We demonstrate broadband optical amplification at 1.3 μm in silicate glass-ceramics containing β-Ga2O3:Ni2+ nanocrystals with 980 nm excitation for the first time. The optical gain efficiency is calculated to be about 0.283 cm-1 when the excitation power is 1.12 W. The optical gain shows similar wavelength dependence to luminescence properties.
© 2007 Optical Society of America
The requirement to improve solid-state lasers, optoelectronic communication devices has renewed the interest in activator doped glass. Glass ceramic (GC) materials are one of the most important hybrid optical materials which can be produced by controlled nucleation and crystallization of glass. This class of nanocomposite materials can interestingly be used as hosts for activators because they can potentially combine the advantages of both glass (such as easy fabrication) and doped crystal (such as optical activity) .
Transition metals (TM) ions such as Ni2+ and Cr4+ are important activators since they show broadband luminescence in the near infrared region when incorporated into crystal matrices . Unfortunately, TM ions only exhibit weak or even no luminescence in amorphous hosts, owing to the strong nonradiative relaxation. GC offers a potential solution to these problems. The design of novel matrix for TM ions is ultimately limited by two mechanisms. The first one is the necessary of certain crystal field environments, e.g. tetrahedral environment for Cr4+ and octahedral environment for Ni2+ [1, 3, 4]. The second one is nonradiative decay processes, which limit the quantum efficiency of TM ions in matrix. The requirement of certain crystal field environment indicates that the well studied transparent oxyfluoride GCs , which often provide 8-fold coordinated environment, are probably not suitable hosts for TM ions luminescence. To our knowledge, the GC matrixes for Cr4+ were mainly focused on forsterite, willemite and mixed orthosilicates and for Ni2+ was limited on spinel . At present, the relative high luminescence quantum efficiency was reported for Ni2+ doped GC materials in Ref. . Thus far, however, there are no reports about the realization of optical amplification in this class of hybrid optical materials. In this letter, we demonstrate an optical amplification at 1.3 μm in silicate GC containing β-Ga2O3:Ni2+ nanocrystals.
Ni2+-doped silicate glass sample was prepared by the conventional melting-quenching technique. The glass composition was 66.5SiO2-19.5Ga2O3-6.5Al2O3-7.5Na2O (mol%), doped with 0.15 mol% NiO. A mixture of SiO2 (A.R.), Al2O3 (A.R.), Na2CO3 (A.R.), Ga2O3 (5N) and NiO (4N) was melted in a corundum crucible at 1680 °C for 2 h at ambient temperature. The glass samples were then obtained by rapidly quenching the melt to room temperature and annealing at 400 °C for 10 h. Characteristic temperatures were measured by differential thermal analysis (DTA) at a heating rate of 10 °C/min under N2 atmosphere. The Ga2O3 nanocrystals were precipitated by heat treatment of the glass samples at 900 °C for 2 h. The sample was then cut and polished.
3. Results and discussion
From DTA measurement, glass transition temperature (Tg), first crystallization temperature (T x1) due to the precipitation of β-Ga2O3 crystals, and second crystallization temperature (T x2) corresponding to the crystallization of the glass matrix is 650 °C, 900 °C and 1150 °C, respectively. The XRD results show that the as-made glass is amorphous, while sharp diffraction peaks appear in the XRD pattern of the glass ceramic heat-treated at 900 °C for 2 h and all of them can be indexed by the diffraction peaks of monoclinic β-Ga2O3 structure. TEM results show that β-Ga2O3 particles have a relatively uniform diameter of about 50 nm. The detailed structural characterization of the glass ceramics can be found in Ref. .
Figure 1 shows the transmittance and absorption spectra of as-made glass and GC measured by a double-beam spectrophotometer (JASCO FP-6500). The transmittance of the GC is more than 90%. As shown in the inset of Fig. 1, the crystallization process has changed the absorption spectrum of glass. Just like other TM ions, Ni2+ is sensitive to crystal field because their valent electrons lack of shielding from surrounding crystal fields and the phenomenon is ascribed to the change of Ni-ion environment . The absorption bands centered at 440, 880 and 1750 nm in the as-made glass can be ascribed to the trigonal bipyramid fivefold and tetrahedral fourfold Ni2+ ions in silicate glasses . The broad absorption bands at 1011 and 606 nm in GC can be attributed to transitions from 3 A 2(F) ground state to the 3 T 2(F) and 3 T 1(F) excited states of octahedral Ni2+ . It is supposed that Ni2+ substituted Ga3+ since Ga atoms occupy both octahedral and tetrahedral sites in β-Ga2O3 . Pumped by a common commercial semiconductor at 980 nm, GC shows intense broadband emission centered around at about 1200 nm. The fluorescent lifetime is more than 665 μs at room temperature. On contrast, the as-made glass shows no emission in the near-infrared wavelength region.
A configuration resembling a traditional two-wave mixing geometry was chosen in gain measurements (shown in the Fig. 2). The thickness of sample for gain measurement was 4 mm. A tunable laser diode (New Focus Inc. 6324, 0.3 nm width) was used as seed source, in which the tunable region is from 1272 to 1348 nm. The seed beam was chopped at 200 Hz by a chopper (Stanford Research System Inc.: SR 540). A thermoelectrically cooled laser diode (Suwtech LDC-1500) of 980 nm center wavelength, which was followed by a focus lens to control the pump beam size for mode matching, was used for excitation. The amplified seed beam was detected by an InGaAs detector (Newport 818-BB-30). The detected electric pulses were displayed by a digital oscilloscope (Agilent, infiniium 54833A DSO).
Figure 3 shows optical gain properties at 1300 nm. The optical gain increases linearly with excitation power up to 1.12 W. The inset shows an oscilloscope image of the amplification phenomenon with the excitation power of 1.12 W and the optical gain is calculated at 1.12. Optical gain coefficient is defined as:
Where l is the thickness of glass sample and I/I0 is optical gain. The optical gain coefficient is calculated to be about 0.283 cm-1. The gain operation at 1300 nm has much significance since it is at the second telecommunication window and it has minimum dispersion for silica optical fiber. So it has potential application in optical communication system.
Ni2+-doped GC shows broadband near-infrared luminescence (see the inset of Fig. 4). In order to investigate its broadband amplification properties, the spectral dependence of optical gain was investigated. Figure 4 shows the results of optical gain as a function of different seed beam wavelength from 1272 to 1348 nm under excitation with 980 nm (the power is 1.12 W). Points and curve represents experimental measurements and fluorescence spectrum at 980 nm excitation. It can be seen that the measured spectral dependence of the optical gain appears to closely resemble the fluorescence spectrum. The optical amplification with 77 nm bandwidth indicates the present Ni2+-doped GC may be potential for developing broadband optical amplifier and tunable laser. Additionally, the spectral dependence of optical gain is an important consideration if a fiber amplifier is to be incorporated in a broadband WDM system.
The realization of optical amplification in Ni2+-doped β-Ga2O3 GC might be attributed to the improvement of the optical properties. Numerous researches have showed that TM ions luminescence is strongly baffled by nonradiative process and the local arrangement around TM ions, i.e. crystal field environments, take an important role [9–11]. The variation of crystal field environment can be explained by utilizing the crystal field parameters Dq and the Racah parameters B and C, which were determined by solving the Tanabe-Sugano matrix according to absorption spectrum . The comparison of crystal field parameters and luminescent decay time for Ni2+ in β-Ga2O3 GC and ever reported GC materials are summarized in table 1. The variation of Dq/B values and fluorescent lifetime follows the order:
To explain the experimental results, we employed the simplified Mott model about the probability of nonradiative transition , which obeys the activation energy relation:
Where W 0 is a constant, Δ is the energy gap and for 3 T 2(F)-3 A 2(F) transition it is the energy difference between the bottom of the 3 T 2(F) level and the intersection of the parabolas of the 3 T 2(F) and 3 A 2(F) levels, k is the Boltzmann constant, and T is the temperature. According to the Tanabe-Sugano energy levels diagram the 3 A 2(F) level is independent on the Dq/B value but the 3 T 2(F) level changes with the Dq/B parameter. As a result, the larger Dq/B value, the larger the activation energy Δ will be. This means that if Ni2+ occupies the strong crystal field sites, the nonradiative rates will be smaller, which results in intense luminescence and long fluorescent lifetime. The same conclusion can be acquired if a tunneling process between the 3 T 2(F) and 3 A 2(F) levels accounts for the nonradiative transitions. On the other hand, the phonon energy of matrix might affects the probability of nonradiative decay of active ion dopants . The moderate phonon energy of β-Ga2O3 (767 cm-1) also favors radiative transition process of Ni2+.
The gain system has many attractive features. In comparison with crystal, TM-doped GC can be easily fabricated into optical fibers keeping the optical active of TM ions . In present system, the silicate-based GC with broadband amplification at 1300 nm telecommunication window when it is excited with a common commercial semiconductor laser (980 nm), which implies this materials can overcome the disadvantages of the present Praseodymium-doped fluoride fiber amplifiers (PDFFAs) such as brittle and narrow amplification bandwidth (25 nm) [17, 18]. The broadband amplification and long fluorescent lifetime have potential applications in ultrashort-pulses generation and high-power laser. The optical gain can be improved by increase the pump power. At present, we have got twofold gain in bulk GC when the pump power is 7 W.
In summary, we demonstrated broadband optical amplification near the 1300 nm region for silicate GC containing β-Ga2O3:Ni2+ nanocrystals. The amplification bandwidth was 77 nm and it can be expected to be broader. The optical gain shows similar wavelength dependence to luminescence properties. The available spectroscopic data such as long fluorescent lifetime and broadband amplification shows the present GC system has potential applications in broadband optical fiber amplifiers and tunable lasers.
The authors would like to acknowledge the financial support provided by the National Natural Science Foundation of China (Grant No. 50672087) and National Basic Research Program of China (2006CB806000b).
References and links
1. L. R. Pinckney and G. H. Beall, “Transition element-doped crystals in glass,” Proc. SPIE 4452, 93–99 (2001). [CrossRef]
2. S. Kück, “Laser-related spectroscopy of ion-doped crystals for tunable solid-state lasers,” Appl. Phys. B 72, 515–562 (2001). [CrossRef]
3. T. Suzuki, K Horibuchi, and Y. Ohishi, “Structural and optical properties of ZnO-Al2O3-SiO2 system glass-ceramic containing Ni2+-doped nanocrystals,” J. Non-Crys. Solids 351, 2304–2309 (2005). [CrossRef]
4. T. Suzuki, G. S. Murugan, and Y. Ohishi, “Optical properties of transparent Li2O-Ga2O3-SiO2 glass-ceramics embedding Ni-doped nanocrystals,” Appl. Phys. Lett. 86, 131903 (2005). [CrossRef]
5. Y. Wang and J. Ohwaki, “New transparent vitroceramics codoped with Er3+ and Yb3+ for efficient frequency upconversion,” Appl. Phys. Lett. 63, 3268–3270 (1993). [CrossRef]
6. S. F. Zhouet al, J. Phys. Chem. C, accepted.
7. L. Galoisy and G. Calas, “Structural environment of nickel in silicate glass/melt systems: Part 1. Spectroscopic determination of coordination states,” Geochim. Cosmochim. Acta 57, 3613–3626 (1993). [CrossRef]
8. N. Jianget al, submitted to Appl. Phys. Lett. [PubMed]
9. M. V. Iverson, J. C. Windscheif, and W. A. Sibley, “Optical parameters for the MgO:Ni2+ laser system,” Appl. Phys. Lett. 36, 183–184 (1980). [CrossRef]
10. Y. Suzuki, W. A. Sibley, O. H. El Bayoumi, T. M. Roberts, and B. Bendow, “Optical properties of transition-metal ions in zirconium-based metal fluoride glasses and MgF2 crystals,” Phys. Rev. B 35, 4472–4482 (1987). [CrossRef]
11. C. Anino, J. Théry, and D. Vivien, “Cr4+ doped Li2MgSiO4, a new potential tunable laser material with room temperature fluorescence lifetime>100 μs,” Proc. SPIE 3176, 38–41 (1996). [CrossRef]
12. H. Shigemura, M. Shojiya, R. Kanno, Y. Kawamoto, K. Kadono, and M. Takahashi, “Optical property and local environment of Ni2+ in fluoride glasses,” J. Phys. Chem. B 102, 1920–1925 (1998). [CrossRef]
13. S. F. Zhouet al, to be submitted
14. N. F. Mott and R. W. Gurney, Electronic Processes in Ionic Crystals; (Oxford, 1948).
15. M. Hughes, H. Rutt, D. Hewak, and R. J. Curry, “Spectroscopy of vanadium (III) doped gallium lanthanum sulphide chalcogenide glass,” Appl. Phys. Lett. 90, 031108 (2007). [CrossRef]
16. B. N. Samson, L. R. Pinckney, J. Wang, G. H. Beall, and N. F. Borrelli, “Nickel-doped nanocrystalline glass-ceramic fiber,” Opt. Lett. 27, 1309–1311 (2002). [CrossRef]
17. Y. Miyajima, T. Sugawa, and Y. Fukasaku, “38.2 dB amplification at 1.31 μm and possibility of 0.98 μm pumping in Pr3+-doped fluoride fibre,” Electron. Lett. 27, 1706–1707 (1991). [CrossRef]
18. T. J. Whitley, “A review of recent system demonstrations incorporating 1.3 μm praseodymium-doped fluoride fiber amplifiers,” J. Lightw. Technol. 13, 744–760 (1995). [CrossRef]