The suitability for optical amplification in the C + L band of a new zirconia-yttria–alumina-baria silica glass fiber is evaluated. The gain and noise figure are characterized using this fiber as the gain medium. A flat gain of 25 dB with a variation of less than 3 dB in the range of 1525 to 1565 nm with a significantly low noise figure less than 4.2 dB at small signal input power, are achieved using a short length of fiber with only 1 m. The performance of the amplifier can be improved with higher pump powers and longer fiber lengths.
© 2019 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
Erbium-doped fiber amplifiers (EDFAs) are used in most current communication systems [1,2]. The spreading in the use of EDFAs is due to their intriguing characteristics such as high gain [1–5], high saturation output power, polarization independent gain , crosstalk absence , low noise figure , and low insertion loss. Moreover, EDFAs exhibit a large gain bandwidth, which enable their use to amplify several channels simultaneously, as required for dense wavelength division multiplexing (WDM). The same feature could allow their use also in the L-band of the fiber transmission window. However, the non-uniform gain spectrum of EDFAs causes problems when multiple EDFAs are cascaded in the system. This feature could limit the usable bandwidth of EDFAs and, hence, the amount of data transmission by the system. Actually, some approaches have been used to mitigate this limitations, namely by using gain equalizing ﬁlters, long-period ﬁber grating, chirped ﬁber Bragg grating, and so on [6–8]. Obviously, these schemes increase the system overall cost and complexity.
Extensive research has been done on EDFAs using various glass host and co-dopant materials such as silica, ﬂuorozirconate, chalcogenides, bismuth, aluminum, phosphorous, lanthanum, etc. [9–11]. Comparison shows that these materials demonstrate different qualities that have a signiﬁcant impact on the overall performance and applicability of an optical ampliﬁer. Some materials have wider emission bandwidth that is suitable for WDM systems. Others, allow higher erbium concentrations before detrimental effects such as concentration quenching  and cluster formation  occur, which can translate to equal gain in a shorter fiber length that results in a more compact device.
In this work, we use nano-engineered Er3+-doped zirconia-yttria-alumino silicate glass with increasing the doping levels of Er2O3 along with ZrO2 in presence of small amount of BaO as the fiber host. The introduction of more Zr4+ ions into yttria-alumino-silicate glass host avoids the cluster formation of Er3+ in silica and consequently exacerbates the luminescence . Moreover, replacing the intermediate Al2O3 by a modifier ZrO2, the number of non-bridging oxygen is expected to increase which makes the silica network structure more open. As a result of it, in this host, Er3+ shows wider emission spectra as compared to silica-EDF, especially at the longer wavelengths above 1580 nm because of its larger emission cross-section . The motivation for the incorporation of BaO is not only to modify the local field environment of the Er ion [16,17] in order to reduce the clustering effect of rare-earths just like to Al and P but also to increase the glass forming zone of the doping host of Er2O3 in order to reduce the phase-separation of the doping host [18,19]. Thus, the addition of BaO serve as modifier lead to some structural modifications and local electrostatic field variations near the rare earth ion due to differences in their ionic radii in the glass matrix. So, the symmetry and (or) covalency of the glass at the rare earth ion will be different from that of silica glass. Such variations should have strong bearing on various luminescence transitions and as well on the upconversion, as shown by Courrola and Kassab et al . The inclusion of the Y2O3 into the host matrix serves the additional purpose of slowing down or eliminating changes in the ZrO2 crystal structure. This is a crucial factor in the fabrication process . By combining C- and L-band EDFAs, the main goal of this kind of novel host for Er3+ is to achieve a broad band optical amplification, a better flat-gain value together with a lower noise figure, using short lengths of erbium doped fiber.
2. Fiber fabrication and characterization
The erbium doped nano-engineered zirconia-yttria–alumina-baria (Zr-Y-Al-Ba) silica glass based fiber (ZYAB-EDF) was made using an optical preform fabricated by modified chemical vapour deposition (MCVD) process in combination with solution doping technique. The preform is then subjected to thermal annealing under suitable heating condition. The doping of Er2O3 into the zirconia yttria-alumina-baria-silica host was done via solution doping process in a porous silica layer that was deposited by MCVD process at ∼1550 °C, followed by presintering at ∼1500 °C. The porous layer deposited into inner surface of silica tube was soaked with alcoholic solution of a mixture of certain strength of ErCl3.6H2O, AlCl3.6H2O, YCl3.6H2O, ZrOCl2 8H2O and BaCl2 2H2O for a period of 1 hour. Then, the solution was drained out from the inner side of silica tube and dried by air to remove the solvent from the soaked layer. After that, the layer was sintered with gradually increasing the heating temperature of soaked porous layer deposited tube from 1500 °C to 1850 °C in order to convert it into a transparent layer. Finally, the tube was collapsed by heating at ∼2000°C, where it was converted into solid rod known as preform. A small length of the preform (around 10 cm) was thermally annealed at 1250 °C for 3 hours with heating and cooling rate of 10°C/min in a closed furnace that maintained an optimized heating cycle. The fiber was drawn from such annealed preform using fiber drawing tower with on-line resin coating after heating the preform at 2000 °C. Here the thermal annealing was a determinant step in the formation of the nano-engineered zirconia-yttrium alumina silica glass based optical fiber. The nano-crystalline host of ZrO2 is retained into the silica glass matrix at fiber drawing stage as confirmed from transmission electron microscope (TEM) analyses with EDX spectra and electron diffraction patterns as described below.
The sizes of nanophases changes from the annealed preform heated at 1250 °C for 3 hours to the drawn fiber. We have taken the TEM picture of both the annealed preform and the fiber drawn from the preform using the high resolution transmission electron microscope Model: Tecnai G2 30ST (FEI Company, USA). The TEM analyses of the annealed optical preform sample is shown in Fig. 1 where sizes of the nanoscale particles become around 7 to 8 nm. The high resolution TEM (HTEM) (Fig. 1(b)) along with electron diffraction pattern (Fig. 1(c)) in the inset shows the crystalline nature of the particles. When the fiber was drawn from such annealed preform, the sizes of the particles reduced to 3 to 4 nm as shown in Fig. 2 (a). The larger sized particles present within the core of annealed preform, having sizes from 7 to 8 nm, transform into very fine sized particles and tend to dissolve into the parent glassy phase with average size of around 3 to 4 nm and this particle size does not lead to significant scattering loss. A possible reason of decrease in particle size is re-melting and stretching of the separated phase during the fiber drawing [21,22].
The shapes of the particles are very irregular circular structure when drawn the fiber from the annealed preform. At high temperature, during drawing the crystalline nature of the particles is destroyed partially as confirmed from the electron diffraction pattern shown in Fig. 2(c). The presence of surrounding crystalline environment of Er ions will influence the erbium spectroscopy and the EDFA performance through the increase of the broadening of emission spectra along with more flatness of the optical gain spectra.
The core appears homogeneous at optical wavelengths, but not at a nanoscale level. The TEM picture of optical fiber sample displayed in Fig. 2(a) show that sizes of the nanoscale particles become around 3 to 4 nm. The HTEM (Fig. 2(b)) along with electron diffraction pattern in the inset (Fig. 2(c)) shows the partial crystalline nature of the particles.
We have made the optical preform with and without BaO keeping all the others fabrication parameters, including the solution strength of the precursors of ZrO2, Y2O3, Al2O3 and Er2O3. The core glass becomes slightly opaque in nature in absence of BaO, which mainly arises due to formation of phase-separated region as shown in the FESEM (field emission scanning electron microscopy) picture of the optical fiber preform sample (Fig. 3 (a)). This phase-separated region will induce high background loss. On the other hand, the addition of a minor amount of BaO in erbium doped zirconia-yttria–alumina silica glass based fiber preform does not form any phase-separated zones as shown in Fig. 3 (b).
The core and cladding geometry of the ﬁber was inspected using an optical microscope (Olympus BX51). The core was homogeneous and had no observable defects at the interface between the core and the silica cladding. The fiber cross-section is shown in Fig. 4 (a), and the core and cladding diameters of the fiber were measured to be 10.84 and 125.75 µm, respectively.
The content of Al, Zr, Y, Ba and Er within the core region of fiber was measured by electron-probe micro-analysis (EPMA) using a JEOL EPMA instrument. The fiber core glass contain 9.0 wt% Al2O3, 4.5 wt% ZrO2 , 0.225 wt% Y2O3 , 0.10 wt% BaO, and 1.5 wt% Er2O3. The elemental distribution curve of different dopants along the diameter of the core of fiber is presented in Fig. 4 (b). The high erbium ion concentration (1.5 wt%) was possible due to the co-doping with Zr and Al ions, which reduces ion clustering effects. Here a minor amount of BaO (0.10 wt%) was added to increase the glass forming zone of the doping host of Er2O3. The effect of BaO on the glass network as modifiers results from the weakening or breaking of Si-O and Al-O bonds and the formation of non-bridging oxygen (NBO) . Generally, BaO tends to reinforce the three-dimensional nature of the network and make it more dense, due to the large size of the Ba ion, leading to a higher glass-transition temperature (Tg) .
The attenuation spectrum, displayed in Fig. 5, was measured in a 1 m of fiber spliced to single mode pigtails with FC-APC connectors, using a broadband source with a spectral range from 410 to 2400 nm (model SuperK EXTREME from Fianium) and an optical spectrum analyzer (OSA), model Q8384 from Advantest. The attenuation spectrum present two main absorptions bands, one around 980 nm with a maximum value of 21.5 dB and a broad band, roughly flat, between 1450-1570 nm with a maximum value of 24.7 dB. Due to the instability of the broadband source, the background loss is very noisy, thus we can’t determine it accurately. However, this attenuation spectrum shows that the background loss is lower than 1 dB/m.
The fiber parameters are summarized in Table 1.
The emission spectrum of the ZYAB-EDF fiber was measured using a laser diode of 980 nm wavelength as excitation source for two different powers: 100 and 398 mW. The light emitted by the source was injected into the fiber which was directly connected with an OSA (model Q8384 from Advantest). The results for different fiber lengths are shown in Fig. 6. We can see that the fiber shows a broad emission spectrum with a strong emission around 1550 nm. The maximum emission intensity is obtained for a fiber length of 1 m. The emission spectrum became flatter with the increase of power which indicate that this fiber will produce more uniform gain at C-band. With the 2 and 4 m fiber, we can see a deviation of the maximum intensity peak towards the L-band region, which suggests that with longer fibers and higher pump powers values it may be possible to produce gain in the L-band.
For comparison propose, we measured the emission spectrum of a commercial EDF with very high absorption and optimized for short length L-band EDFAs (I-25 from Fibercore), the results are displayed in Fig. 7. We can see that the emission performance of the fiber is similar to the ZYAB-EDF, however, the maximum emission intensity and a flat gain is achieved with a length of 2 m, twice the length needed with the ZYAB-EDF. Furthermore, 4 m of I-25 fiber still doesn’t present emission in the L-band. Therefore, we need shorter ZYAB-EDF lengths to achieve similar results to those of I-25 fiber.
3. Optical amplification
To analyze the optical amplification performance of the proposed fiber we measure the gain and noise figure for amplifiers using short pieces of the ZYAB-EDF as the gain medium. Since the emission was maximum for a length of 1 m, we choose to test two different lengths of fiber 1 m and 1.5 m. The experimental setup used is displayed in Fig. 8, in which the amplifier is forward pumped by a 980 nm laser diode via a 980/1550 nm WDM coupler. A tunable laser (model OSICS 1560 from Anritsu), with a maximum output power of 13 dBm and a wavelength range from 1500 nm to 1620 nm, was used as the input signal. The power of the input signal was adjusted by a variable optical attenuator (VOA) and controlled using a 99-1% coupler, placed between the laser output and the WDM coupler, and an optical power meter (OPM), model AXS-100 from EXFO. The gain and noise figure where characterized using an OSA (model FTB-500 from EXFO).
First, we analyze the gain variation with the pumping power for the two lengths of fiber using an input signal with -25 dBm of power (small signal) and wavelengths of 1550 and 1590 nm. The results are displayed in Fig. 9. As expected, the gain initially increases exponentially with the pump power, however the gain limit is not achieved, which means that the near-complete inversion regime is not achieved with the used pump power values. Therefore, it is possible to improve the performance of this amplifier with higher pumping powers values. Nevertheless, for the signal of 1550 nm a gain higher than 27 dB was achieved for the 1 m fiber and for the signal of 1590 nm the fiber with 1.5 m of lengths show better performance achieving/providing a gain of 13 dB.
The gain and noise performance were characterized in the C-band and L-band for small-signal. We used 20 signals in the range from 1515 to 1610 nm with -25 dBm of power and a fixed pump power of 100 mW. The results are shown in Fig. 10, where we can see that the maximum gain of 25.6 dB is obtained for the 1 m fiber. It is also evident that the gain is higher in the C-band and have a flat band of high amplification that goes from 1525 to 1565 nm, with a variation less than 3 dB. In the L-band, the gain is slightly higher for the 1.5 m fiber, which suggests that the gain performance of the L-band can be improved with a longer fiber and higher pump power. From Fig. 10, we can also observe that the noise has good performance in the two lengths of fiber, however, the 1 m fiber shows best results for the noise figure with values lower that 4.2 dB.
The performance characteristics in terms of signal amplification and noise of a new zirconia-yttria–alumina-baria (Zr-Y-Al-Ba) silica glass based fiber were measured. Using only 1 m of YZAB-EDF as gain medium, a flat gain band of 25 dB was achieved with variation of less than 3 dB and a width of 40 nm within the C band, for the small-signal gain region and a pump power of 100 mW. A low noise figure less than 4.2 dB was achieved in these conditions. In addition, the results suggest that the performance of the amplifier can be improved with higher pump powers and the gain in band L can be enhanced by using longer lengths ZYAB-EDFs. These results demonstrate that the ZYAB-EDF is suitable for optical amplification in C + L band with cost-effectiveness, since it just require a very short lengths of fiber.
Fundação para a Ciência e a Tecnologia (FCT) (POCI-01-0145-FEDER-007688, UID/CTM/50025/2013, UID/EEA/50008/2019).
This research was funded by FEDER funds through the COMPETE 2020 Programme and National Funds through Fundação para a Ciência e a Tecnologia (FCT) under the projects UID/CTM/50025/2013, POCI-01-0145-FEDER-007688, UID/EEA/50008/2019 and Ana M. Rocha contract program 1337. The funding from the project: Development of ultra-broadband (1100-2200 nm) light sources based on modified nano-engineered silica glass optical fibers doped with bismuth and multiple rare-earths toward OCT applications from the Indo-Portuguese Programme for Cooperation in Science and Technology, is also acknowledged.
1. G. J. Foschini and I. M. I. Habbab, “Capacity of broadcast channels in the near-future CATV architecture,” J. Lightwave Technol. 13(3), 507–516 (1995). [CrossRef]
2. T. Otani, K. Goto, T. Kawazawa, H. Abe, and M. Tanaka, “Effect of span loss increase on the optically amplified communication system,” J. Lightwave Technol. 15(5), 737–742 (1997). [CrossRef]
3. C. R. Giles, E. Desurvire, J. R. Talman, J. R. Simpson, and P. C. Becker, “2-Gbit/s signal amplification at lambda = 1.53 mu m in an erbium-doped single-mode fiber amplifier,” J. Lightwave Technol. 7(4), 651–656 (1989). [CrossRef]
4. E. Desurvire, C. R. Giles, and J. R. Simpson, “Gain saturation effects in high-speed, multichannel erbium-doped fiber amplifiers at lambda = 1.53 µm,” J. Lightwave Technol. 7(12), 2095–2104 (1989). [CrossRef]
5. O. Lumholt, J. H. Povlsen, K. Schusler, A. Bjarklev, S. Dahl-Petersen, T. Rasmussen, and K. Rottwitt, “Quantum limited noise figure operation of high gain erbium doped fiber amplifiers,” J. Lightwave Technol. 11(8), 1344–1352 (1993). [CrossRef]
6. M. Pal, S. Bandyopadhyay, P. Biswas, R. Debroy, M. C. Paul, R. Sen, K. Dasgupta, and S. K. Bhadra, “Study of gain flatness for multi-channel amplification in single stage EDFA for WDM applications,” Opt. Quantum Electron. 39(14), 1231–1243 (2007). [CrossRef]
7. A. M. Vengsarkar, N. S. Bergano, C. R. Davidson, J. R. Pedrazzani, J. B. Judkins, and P. J. Lemaire, “Long-period fiber-grating-based gain equalizers,” Opt. Lett. 21(5), 336 (1996). [CrossRef]
8. H. S. Kim, S. H. Yun, H. K. Kim, N. Park, and B. Y. Kim, “Actively gain-flattened erbium-doped fiber amplifier over 35 nm by using all-fiber acoustooptic tunable filters,” IEEE Photonics Technol. Lett. 10(6), 790–792 (1998). [CrossRef]
9. B. Wang, G. Pub, R. Osnato, and B. Palsdottir, “Characterization of gain spectral variation of erbium-doped fibers codoped with aluminum,” Proc. SPIE 5280, 161 (2004). [CrossRef]
10. M. Yamada, T. Kanamori, Y. Terunuma, K. Oikawa, M. Shimizu, S. Sudo, and K. Sagawa, “Fluoride-based erbium-doped fiber amplifier with inherently flat gain spectrum,” IEEE Photonics Technol. Lett. 8(7), 882–884 (1996). [CrossRef]
11. A. Jha, S. Shen, and M. Naftaly, “Structural origin of spectral broadening of 1.5-µm emission in Er3+-doped tellurite glasses,” Phys. Rev. B 62(10), 6215–6227 (2000). [CrossRef]
12. E. Snoeks, P. G. Kik, and A. Polman, “Concentration quenching in erbium implanted alkali silicate glasses,” Opt. Mater. (Amsterdam, Neth.) 5(3), 159–167 (1996). [CrossRef]
13. D. M. Gill, L. McCaughan, and J. C. Wright, “Spectroscopic site determinations in erbium-doped lithium niobate,” Phys. Rev. B 53(5), 2334–2344 (1996). [CrossRef]
14. V. C. Costa, M. J. Lochhead, and K. L. Bray, “Fluorescence Line-Narrowing Study of Eu 3+ -Doped Sol−Gel Silica: Effect of Modifying Cations on the Clustering of Eu3+,” Chem. Mater. 8(3), 783–790 (1996). [CrossRef]
15. M. C. Paul, S. W. Harun, N. A. D. Huri, A. Hamzah, S. Das, M. Pal, S. K. Bhadra, H. Ahmad, S. Yoo, M. P. Kalita, A. J. Boyland, and J. K. Sahu, “Performance comparison of Zr-based and Bi-based erbium-doped fiber amplifiers,” Opt. Lett. 35(17), 2882 (2010). [CrossRef]
16. L. C. Courrol, L. R. P. Kassab, M. E. Fukumoto, N. U. Wetter, S. H. Tatumi, and N. I. Morimoto, “Spectroscopic properties of heavy metal oxide glasses doped with erbium,” J. Lumin. 102–103, 91–95 (2003). [CrossRef]
17. A. Dhar, M. C. Paul, S. Das Chowdhury, M. Pal, A. Pal, and R. Sen, “Fabrication and properties of rare-earth-doped optical fiber using barium as an alternate codopant,” Phys. Status Solidi A 213(11), 3039–3045 (2016). [CrossRef]
18. L. Yuliantini, M. Djamal, R. Hidayat, K. Boonin, and J. Kaewkhao, Spectroscopy Properties of Er 3+ Ion Doped ZnO-Al2O3-BaO-B2O3 Glass for Photonic Application (Wiley,2018), Vol. 5.
19. M. Pokhrel, G. A. Kumar, S. Balaji, R. Debnath, and D. K. Sardar, “Optical characterization of Er 3 þ and Yb 3 þ co-doped barium fluorotellurite glass,” J. Lumin. 132(8), 1910–1916 (2012). [CrossRef]
20. M. Pal, M. C. Paul, S. K. Bhadra, S. Das, S. Yoo, M. P. Kalita, A. J. Boyland, and J. K. Sahu, “Study of Multichannel Amplification in Erbium-Doped Zirconia-Yttria- Alumino-Silicate Fiber,” J. Lightwave Technol. 29(14), 2109–2115 (2011). [CrossRef]
21. M. C. Paul, S. Bysakh, S. Das, M. Pal, S. K. Bhadra, S. Yoo, A. J. Boyland, and J. K. Sahu, “Nano-Engineered Yb 2 O 3 Doped Optical Fiber: Fabrication, Material Characterizations, Spectroscopic Properties and Lasing Characteristics: A Review,” Sci. Adv. Mater. 4(2), 292–321 (2012). [CrossRef]
22. D. Dutta, A. Dhar, A. V. Kir’yanov, S. Das, S. Bysakh, and M. C. Paul, “Fabrication and characterization of chromium-doped nanophase separated yttria-alumina-silica glass-based optical fibers,” Phys. Status Solidi A 212, 1836–1844 (2015). [CrossRef]
23. S. X. Shen and A. Jha, “Raman Spectroscopic and DTA Studies of TeO2-ZnO-Na2O Tellurite Glasses,” Adv. Mater. Res. 39–40, 159–164 (2008). [CrossRef]
24. N. H. Ray, “Composition—property relationships in inorganic oxide glasses,” J. Non-Cryst. Solids 15(3), 423–434 (1974). [CrossRef]