation. This figure shows the evolution of the transmittance of Bragg arrays 2.5 mm long written at the wavelength of 1 335 nm with a dose of 250 mj/cm2, as a function of the number of pulses, for wavelengths of 245 nm (measured values represented by triangles), 240 nm (measured values represented by squares) and 250 nm (measured values represented by circles).

Fibres E3 The attempts to write arrays in the hydrogenated fibre F3 failed (pump laser operating in pulsed mode or continuous mode).

CONCLUSION5

The results are summarised in table II.

This table shows that the photosensitivity of aluminosilicate fibres is increased by hydrogenation of the fibre. The increase is greater than 4 times for fibres doped with cerium and greater than 10 times for fibres doped with terbium. The presence of rare earth (cerium, terbium or other) ions is necessary since no arrays were detected during attempts to write them in the undoped hydrogenated aluminosilicate fibre. Likewise, the use of a pulsed source proved to be necessary to achieve writing as no arrays were written using continuous sources.

PROPERTIES OF THE ARRAYS WRITTEN IN ALUMINOSILICATE FIBRES DOPED WITH Tb3+ AND Ce3+ IONS Durabilitv of the arravs The long-term durability of the arrays inscribed in the fibre F2 can be estimated from the following observations.

An array 3 mm long was inscribed in March 1995 using 13 000 pulses with a flux per pulse equal to 257 mj/CM2. Its reflectivity as measured a few minutes HYDROGENATED 12 after stopping the pulses was equal to 0.76 + 0.03. One hour after the pulses were stopped the reflectivity of the array was the same. The reflectivity of the array measured in June 1995 was equal to 0.77 + 0.03. The spectral transmittance of "saturated" (R = 0.99) arrays was measured regularly for 15 months with no significant change in the reflectivity or in the shape of the spectral response.

To specify the durability of the arrays, isochronal thermal destruction experiments were carried out on written in hydrogenated and non-hydrogenated The experiment entailed submitting the arrays to pulses of increasing amplitude for 30 minutes. the pulses the temperature of the fibre was to room temperature and the spectral reflectivity of the array was then measured. At the time of writing, these experiments have not been completed. It appears that the dose impinging on the fibre constitutes one parameter of the thermal resistance of the arrays.

Figure 6 shows, as a function of the temperature to which the fibres were exposed, the evolution of the transmittance of two arrays written by 1.5 x 104 pulses (measured values corresponding to the triangles) and by 4 x 104 pulses (measures values corresponding to the circles) with a dose of approximately 200 mj/cm2. These results are comparable to those that can be observed for type I arrays written in germanosilicate fibres strongly doped with germanium oxide.

The arrays written in the fibre F2 were gamma irradiated. The radiation was obtained from a cobalt source. The irradiation dose and rate were respectively equal to 3.3 x 105 Gy and 1.3 kGy/h. As shown by the spectra reproduced in figures 7a and 7b, on which the transmittance of the fibres before and after irradiation is shown (by the full line and dashed line curves, arrays f ibres. thermal Between reduced 13 respectively) f or a hydrogenated f ibre (f igure 7a) and a non- hydrogenated fibre (figure 7b), the reflectivity of the arrays was not substantially modified by gamma irradiation, whether the fibre was photosensitised by hydrogenation or not.

Isochronal thermal destruction experiments similar to those previously described provided an estimate of the durability of the arrays written in the fibre Fl. Figure 8 shows, as a function of the temperature to which the fibre was exposed, the evolution of the index modulation of Bragg arrays GI, G2 and G3 from figure 1 exposed to thermal pulses for 30 minutes. Raising the temperature of the fibre to less than IOOOC did not significantly modify the value of the index modulation.

Note that heating the fibre to 6000C did not completely destroy the arrays written with high doses of radiation.

Photochrognism created in the fibres bv writina arr_avs UV exposure of germanosilicate fibres photosensitised by hydrogenation produces excess losses in the ultraviolet region 1 < 300 nm and in the infrared region towards 1.4 gm, 1.25 gm and 0.95 gm. The loss excess in the infrared results from the formation of OH bonds.

The losses can be as high as 7 dB/cm at 1.41 gm, which corresponds to a concentration of OH bonds equivalent to that of the germanium introduced to dope the core of the fibre.

The inventors therefore investigated whether the exposure of the hydrogenated aluminosilicate fibres modifies their absorption spectra.

A 21 cm length of fibre F2 was exposed to 50 000 incident pulses with a flux per pulse equal to 250 mj/cm2 (X = 248 mm). The fibre had previously been kept for 47 days in hydrogen at a pressure of 120 bars. The 14 is absorption spectrum induced around 1.4 gm is shown in figure 9. The loss excess remained below 0.045 cm-1 (at 0.2 dB/cm). The concentration in OH bonds that leads to this numerical value of the excess losses was equal to 500 ppm. The concentration of [OHI bonds therefore remained significantly below that of the cerium, estimated at approximately 5 000 ppm 7 000 ppm.

However, note that the conditions for exposure of the fibre did not correspond in this experiment to saturation of the photosensitivity (see figure 4).

Moreover, the plates from preform F2 were placed in a hydrogen oven for 68 days (PH2 = 110 bars). The transmission spectrum of the core of a plate 70 gm thick is shown in figure 10 in the spectral domain {200 nm - 400 nm}.

Curve A shows the spectrum bef ore exposure. The transmittance of the plate was minimal at 292 nm. The core of the preform was then exposed to 9 x 104 light pulses at a wavelength of 250 nm. The pulses impinged on the plate with a flux per pulse equal to 220 mj/cm2.

Curve B shows the transmission spectrum after 1 500 pulses. Curve C shows the spectrum obtained one hour after the end of exposure. The spectrum did not evolve significantly during exposure after 1 500 pulses.

The spectra B and C show that exposure improved the transmittance of the plate in the ultraviolet and that the transmittance continued to increase long after the pulses were stopped.

The transmittance corresponding to the curve C was minimal towards 298 mm. Throughout the duration of the experiment the light flux from the deuterium lamp used to record the transmission spectrum impinged on the core of the preform.

it therefore appears that exposing an aluminosilicate glass doped with cerium and treated by hydrogenation improves the transmittance of the glass in the spectral region explored and displaces the wavelength at which the minimum transmittance is obtained towards the red. This observation was specific to the glass treated by hydrogenation. In a non-hydrogenated glass, exposure under similar conditions to those just described reduced the transmittance and did not modify the value of the wavelength at which the transmittance was minimal. The loss excess healed spontaneously. It had halved 20 hours after firing of the laser was stopped.

An aluminosilicate fibre doped with cerium and treated by hydrogenation was exposed laterally to pulses at a wavelength of 244 nm impinging on the fibre with a dose per pulse equal to 150 mj/cm2.

The absorption spectrum induced in the spectral domain (480 = - 750 nm} is shown in figure 11a. The various curves are for different numbers of pulses impinging on the fibre. The exposure created a loss excess that saturated at a maximal value towards 1 000 pulses, and then decreased slightly to a stationary value.

Figure 11b shows that the loss excess healed spontaneously after stopping the pulses. Figure 12 enables a comparison between the behaviour of a hydrogenated fibre (curves c and D respectively corresponding to the absorption spectrum after stopping exposure and 60 minutes later) and that of an untreated fibre (curves A and B respectively corresponding to the absorption spectrum after stopping exposure and 120 minutes afterwards).

In conclusion, it appears that hydrogenation of the aluminosilicate fibre doped with cerium ions increases the photosensitivity of the fibre without creating loss excess in the visible region. The transmittance of the glass in the spectral region (220 = - 400 nm} was

16 improved by exposure.

Lateral exposure of aluminosilicate fibres doped with terbium lead to a modification of the absorption spectrum in the visible region. When the fibre was hydrogenated (and therefore photosensitive) the excess losses healed spontaneously after the pulses were stopped, with a dynamic similar to that observed on exposing fibres doped with cerium. Absorption spectra induced in the ultraviolet region were recorded by exposing the core of the plate from the pref orm. The experiments proved to be difficult because of the high diffusion in the glass- Exposure of the plate by radiation at a wavelength of 240 = did not significantly modify its transmittance. An induced absorption band was created around 240 nm. This band has not yet been identified and could reveal a modification to the environment of the Tb3+ ions.

GENERAL CONCLUSION ON ALUMINOSILICATE FIBRES DOPED WITH _Ce3+ OR Tb3+ IONS

The hydrogenation of aluminosilicate fibres doped with Ce3+ or Tb3+ ions lead to an increase in the photosensitivity of the fibres by a factor greater than 4 or 10 depending on the ions used to dope the fibre. Heating the fibre during writing of the array can also further increase the photosensitisation produced by hydrogenation. These results can be generalised to fibres doped by other rare earth ions.

In hydrogenated fibres the writing of the arrays was accompanied by a slight degree of photochromism (0. 5 cm-1 = Aa at the start of exposure) which healed during the pulses and disappeared spontaneously after writing was stopped). At 1.39 pm the creation of OH bonds caused a loss excess in the order of 0.05 cm-1.

Treatment of the fibres with deuterium eliminates this effect if it proves to be undesirable.

17 GENERALISATION TO DOPING WITH OTHER RARE EARTH IONS As already mentioned, the inventors consider that the results they have obtained on aluminosilicate fibres doped with Ce3+ or Tb3+ ions can be generalised to any aluminosilicate fibre doped or co-doped with other rare earth ions having a transition in the ultraviolet radiation band used.

First results have been obtained for fibres doped with Er3+ or Tm3+ ions subjected to hydrogenation by molecular hydrogen diffusion at a pressure of 100 bars for 30 days.

For exposure at the wavelength of 235 nm, a refractive index modulation has been obtained of 4.3 x 10-5 for aluminosilicate fibres doped with Er3+ ions and of 6.6 x 10-5 for aluminosilicate fibres doped with Tm3+ ions.

The curves A and B in figure 13 show the refractive index modulation as a function of the exposure dose for aluminosilicate fibres doped with Tm3+ ions subjected to hydrogenation at a pressure of 100 bars for 42 days and exposure at a wavelength of 235 nm with pulses of mj/cm2 (curve A) and at a wavelength of 240 nm with pulses of 150 mj1CM2 (curve B).

it will of course be understood that the present invention has been described above purely by way of example, and that modifications of detail can be made within the scope of the invention.

TABLE I

OPTOGEOMETRIC CHARACTERISTICS OF THE ALUMINOSILICATE FIBRES USED IN THE EXPERIMENTS CNET Core Nature and cladding LP11 mode reference diameter concentration dopants cut-off of core dopanto wavelength 1 F1 5. 2 pm A1203 F 1. 05 JAM Tb203 [Tbl - 5 000 ppm F2 4. 7 gm A1203 7% mole F, P0.74 pm Ce203 000 ppm <[Cel< 8 000 ppm F3 3.7 gm 1 A1203 16% mole 1 F, P 1 1.26 pm 1 C0 TABLE II

MAIN RESULTS OF EXPERIMENTAL WRITING OF ARRAYS IN HYDROGENATED AND NON-HYDROGENATED ALUMINOSILICATE FIBRES Fibre Non-hydrogenated fibre Treated fibre: typically 60 days at 100 bars source Pulsed source Continuous source Pulsed source Continuous source F1 Failed; experimental Succeede conditions: 238 nm < lp < 244 nm 1P = 240 nm; 8nmax = 1.8 x 10-4 pulses; Not attempted Not attempted F=160 mj/cm2, Ll=8 mm 1B = 1 335 nm 8n < 1.3 x 10-5 F2 Succeeded Faile Succeeded Faile 244 nm < lp.5 300 nm lp = 244 nm, 275 nm, Pumping was more lp = 244 nm, 257 nm 8nma X 10-4 2 x = 3.6 308 nm effective at 240 nm - I = 100 W/cm at PLP = 265 nm, I = 50 W/cm2 245 nm than at 250 nm t = 60 min 1.5 x 105 pulses t = 120 min 8nmax = 1.2 X 10-3 XB = 1 550 nm; F = 250 mj/CM2 L, = 10 mm - 30 mm (8 X 104 pulses at L' = 10 mm U=1 550 nm, 1 330 nm 250 mj/cm2) 8n < 2.5 x 10-5 8n < 10-5 F3 Failed Faile Faile lp = 244 nm lp = 244 nm; 2 x 105 lp 244 nm 2 x 105 pulses pulses at 230 mj/cm2 1 50 W/cm2 F = 250 mj/cm2 Not attempted XB = 1 335 nm t 60 min L' = 7 mm L 1 = 7 mm L' = 1 cm 8n < 1.3 x 10-5 8n < 1.3 x 10-5 1B = 1 5 5 0 nm 8B = 1.2 x 10-5 F-& W

Claims (15)

1. Method of photosensitising an optical waveguide made from an aluminosilicate material for the purpose of writing a Bragg array on the waveguide, wherein the waveguide is hydrogenated by diffusion of molecular hydrogen at high pressure, the aluminosilicate material of the waveguide being doped with rare earth ions.

2. Method according to claim 1 wherein the rare earth ions are Ce3+ or Tb3+ ions.

3. Optical waveguide made by the method of either of the preceding claims from a hydrogenated aluminosilicate material doped with rare earth ions.

4. Method of making a Bragg array on a waveguide wherein the waveguide is of the type def ined in claim 3 and is written by exposure to ultraviolet radiation.

5. method according to claim 4 wherein exposure is effected by means of a pulsed laser source.

6. Method according to claim 5 wherein the wavelength of the source is substantially between 240 and 245 nm.

7. Method according to any one of claims 4 to 6 wherein the waveguide is heated during writing of the array.

8. optical waveguide made by the method according to any one of claims 4 to 7.

9. Optical f ibre according to claim 8.

10. Optical component according to claim 8.

constituting a waveguide incorporating a waveguide

11. Method of photosensitising an optical waveguide made from an aluminosilicate material substantially as hereinbefore described with reference to the accompanying drawings.

12. Optical waveguide substantially as herein- before described with reference to the accompanying 21 drawings.

13. Method of making a Bragg array on a waveguide substantially as hereinbefore described with reference to the accompanying drawings.

14. optical fibre substantially as hereinbefore described with reference to the accompanying drawings.

15. Optical component substantially as hereinbefore described with reference to the accompanying drawings.