The radiation sensitivity of Bragg gratings written with a femtosecond IR laser was measured for the first time.
Type I-IR and type II-IR gratings were written into hydrogen loaded as well as unloaded fibers of distinctly different
radiation sensitivity with the intention to find extremely radiation resistant gratings for temperature or stress
measurements in radiation environments, as well as very radiation sensitive ones for radiation dose measurements. With
a highly radiation-hard F-doped fiber we found a radiation-induced wavelength shift between about 3 and 7 pm after a
dose of 100 kGy. These are the lowest shifts observed so far. In such fibers it is very difficult to write gratings with an
UV laser. However, gratings made of the highly radiation-sensitive fibers only showed shifts of about the same size as
those made of the quite radiation-insensitive Corning SMF-28e fiber. This was already observed with UV laser gratings
written in such fibers.
We have measured for the first time the shift of the Bragg wavelength and other parameters of fibre Bragg grating (FBGs) exposed to γ-radiation at different wavelengths up to a dose of 100 kGy. The results show that the FBG sensitivity to radiation increases from 820 nm to 1516 nm so that FBGs with higher Bragg wavelength, e.g. 1.55 μm, are good candidates for high dose radiation sensing: no saturation was observed up to 100 kGy and the measured wavelength shift was fairly independent on the radiation dose rate.
Three different fibre optic radiation sensor systems are described. Two are based on the radiation-induced attenuation increase of radiation sensitive doped fibres, whereas with the third system the Cerenkov light generated by relativistic electrons in radiation hard undoped fibres is detected. All three systems are successfully used for beam optimization and dose measurement at three German electron accelerators.
It is shown that fibers doped with (Ge + P) or with (GE + B) can be used with good accuracy for fluence determinations of fast or slow neutrons, respectively, when they are calibrated at sources of known flux. Calibration at easily available gamma sources, as well as fast neutron dose determination, are complicated by the facts that energetic recoil protons generated by fast neutrons in H-containing fiber coatings contribute to the dose in the fiber core, and that the highly ionizing secondary particles released by fast and slow neutrons will cause significantly lower light absorption than beta or gamma rays of the same dose.
The radiation-induced loss of single mode fibers with undoped and Ge-doped core material was measured at 1300 nm and 1550 nm in the time range from 0.1 seconds to approximately less than 500,000 seconds at room temperature (plus 25 degrees Celsius). With the Ge-doped fiber, measurements were also made at minus 50 degrees Celsius and plus 80 degrees Celsius. The ratio of the losses at 1300 nm and 1550 nm increased at plus 25 degrees Celsius with both fiber types within about 10 seconds to a maximum value of approximately equal to 2 (Ge-doped) or approximately equals 2.7 (undoped). Then the ratio decreased continuously and became less than 1 (equals higher loss at 1550 nm) after about 35,000 - 70,000 seconds (Ge-doped) or approximately greater than 350,000 seconds (undoped). After approximately equal to 500,000 seconds a value of approximately equal to 0.75 was reached with the Ge-doped fiber, with an observable tendency to fall further. At minus 50 degrees Celsius the ratio increased up to about 2.2 and remained constant (as if frozen) during the whole irradiation time of 500,000 seconds. At plus 80 degrees Celsius, however, the radiation- induced loss at 1550 nm was higher at the beginning and became lower than the one at 1300 nm only after an irradiation time of approximately greater than 10<SUP>4</SUP> seconds. Additionally the annealing time of loss was measured after the end of irradiation for varying irradiation times between about 3 seconds and 400,000 seconds. The results can qualitatively explain the radiation-induced loss curves at 1300 nm and 1550 nm as well as their ratio.
Optical fibers find rapidly growing use also in the nuclear industry. The dependence of their radiation-induced loss on fiber type, wavelength, temperature, light power, dose rate, and radiation type (gamma rays, neutrons) is pointed out and test results of modern (1989 - 1993) single mode (SM), graded index (GI), multimode stepindex (MM SI), and polymer optical fibers (POF) are presented. Continuous <SUP>60</SUP>Co gamma irradiation of the SM fibers with a dose rate of about 1.5 Gy/s up to a final dose of 10<SUP>6</SUP> Gy led to radiation-induced losses of only 0.85 to 1.3 dB/10 m at 1300 nm wavelength and temperatures around 30 degree(s)C, whereas the GI fibers had losses of 1.3 to 2 dB/10 m under the same conditions. The lowest radiation-induced loss show MM SI fibers with pure SiO<SUB>2</SUB> core of high OH-content: about 0.15 dB/10 m around 850 nm and about 0.1 dB/10 m around 1060 nm (10<SUP>6</SUP> Gy, equals 30 degree(s)C). POF with a core made of polymethyl methacrylate also have loss increases of <EQ 0.1 dB/10 m (670 nm, room temperature), but only up to gamma dose values <EQ 800 Gy. The breaking stress of allglass fibers after 10<SUP>6</SUP> Gy increased by about 2% (MM SI) up to about 10% (SM). 14 MeV neutron irradiations seem to cause higher losses and reduced breaking stress, compared with <SUP>60</SUP> Co gamma irradiations up to the same total dose.
Active optoelectronic devices such as light emitting diodes (LEDs), laser diodes (LDs), photodiodes (PDs), and optocouplers (OCs) were evaluated for degradation under gamma, 14 MeV neutron, and flash X-ray irradiation. Dose rates, total dose values, and neutron fluences were chosen such that we get estimates of the behavior especially in space environments and nuclear engineering. The devices are designed for wavelengths from 660 to 1550 nm. LEDs showed a decrease of output power between 0.1 and 28 dB after a total gamma does of 10<SUP>6</SUP> Gy or neutron fluences of 4 X 10<SUP>14</SUP> cm<SUP>-2</SUP> (1 MeV), respectively. The threshold current of LDs shifted to higher values with increasing dose (<SUP>60</SUP>Co) and neutron fluence. Irradiations of PDs with <SUP>60</SUP>Co gammas up to a total dose of 10<SUP>6</SUP> Gy as well as irradiations with neutrons up to fluences of 4 X 10<SUP>14</SUP> cm<SUP>-2</SUP> (1 MeV) lead to a strong increase of dark current. OCs show a significant reduction of the current transfer ratio at dose values between 10<SUP>3</SUP> and 10<SUP>4</SUP> Gy. Except two types, the optocouplers did not survive neutron fluences >= 8 X 10<SUP>11</SUP> cm<SUP>-2</SUP> (1 MeV).
Test results of residual passive components of fiber optic systems (apart from optical fibers), such as connectors, couplers, wavelength division multiplexers (WDMs), and optical glues are presented. Most of the components were only irradiated at a <SUP>60</SUP>Co gamma ray source with a dose rate of about 0.2 Gy/s up to a final dose of 100 (integrated optic devices) or 10<SUP>4</SUP> Gy (fiber optic components). Direct connectors can contribute <EQ 0.3 dB to the total loss (10<SUP>4</SUP> Gy), compared with 0.3 to 0.6 dB for lens connectors. The measured coupler inherent losses were between about 0.01 dB (fused coupler, high OH multimode stepindex fibers, (lambda) equals 865 nm) and about 50 dB (coupler made of GRIN lenses, (lambda) equals 865 nm). Fused WDMs showed negligible loss increase at the high power outputs and an increase of output power of up to several dB at the isolating output, as well as shifts of the isolating and throughput regions between about 5 and 14 nm (10<SUP>4</SUP> Gy). Optical glues of 20 to 40 micrometers thickness showed noticeable loss increase only for wavelengths <EQ 500 nm (10<SUP>4</SUP> and 10<SUP>5</SUP> Gy).
We have measured radiation-induced losses of all kinds of optical fibers as well as of different kinds of connectors, couplers, and multiplexers. Fiber irradiations were performed at <SUP>60</SUP>Co sources with dose rates ranging from <EQ 1 Gy/d up to about 10<SUP>5</SUP> Gy/d and temperatures between -195 degree(s)C and +100 degree(s)C, as well as at a flash x-ray facility. We present typical results for all test objects. <SUP>60</SUP>Co irradiations up to 10<SUP>6</SUP> Gy within about one week showed that there exist high bandwidth graded index and single mode fibers with induced losses of only about 5 dB in lengths of 50 m that are typical, e.g., for satellites or nuclear power plants. Pure silica core fibers with high OH content can even show less than 0.5 dB under the same conditions.
The radiation sensitivity of different integrated optic (IO) devices was compared under standardized test conditions. We investigated four relatively simple device types made by four different manufacturers. The waveguide materials were proton exchanged LiTaO<SUB>3</SUB>, LiNbO<SUB>3</SUB>:Ti, Tl-diffused glass, and Ag-diffused glass, respectively. In order to standardize the irradiation parameters we followed the 'Procedure for Measuring Radiation-Induced Attenuation in Optical Fibers and Optical Cables' proposed by the NATO NETG as close as possible. In detail we made pulsed irradiations with dose values of about 500 rad*, 10<SUP>4</SUP> rad, and 10<SUP>5</SUP> rad, as well as continuous irradiations at a <SUP>60</SUP>Co source with a dose rate of 1300 rad*/min up to a total dose of 10<SUP>4</SUP> rad. Device temperatures were about 22 degree(s)C, -50 degree(s)C, and +80 degree(s)C.
In order to estimate the radiation sensitivity of a data transmission system based on optical fibers, the radiation-induced loss of other included passive components has to be known, too. We have irradiated (at room temperature) a variety of couplers made by different manufacturers at a <SUP>60</SUP>Co source with a dose rate of 1300 rad(SiO<SUB>2</SUB>)/min up to a total dose of 10<SUP>6</SUP> rad(SiO<SUB>2$</SUB>. Most of them were fabricated by the Fused Biconical Taper process of 200/280 micrometers multimode step index fibers (tested at (lambda) equals 850 nm), and of 10/125 micrometers single mode fibers (tested at (lambda) equals 1300 nm and (lambda) equals 1550 nm). Couplers with 50/125 micrometers gradient index fibers (tested at (lambda) equals 850 nm and (lambda) equals 1300 nm) were also among them. The measured coupler inherent losses were between about 0.01 dB (fused coupler, high OH MM step index fibers, (lambda) equals 865 nm) and about 50 dB (coupler made of GRIN lenses, (lambda) equals 865 nm).
The radiation-induced loss of multimode step-index fibers with undoped synthetic silica core has been reduced significantly over the years. Most of the efforts focussed on material modifications and improvements of the undoped core material having either high or low OH- content. With increasing radiation resistance due to core material improvements, influences of the core-cladding interface and of the cladding material became more important. These effects can be analyzed with an improved measurement system using different excitation conditions at the frontface of the step-index fibers: low order meridional rays have high intensity in the center of the core material and low intensity in the core-cladding transition region, in contrast to high order skew or helix rays having opposite intensity profiles. The measurement system including the specific excitation conditions is described. Applying this setup, radiation-induced losses during and after continuous gamma irradiation were investigated with regard to differences of the specific excitation conditions. The measured values at 10 krad total dose differ by more than a factor of two.
Optical fibers with undoped silica core and fluorine doped silica cladding are known to be radiation resistant. Especially fibers with high OH-content core material show very low radiation induced losses which recover within a few hundred seconds after irradiations up to 100 krd dose. Even with these promising properties radiation resistance is a limiting factor for some applications. Optical fibers with different grades of high OH-content undoped silica core materials were produced by using the same preform manufacturing and fiber drawing parameters. The UV-attenuation spectra of those fiber samples were measured. Then they were irradiated at a <SUP>60</SUP>-Co-source, and the induced losses were recorded in the spectral range between 200 nm and 1600 nm wavelength. A correlation between the UV-attenuation characteristics of the unirradiated fiber and its radiation induced loss was found. The results indicate that defects already existing in the unirradiated fiber ('precursor defects') are responsible for the differences in radiation resistance at dose values below 100 krd. These precursor defects are transformed into color centers by ionizing radiation.
To determine systematic trends that possibly allow extrapolation of the results of continuous irradiations to low dose rates four different fibers were exposed to dose rates between 0.05 rd/s and 180 rd/s of a Co-60 source and to the pulsed radiation of a flash X-ray facility. Data obtained show that in some cases extrapolation of the induced losses measured in this dose rate range to higher dose rates yields nearly exactly the results of pulsed irradiations with about 10 exp 11 to 3 x 10 exp 12 rd/s so that the extrapolation down to values of 10 exp -5 seems to be allowed too.
Three SM-preforms with the same undoped OH-rich core material  and different fluorine content in the cladding are produced. From these preforms SM-fibers with different diameter are drawn and characterized by fiber diameter, cut-off wavelength, and effective V-parameter at 850 nm. The fibers are tested under continuous <sup>60</sup>Co-irradiation and pulsed irradiation. All measurements are performed at about 850 nm wavelength and room temperature. Due to different power distributions in SM-fibers with different V-parameters, the influence of radiation induced attenuation in the fluorine doped silica cladding is examined and the dependence of this attenuation on the fluorine concentration is shown. In addition MM-fibers with same core material are irradiated as a reference for the core material.