Open Access
Add to BibSonomy
Abstract
Glasses doped with rare-earth elements represent a key-enabling amplification technology for fiber-based systems and integrated devices. Even if silica has been the preferred glass matrix for optical amplifiers, some other glasses have been proposed as an alternative. Phosphates, in particular, present a lower glass transition temperature and have a structure that envisages a capacity to host erbium and ytterbium dopants. In this paper, we characterize the gain of zinc and sodium zinc phosphate doped either only with erbium or with erbium and ytterbium. The experimental results reveal that, among the different considered configurations, the sodium zinc phosphate with erbium-ytterbium co-doping leads to an on-off gain coefficient as large as 40 dB/cm.
© 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Rare earth-doped glasses are essential in many optical applications, particularly as a gain medium in fiber lasers and optical amplifiers [1,2]. Indeed, doped-fiber amplifiers are widely claimed as the main key-enabling technology that paved the way for the deployment of multi-span metropolitan and intercontinental links that resulted in the blossoming of the gold-era of optical communications experienced in the 2000s [3]. Amongst the different rare earth elements, Erbium is especially interesting because its 4I13/2-4I15/2 transition results in strong fluorescence in the 1500 nm to 1600 nm. This band corresponds to the wavelength range at which the standard single-mode fiber presents minimum attenuation and, consequently, it is the preferred transmission window in most optical communication systems [4]. One of the challenges of doping with erbium is the limited concentration that a certain glass matrix can accept. This issue was partially overcome by adding ytterbium as a co-doping material, which in addition to preventing the formation of ion clusters, increases the capture/emission cross-sections and extends the gain bandwidth [5,6].
Erbium and ytterbium ions require a glass matrix that strongly impacts the way in which they interact with light. The glass traditionally used to host the Er3+ and Yb3+ ions is fused silica. However, some other host materials have been explored as an alternative to silicates with the aim of increasing the solubility of the doping agents and enhancing the optical gain [7]. For instance, in [8,9] borate glasses are analyzed, whereas in [10] and [11], fluorides [12] and tellurite [13] glasses are studied. More recently, phosphate glasses have emerged as a promising material because they present high transparency in the infrared region, they have a slightly lower fusion temperature than silicates (1100$^\circ$C for phosphates versus 1200$^\circ$C for silicates) and they show high thermal stability, as well as low refractive index and dispersion coefficient [14,15].
Furthermore, the presence of phosphorus atoms in the phosphates introduces oxygen bridges that confer a chain structure, thus increasing saturation rare-earth doping concentrations up to 1021 ions/cm-3 [16]. Additionally, the adoption of a phosphate glass matrix enlarges the emission cross-section of Er3+ and Yb3+ and, consequently, leads to an improved gain coefficient [17]. Due to these advantageous characteristics, phosphate glasses have been successfully employed to build high-gain compact optical amplifiers [18].
The research on novel phosphate matrices is still an open topic. Among the different phosphate glasses, zinc phosphate, Zn3(PO4)2, is specially attractive because of its relatively simple synthesis [19,20]. In addition, zinc acts as a network modifier that increases ion solubility, improves chemical durability, and reduces the glass transition temperature [15]. Another interesting phosphate is the sodium zinc phosphate, NaZn(PO3)3, which has been employed as a matrix in screens and white light emitters [21]. The presence of sodium further enhances the ion solubility that enables even higher doping ion concentrations [19]. Both zinc and sodium-zinc phosphates show high chemical and temperature stability, as well as good environmental resistance [22]. In addition, they have a relatively low glass transition temperature that allows the construction of fibers. Nevertheless, the most notorious feature of these phosphate matrices is their potential as a host matrix for improving the optical gain due to the aforementioned elevated ion solubility and the large emission cross-section. This elevated gain is an extremely desirable property as it enables compact devices and increases energy efficiency, which ultimately results in an overall power consumption reduction and lower environmental impacts during the use stage. Therefore, Er3+-Yb3+ co-doped zinc and sodium-zinc phosphates may find applications not only in fiber-based devices but also for integrated photonics, which is subject to tight footprint constraints. Er3+-Yb3+ co-doped zinc phosphates have been studied from the point of view of their spectroscopic properties, whereas in [23], rate-equations-based simulations are reported. Therefore, experimental characterization of optical gain in the 1550 nm band is still missing.
In this paper, we characterize the optical gain at 1550 nm of Er3+-Yb3+ co-doped zinc phosphate and sodium-zinc phosphate. Experimental results reveal that sodium-zinc phosphate matrix glass outperforms zinc phosphate in terms of gain, showing that this material may be employed as an efficient gain medium. The rest of the document is organized as follows: in Section 2, we describe the experiment, including material synthesis and the employed characterization setup. The experimental results are reported in Section 3 and, finally, in Section 4, the main conclusions are drawn.
2. Experiment
2.1 Material synthesis
For the synthesis of zinc phosphate, the following chemical reaction was implemented:
In both cases, the reagents in powder form were first weighed and mixed with 10 ml of distilled water. The resultant mixture was dried in an oven at 100$^{\circ }$C for 24 hours, before being transferred to a crucible and heated up to 600$^{\circ }$C in a muffle furnace for 3 hours and left to cool. The acquired powder was weighed to ensure that 10 g of phosphate was obtained. After weighing the amount of dopant, the erbium and ytterbium oxides are subsequently added to the phosphate powder, which is then mixed with 15 ml of distilled water and stirred in a beaker. The mix was kept at 80$^{\circ }$C for 24 hours, deposited in a crucible, and heated for 5 hours in a muffle furnace set at 1100$^{\circ }$C. The stoichiometric quantities of P2O5 for the zinc phosphate and sodium zinc phosphate were 39.04% and 58.00%, respectively. Thus, by keeping the percentage of this compound below 60%, we avoid the presence of P-O-P bonds, which can be degraded via moisture absorption [24,25].
In total six samples were fabricated. Half of them, identified as Z1 to Z3, were built of Zn3(PO4)2, whereas the other three, Z4 to Z6, were composed of NaZn(PO3)3. Regarding the doping agent and concentrations, one of the samples of Zn3(PO4)2, Z1, and another sample of NaZn(PO3)3, Z4, were left undoped for reference. The other samples were doped either with erbium or with a combination of erbium and ytterbium according to the percentages listed in Table 1. The fabricated samples had semi-spherical shapes with a diameter of 15 mm and a thickness at the widest point of 4 mm.
Table 1. Produced samples including the sample identifier, the glass matrix, the doping agents, and their concentrations.
In order to confirm that the fabricated samples are not crystalline but amorphous, in Fig. 1, we show the X-ray diffractogram of sample Z6.
2.2 Setup for optical gain characterization
The experimental setup employed to characterize the optical gain coefficient is shown in Fig. 2(a). A continuous wave laser operating at 980 nm with a variable output was used as a pump source. Another continuous wave laser, in this case emitting at a nominal wavelength of 1550 nm, was used as a signal source. The pump and signal beams were combined in a fiber-based wavelength division multiplexing (WDM) multiplexer, whose output fiber was finished with a Physical Contact (PC)-type connector. The combined beam diverges and propagates through the sample under test. After passing through the sample, part of the light is captured in an optical spectrum analyzer (OSA), mod. Anritsu MS9740A, which allows the discrimination of the pump and signal.
Fig. 2. Experimental setup employed to characterize the gain coefficient of the manufactured Er-Yb co-doped Zn and Na-Zn phosphates.
2.2.1 Measurement of the gain coefficient
In order to quantify the gain coefficient, we calculated the gain by dividing the measured signal power (at 1550 nm) when the beams traverse a doped sample, $P|_\text {doped}$, and an undoped sample with the same matrix material, $P|_\text {undoped}$. The gain then is calculated as follows:
When we divide the power at the output of a doped sample by that at the output of an undoped sample, we do not only eliminate the effect of the beam divergence but also the impact of reflection losses present in the sample surfaces. However, this gain still depends on the sample thickness. In order to avoid this dependency, we normalized the computed gain by the sample thickness, $L$, thus obtaining the gain coefficient:
2.2.2 Measurement of the on-off gain coefficient
The measurement of the gain coefficient using an undoped sample as a reference is valid only if the refractive index of the doped and undoped samples, as well as their widths, are equal. If these conditions are not fulfilled, Eq. (1) may not be accurate. An alternative metric is the so-called on-off gain, $G_\text {on-off}$, which is calculated using only the doped sample considering the ratio of the power of the signal at the output of the sample when the pump is on, $P^\text {on}$, and off $P^\text {off}$. Thus, in mathematical terms, the on-off gain is given by:
An additional advantage of the on-off gain is the fact that it considers the losses at the signal band of the dopants when there is no pump power, which is important in the design of amplifiers. As in the case of the net gain, the on-off gain can be normalized to get the on-off gain coefficient:
3. Results and discussion
In Fig. 3 we show the captured spectra for a particular sample, i.e. the sample Z6, which is composed of a NaZn(PO$_3$)$_3$ matrix co-doped with erbium and ytterbium. The traces correspond to the spectra for three cases: (i) the pump is on and the signal is off, (ii) the pump is off and the signal is on, and (iii) both the signal and the pump are on. Looking at the spectra in Fig. 3, the pump and signal can be clearly identified at 980 nm and 1550 nm, respectively. The differences between the three cases, however, cannot be clearly observed due to the large frequency span. In order to discriminate the different spectra at the pump and signal wavelengths, we included two insets. Looking at the signal wavelength, if we compare case (ii) and case (iii), that is the signal power when the pump is off and when it is on, we can observe that the signal power increases when the pump is switched on. This power increase, indeed, represents the on-off gain. The signal gain is accompanied by a small pump depletion, which can be perceived when we compare case (i) and case (iii) at the pump frequency. Alongside the spectra for the doped samples, the output spectra of the undoped fiber were also measured. The power at the signal frequency when the pump is off and when the undoped samples were placed were used as a reference to calculate the on-off gain and the gain coefficients according to Eqs. (3,4) and Eqs. (1,2), respectively. In Table 2, we list the measured power levels at the signal wavelength, around 1550 nm, for the two undoped samples and the doped samples when the pump is off. If we compare the output power levels of undoped samples and doped samples for each glass matrix, it is possible to perceive the latter presents a larger transmission loss, which was expected since the absorption of erbium and ytterbium represents an additional source of loss.
Fig. 3. Captured spectra when the pump is on and the signal is off (blue), when the pump is off and the signal is on (red), and when both the pump and signal are on (black). The insets show the spectra around the pump wavelength (980 nm) and the signal wavelength (1550 nm), revealing some pump depletion and signal on-off gain. The employed wavelength resolution was 0.7 nm.
Table 2. Measured power at 1550 nm for the undoped samples and for the doped samples when the pump is off.
Once the signal power (at 1550 nm) was measured for the undoped samples and for the doped samples in absence of a pump, the pump was switched on and the power levels at the signal wavelength for different pump power levels were measured. The measured power levels for zinc and sodium-zinc matrices considering Er and Er-Yb doping are shown in Fig. 4(a) and Fig. 4(b), respectively. In both cases, the measured power is significantly higher for the samples with sodium-zinc phosphate matrix than for those with zinc phosphate samples. When confronting the measured signal power levels at the output of samples with erbium and erbium-ytterbium doping, it is possible to see that co-doping significantly increases the signal power for the sodium think phosphate matrix, whereas for the zinc matrix, a more moderated increase is experimented. In addition, it is possible to observe a power saturation effect in the case of the co-doping of the zinc phosphate sample.
Fig. 4. Measured signal power levels in terms of the pump power for (a) erbium-doped zinc and sodium zinc phosphate matrix samples (b) erbium ytterbium-doped zinc and sodium zinc phosphate matrix samples. Gain and on-off gain for (c) erbium-doped zinc and sodium zinc phosphate matrix samples and (b) erbium ytterbium-doped zinc and sodium zinc phosphate matrix samples. In (c) and (d) we also included the gain coefficients for the different samples.
As mentioned in Subsection 2.2, we consider two different gains: on the one hand, we calculate a gain referred to the undoped sample and, on the other hand, the on-off gain, which normalized the signal power with respect to the same sample when the pump power is switched off. The gains for the samples with erbium and erbium-ytterbium doping are shown in Fig. 4(c) and Fig. 4(d), respectively. In both figures the gains and the on-off gains for zinc and sodium-zinc phosphate matrices are presented. Analyzing the case of erbium-doped samples it can be observed that the on-off gain is higher than the net gain for the whole analyzed pump power range. Comparing the results for different glass matrices, the on-off gain is higher for the sodium-zinc phosphate than for the zinc phosphate. However, the gain of the sodium zinc phosphate is higher than that presented by the zinc phosphate only for pump power levels exceeding -40 dBm. Regarding the samples with erbium-ytterbium coding, the gain curves present a different behavior. The first noticeable difference between the gain curves of the zinc and sodium zinc matrices is the significant separation. As can be observed, sodium zinc phosphate presents much higher net and on-off gain values than zinc phosphate. Finally, when we set side-by-side the curves for the erbium-doped and erbium-ytterbium co-doped samples, it is possible to note that the gains and on-off gains of erbium-ytterbium co-doped samples are higher, especially for the sodium zinc phosphate matrix. Regarding the gain coefficients, since all the samples have a similar thickness, the curves follow the same tendencies as the gains. Nevertheless, it is worth noting the large on-off gain coefficient, particularly for the erbium-ytterbium co-doped sodium zinc phosphate.
Compared to other phosphate glasses reported in the literature, in [26], the authors reported a net gain of 4.1 dB at 1550 nm in a 10 mm-long phosphate waveguide. Considering that the losses due to absorption were measured to be 8.7 dB/cm, the on-off gain of the waveguide was 12.8 dB/cm. Therefore the gain measured for the sodium zinc phosphate matrix is sensibly larger, whereas the zinc phosphate matrix is slightly lower. More recently, a fiber with heavily doped core and silica cladding fiber with a 3.1 dB/cm net gain was reported in [27]. It is expected that the gain of doped material to be larger than this value. Regarding doped silica, in [28], a gain of 25 dB was achieved at 1538 nm for a 30 cm-long fiber. This results in an on-off gain of 0.83 dB/cm. However, as in the previous case, part of the mode overlaps with the undoped cladding, and therefore, the on-off gain of the doped material should be higher. Therefore, the proposed matrices present high potential as an alternative to silica.
4. Conclusions
In this paper, we characterized the gain and the gain coefficient of zinc and sodium zinc phosphate doped either with erbium only or co-doped with erbium and ytterbium. The experimental results with a pump at 980 nm and a signal at 1550 nm, show that the sodium zinc matrix outperforms the zinc matrix in terms of on-off gain. In addition, we found the gain of sodium zinc matrix with erbium doping saturates for moderate pump power levels. When co-doped with erbium and ytterbium, on the other hand, the saturation power increases, allowing on-off gain levels of up to 18 dB and gain coefficients exceeding 45 dB/cm. Such a high gain coefficient makes sodium zinc matrix with erbium and ytterbium co-doping a high-potential candidate for the implementation of compact integrated waveguide amplifiers and lasers.
Funding
Consejo Nacional de Ciencia y Tecnología (285600); Conselho Nacional de Desenvolvimento Científico e Tecnológico (313378/2021-5); Fundação de Amparo à Pesquisa do Estado de São Paulo (15/24517-8).
Disclosures
The authors declare no conflicts of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. P. D. Dragic, M. Cavillon, and J. Ballato, “Materials for optical fiber lasers: a review,” Appl. Phys. Rev. 5(4), 041301 (2018). [CrossRef]
2. M. J. Digonnet, Rare-Earth-Doped Fiber Lasers and Amplifiers, Revised and Expanded (CRC Press, 2001).
3. G. P. Agrawal, “Optical communication: its history and recent progress,” in Optics in our Time (Springer, 2016), pp. 177–199.
4. G. P. Agrawal, Fiber-Optic Communication Systems, vol. 222 (John Wiley & Sons, 2012).
5. Y.-H. Wang, C.-S. Ma, D.-L. Li, and D.-M. Zhang, “Formulized analytical technique for gain characteristics of phosphate glass Er3+/Yb3+ co-doped waveguide amplifiers,” Opt. Appl. 10(11), 116006 (2021). [CrossRef]
6. V. Gapontsev, S. Matitsin, A. Isineev, and V. Kravchenko, “Erbium glass lasers and their applications,” Opt. Laser Technol. 14(4), 189–196 (1982). [CrossRef]
7. K. Mokhtar, K. Mohamed, G. Lakhdar, B. Sébastien, and A. Hamza, “Electrical conductivity and dielectric properties of rare earth ions (Ce3+, Pr3+ and Eu3+) doped in zinc sodium phosphate glass,” J. Non-Cryst. Solids 567, 120933 (2021). [CrossRef]
8. N. A. Zaini, S. N. Mohamed, and Z. Mohamed, “Synthesis and optical properties of borate glass co-doped erbium/vanadium,” in AIP Conference Proceedings, vol. 2368 (AIP Publishing LLC, 2021), p. 030009.
9. M. Abdullah, K. Azman, H. Azhan, and W. Razali, “Optical characterization of erbium doped sodium borate glass,” in Advanced Materials Research, vol. 622 (Trans Tech Publ, 2013), pp. 191–194.
10. D. Ehrt, “Structure, properties and applications of borate glasses,” Glass technology 41(6), 182–185 (2000).
11. M. N. R. Jauhariyah and A. Marzuki, “Optical properties of erbium doped tellurite glasses,” in Materials Science Forum, vol. 864 (Trans Tech Publ, 2016), pp. 37–41.
12. D. Ehrt, “Phosphate and fluoride phosphate optical glasses–properties, structure and applications,” Phys. Chem. Glasses: Eur. J. Glass Sci. Technol., Part B 56(6), 217–234 (2015). [CrossRef]
13. X. Li and W. Zhang, “Temperature-dependent fluorescence characteristics of an ytterbium-sensitized erbium-doped tellurite glass,” Phys. B 403(18), 3286–3288 (2008). [CrossRef]
14. S. Hraiech, M. Ferid, Y. Guyot, and G. Boulon, “Structural and optical studies of Yb3+, Er3+ and Er3+/Yb3+ co-doped phosphate glasses,” J. Rare Earths 31(7), 685–693 (2013). [CrossRef]
15. N. G. Boetti, G. C. Scarpignato, J. Lousteau, D. Pugliese, L. Bastard, J.-E. Broquin, and D. Milanese, “High concentration Yb-Er co-doped phosphate glass for optical fiber amplification,” J. Opt. 17(6), 065705 (2015). [CrossRef]
16. A. A. Reddy, S. S. Babu, and G. V. Prakash, “Er3+-doped phosphate glasses with improved gain characteristics for broadband optical amplifiers,” Opt. Commun. 285(24), 5364–5367 (2012). [CrossRef]
17. A. Langar, C. Bouzidi, H. Elhouichet, and M. Férid, “Er–Yb codoped phosphate glasses with improved gain characteristics for an efficient 1.55 µm broadband optical amplifiers,” J. Lumin. 148, 249–255 (2014). [CrossRef]
18. X. Shan-Hui, Y. Zhong-Min, F. Zhou-Ming, Z. Qin-Yuan, J. Zhong-Hong, and X. Wen-Cheng, “Efficient fibre amplifiers based on a highly Er3+/Yb3+ codoped phosphate glass-fibre,” Chin. Phys. Lett. 26(4), 047806 (2009). [CrossRef]
19. U. Caldi no, A. Lira, A. Meza-Rocha, I. Camarillo, and R. Lozada-Morales, “Development of sodium-zinc phosphate glasses doped with Dy3+, Eu3+ and Dy3+/Eu3+ for yellow laser medium, reddish-orange and white phosphor applications,” J. Lumin. 194, 231–239 (2018). [CrossRef]
20. R. K. Brow, D. R. Tallant, S. T. Myers, and C. C. Phifer, “The short-range structure of zinc polyphosphate glass,” J. Non-Cryst. Solids 191(1-2), 45–55 (1995). [CrossRef]
21. W. Mao, M. Cai, W. Xie, P. Li, S. Xu, and J. Zhang, “Enhanced luminescence of manganese in singly doped white light zinc phosphate glass,” J. Am. Ceram. Soc. 103(3), 1864–1870 (2020). [CrossRef]
22. Z. Černošek, M. Chládková, and J. Holubová, “The influence of TiO2 on the structure and properties of sodium-zinc phosphate glasses,” J. Non-Cryst. Solids 531, 119866 (2020). [CrossRef]
23. F. L. F. Rivera, G. G. Pérez-Sánchez, M. A. Barron-Meza, J. R. Miranda, and D. Y. M. Velázquez, “Mathematical modelling of einstein’s rate equations for zinc phosphate glass with Er3+-Yb3+,” Open J. Appl. Sci. 11(09), 1038–1045 (2021). [CrossRef]
24. R. Zallen, The physics of amorphous solids (John Wiley & Sons, 2008).
25. G. K. Marasinghe, M. Karabulut, C. Ray, D. Day, M. G. Shumsky, W. B. Yelon, C. H. Booth, P. G. Allen, and D. K. Shuh, “Structural features of iron phosphate glasses,” J. Non-Cryst. Solids 222(1-2), 144–152 (1997). [CrossRef]
26. Y. Yan, A. d. Faber, H. De Waal, P. Kik, and A. Polman, “Erbium-doped phosphate glass waveguide on silicon with 4.1 dB/cm gain at 1.535 µm,” Appl. Phys. Lett. 71(20), 2922–2924 (1997). [CrossRef]
27. B. Denker, B. Galagan, V. Kamynin, A. Ponosova, K. Riumkin, S. Semjonov, S. Sverchkov, and V. Tsvetkov, “Gain characteristics of fibers with a heavily erbium-doped phosphate-based core and silica cladding,” J. Opt. Soc. Am. B 36(10), 2705–2711 (2019). [CrossRef]
28. A. Belov, E. Dianov, G. Devy’atykh, V. Khopin, A. Gur’yanov, and Y. B. Zverev, “Erbium-doped fibers based on cesium–silicate glasses,” Opt. Fiber Technol. 6(1), 61–67 (2000). [CrossRef]
