Rare earth doped sol-gel thin films

The ongoing fast evolution of the optical telecommunication network has generated an increasing demand for optically active devices. Rare earth (RE)-doped glasses are very well suited for the fabrication of such devices. While early work was directed towards fibre lasers and amplifiers, researchers in later years attempted to explore waveguide amplifiers in integrated optics form. Fibre-based amplifiers offer advantages in terms of easy and efficient coupling to the optical transmission network. The development of the erbium-doped fibre amplifier (EDFA) in the eighties had an important influence on the evolution of the optical fibre network [1-3]. The EDFA allowed optical signal amplification and regeneration without the former necessary transformation of the optical signal to an electrical signal and vice versa. However, the same amplification device in a planar waveguiding geometry is advantageous in terms of compactness and ability to integrate other functional elements on the same chip. The lossless splitter is a good example of a device that can be easily integrated on a chip. The size of waveguide amplifiers should be in the order of centimetres, which is small compared to fibre devices. Thus, in order to achieve comparable gain, waveguide amplifier materials have to be heavily doped. These required high doping rates are the main challenge of waveguide amplifiers. Indeed, the solubility of RE ions in glass is very limited. Furthermore, ion-ion interactions such as cooperative up-conversion or energy transfer are very likely to occur at high RE ion concentration. Achievement of significant optical gain is not straightforward. These problems could, however, be overcome and erbium-doped amplifiers and lossless splitters were obtained by ion exchange in silicate [4] or phosphate glass [5] and by ion implantation into Al2O3 [6]. Today, erbium-doped waveguide amplifiers (EDWA) are commercially available [7]. The realisation of ion-exchanged waveguide DFB (Distributed FeedBack) lasers on phosphate Er-Yb-codoped glass was also reported recently [8].
The most studied RE element is erbium. Its photoluminescence (PL) emission is situated at 1.55 µm, in the centre of the third telecommunication window, corresponding to the minimum absorption region of optical fibres. The first report on a sol-gel erbium-doped glass dates back to 1987[9]. Praseodymium is a further RE element of interest for amplification in the second telecommunication window at 1.32 µm, corresponding to zero dispersion wavelength of optical fibres. The development of the EDFA and serious problems found with efficient Pr doping have, however, limited its interest for telecommunications. But the application field of active integrated devices is not limited to telecommunication devices. Light sources in the visible and near infrared are needed for lasers [10], sensors, medical applications or displays [11]. Actually, the great number of RE elements allows to cover most wavelengths from blue to near infrared.
Compared to other techniques, such as melt quenching, sol-gel processing is a very recent technique to produce glasses or ceramics. Because of its inherent advantages, it has become a very popular technique. This is partially due to the fact that it allows the elaboration of high quality materials without any important investment. The equipment needed for laboratory-scale production consists of a narrow-necked glass flask with stir bar on a stirring hot plate. For thin film production a dip-coating device or a spin-coater together with a furnace has to be added. Compared to vacuum deposition machines, the investment is negligible. Moreover, the specificity of its chemistry offers several advantages. The use of high purity chemicals instead of minerals allows obtaining very pure materials. The stoichiometry of the glass can be easily controlled by mixing the different precursors in the liquid solution. The low viscosity of the sols allows to achieve high homogeneity at a molecular level. Due to these specific features, sol-gel allows to elaborate an enhanced variety of amorphous or crystalline materials. One example of materials, which are exclusively fabricated by sol-gel, consists in hybrid organic-inorganic ormocers or ormosils, which are of great interest for optical thin film devices. Thus, RE-doping becomes straightforward. Finally, sol-gel reactions take place at room temperature. It does not include any high temperature heat treatment. Heat treatment are usually performed after film formation for densification or pyrolysis of residual impurities. These heat treatments are, however, performed at relatively low temperatures. For example, silica glass can be prepared at 800-1000°C compared to 2000°C required for melting crystalline silica into glassy form.
There are, however, a number of weak issues in the sol-gel process. This weakness can be mostly attributed to the many process parameters which must be optimised and maintained reproducibly, the unavoidable material shrinkage during processing and the need to introduce and expel water during the sol-gel reaction. The last point is even more crucial for RE-doped materials as hydroxyl groups are responsible for very efficient luminescence quenching. Also, porosity must be removed, both to suppress waveguide losses due to light scattering and internal adsorption of atmospheric water.
The great amount of research which has been done over the last years, results in improved understanding of sol-gel chemistry [12]. Thus, the improved control of the process parameters allows the elaboration of high quality thin films of various compositions. Moreover, the adjustment of standard thin film patterning techniques, enables the realisation of integrated optics devices [13]. A very good introduction into the field of sol-gel and recent research activities can be found in Ref. [14].

1. Mears, R.J., Reekie, L., Jauncey, I.M., and Payne, D.N. 1987, Electron. Lett., 23, 1026.
2. Desurvire, E. 1994, Erbium-Doped Fibre Amplifiers, Wiley, New York.
3. Dejneka, M., and Samson, B. 1999, MRS Bul., 24, 39.
4. Camy, P., Román, J.E., Willems, F.W., Hempstead, M., van der Plaats, J.C., Prel, C., Beguin, A., Koonen, A.M.J., Wilkinson, J.S., and Lerminiaux, C. 1996, Electron. Lett., 32, 321.
5. Delavaux, J.M.P., Granlund, S., Mizuhara, O., Tzeng, L.D., Barbier, D., Rattay, M., Saint André, F., and Kevorkian, A. 1997, IEEE Photon. Tech. Lett., 9, 247.
6. van der Hoven, G.N., Koper, R.J.I.M., Polman, A., van Dam, C., van Uffeien, J.W.M., and Smit, M.K. 1996, Appl. Phys. Lett., 68, 1886.
7. Yenah, A., Delaveaux, J.M.P., Toulouse, J., Barbier, D., Srasser, T.A., and Pedrazzani, J.R. 1997, IEEE Photon. Tech. Lett., 9, 1099.
8. Blaize, S., Bastard, L., Cassagnetes, C., Vitrant, G., and Broquin, J.-E. 2002, SPIE Vol., 4640, 218.
9. Sun, K., Lee, W.-H., and Risen Jr, W.M. 1987, J. Non-Cryst. Solids, 92, 145.
10. Moncorgé, R., Merkle, L.D., and Zandi, B. 1999, MRS Bul., 24, 21.
11. Ballato, J., and Lewis, J. 1999, MRS Bul., 24, 51.
12. Brinker, C.J., and Sherer, G.W. 1990, Sol-Gel Science, Academic Press, San Diego.
13. Yeatman, E.M. 1997, SPIE Vol., CR68, 1.
14. Najafi, S.I. 1998, Selected Papers on Sol-Gel for Photonics, SPIE, Washington.