Germania-silica compositions can have good recovery from a short radiation dose, but recovery is incomplete and there is generally a residual increase in attenuation.  Increments decrease significantly at longer wavelengths.

What combinations of total dose, exposure time, operating wavelength and temperature are you considering? Initial increments are higher and recovery is slower low temperatures.

For 850 nm transmission, boron co-doping of germanosilicate core fibre can increase the transient response while still showing good long term recovery.  The permanent loss increases for wavelengths greater than 1100 nm.  At 1300 nm and longer wavelengths there is a significant increase in intrinsic attenuation before any radiation exposure.

Doping with phosphorus suppresses the transient but increases the long term or permanent damage, so is probably not what you are looking for.

Long term recovery tends to be better in un-doped silica core fibres.  The transient is generally higher and recovery slower in low hydroxyl ('dry') silica than in high OH flame-hydrolysed silica. The increase in attenuation with dose is highly non-linear in silica core fibres, and photobleaching effects (accelerated recovery at high optical intensity) can be significant.

Article Interlaboratory Comparison of Radiation-Induced Attenuation ...

i try to develop a distributed dosimeter with optical fiber with fading. i dont know exactly the range of radiation i consider, but its application is for nuclear power plant.

Have you some chemical numbers for  make a boron-codoping germanosilicate?

this way seems very interesting, thx.

I don't have exact composition data for borosilicate co-doped fibre.  Boron doping reduces the refractive index of silica, so it may not make sense to add large quantities.  You can get some idea of how much to increase the germania concentration to compensate from Hammond & Norman's 1977 OQE paper.  High concentrations of B203 and GeO2 will reduce glass viscosity, increase expansion coefficient mismatch, and may make processing more difficult.

Most of the telecommunications community lost interest in boron doping in the 1980's when 1300 and 1550 nm transmission became more important.

Check out E. J. Friebele & M. E. Gingerich "Radiation damage in optical fiber waveguides at long wavelengths", Adv. Ceram. vol 2, (Phys. Fiber Optics) pp387-92 (1981) for more details of the radiation response - I no longer have a copy to hand.

You need to think carefully about the dose, fibre length, exposure and recovery times.  Much of the work on transient exposure and recovery only considered timescales of microseconds to minutes or hours.  In typical doped silica fibres at room temperature, most of the recovery occurs within a few seconds, and there is limited change after that.  This may be too rapid for the application you are considering.

There are publications which report changes over longer time scales, but I have paid little attention to this area since 1990.

http://www.orc.southampton.ac.uk/publications/00xx/73.pdf

You are very helpful Alan. Thanks a lot.