Open Access
Add to BibSonomyAbstract
Writing and thermal annealing of fiber Bragg gratings (FBGs) in an optical fiber containing 75 mol.% GeO2 in the core have been studied by analyzing the first three diffraction orders of the FBGs. Double oscillations of the grating reflectivity has been observed during the FBG formation in an H2-loaded fiber, and the corresponding three grating types revealed have been labeled as type I(H2), type IIa(H2)-, and type IIa(H2)+. The results obtained have shown that the negative index change related to the type IIa photosensitivity cannot be described by the accumulated UV-dose only, but strongly depends on the UV-radiation intensity for both pristine and H2-loaded fibers, unlike the type I and type I(H2) photosensitivities.
©2012 Optical Society of America
1. Introduction
It is well-known that during UV-irradiation of germanosilicate optical fibers and glasses, several physical processes proceed simultaneously and lead to a refractive index change. These processes are usually referred to as different photosensitivity types. Joint action of two or more photosensitivity types leads to a complex dynamics of the index evolution during the irradiation. In particular, during the inscription of fiber Bragg gratings, successive phases of increasing and decreasing of the grating reflectivity can be observed. Each phase is caused by the domination of one of the photosensitivity types, and, therefore, the FBG type is commonly labeled in accordance with the main photosensitivity type.
Low-temperature hydrogen loading of germanosilicate fibers is the most common way to enhance photosensitivity and is widely used for fiber grating fabrication. If the concentration of molecular hydrogen dissolved in the glass network is ~1 mol.%, the photoinduced index in a fiber with a relatively small germanium concentration (≤ 10 mol.% GeO2) increases approximately ten-fold with respect to an unloaded fiber under the same UV-irradiation conditions [1]. Such a significant increase is due to the formation of hydrogen-containing groups, mainly hydroxyls (Ge-OH and Si-OH), which leads to a buildup of the glass polarizability [2]. The photosensitivity caused by this mechanism is commonly named type I(H2), by analogy with the positive index change in pristine fibers (type I), which is thought to result from the photoinduced densification of the glass network [3]. Type I FBGs, in contrast to type I(H2) ones, do not exhibit temperature or strain-induced reversible change of the index modulation [4,5], which confirms the fact that the mechanisms of photoinduced index change are different.
Bragg gratings of type IIa are usually observed in pristine fibers with a relatively high germanium concentration (≥ 20 mol. % GeO2) at large UV-exposure. This photosensitivity type is supposed to be due to glass network dilation leading to an index decrease in the irradiated regions [6]. A type IIa FBG arises after the complete disappearance of the original type I FBG, and its spectral properties are similar to those of the type I FBG. Nevertheless, these two FBG types possess different thermal stability [7], as well as different coefficients of temperature- and strain–induced reversible changes of their spectral properties [5,8]. Some other types of FBGs can also be formed in silica-based fibers (type Ia FBGs, type II FBGs, regenerated gratings etc.) [9].
Photosensitivity of H2-loaded fibers with a high germanium concentration has a number of peculiarities requiring further study. For example, in [10] it was shown that the presence of H2 in glass leads to a noticeable change in the inscription dynamics of type IIa FBG. In particular, in a fiber containing ~30 mol.% GeO2 type IIa FBG was not observed at the H2 concentration СН2 ~2 mol.%, whereas with СН2 ~0.02 mol.%, type IIa grating appears even faster than in the pristine fiber.
In this paper, we represent the results of our study (started in [11]) of the writing and thermal annealing processes in a fiber containing 75 mol.% GeO2 in the core. The first three FBG diffraction orders have been investigated both in pristine and H2-loaded fibers. Fourier analysis has been applied to obtain the FBG pitch profiles as well as the dose dependences of the local photoinduced refractive index change.
2. Experiment
The single-mode optical fiber used in the experiments was fabricated at FORC RAS in collaboration with ICHPS RAS by the MCVD technique [12]. The refractive index profile of the fiber was nearly step-like. The core composition was evaluated with the help of X-ray microanalysis to be 0.75GeO2-0.25SiO2. The fiber had the core-cladding index difference Δn ≈0.104, the cut-off wavelength λс ~1.0 μm, and the core diameter of about 1.4 μm. The fiber cladding contained no dopants.
The fiber was loaded with molecular hydrogen at pressure of 75 bar and temperature of 100°C for 12 hours. This loading conditions give the concentration of H2 in the glass network СН2 ~0.5 mol.%. To outdiffuse the remaining molecular hydrogen the fiber was kept in a furnace at 100°C for 12 hours after the irradiation procedure.
The spectral properties of the first three FBG diffraction orders were investigated. Simultaneous registration of two or more FBG orders is complicated owing to a large spectral spacing between them, which results in essential experimental errors. Therefore, the 1550 nm – wavelength range was used for measuring all the orders, the period of the grating being increased properly for high-order FBGs, similar to the procedure in paper [13]. The gratings were written with a cw frequency-doubled argon-ion laser radiation (244 nm) by means of a Lloyd interferometer [14], the visibility of the interference pattern being slightly less than unity. The average power density of the UV radiation on the fiber surface (I0 ~30 W/cm2) was uniform along the grating length, which was confirmed by a good coincidence of the theoretical and experimental bandwidths. All the FBGs had a length of about 3 mm.
Investigation of temperature stability of photoinduced FBGs allows one to compare thermal stability of photoinduced changes governed by different physical mechanisms, to reveal the contribution of different defect centers and hydrogen-containing groups to the induced index, and to define other thermoinduced processes influencing the FBG properties. One of the most informative ways of the thermal annealing analysis is the consideration of temperature derivatives which can be obtained relatively easily in the case of linear annealing (tempering) [15,16]. In our experiments, the FBGs were annealed in an electric furnace, which temperature was increased at a constant rate dT/dt = 0.25 K/sec [17]. The FBG annealing procedure consisted of two steps. At the first step the grating was annealed up to a temperature at which its reflection coefficient is small, but still detectable (RBG ~0.1 ÷ 0.2 dB). After cooling the furnace down to the room temperature, the second annealing step was performed with the same temperature rate, in order to determine the reversible change of the grating resonance wavelength.
During the writing and annealing processes the FBG transmission spectra were registered with the help of an ANDO-6317B spectrum analyzer with a resolution of 0.02 nm, each spectrum being measured during 15 sec. Then the FBG reflection coefficient R and a resonance wavelength λBG have been determined.
To analyze the results obtained, we used the spatial distribution of the UV radiation intensity along the fiber z-axis as follows:
where I0 is the average UV intensity, V is the interference pattern visibility, and Λ is the grating period. Next we can represent the spatial distribution of photoinduced index Δnind(z) of a uniform grating as a Fourier series:where Δnm(z) is the m-th Fourier component of the induced index, Δnavr is the average induced index and Am is the Fourier expansion coefficient determining the index modulation for the m-th diffraction order with . Note that in our case it is better to use alternating-sign coefficients Am instead of , because the induced index is non-monotonic. The absolute value of Am was calculated using the following equation [18]:where η is the power of the fiber fundamental mode propagating in the fiber core, L is the grating length. The typical Am error was about 10−5. The Am sign was derived from the sign of the m-th Fourier component at the intensity maxima Δnm(z = kΛ), where k∈¢. For example, for positive and monotonic induced index change (type I and type I(H2) photosensitivity), coefficient A1 is positive, while for type IIa gratings it becomes negative. If the dose dependence of the induced refractive index is nonlinear, the higher FBG orders appear and the sign of their Fourier coefficients Am can be uniquely determined from the shape of the dose dependence. For instance, if the dose dependence is sublinear, which is typical for type I and type I(H2) photosensitivity, the higher orders arise with A2<0 and A3>0. Note that coefficients Am can change their signs during irradiation or annealing processes, which is accompanied by a decrease of the FBG reflection coefficient down to zero and its subsequent increase.Average index change Δnavr induced during the UV irradiation was calculated using the FBG resonance wavelength shift ΔλBG according to the following equation:
where is the FBG resonance wavelength measured at the very beginning of the irradiation process and neff is an effective index of the fiber fundamental mode. To calculate Δnavr in the annealing process, Eq. (4) was used, taking into account the reversible change of the grating resonance wavelength: and . The typical error of the Δnavr measurements was about 10−4.3. Inscription of the FBGs
3.1 Pristine fiber
Figure 1(a) shows the dependence of Am (m = 1, 2, 3) and Δnavr on the average UV dose (D = I0t, where t is the irradiation time) measured in the pristine fiber during the FBG writing process.
Fig. 1 (a) Dose dependences of Am and Δnavr, measured in pristine fiber for first-order and high-order FBGs during the writing process. The inset depicts the FBG pitch profiles calculated for different UV doses. (b) Dose dependences of induced index for different coordinates along the fiber.
One can see that the first-order FBG writing dynamics (A1) is typical for type IIa grating: the magnitude of the induced index is determined by the competition of two processes with different index signs, namely type I photosensitivity (Δnind>0) and type IIa photosensitivity (Δnind<0) [19,20]. Because of a high germanium concentration in the fiber core, the negative induced index due to type IIa photosensitivity exceeds the positive index due to type I photosensitivity even at relatively low irradiation doses D ≥ 10 kJ/cm2 [12]. The maximum positive A1 value of amounted to only 1.3 × 10−4, whereas its negative value to −2.6 × 10−3 at D ≈57 kJ/cm2 and was far short of a saturation.
It is important to note that the reflection coefficient of the second- and third-order gratings oscillates during the irradiation. The A2 value in the negative half-plane reached −6.3 × 10−4 at D ≈30 kJ/cm2, later it became positive at D ≈70 kJ/cm2 and increased to 6.3 × 10−4 at the end of the irradiation. The A3 dynamics is shifted to higher doses, its maximum value being 5.4 × 10−4 at D ≈70 kJ/cm2. A3 becomes zero at D ≈170 kJ/cm2 and reaches −2 × 10−4 at the end of irradiation. It is interesting that owing to the fast development of type IIa photosensitivity in the fiber tested, the maximum Δnmod values for second- and third-order gratings are several times higher than that for the first-order grating.
Figure 1(a) also shows the dose dependences of average index Δnavr(D) measured for all the FBGs. It should be noted that high-order FBGs cannot be measured until a certain Δnavr value has been induced; therefore, we shifted Δnavr for these gratings in Fig. 1(a) along the x-axis to achieve the best coincidence with Δnavr of the first-order grating at high UV doses. Note that in spite of different FBG periods, a very good agreement of the Δnavr dependences for all the three gratings is observed. This means that the dose dependence of the induced index does not strongly depend on the grating period. We verified this fact additionally by measuring two diffraction orders during FBG writing simultaneously. In particular, when the first and second orders were simultaneously monitored, their resonance wavelengths were ~1614 nm and ~830 nm, whereas simultaneous registration of the second and third orders yielded ~1550 nm and ~1051 nm, respectively. The values of Am and Δnavr measured in both the synchronous experiments proved to agree well with the results shown in Fig. 1(a).
It is noteworthy that during the FBG writing process, change of the Am sign is accompanied by a significant displacement of the Bragg wavelength, which manifests itself as a sharp Δnavr increase by 1 - 2 × 10−4 (Fig. 1). This phenomenon requires further investigations.
The spatial distributions of the induced index along the grating pitch shown in the inset in Fig. 1(a) were calculated by Eq. (2) for several irradiation doses using the data for Am (m = 1, 2, 3) and Δnavr. We see that the pitch profile varies significantly during the irradiation process, the induced index being negative in the vicinity of the UV intensity maximum. Note that the high contrast of the interference pattern in our experiments is responsible for significant index modulation amplitude in the higher diffraction orders even at large irradiation doses. It is seen that the pitch profile is non-sinusoidal even at very small doses. As a result, Δnmod significantly differs from Δnavr.
Using the FBG pitch profiles shown in the inset of Fig. 1(a), dose dependences of the photoinduced index at different points along the pitch were calculated for V = 0.95 and depicted in Fig. 1(b). It is clearly seen that the refractive index at low doses (type I photosensitivity) is uniquely defined by the irradiation dose, whereas at high doses (type IIa photosensitivity), the total induced index strongly depends on the coordinate along the pitch. The latter testifies that the negative induced index change related to type IIa photosensitivity is defined not only by the irradiation dose, but also depends on the local intensity of the UV radiation, being larger in the places with a higher UV intensity. Note that the same conclusion can be drawn from closer inspection of the experimental results presented in paper [20], where pulsed radiation of an excimer laser was used to inscribe gratings. Moreover, the observed intensity dependence calls for an appropriate modification of the three-level model suggested in [20]. The exact mechanism of the intensity dependence of the type IIa photosensitivity requires further investigation.
3.2 H2-loaded fiber
The dependences shown in Fig. 2 are similar to those in Fig. 1, but were obtained for the fiber subjected before irradiation to low-temperature hydrogen loading. The presence of hydrogen dissolved in glass leads to essential modification of both the dose dependences, Am(D) and Δnavr(D). The value of А1 grew up to ~5 × 10−4 at the beginning of irradiation (D ≤ 1 kJ/cm2) owing to type I(H2) photosensitivity, and then decreased down to zero, became negative at D ~2.5 kJ/cm2, and reached −4 × 10−4 at about 5.5 kJ/cm2 as a result of predominance of the type IIa photosensitivity. Surprisingly, subsequent irradiation led to one more oscillation of the reflection coefficient, which was accompanied by the formation of a new FBG type with a positive А1. Reflection of this new grating increased steadily without noticeable saturation up to the end of the irradiation (D ~55 kJ/cm2), where the value of А1 was as great as ~1.7 × 10−3. Note that during the whole irradiation process, the grating reflection spectrum retains its bandwidth and corresponds well to the spectrum of a uniform grating, which means that the reflection oscillation is caused by the grating type change rather than the experimental errors.
Fig. 2 (a) Dose dependences of Am and Δnavr, measured in an H2-loaded fiber for the first- and high-order gratings during their writing process. The inset depicts the FBG pitch profiles calculated for different UV doses. (b) Dose dependences of the induced index for different coordinates.
The two-fold oscillation of the grating reflection coefficient during the writing process in an H2-loaded germanosilicate fiber has been observed for the first time, to our knowledge. We suppose that the cause of such a complicated dose dependence of RBG is the competition of two photosensitivity processes with essentially different characteristic times, namely, the type I(Н2) photosensitivity (positive index change) and the type IIa photosensitivity (negative index change). The last stage of the grating formation is, therefore, the result of relatively fast saturation of the type IIa photosensitivity and continuous monotonic index growth related to the type I(Н2) photosensitivity. The latter is additionally confirmed by a significant increase of the mean index change at high irradiation doses (see Fig. 2(a)). Comparison of Fig. 1(a) and Fig. 2(a) also shows that the presence of molecular hydrogen fastens the development of the type IIa photosensitivity, which was also been observed in [10] in a germanosilicate fiber with a relatively low H2 concentration.
According to the assumption of the competition of two photosensitivity mechanisms, we suggest the following abbreviations for the three FBG types observed:
- • Type I(H2) FBG arises at the beginning of irradiation, when the main part of the induced index is caused by type I(H2) photosensitivity;
- • In type IIa(H2)- FBG, type IIa photosensitivity predominantly contributes to A1 and makes it negative;
- • Type IIa(H2)+ FBG is a high UV-dose FBG in which A1 becomes positive again as a result of the factors described above. For that grating type, the index change is mainly caused by the hydrogen-related photosensitivity type, but the glass network in the regions with high UV intensity experiences structural changes due to the type IIa photosensitivity (see Fig. 2(b), which is discussed below).
Here we add the designation “Н2” in the last two cases, in order to underline the fact that the type IIa mechanism acts in the presence of molecular hydrogen. This can, in general, lead to the FBG properties different from those of “classical” type IIa gratings. For example, this FBG type can have different temperature sensitivity of the Bragg wavelength, different temperature stability, different temperature- or strain-induced reversible reflectivity changes, etc. Additional signs «+» и «ˉ» are added to separate two latter FBG types with different signs of the A1 coefficient due to different prevailing mechanisms photosensitivity.
It should be mentioned that the relative contribution of the mechanisms discussed can vary depending on the fiber parameters, hydrogen concentration, and irradiation conditions. Therefore, the A1 dose dependence can differ from that shown in Fig. 2(a). In particular, it can cross the x-axis only once, if the type IIa photosensitivity exceeds the type I(H2) photosensitivity at high doses (as was observed in [10] at low hydrogen concentration), or even does not cross the x-axis if type I(H2) photosensitivity prevails at all doses.
The general behavior of the Am dose dependences for the second- and third-order gratings measured in the H2-loaded fiber (Fig. 2(a)) is similar to that observed in the pristine fiber (Fig. 1(a)). At the same time comparison of this two figures shows that both the second- and third-order gratings in the H2-loaded fiber arise and reach their maxima at 5 - 6 times lower UV-doses. In addition, the Am coefficients (for both the high-order gratings) change their signs in the pristine fiber and retain their initial signs for the H2-loaded fiber.
Note that the dose dependences of the average index change coincide well for all the three FBGs written in the H2-loaded fiber (Fig. 2(a)), as was found for the pristine fiber (Fig. 1(a)). In the former case the average induced index experiences approximately linear growth in a broad dose range, except the low-dose part (D < 3 kJ/cm2) and Δnavr is as large as 5.5 × 10−3 at the end of irradiation. Such a linear Bragg wavelength growth is typical for the formation of type Ia gratings after prolonged exposure of an H2-loaded fibers [21,22] and can be attributed to the formation of a great quantity of hydroxyl groups in the fiber core [23]. According to the results described above, we believe that the main cause of the formation of the type Ia grating is a decrease of the induced index at the maxima of the interference pattern, which takes place due to type IIa photosensitivity. The A1 coefficient becomes negative (formation of a type Ia FBG) if the induced index related with type I(H2) photosensitivity is saturated in the places both with high and low UV-intensity. If so, we can conclude that type Ia FBG is nothing more nor less than the type II(H2)- in the classification described above. The difference between our dose dependences and those obtained in [21] can be explained by the different core glass compositions (mainly, by different germanium concentrations).
The inset in Fig. 2 demonstrates how the pitch profile of an FBG evolves with UV-dose in the case of the H2-loaded fiber. The profiles were calculated in accordance with Eq. (2) using the experimental data of Δnavr and Am (m = 1, 2, 3) just as it was done above for the case of pristine fiber (inset of Fig. 1(a)). The joint action of type I(H2) and type IIa photosensitivity mechanisms makes the pitch dynamics even more complicated as compared to that of the pristine fiber. The distinctive feature of the profiles shown is a substantial increase of the induced index in the minima of the interference pattern (z/Λ = ± 0.5), which testifies to relatively large type I(H2) photosensitivity even at low UV-doses.
The dose dependences of the local induced index calculated for different coordinates along the grating pitch for the H2-loaded fiber are shown in Fig. 2(b). The calculation was performed in the same way as for the pristine fiber (Fig. 1(b)). We see that in both the pristine and H2-loaded fiber, a negative contribution of the type IIa photosensitivity to the total induced refractive index increases with an increase of UV-intensity. One can also see that the contribution of type IIa photosensitivity leaves a trace in the total induced index near the UV intensity maximum even at maximal studied UV doses, which means that the structural change in the glass network caused by type IIa photosensitivity remains at that doses. Note that, according to our calculations, for FBGs written in the H2-loaded fiber the resultant index change is always positive even in the central part of the pitch, in spite of the significant negative contribution of the type IIa photosensitivity.
4. Annealing of FBGs written in H2-loaded fiber
Annealing of the type IIa FBG written in the pristine fiber showed that the index modulation amplitude remained almost constant up to 730 K, thereafter Δnmod decreased monotonically down to zero at 990 K. Such a behavior is common for type IIa gratings [7] and testifies that the induced index is mainly caused by this type of photosensitivity.
To study the annealing properties of gratings written in the H2-loaded fiber, five FBGs were written with several UV-doses (shown by arrows in Fig. 2(a)). Three doses were selected so as to provide the maxima of three FBG types introduced above and two doses were intermediate. Temperature dependences of the A1 and dA1/dT values measured during the annealing of these FBGs are plotted in Fig. 3(a) and Fig. 3(b) respectively. For convenience, in both cases the curves are shifted with respect to each other by one division of the y-axis, whereas zero levels are indicated by dotted lines. It should be noted that keeping of the FBGs at 100°C prior to the annealing experiment to let the remaining molecular hydrogen outdiffuse from the glass resulted in the erasure of the less stable part of the induced refractive index and, as a consequence, in some distortion of the dependences in the 400 – 500 K temperature range. This can be partially associated with the increased solubility of molecular hydrogen in the irradiated parts of the gratings [24].
Fig. 3 Temperature dependences of A1 (a) and dA1/dT (b) obtained for FBGs written with different UV-doses in H2-loaded fiber (1 - 5 are the FBG numbers).
It is seen that the A1 annealing dependences differ significantly for different irradiation doses, having, at the same time, a number of common features. The most complicated annealing behavior is observed for FBGs 2, 3, and 4 written at intermediate UV-doses, in which the contributions of the type I(H2) and the type IIa photosensitivities are comparable.
The annealing of the refractive index induced in the H2-loaded germanosilicate fibers is usually characterized by a steady decrease of the index up to high temperatures, with the temperature derivative exhibiting no distinct annealing bands [25]. The key issue is the formation of various hydrogen-containing groups (hydroxyl groups, hydride groups, water molecules etc.) in the fiber core during the irradiation [26–28]. It was shown in [29] that the anneling behavior can be explained by the contribution of hydroxyl groups, both Si-OH and Ge-OH, as well as water molecules, which have broad overlapping annealing bands covering a large temperature range. One can see a monotonic decrease of the induced index in Fig. 3(a) corresponding to FBGs no. 1 and no. 5. Nevertheless, all the dependences in Fig. 3(a), except for FBG no. 1 exhibit a noticeable increase at about 600 K, which manifests itself as a dip at this temperature in Fig. 3(b). It is interesting that for FBG no. 3, A1 even changed its sign. We believe that this variation of the index modulation amplitude is caused by thermoinduced interaction of water molecules with regular glass network:
Such a reaction can result in an increase of the glass refractive index owing to higher polarizability of two hydroxyl groups with respect to that of a water molecule. This assumption is confirmed by a decrease of the infrared absorption at 3200 - 3300 cm−1 ascribed to water [26,27,30], as well as by a noticeable increase of the OH-related absorption [29,31] at these temperatures.
The thermoinduced hydroxyl groups formed by reaction (5) are then annealed at a higher temperature, as it takes place for the photoinduced groups. Both photoinduced and thermoinduced hydroxyls are located in the glass network in pairs neighboring each other. This circumstance explains relatively low thermal stability of these groups: they can decay rather easily to form a hydrogen molecule, which, in turn, outdiffuses from the glass. Note that unpaired hydroxyls in wet silica are much more thermostable and cannot be ruptured at the temperatures of our experiment [27].
In the temperature range 800 - 1000 K, the index modulation amplitude exhibits some increase for most of the FBGs in Fig. 3(a). It is possible that this growth is caused by the annealing of the negative part of the induced refractive index caused by type IIa photosensitivity. Indeed, in this temperature range, the saturated type IIa FBGs written in the pristine fiber were also erased.
A further temperature increase leads to the formation of regenerated gratings [32] in the range of 1000 – 1200 K. Gratings of this type are characterized by a relatively fast A1 drop down to zero, which is followed by a buildup in the negative half-plane. Note that the modulation amplitude of the resultant regeneration gratings is saturated at 1 - 2 × 10−4, the value almost independent of the irradiation dose. The exact mechanism of the regenerated grating formation remains unclear at the moment. This grating type can be observed in both hydrogen-loaded and pristine fibers, if the latter are strained in the process of the FBG inscription or annealing [33]. It should be noted that regenerated long-period fiber gratings have not been observed, these fact may provide a key for the understanding of regeneration process.
Temperature dependences of Δnavr and d(Δnavr)/dT measured during the annealing of the five gratings are depicted in Fig. 4(a) and Fig. 4(b), respectively. The curves are displaced by one division of the vertical axis as in Fig. 3.
Fig. 4 Temperature dependences of the average induced index (a) and its temperature derivative (b) for the FBGs 1-5.
For all the gratings tested, most of the average induced index is erased in the temperature range of 600 – 900 K (Fig. 4). This temperature range corresponds well to the annealing range of hydroxyl groups measured in the same annealing conditions [29]. For the fiber considered, the concentration of photoinduced Ge-OH groups, which are less thermostable, is larger than that of highly stable Si-OH groups; and therefore, the low-temperature part (600 - 800 K) of the annealing curves is more pronounced than the high-temperature one (800 - 1000 K) (see Fig. 4(b)).
Some Δnavr increase is observed for most gratings at temperatures 500 – 600 K being more prominent for higher irradiation doses (Fig. 4(a)). A similar increase has already been noticed for the A1(Т) dependences shown in Fig. 3(a); therefore, the increase of the average induced index in this temperature range can be also attributed to reaction (5).
Figure 5 gives the annealing dependences of the Am coefficients for the second- and third-order FBGs. The gratings annealed in this experiment were obtained at the end of the irradiation shown in Fig. 2. As in the case of annealing of the first-order FBG (Fig. 3), annealing of high-order FBGs resulted in a significant (almost two-fold) increase of the index modulation amplitude in the temperature range of 500 – 600 K, which we explained by the effect of thermoinduced reaction (5). The essential contribution of this reaction to the high-order FBGs is attested by a saturable dose dependence of the photoinduced H2O concentration confirmed by the results obtained in [2]. A major part of Δnmod for both the high-order gratings is annealed at 600 - 1000 K, and the annealing behavior of Δnmod is very similar to that of Δnavr (see Fig. 4(a)) in this temperature range.
5. Conclusion
Peculiarities of the processes of writing and annealing of FBGs in a single-mode fiber heavily doped with germanium (75 mol.% GeO2 in the core) have been revealed. Among these peculiarities are essentially nonmonotonic dose dependences of the index modulation amplitude in the first three diffraction orders of FBGs being written in both pristine and H2-loaded fibers. In the latter case, a two-fold oscillation of the FBG reflectivity was observed, which means the appearance of three grating types in the first-order grating. In accordance with the most influencing photosensitivity types we have proposed the following designation of the grating types successively formed in this case: type I(H2), type IIa(H2)- and type IIa(H2)+ FBGs, respectively. Using the measured dose dependences of the index modulation amplitude in the first three orders as well as the average index change we have reconstructed the spatial distributions of the induced index along the grating pitch for different UV-irradiation doses, which are difficult to obtain by any other techniques. The calculated index profiles show that the negative index change related to the type IIa photosensitivity cannot be described by the accumulated UV-dose only, but depends strongly on the intensity of the UV-radiation for both the pristine and H2-loaded fibers: the higher the intensity, the larger the negative index change. It should be noted that, according to our data, the type I as well as type I(H2) photosensitivities do not exhibit any dependence on the UV-intensity and are fully defined by the total irradiation dose.
The proposed investigation method of the induced index change has allowed us to separate the contributions of the different photosensitivity mechanisms. In particular, it became clear that a significant part of the induced index change in H2-loaded fibers with a high germanium concentration is caused by the type IIa photosensitivity. We found that the type IIa photosensitivity is enhanced by the presence of molecular hydrogen in the glass network, even in the case of a relatively large H2 concentration.
The results of this investigation have allowed us to assume that the formation of the type Ia FBGs is associated with the contribution of the type IIa photosensitivity, which is responsible for most of the index modulation amplitude of the first-order grating at high irradiation doses. Therefore, the type Ia grating is nothing but type IIa(H2)- grating in terms of our classification.
Annealing of the first- and high-order FBGs written in an H2-loaded fiber performed with the help of a linear heating technique has shown that several thermoinduced processes take place. It has been concluded that most of the induced index in an H2-loaded fiber heavily doped with germanium is caused by the formation of hydroxyl groups, mainly Ge-OH groups. A significant increase of both index modulation and average index at about 600 K has been explained by the thermoinduced reaction of the photoinduced water molecules with the regular bonds of the glass resulting in the formation of two hydroxyl groups. In addition, the effect of annealing of the negative index change caused by the type IIa photosensitivity and the formation of regenerated gratings have been observed in gratings written in an H2-loaded fiber heavily doped with germanium.
Acknowledgments
The authors are very grateful to V. M. Mashinsky, A. N. Guryanov, and V. F. Khopin, who have developed and provided the unique optical fiber used in our study.
References and links
1. P. J. Lemaire, R. M. Atkins, V. Mizrahi, and W. A. Reed, “High pressure H2 loading as a technique for achieving ultrahigh UV photosensitivity and thermal sensitivity in GeO2 doped optical fibres,” Electron. Lett. 29(13), 1191–1193 (1993). [CrossRef]
2. M. Lancry, B. Poumellec, P. Niay, M. Douay, P. Cordier, and C. Depecker, “VUV and IR absorption spectra induced in H2-loaded and UV hyper-sensitized standard germanosilicate preform plates through exposure to ArF laser light,” J. Non-Cryst. Solids 351(52-54), 3773–3783 (2005). [CrossRef]
3. B. Poumellec, P. Guenot, I. Riant, P. Sansonetti, P. Niay, P. Bernage, and J. F. Bayon, “UV induced densification during Bragg grating inscription in Ge:SiO2 preforms,” Opt. Mater. 4, 401–409 (1995).
4. A. Hidayat, Q. Wang, P. Niay, M. Douay, B. Poumellec, F. Kherbouche, and I. Riant, “Temperature-induced reversible changes in the spectral characteristics of fiber Bragg gratings,” Appl. Opt. 40(16), 2632–2642 (2001). [CrossRef] [PubMed]
5. P. I. Gnusin, S. A. Vasiliev, O. I. Medvedkov, and E. M. Dianov, “Reversible changes in the reflectivity of different types of fibre Bragg gratings,” Quantum Electron. 40(10), 879–886 (2010). [CrossRef]
6. N. H. Ky, H. G. Limberger, R. P. Salathe, F. Cochet, and L. Dong, “UV-irradiation induced stress and index changes during the growth of type-I and type-IIA fiber gratings,” Opt. Commun. 225(4-6), 313–318 (2003). [CrossRef]
7. S. Pal, “Characterization of thermal (in)stability and temperature-dependence of type-I and type-IIA Bragg gratings written in B–Ge co-doped fiber,” Opt. Commun. 262(1), 68–76 (2006). [CrossRef]
8. X. Shu, D. Zhao, L. Zhang, and I. Bennion, “Use of dual-grating sensors formed by different types of fiber Bragg gratings for simultaneous temperature and strain measurements,” Appl. Opt. 43(10), 2006–2012 (2004). [CrossRef] [PubMed]
9. J. Canning, “Fibre gratings and devices for sensors and lasers,” Laser Photon. Rev. 2(4), 275–289 (2008). [CrossRef]
10. H. Poignant, J. F. Bayon, E. Delevaque, M. Monerie, J. L. Mellot, D. Grot, P. Niay, P. Bemage, and M. Douay, “Influence of H2 loading on the kinetics of Type IIA fibre Bragg grating photoinscription,” in IEE Colloquium on Optical Fibre Gratings (London, UK, 1997), pp. 2/1-2/7.
11. O. I. Medvedkov, S. A. Vasiliev, P. I. Gnusin, and E. M. Dianov, “Three Bragg grating types in hydrogen-loaded heavily germanium- doped fibers,” in Bragg gratings, Photosensitivity and Poling in Glass Waveguides conference, Technical Digest (CD) (Optical Society of America, 2012), paper BM4D.2.
12. V. M. Mashinsky, V. B. Neustruev, V. V. Dvoyrin, S. A. Vasiliev, O. I. Medvedkov, I. A. Bufetov, A. V. Shubin, E. M. Dianov, A. N. Guryanov, V. F. Khopin, and M. Yu. Salgansky, “Germania-glass-core silica-glass-cladding modified chemical-vapor deposition optical fibers: optical losses, photorefractivity, and Raman amplification,” Opt. Lett. 29(22), 2596–2598 (2004). [CrossRef] [PubMed]
13. C. W. Smelser, S. J. Mihailov, and D. Grobnic, “Impact of index change saturation on the growth behavior of higher-order type I ultrafast induced fiber Bragg gratings,” J. Opt. Soc. Am. B 25(5), 877–883 (2008). [CrossRef]
14. S. A. Vasiliev, O. I. Medvedkov, I. G. Korolev, A. S. Bozhkov, A. S. Kurkov, and E. M. Dianov, “Fibre gratings and their application,” Quantum Electron. 35(12), 1085–1103 (2005). [CrossRef]
15. W. Primak, “Large temperature range annealing,” J. Appl. Phys. 31(9), 1524–1533 (1960). [CrossRef]
16. J. Rathje, M. Kristensen, and J. E. Pedersen, “Continuous anneal method for characterizing the thermal stability of ultraviolet Bragg gratings,” J. Appl. Phys. 88(2), 1050–1055 (2000). [CrossRef]
17. A. S. Bozhkov, S. A. Vasiliev, O. I. Medvedkov, M. V. Grekov, and I. G. Korolev, “A setup for investigating induced refractive index change in optical fibers at high temperatures,” Instrum. Exp. Tech. 48(4), 491–497 (2005). [CrossRef]
18. R. Kashyap, Fiber Bragg Gratings, P.L. Kelly, I. Kaminow, and G. Agrawal, eds. (Academic Press, 1999).
19. W. X. Xie, P. Niay, P. Bernage, M. Douay, J. F. Bayon, T. Georges, M. Monerie, and B. Poumellec, “Experimental evidence of two types of photorefractive effects occurring during photoinscriptions of Bragg gratings within germanosilicate fibres,” Opt. Commun. 104(1-3), 185–195 (1993). [CrossRef]
20. L. Dong, W. F. Liu, and L. Reekie, “Negative-index gratings formed by a 193-nm excimer laser,” Opt. Lett. 21(24), 2032–2034 (1996). [CrossRef] [PubMed]
21. Y. Liu, J. A. R. Williams, L. Zhang, and I. Bennion, “Abnormal spectral evolution of fiber Bragg gratings in hydrogenated fibers,” Opt. Lett. 27(8), 586–588 (2002). [CrossRef] [PubMed]
22. A. G. Simpson, L. Zhang, K. Zhou, and I. Bennion, “Abnormal photosensitivity effects and the formation of type IA FBGs,” in Bragg gratings, Photosensitivity and Poling in Glass Waveguides conference, Technical Digest (Optical Society of America, 2003), paper MD-32.
23. K. Kalli, A. G. Simpson, K. Zhou, L. Zhang, D. Birkin, T. Ellingham, and I. Bennion, “Spectral modification of type IA fibre Bragg gratings by high-powernear-infrared lasers,” Meas. Sci. Technol. 17(5), 968–974 (2006). [CrossRef]
24. S. A. Vasiliev, O. I. Medvedkov, V. G. Plotnichenko, E. M. Dianov, and A. O. Rybaltovsky, “Increased solubility of molecular hydrogen in UV-exposed germanosilicate fibers,” Opt. Lett. 31(1), 11–13 (2006). [CrossRef] [PubMed]
25. H. R. Sørensen, H. J. Deyerl, and M. Kristensen, “Thermal stability of UV-written gratings in low- and high Ge content fibers,” in Bragg gratings, Photosensitivity and Poling in Glass Waveguides conference, Technical Digest (Optical Society of America, 2003), paper MD-30.
26. C. Dalle, P. Cordier, C. Depecker, P. Niay, P. Bernage, and M. Douay, “Growth kinetics and thermal annealing of UV-induced H-bearing species in hydrogen loaded germanosilicate fibre performs,” J. Non-Cryst. Solids 260(1-2), 83–98 (1999). [CrossRef]
27. K. M. Davis and M. Tomozawa, “An infrared spectroscopic study of water-related species in silica glasses,” J. Non-Cryst. Solids 201(3), 177–198 (1996). [CrossRef]
28. B. I. Greene, D. M. Krol, S. G. Kosinski, P. J. Lemaire, and P. N. Saeta, “Thermal and photo-initiated reactions of H2 with germanosilicate optical fibers,” J. Non-Cryst. Solids 168(1-2), 195–199 (1994). [CrossRef]
29. P. I. Gnusin, S. A. Vasiliev, O. I. Medvedkov, and E. M. Dianov, “Temperature-resolved spectroscopy of UV-induced absorption in H2-loaded germanosilicate fiber,” in Bragg gratings, Photosensitivity and Poling in Glass Waveguides conference, Technical Digest (CD) (Optical Society of America, 2012), paper BM4D.3.
30. M. Lancry, P. Niay, M. Douay, C. Depecker, P. Cordier, and B. Poumellec, “Isochronal annealing of BG written either in H2-loaded, UV hypersensitized or in OH-flooded standard telecommunication fibers using ArF laser,” J. Lightwave Technol. 24(3), 1376–1387 (2006). [CrossRef]
31. V. Grubsky, D. S. Starodubov, and J. Feinberg, “Photochemical reaction of hydrogen with germanosilicate glass initiated by 3.4 5.4-eV ultraviolet light,” Opt. Lett. 24(11), 729–731 (1999). [CrossRef] [PubMed]
32. S. Bandyopadhyay, J. Canning, P. Biswas, M. Stevenson, and K. Dasgupta, “A study of regenerated gratings produced in germanosilicate fibers by high temperature annealing,” Opt. Express 19(2), 1198–1206 (2011). [CrossRef] [PubMed]
33. S. A. Vasiliev, O. I. Medvedkov, A. S. Bozhkov, and E. M. Dianov, “Annealing of UV-induced fiber gratings written in Ge-doped fibers: investigation of dose and strain effects,” in Bragg gratings, Photosensitivity and Poling in Glass Waveguides conference, Technical Digest (Optical Society of America, 2003), paper MD-31.
