Enhanced radiation resistance of ytterbium-doped silica fiber by pretreating on a fiber preform

Chongyun Shao; Chongyun Shao; Yan Jiao; Yan Jiao; Fengguang Lou; Meng Wang; Lei Zhang; Suya Feng; Shikai Wang; Danping Chen; Chunlei Yu; Chunlei Yu; Chunlei Yu; Lili Hu; Lili Hu; Lili Hu

doi:10.1364/OME.384362

1. Introduction

Ytterbium-doped silica fiber (YDF) lasers offer a very attractive technology to implement space laser communication, and space trash disposal, owing to their reduced low weight, small size, and high peak power combined with good beam quality [1,2]. However, the YDFs will face a harsh ion radiation environment (such as proton, electron, X- and γ-ray) during their space mission. The radiation-induced darkening (RD) effect, which is mainly due to the generation of color centers, can seriously deteriorate the laser performance of YDF [3–5]. Loading H₂ or D₂ has been proven to be the most effective method to inhibit the RD effect [6,7]. However, more than 70% of H₂ or D₂ molecules will easily diffuse out from the fiber core within three months at room temperature. To prevent the out-diffusion of gas, two physical methods have been proposed, covering the fiber with a hermetic coating [8] and developing a new fiber structure [9]. Although effective, the processes of these two methods are very complicated, and these two methods are not suitable for the common double-clad fibers.

In this work, a new and chemical method by pretreating on fiber preform has been proposed to inhibit the out-diffusion of gas and to improve the radiation resistance of optical fiber. This pretreatment method involves three steps: loading deuterium (D₂), pre-irradiation and thermal annealing. The purpose of each step is as follows: 1) Loading D₂ under high-temperature and high-pressure is used to promote the diffusion of D₂ molecules into preform core glass; 2) Pre-irradiation is aimed to provide activation energy for the chemical reaction between D₂ molecule and core glass, thus to prevent D₂ molecules from diffusing out of core glass; 3)Thermal annealing helps to bleach color centers formed in the pre-irradiation process, and to release residual D₂ molecules that are not chemically bonded to the core glass. It helps to reduce the background loss and improve the laser slope efficiency of the active fiber. Effects of pretreatment condition on the optical loss at 1200 nm and laser slope efficiency of YDFs before and after γ-radiation were comparatively investigated. And the related mechanisms were revealed by combining the optical absorption, continuous wave electron paramagnetic resonance (CW-EPR), Raman and Fourier transform infrared (FTIR) spectroscopies.

2. Experimental details

2.1 Sample preparation

Figure 1 shows the experimental flow chart. Yb³⁺/Al³⁺/P⁵⁺/Ce³⁺/F⁻ co-doped preform with P/Al molar ratio more than one was prepared via traditional modified chemical vapor deposition (MCVD) combined with solution-doping method. The detailed fabrication process of fiber preform has been described in our previous patent [10]. The preform was cut into two sections: one undergoes loading gas, pre-irradiation and thermal annealing pretreatment in turn (named pretreated), the other one without any pretreatment was used as a reference (named pristine). Detailed pre-treatment condition is shown in Table 1.

Fig. 1. Flow chart of sample preparation and test

Download Full Size | PDF

Table 1. Detailed pretreatment condition for fiber preform

View Table | View all tables in this article

Two optical fibers with a core diameter of 20 µm and inner cladding of 400 µm were drawn from pristine and pretreated preforms, respectively. The fiber was coated by acrylate with thickness of 60 um as outer cladding. For convenience, the two fibers are called as pristine and pretreated fibers, respectively. The numerical aperture (N_A) of two optical fibers is almost the same, 0.068. The absorption coefficient at 970 nm of pristine and pretreated fibers is 0.92 and 0.9 dB/m, respectively.

2.2 Tests for preforms

In order to study the effect of pretreatment condition on the OD content and glass structure, the preform was cut and polished to 2 mm slice for Fourier transform infrared (FTIR) and Raman measurements. FTIR spectra were recorded via a FTIR spectrophotometer (Nicolet 6700). Raman spectroscopy was performed using a Renishaw inVia Raman microscope. The OD content was determined based on the transmission of fundamental OD band located at 2600 cm^-1 in FTIR spectrum, using Lambert-Beer law [11]:

(1)$$\def\upmu{\unicode[Times]{x00B5}}{C_{OD}} = \frac{{{M_{OD}}}}{{\varepsilon \rho }} \times \frac{1}{L} \times lg\frac{{{T_0}}}{T} \times {10^6}$$

Where C_OD is the OD content (weight ppm), M_OD is the molecular weight of the OD species, L is the sample thickness (cm), ρ is the sample density (g/cm³), and ɛ is the molar extinction coefficient, here ɛ=77.5 L/(mol*cm). T₀ and T are the transmission of baseline and OD absorption peak, respectively.

In order to study the effect of pretreatment condition on the radiation-induced color centers, a part of pristine and pretreated preforms were re-irradiated by γ-ray. Considering the shielding of cladding glass to core glass, the total γ-radiation dose is accumulated to 1700Gy at a dose rate of 1.67Gy/min. Preform slice with thickness of 2 mm was used for absorption test. The absorption spectra were recorded using a UV−vis−NIR spectrophotometer (Lambda 950). Power sample of core glass weighting about 200 mg was used for continuous wave electron paramagnetic resonance (CW-EPR) test. The CW-EPR experiments were carried out on a Bruker X-Bond EPR spectrometer (ELEXSYS E 580).

2.3 Tests for fibers

In order to study the effect of pretreatment condition on the laser performance, the loss spectrum and laser slope efficiency were recorded. The optical loss of optical fibers was tested by cut off method. The laser slope efficiency of optical fibers was measured by a simple fiber laser setup as shown in Fig. 2. The pump source is a 970 nm laser diode (LD) with beam spot of 200µm and N_A of 0.2. The pump light was coupled into the cladding of optical fiber by a pair of aspherical lenses. A butt-coupled dichroic mirror (transmission more than 99% at 970 nm, reflectivity more than 99% from 1040 to 1130 nm) and a fiber output end face with ∼4% Fresnel reflectivity was used to form the laser cavity. A filter with transmission less than 0.5% at 970 nm was placed in front of a power meter to make sure that only laser was detected. The length of all optical fibers under laser slope efficiency test was about 25m. Since the absorption coefficient at 970 nm is more than 0.9 dB/m, such fiber length (25m) ensures that more than 99% of the pump laser is absorbed by the fiber, but this fiber length was not optimized. If the optimal fiber length is determined, the laser performance and radiation resistance may be better, as reported by Ladaci et al. [12] in Er³⁺-doped silica fiber.

Fig. 2. Schematic diagram of the fiber laser experimental setup

Download Full Size | PDF

In order to evaluate the radiation resistance behaviors of optical fiber, it was exposed to γ-ray to a total dose of 700Gy at a dose rate of 0.34Gy/min. Such a dose corresponds to the total radiation dose absorbed by the satellite during its ten-year service in geosynchronous orbit (GSO) [13].

In order to accelerate the evaluation of radiation resistance stability of the pretreated fiber under vacuum environment, a vacuum drying oven was used to heat and vacuum the pretreated fiber. The vacuum treated conditions are as follows: the pressure was about 100Pa, the temperature was constant at 70℃, and the duration time was 30 days.

3. Results and discussion

3.1 Effect of the pretreatment condition on the OD content and glass structure

Figure 3(a) shows the FTIR spectra of liquid nature water (H₂O), heavy water (D₂O), and their mixed solution (H₂O & D₂O). Only 3400 cm⁻¹ band is observed in H₂O. And only 2500 cm⁻¹ band is observed in D₂O. Both 3400 and 2500 cm⁻¹ bonds are observed in H₂O and D₂O mixed solution. The 3400 cm⁻¹ band is due to the fundamental vibration of OH groups in H₂O. The 2500 cm⁻¹ band is due to the fundamental vibration of OD groups in D₂O.

Fig. 3. Fourier transform infrared (FTIR) spectra of (a) liquid nature water (H₂O), heavy water (D₂O), and their mixed solution (H₂O & D₂O), (b) pristine, D₂ loaded and pretreated preform core glasses

Download Full Size | PDF

Figure 3(b) shows the FTIR spectra of pristine and pretreated preform core glasses. The FTIR of D₂ loaded sample is also added into Fig. 3(b) for comparison. For pristine sample, two absorption bands located at 2250 and 2660 cm⁻¹ can be observed, they are due to silica matrix related vibration (See Table 2 for more detail). For D₂ loaded and pretreated samples, in addition to the two absorption bands, a new absorption band located at 2640 cm⁻¹ can be observed. It's worth pointing out that the D₂ molecule in glass is infrared inactive due to its symmetric diatomic structure. In addition, the pretreated sample has been annealed at 900 ℃ for 10 hours before FTIR test, so it is unlikely that D₂ molecules still exists in pretreated sample. Therefore, the 2640 cm⁻¹ band has no direct relationship with D₂ molecules. Stone et al. [14,15] show that the fundamental vibration of OD groups in pure SiO₂ glass is located at 2650 cm⁻¹. The OD groups in pure SiO₂ glass is mainly bonded with silicon to form Si-OD bond. For Yb³⁺/Al³⁺/P³⁺ co-doped SiO₂ glass in this work, the OD groups may bond to silicon, aluminum and phosphorus atoms to form Si-OD, Al-OD and P-OD bonds. The 2640 cm⁻¹ band in this work may originate from the overlap of the fundamental vibrations of Si-OD, Al-OD and P-OD bonds. While the OD groups in liquid heavy water is primarily bonded to the deuterium atoms to form D-OD bond. The position change of OD fundamental vibration band in liquid water (H₂O/D₂O), in pure SiO₂ glass and Al³⁺/P³⁺ co-doped SiO₂ glass is caused by the local structural change of OD bond. Compared with the pristine and D₂ loaded samples, the OD absorption band is particularly obvious in the pretreated sample. According to Eq. (1), there are about 0, 12 and 79 ppm OD groups in pristine, D₂ loaded and pretreated samples, respectively. This result suggests that only a small amount of OD groups had been formed during high-temperature and high-pressure loading D₂ process, and a large number of OD groups were formed in the subsequent pre-irradiation process.

Table 2. Position and assignment of main OH, OD and Si-O-Si vibration absorption bands in silica glass

View Table | View all tables in this article

Izawa et al. [16] systematically studied the silica matrix related vibration absorption in silica glass. Stone et al. [14,15] systematically studied OH and OD related vibration absorption in silica glass. For comparison, the primary absorption bands due to OH, OD and Si-O-Si vibration in silica glass are listed in Table 2.

As well known the absorption of glass in the infrared region primarily originates from the molecular vibration. The frequency (ν) of molecular vibration depends on the reduced mass (µ) and force constant (K) of anions and cations in a vibrating group. The relationship between them is as follows [17]:

(2)$$\nu = \; \frac{1}{{2\pi c}}\; \sqrt {\frac{K}{\upmu }} $$

(3)$${\upmu } = \frac{{{m_1}{m_2}}}{{{m_1} + {m_2}}}$$

Here, c is the speed of light. µ is the reduced mass of anions and cations. m₁ and m₂ are the molar mass of anions and cations, respectively. K is approximately equal to the bond energy between anions and cations.

According to formula (2) and (3), since the K values of OD and OH are nearly the same, the vibration frequency of OH is about 1.374 times of that of OD due to the change in reduced mass as shown in Table 2.

Figure 4 shows the Raman spectra of pristine and pretreated preform core glasses. The Raman spectrum of pure silica glass (α-SiO₂) is also added into Fig. 4 for comparison. The Raman of pure silica glass presents a strong and broad band at 440 cm⁻¹ as well as five weak bands peaked at 490, 606, 800, 1160 and 1200 cm⁻¹, respectively. Among them, the 440, 800, 1060 and 1200 cm⁻¹ are ascribed to symmetric bending (ω₁), symmetric stretching(ω₃), TO and LO asymmetric stretching (TO-ω₄ and LO-ω₄) vibration modes of Si-O-Si bonds, respectively [19]. The positions of these silica matrix related vibration modes in Raman spectrum are basically the same as those in FTIR spectrum (See Table 2). The 490 and 600 cm⁻¹ bands are due to the planar four- and three-fold ring units, respectively [19]. In addition to these featured Raman bands, two new Raman bands peaked at 1150 and 1325 cm⁻¹ can be also observed in pristine and pretreated preform core glasses. The broad band at 1000-1250 cm⁻¹ is derived from the superposition of Al-O-P, Al-O-Si, P-O-Si and Si-O-Si stretching vibration modes [20,21]. The strong peak at 1150 cm⁻¹ is generally regarded as the evidence of the existence of like-AlPO₄ structure which consists of Al-O-P linkages [20,21]. The weak and narrow band peaked at 1326 cm⁻¹ is ascribed to the stretching vibration of P = O bond in P⁽³⁾ unit [20,21]. Compared to the pristine sample, there was no significant change in the intensities of other vibration peaks, but an obvious decrease in the intensity of 1326 cm⁻¹ band and a weak decline in 490 cm⁻¹, which suggests that a large number of P = O bonds and parts of four-fold ring units have been destroyed during pretreatment process.

Fig. 4. Raman spectra of pristine and pretreated preform core glasses as well as pure silica glass (α-SiO₂), the inset in Fig. 4 shows a magnified view of Raman spectrum from 1275 to 1375 cm⁻¹.

Download Full Size | PDF

3.2 Effect of the pretreatment condition on the radiation-induced color centers

Figure 5(a) shows the radiation-induced absorption (RIA) spectra of pristine and pretreated preform core glasses, the RIA spectra were obtained by subtracting the absorption spectra of the non-irradiated samples from those of the 1.7 kGy gamma irradiated samples. The RIA intensity of pretreated sample is obviously lower than that of pristine sample, which suggests that the radiation resistance of pretreated sample is much better than that of pristine sample.

Fig. 5. Radiation-induced absorption (RIA) (a) and electron paramagnetic resonance (EPR) spectra (b) of pristine and pretreated preforms before and after 1700 Gy γ-irradiation.

Download Full Size | PDF

Figure 5(b) shows the CW-EPR spectra of pristine and pretreated preform core glasses before and after 1.7 kGy gamma radiation. Before irradiation, no obvious CW-EPR signal can be detected in pristine and pretreated samples, it suggests that the color centers formed in the pre-irradiation process of pretreated sample have been completely bleached during subsequent thermal annealing. After 1700 Gy irradiation, a hyperfine doublet with a 50 Gauss separation is only observed in the pristine sample, but it is absent in the pretreated sample, as shown in Fig. 4(b). This doublet CW-EPR signal is due to the phosphorus oxygen hole color (P-OHC) according to the study of Griscom et al. [22] in γ-irradiated phosphosilicate glass.

In Fig. 5(a), the RIA intensity in pretreated sample is greatly decreased compared to that of pristine sample, which is due to the inhibition of P-OHC as shown in Fig. 5(b). Based on the early studies [21,22], the absorption peaks of P-OHC are located at about 500 and 570 nm, whose FWHM are 130 and 100 nm, respectively. The broadband absorption band of P-OHC is primarily responsible for the radiation-induced loss in aluminum and phosphorus co-doped silica fiber with P/Al ratio more than one [21].

Early studies [21,23–25] show that the radiation-induced Al- and/or P-related defects in Al/P co-doped silica glasses are highly dependent on the structural changes, which are controlled by the P/Al molar ratio. When P/Al ≤ 1, P preferentially connects with Al to form like-AlPO₄ structure, and the excess Al primarily exists in the form of four-coordinated aluminum, which is the precursor of aluminum hole centers (Al-OHC). when P/Al > 1, Al preferentially connects with P to form like-AlPO₄ structure, and the excess P mainly exists in the form of P⁽³⁾ unit (P = O), which is the precursor to form phosphorus related defects. The P/Al molar ratio in preform core is slightly more than one in this work. The like-AlPO₄ structure and P⁽³⁾ unit have been observed by Raman (See Fig. 4). Since like-AlPO₄ structure is very stable and not easy to be damaged by high-energy rays, neither new Al-related groups (e.g. Al-OH) will be formed in the process of pretreatment, nor new Al-related defects (e.g. aluminum oxygen hole color) will be formed in the subsequent gamma re-irradiation process. However, the P = O double bonds are more likely to be destroyed than other single bonds (e.g. P-O, Al-O and Si-O bonds) in the processes of pretreatment and subsequent gamma re-irradiation [21].

Figure 6(a-d) show the structural diagram of mutual evolution of P-related groups and P-related defects under irradiation. For the preform without any pretreatment, POHC and P₂ color centers can be formed via ionizing P = O bond under irradiation [21,26]. The reaction model is shown in Fig. 6(a) and expressed as follows:

(4)$$2 \;{\ast \; }\begin{array}{c} {{{[{\textrm{O} = \textrm{P}{\textrm{O}_{3/2}}} ]}^0}}\\ {({\textrm{P} = {\textrm{O}\; \textrm{bond}}} )} \end{array}\mathop \to \limits^{{\; \textrm{h}\upsilon \; }} \; \begin{array}{c} {{{[{\textrm{P}{\textrm{O}_{5/2}} + {\textrm{h}^ + }} ]}^ + }\; }\\ {({\textrm{POHC}} )} \end{array} + \; \begin{array}{c} {{{[{\textrm{P}{\textrm{O}_{4/2}} + {\textrm{e}^ - }} ]}^0}\; }\\ {({{\textrm{P}_2}} )} \end{array} + {\textrm{e}^ - }$$

Here, e⁻ and h⁺ represent an electron and a hole, respectively.

Fig. 6. structural diagram of mutual evolution of P-related groups and P-related defects under irradiation for (a) pristine preform, (b) D₂ loaded preform, (c-d) pretreated preform

Download Full Size | PDF

If the preform is loaded with D₂ gas before gamma radiation, the P = O bond will preferentially react with D₂ gas under pre-irradiation process to form P-OD and P-D bonds, resulting in the decrease of P = O vibration intensity in Raman spectrum (See the inset of Fig. 5). The reaction model is shown in Fig. 6(b) and described as follows:

(5)$$2 \;{\ast \; }\begin{array}{c} {{{[{\textrm{O} = \textrm{P}{\textrm{O}_{3/2}}} ]}^0}\; }\\ {({\textrm{P} = {\textrm{O}\; \textrm{bond}}} )} \end{array} + {\; }{\textrm{D}_2}\mathop \to \limits^{{\; \textrm{h}\upsilon \; }} {\; }\begin{array}{c} {{{[{\textrm{DO} - \textrm{P}{\textrm{O}_{3/2}}} ]}^ + }\; }\\ {({\textrm{P} - \textrm{OD}} )} \end{array} + \; \begin{array}{c} {\; {{[{\textrm{D} - \textrm{P}{\textrm{O}_{4/2}}} ]}^{2 + }}\; }\\ {({\textrm{P} - \textrm{D}} )} \end{array} + 3{\textrm{e}^ - }$$

Molecular orbital calculation suggests that the Si-O-Si bonds in planar four- and three-fold ring units are extremely unstable in thermodynamics due to their higher strain energy [27]. In addition to the P = O bonds, the strain Si-O··Si bands in four- and three-fold ring units will also react with D₂ gases under subsequent pre-irradiation process to form Si-OD and Si-D bands, resulting in the decrease of four-fold ring vibration intensity in Raman spectrum (See Fig. 4). The formation of Si-OD (2640 cm⁻¹) and Si-D (1700 cm⁻¹) bands caused by deuterium loading has been confirmed by Hartwig et al. [28] via in-situ Raman spectroscopy in pure silica glass with high OD content (∼ 1200 ppm). It can be described as follows:

(6)$$\begin{array}{c} {[{\textrm{O}_{3/2}{\textrm{Si}} - \textrm{O} \cdot\cdot {\textrm{Si}}{\textrm{O}_{3/2}}]^0}\; }\\ {({\textrm{Si} - \textrm{O} \cdot\cdot {\textrm{Si}}} )} \end{array} + \; {\textrm{D}_2}\mathop \to \limits^{{\textrm{h}\upsilon \; }} {\; }\begin{array}{c} {{{[{{\textrm{O}_{3/2}}\textrm{Si} - \textrm{OD}} ]}^0}\; }\\ {({\textrm{Si} - \textrm{OD}} )} \end{array} + {\; }\begin{array}{c} {{{[{\textrm{D} - \textrm{Si}{\textrm{O}_{3/2}}} ]}^{2 + }}\; }\\ {({\textrm{Si} - \textrm{D}} )} \end{array} + 2{\textrm{e}^ - }$$

The existence of OD groups containing P-OD and Si-OD bonds have been confirmed by FTIR spectrum (See Fig. 3(b)). However, neither Raman nor FTIR can detect the vibration peak (which is located at 1700 cm⁻¹) of Si-D or P-D bond in this work (not shown). This result indicates that the content of Si-D and P-D bonds is far less than that of Si-OD and P-OD bonds.

In fact, since two positively charged atoms (e.g. D⁺ and Si⁴⁺, or D⁺ and P⁵⁺) tend to repel each other, the wrong Si-D and P-D bonds are very unstable and easily ionized during irradiation (See Fig. 6(c)), they can be described as follows:

(7)$$\begin{array}{c} {{{[{\textrm{D} - \textrm{Si}{\textrm{O}_{3/2}}} ]}^{2 + }}\; }\\ {({\textrm{Si} - {\textrm{D}\; \textrm{bond}}} )} \end{array}\mathop \leftrightarrow \limits^{{\textrm{h}\upsilon }} \; \begin{array}{c} {{{[{\textrm{D}\cdot} ]}^0}\; }\\ {({{\textrm{Deuterium}\; \textrm{radical}}} )} \end{array} + \begin{array}{c} {{{[{\cdot\textrm{Si} - {\textrm{O}_{3/2}}} ]}^0}\; }\\ {({\textrm{Si} - \textrm{E}^{\prime} } )} \end{array}$$

(8)$$\begin{array}{c} {\; {{[{\textrm{D} - \textrm{P}{\textrm{O}_{4/2}}} ]}^{2 + }}\; }\\ {({\textrm{P} - {\textrm{D}\; \textrm{bond}}} )} \end{array}\mathop \leftrightarrow \limits^{{\textrm{h}\upsilon }} \; \begin{array}{c} {{{[{\textrm{D}\cdot} ]}^0}\; }\\ {({{\textrm{Deuterium}\; \textrm{radical}}} )} \end{array} + \begin{array}{c} {{{[{\textrm{P}{\textrm{O}_{4/2}} + {\textrm{e}^ - }} ]}^0}\; }\\ {({{\textrm{P}_2}} )} \end{array}$$

During re-irradiation, two deuterium radicals tend to dimer into D₂ gas molecule, then can inhibit the formation of POHC and P₂ color center according to the reactions (4) and (5). Deuterium radicals can also react with POHC color center, which has been created by gamma-ray, to form P-OD bond (See Fig. 6(d)), it can be described as follows:

(9)$$\begin{array}{c} {{{[{\textrm{D}\cdot} ]}^0}\; }\\ {({{\textrm{Deuterium}\; \textrm{radical}}} )} \end{array} + \begin{array}{c} {{{[{\textrm{P}{\textrm{O}_{5/2}} + {\textrm{h}^ + }} ]}^ + }\; }\\ {({\textrm{POHC}} )} \end{array}\mathop \to \limits^{{\; \textrm{h}\upsilon \; }} \; \begin{array}{c} {{{[{\textrm{DO} - \textrm{P}{\textrm{O}_{3/2}}} ]}^ + }\; }\\ {({\textrm{P} - \textrm{OD}} )} \end{array}\; $$

Based on the above analysis, there are two main reasons responsible for the radiation hardening of pretreated preform or fiber, one is due to the decrease of precursors (P = O bonds) of POHC color centers, the other one is ascribed to the existence of deuterium radicals (D·) ionized from Si-D and P-D bonds.

3.3 Effect of the pretreatment condition on laser performance and radiation resistance

It is generally known that the laser performance of YDF will be seriously deteriorated by the presence of many hydroxyl groups (OH). In order to evaluate the influence of OD groups brought by pretreatment process on the laser performance and radiation resistance of YDFs, the optical loss at 1200 nm and laser slope efficiency of pristine and pretreated YDFs before and after γ-radiation was comparatively investigated.

Figures 7(a) and (b) present the optical loss spectra of pristine and pretreated optical fibers, respectively. The optical loss at 1200 nm in pristine and pretreated fibers are 6 and 20 dB/km, respectively. Two strong absorption bands located at 1384 and 1263 nm as well as a weak absorption band peaked at 1244 nm in pretreated fiber can be observed, but only one band peaked at 1384 nm in pristine fiber can be observed. According to the early studies [14,15], the 1384, 1263 and 1244 nm bands are attributed to first overtone of OH, second overtone of OD, and combination tone of OH, respectively (See Table 2 for more detail). The 1263nm band further confirms the existence of OD groups in the pretreated fiber. Compared with pristine fiber, the increase of OH absorption peak (1384 nm) in pretreated fiber indicates that the D₂ used in the pretreatment process may contain a small amount of H₂. This is further supported by the FTIR spectrum, in which the intensity of OH fundamental vibration (3600 cm⁻¹) in pretreated preform is stronger than that in pristine preform (See Fig. 3(b)).

Fig. 7. Optical loss spectra (a, b, c) and laser slope efficiency (d, e, f) of pristine (a, d), pretreated (b, e), and pretreatment combined with vacuum treated (c, f) optical fibers before and after 700 Gy γ-irradiation.

Download Full Size | PDF

Figures 7(d) and (e) show laser slope efficiency of pristine and pretreated optical fibers, respectively. For the non-irradiated optical fibers, the laser slope efficiency in pristine and pretreated fiber are 79% and 75%, respectively. This result suggests that pretreatment in deuterium gas has no obvious negative impact on the laser properties of non-irradiated YDFs. However, our previous work shows that if H₂ is used instead of D₂ to pretreat on the YDF preform using the same method and condition, the laser efficiency will be halved (from 79% to 45%) in H₂ pretreated fiber (not reported). The difference between H₂ and D₂ pretreated fibers is correlated to the absorption positions and intensities of OH and OD groups in the near infrared band (See Table 2). For the same content of OH and OD, the vibration intensity of second overtone of OD (1.26µm) is about 1/60 of that of first overtone of OH (1.38µm), even less [18]. Compared with the first overtone of OH (1.38µm), the first overtone of OD (1.86µm) is further away from the pump and laser wavelength (1µm) of YDFs. Therefore, the negative effect of OD on the laser performance of YDF is much less than that of OH group.

After 700 Gy gamma-irradiation, the optical loss at 1200 nm of the pristine fiber is increased by 88.8 times (6→533 dB/km), while the optical loss of the pretreated fiber is only increased by 3.5 times (20→70 dB/km) as shown in Figs. 7(d) and (e). At the same time, the pristine fiber has no laser output at all, while the slope efficiency of D₂ pretreated fiber is decreased by 21%, from 75% to 59% as shown in Figs. 7(d) and (e). This result suggests that the radiation resistance of pretreated fiber has been greatly improved compared with that of pristine fiber, which is primarily ascribed to the inhibition of P-OHC in pretreated fiber (See Fig. 4).

As is well known, space is a complex environment with high vacuum, strong radiation and large temperature difference. Under the condition of vacuum and heating, the H₂ or D₂ molecules in common double clad fiber will be accelerated to diffuse out of the fiber, resulting in the failure of the radiation resistance of the fiber. In order to verify that the pretreated fiber does have no the problem of gas out-diffusion, but may have a long-term anti-radiation stability, the pretreated fiber was further treated in a vacuum drying oven. The vacuum treated conditions are as follows: the pressure was about 100Pa, the temperature was constant at 70 ℃, and the duration time was 30 days.

Figures 7(c) and (f) present the optical loss and laser slope efficiency of pretreatment combined with vacuum treated fiber, respectively. Before irradiation, the optical loss at 1200 nm and laser slope efficiency of this fiber are 22dB/km and 74%, respectively. After 700 Gy irradiation, the optical loss and laser slope efficiency of this fiber are 74dB/km and 56%, respectively. The optical loss at 1200 nm and laser slope efficiency of pretreatment combined with vacuum treated fibers are almost identical to these of pretreated fibers, regardless of before and after irradiation. This result further confirms that the pretreatment method proposed in this paper causes the D₂ molecule chemically reacts with the glass network to form stable chemical bonds (such as Si-OD), which can inhibit the diffusion of deuterium out of the fiber even in vacuum and heating condition. It ensures that the pretreated fiber may have long-term radiation resistance stability when it is used in a vacuum environment (such as space).

4. Conclusion

The radiation resistance of ytterbium-doped silica fiber (YDF) can be significantly improved by a new and chemical pretreatment method on fiber preform. It consists of loading deuterium (D₂), pre-irradiation and thermal annealing pretreatment on the YDF preform. After an irradiation dose level of 700 Gy, the degradation of laser output power in the YDF made from the pretreated preform (named pretreated YDF) remains below 21% (from 79% to 59%), whereas the YDF made from the pristine preform without any pretreatment (named pristine YDF) has no laser output at all. Electron paramagnetic resonance testing shows that the phosphorus oxygen hole centers (POHCs), which is primarily responsible for the radiation-induced darkening effect, can be inhibited in pretreated YDF during the subsequent re-irradiation process. Raman and FTIR spectra show that the inhibition of POHCs in pretreated YDF is correlated to the decrease of precursors (P = O bonds) of POHCs and the existence of deuterium radicals (D·). Laser experiments show that this new pretreatment method has no obvious negative impact on the optical loss at 1200 nm (6 vs. 20 dB/km) and laser slope efficiency (79% vs. 75%) of non-irradiated YDFs, which is ascribed to the first overtone of OD (1.87 µm) is farther away from the pump and laser wavelength of YDF (1µm) compared with that of OH (1.38 µm). Vacuum experiment on the pretreated YDF predicts that the pretreated YDF may have a long-term radiation resistance stability when it is used in a vacuum environment (such as space).

Funding

National Natural Science Foundation of China (61775224, 61875216).

Disclosures

The authors declare no conflicts of interest.

References

1. M. W. Wright and G. C. Valley, “Yb-doped fiber amplifier for deep-space optical communications,” J. Lightwave Technol. 23(3), 1369–1374 (2005). [CrossRef]

2. Y. Wang, G. Chen, and J. Li, “Development and prospect of high-power doped fibers,” High Power Laser Sci. Eng. 6, e40 (2018). [CrossRef]

3. J. Duchez, F. Mady, Y. Mebrouk, N. Ollier, and M. Benabdesselam, “Interplay between photo- and radiation-induced darkening in ytterbium-doped fibers,” Opt. Lett. 39(20), 5969–5972 (2014). [CrossRef]

4. F. X. Fenghou Xie, C. S. Chongyun Shao, F. L. Fengguang Lou, M. W. Meng Wang, C. Y. Chunlei Yu, S. F. Suya Feng, and L. H. Lili Hu, “Effect of power scale of 974 and 633 nm lasers on the induced loss at 633 nm of Yb³⁺/Al³⁺ co-doped silica fiber,” Chin. Opt. Lett. 16(1), 010603 (2018). [CrossRef]

5. M. L. Miao Li, G. C. Gui Chen, J. L. Jinyong Leng, J. C. Jinbao Chen, and J. L. Jinyan Li, “Non-isothermal bleaching of Yb-Li co-doped silica fibers,” Chin. Opt. Lett. 17(3), 031401 (2019). [CrossRef]

6. N. Zhao, W. Li, J. Li, G. Zhou, and J. Li, “Elimination of the Photodarkening Effect in an Yb-Doped Fiber Laser With Deuterium,” J. Lightwave Technol. 37(13), 3021–3026 (2019). [CrossRef]

7. S. Girard, A. Morana, A. Ladaci, T. Robin, L. Mescia, J. Bonnefois, M. Boutillier, J. Mekki, A. Paveau, B. Cadier, E. Marin, Y. Ouerdane, and A. Boukenter, “Recent advances in radiation-hardened fiber-based technologies for space applications,” J. Opt. 20(9), 093001 (2018). [CrossRef]

8. K. V. Zotov, M. E. Likhachev, A. L. Tomashuk, M. M. Bubnov, M. V. Yashkov, A. N. Guryanov, and S. N. Klyamkin, “Radiation-Resistant Erbium-Doped Fiber for Spacecraft Applications,” IEEE Trans. Nucl. Sci. 55(4), 2213–2215 (2008). [CrossRef]

9. S. Girard, A. Laurent, E. Pinsard, T. Robin, B. Cadier, M. Boutillier, C. Marcandella, A. Boukenter, and Y. Ouerdane, “Radiation-hard erbium optical fiber and fiber amplifier for both low- and high-dose space missions,” Opt. Lett. 39(9), 2541–2544 (2014). [CrossRef]

10. F. Lou, C. Yu, L. Hu, M. Wang, X. Xu, L. Zhang, S. Feng, and D. Chen, “Ytterbium aluminum phosphorus fluorine doped silica fiber preform core rod and its preparation method,” (2019.04.19, 2016).

11. K. M. Davis, A. Agarwal, M. Tomozawa, and K. Hirao, “Quantitative infrared spectroscopic measurement of hydroxyl concentrations in silica glass,” J. Non-Cryst. Solids 203, 27–36 (1996). [CrossRef]

12. A. Ladaci, S. Girard, L. Mescia, T. Robin, A. Laurent, B. Cadier, M. Boutillier, Y. Ouerdane, and A. Boukenter, “Optimized radiation-hardened erbium doped fiber amplifiers for long space missions,” J. Appl. Phys. 121(16), 163104 (2017). [CrossRef]

13. M. N. Ott, X. L. Jin, R. Chuska, P. Friedberg, M. Malenab, and A. Matuszeski, “Space flight requirements for fiber optic components: qualification testing and lessons learned,” in Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series (2006).

14. J. Stone, “Interactions of hydrogen and deuterium with silica optical fibers: A review,” J. Lightwave Technol. 5(5), 712–733 (1987). [CrossRef]

15. J. Stone, “Reduction of OH absorption in optical fibers by OH→OD isotope exchange,” Ind. Eng. Chem. Prod. Res. Dev. 25(4), 609–621 (1986). [CrossRef]

16. T. Izawa, N. Shibata, and A. Takeda, “Optical attenuation in pure and doped fused silica in the IR wavelength region,” Appl. Phys. Lett. 31(1), 33–35 (1977). [CrossRef]

17. B. Staurt, Infrared Spectroscopy: Fundamentals and Applications (John Wiley and Sons, Ltd., 2004).

18. B. Kumar, N. Fernelius, and J. A. Detrio, “Deuterium treatment and infrared transmission spectra of fused silica,” J. Am. Ceram. Soc. 64(12), C-178–C-180 (1981). [CrossRef]

19. F. L. Galeener and A. E. Geissberger, “Raman studies of vitreous SiO₂ versus fictive temperature,” Phys. Rev. B 28(6), 3266–3271 (1983). [CrossRef]

20. S. Wang, F. Lou, C. Yu, Q. Zhou, M. Wang, S. Feng, D. Chen, L. Hu, W. Chen, M. Guzik, and G. Boulon, “Influence of Al³⁺ and P⁵⁺ ion contents on the valence state of Yb³⁺ ions and the dispersion effect of Al³⁺ and P⁵⁺ ions on Yb³⁺ ions in silica glass,” J. Mater. Chem. C 2(22), 4406–4414 (2014). [CrossRef]

21. C. Shao, J. Ren, F. Wang, N. Ollier, F. Xie, X. Zhang, L. Zhang, C. Yu, and L. Hu, “Origin of Radiation-Induced Darkening in Yb³⁺/Al³⁺/P⁵⁺ -Doped Silica Glasses: Effect of the P/Al Ratio,” J. Phys. Chem. B 122(10), 2809–2820 (2018). [CrossRef]

22. D. L. Griscom, E. J. Friebele, K. J. Long, and J. W. Fleming, “Fundamental defect centers in glass: electron spin resonance and optical absorption studies of irradiated phosphorus-doped silica glass and optical fibers,” J. Appl. Phys. 54(7), 3743–3762 (1983). [CrossRef]

23. S. Girard, A. Alessi, N. Richard, L. Martin-Samos, V. De Michele, L. Giacomazzi, S. Agnello, D. Di Francesca, A. Morana, and B. Winkler, “Overview of radiation induced point defects in silica-based optical fibers,” Reviews in Physics 4, 100032 (2019). [CrossRef]

24. M. E. Likhachev, M. M. Bubnov, K. V. Zotov, A. Tomashuk, D. S. Lipatov, M. V. Yashkov, and A. Guryanov, “Radiation resistance of Er-doped silica fibers: effect of host glass composition,” J. Lightwave Technol. 31(5), 749–755 (2013). [CrossRef]

25. S. Girard, Y. Ouerdane, B. Tortech, C. Marcandella, T. Robin, B. Cadier, J. Baggio, P. Paillet, V. Ferlet-Cavrois, and A. Boukenter, “Radiation effects on ytterbium-and ytterbium/erbium-doped double-clad optical fibers,” IEEE Trans. Nucl. Sci. 56(6), 3293–3299 (2009). [CrossRef]

26. G. Origlio, F. Messina, M. Cannas, R. Boscaino, S. Girard, A. Boukenter, and Y. Ouerdane, “Optical properties of phosphorus-related point defects in silica fiber preforms,” Phys. Rev. B 80(20), 205208 (2009). [CrossRef]

27. F. L. Galeener, “Planar rings in vitreous silica,” J. Non-Cryst. Solids 49(1-3), 53–62 (1982). [CrossRef]

28. C. M. Hartwig, “The radiation-induced formation of hydrogen and deuterium compounds in silica as observed by Raman scattering,” J. Chem. Phys. 66(1), 227–238 (1977). [CrossRef]

Sample	Loading Gas				Pre-irradiation			Thermal annealing
Sample	Gas	Temperature (℃)	Pressure (MPa)	Time (day)	Rays	Dose (kGy)	Time (h)	Temperature (℃)	Time (h)
Pristine	None	None	None	None	None	None	None	None	None
Pretreated	D₂	280	10	30	γ-ray	100	17	900	10

Vibration group	Frequency (cm⁻¹)	Wavelength (µm)	Assignment		Reference
OH	3620	2.72	ν(OH)	Fundamental vibration	[14,15]
	7250	1.38	2ν(OH)	First overtone	[14,15]
	8060	1.24	2ν(OH) + ν₁	Combination tone	[14,15]
	10530	0.95	3ν(OH)	Second overtone	[14,15]
OD	2650	3.8	ν(OD)	Fundamental vibration	[14,15,18]
	5350	1.87	2ν(OD)	First overtone	[14,15,18]
	6030	1.66	2ν(OH) + ν₁	Combination tone	[14,15,18]
	7940	1.26	3ν(OH)	Second overtone	[14,15,18]
Si-O-Si	280	35.7	ν₂	Symmetric bending	[16]
	430	23.1	ν₄	Asymmetric bending	[16]
	800	12.5	ν₁	Symmetric stretching	[16]
	1060	9.4	ν₃	Asymmetric stretching	[16]
	2660	3.8	2ν₃+ν₄;	Combination vibration	[16]
			2ν₁+2ν₂	Combination vibration
	2250	4.5	2ν₃;	First overtone	[16]
			ν₁+ν₂+ν₃	Combination vibration

Sample	Loading Gas				Pre-irradiation			Thermal annealing
Sample	Gas	Temperature (℃)	Pressure (MPa)	Time (day)	Rays	Dose (kGy)	Time (h)	Temperature (℃)	Time (h)
Pristine	None	None	None	None	None	None	None	None	None
Pretreated	D₂	280	10	30	γ-ray	100	17	900	10

Vibration group	Frequency (cm⁻¹)	Wavelength (µm)	Assignment		Reference
OH	3620	2.72	ν(OH)	Fundamental vibration	[14,15]
	7250	1.38	2ν(OH)	First overtone	[14,15]
	8060	1.24	2ν(OH) + ν₁	Combination tone	[14,15]
	10530	0.95	3ν(OH)	Second overtone	[14,15]
OD	2650	3.8	ν(OD)	Fundamental vibration	[14,15,18]
	5350	1.87	2ν(OD)	First overtone	[14,15,18]
	6030	1.66	2ν(OH) + ν₁	Combination tone	[14,15,18]
	7940	1.26	3ν(OH)	Second overtone	[14,15,18]
Si-O-Si	280	35.7	ν₂	Symmetric bending	[16]
	430	23.1	ν₄	Asymmetric bending	[16]
	800	12.5	ν₁	Symmetric stretching	[16]
	1060	9.4	ν₃	Asymmetric stretching	[16]
	2660	3.8	2ν₃+ν₄;	Combination vibration	[16]
			2ν₁+2ν₂	Combination vibration
	2250	4.5	2ν₃;	First overtone	[16]
			ν₁+ν₂+ν₃	Combination vibration

Enhanced radiation resistance of ytterbium-doped silica fiber by pretreating on a fiber preform

Abstract

1. Introduction

2. Experimental details

2.1 Sample preparation

2.2 Tests for preforms

2.3 Tests for fibers

3. Results and discussion

3.1 Effect of the pretreatment condition on the OD content and glass structure

3.2 Effect of the pretreatment condition on the radiation-induced color centers

3.3 Effect of the pretreatment condition on laser performance and radiation resistance

4. Conclusion

Funding

Disclosures

References

Cited By

Figures (7)

Tables (2)

Equations (9)

Optical Materials Express