A thin film based erbium doped tellurium oxide (TeO2) waveguide amplifier producing gain from 1500nm to 1640nm when pumped at 980nm is demonstrated. At measured internal gains exceeding 14dB lasing due to end facet reflection set in producing the first tellurite waveguide laser. High gains were observed despite significant upconversion, whose impact appears to be mitigated to some extent by residual OH contamination. The device displayed no photosensitive effects from either the high pumping intensities used or the intracavity intensity at 1550nm.
© 2015 Optical Society of America
Erbium doped waveguide amplifiers (EDWAs) provide high gain from a very short device which is advantageous for short pulse amplification, gain sections for on chip loss compensation in complex integrated optics devices, or to enable on-chip CW or high repetition rate on chip mode locked lasers. They are, therefore, key components in integrated optics. EDWAs and EDWA lasers (EDWLs) have previously been realized by differing techniques in various materials , most successfully until now in Al2O3 [2, 3] or multicomponent phosphate glass [4–8]. Devices are now commercially available, and on-chip gains as high as 5.3 dB/cm  have been attained with 980nm pumping in optimized phosphate glass hosts, though with only a ~5nm FWHM gain bandwidth.
Tellurite glasses (based on Tellurium dioxide, TeO2) are well known to have a range of interesting properties for fiber and planar waveguide devices , especially so for optical amplification. Tellurites have a number of advantages as emission hosts for EDWAs and EDWLs over other materials because of their high refractive index (larger emission cross section and more compact devices), large emission bandwidth, low ion to ion cross relaxation, relative independence of the 1550nm Erbium lifetime on concentration, and high Erbium solubility as has been demonstrated in tellurite glass and fiber amplifiers [10–13]. It is clear that tellurite based devices have the potential to deliver higher ultimate gain per unit length and bandwidth than previous demonstrations in other materials . The low phonon energy of around 600-800cm−1 in tellurite also enables the potential use of transitions not possible in other materials commonly used in EDWAs, (for example with Pr3+ gain at 1.3μm).
One of the major reasons for the delay in attaining integrated versions of the high quality fiber based tellurite devices has been the difficulties in fabricating low loss waveguides in an integrated platform. Tellurite EDWAs have been realized in the last few years by femtosecond laser inscription [14, 15] and the discovery of reactive ion etching processes using argon, hydrogen and methane [16–18]. Using these methods tellurite EDWAs have been successfully fabricated in bulk glass plates [14, 15] and by reactive RF sputtering of thin films followed by plasma etching . Gains of up to 2.8dB/cm had been achieved with 1480nm pumping  but only 1.5dB/cm with 976nm pumping in femtosecond inscribed devices  despite the expected higher inversion attainable. Lasing action has not been observed in any tellurite planar waveguide devices. In this paper, we report a high gain amplifier pumped at 980nm and lasing action from both Fresnel reflections and external fiber Bragg gratings.
2. Experiments and results
The thin films required for waveguides were fabricated using reactive RF magnetron co-sputtering techniques using a system from AJA international. The 3” Tellurium and Erbium targets were operated simultaneously in an Oxygen/Argon atmosphere at 5mTorr with 6sccm Ar flow and 9sccm Oxygen flow. The Te and Er target were fed with RF powers of 150W and 75W respectively. A bilayer thin film was fabricated on 100mm diameter thermally oxidized silicon wafers (2μm oxide thickness). The bottom layer had a thickness of 1350nm, Er concentration of 2.2x1020ions/cm3 (1.0%at. Er/Te), and refractive index at 1550nm of 2.075, achieved by slightly reducing the oxygen content of the plasma to index match the following layer. The top layer, had a thickness of 530nm and was comprised of pure TeO2 with no Er yielding a refractive index of 2.075. Rib waveguides were fabricated using standard UV lithography and Reactive Ion Etching (RIE) using a Hydrogen/Methane/Argon gas mix  in the top layer of pure TeO2 using an etch depth of 400nm. The resulting device was clad with spin on UV curable polysiloxane polymer with refractive index of 1.509 at 1550nm. A 5cm long waveguide was prepared. The lifetime of the film measured from the edge  was 1.3ms. A nominal 2μm wide waveguide had an actual width of 1.7μm and was single moded with polarization dependent loss 0.3dB due to mode size variation. The calculated modal overlap of the TE0 and TM0 fundamental modes with the Er doped region at 980nm and 1550nm were 92% and 90%, respectively. Mode areas for TE0 at 980nm and 1550nm were 4.6μm2 and 5.3μm2, respectively; and for TM0, 4.0μm2 and 4.4μm2, respectively. Mode field full widths at 1/e2 intensity were 4.2 x 1.5μm (h x v) for TE, and 3.7 x 1.5μm (h x v) for TM at 1550nm and 4.0 x 1.4μm (h x v) for TE, and 3.5 x 1.4μm (h x v) for TM at 980nm.
Loss and absorption spectrum measurements of the waveguide were made with a pair of lens tipped fibers with 2.5μm beam waist specifically anti-reflection coating at 980 and 1550nm. Measurements were made using a supercontinuum source and optical spectrum analyser from 600nm to 1700nm. The source was attenuated to −30dBm/nm to reduce the pumping effects that can bleach the absorption spectrum. This arrangement was used for all loss and gain measurements. The total measured insertion loss including propagation, coupling, lens, connector losses, etc. is plotted in Fig. 1. The background loss fitting is also plotted, comprising both 1/λ4 and 1/λ2 terms indicating contributions from both sidewall scattering loss and Rayleigh scattering. The 1/λ4 contribution (Rayleigh scattering in the glass film) is only significant at short wavelengths close to the visible whereas the 1/λ2 contribution (representing sidewall roughness induced scattering) is significant across the whole spectrum.
The loss spectrum collected indicated a total connector to connector insertion loss of 8dB at 1400nm, comprising 2.5dB of loss due to the lens tipped fibers, 2.5dB of (calculated) mode overlap, 1.0dB of facet reflections and 2.0dB of background propagation loss for the 5cm device (~0.4dB/cm) due mainly to imperfect etching. It clearly shows that the total background loss from 1100nm to 1500nm was relatively flat at around 8dB. Below 1000nm, the loss increases as sidewall and/or Rayleigh scattering mechanisms become stronger. The peak Erbium absorption at 1532nm was 35dB while the peak absorption at 979nm was 15dB.
Gain measurements were performed using two fiber pigtailed 980nm pumps in the bidirectional configuration. Each pump produced up to 300mW of power at the output of the fiber pigtail connector. The pumps and signal were multiplexed and coupled to the chip using broadband WDMs and lens tipped fibers as shown in Fig. 2(a). The gain spectrum was measured at various input pump powers.
Figures 3(a) and 3(b) show the internal gain spectra, which was measured with respect to the insertion loss in the non-absorbing region at 1400nm in the spectrum. Intense green upconversion was also observed as will be further discussed. At high pump powers, positive internal gain was obtained from 1500nm to 1640nm which covers S, C and L band. The signal at 1535nm started to spike, indicating that the gain at this wavelength was strong enough to overcome the round trip losses. The peak single pass gain just before this spiking was just over14dB. Further increases in pump power resulted in lasing off the facets. The reflectivity of each end facet was estimated to be 8.5% (−10.7dB) using the Fabry-Perot loss measurement technique on a very short piece (2mm) of waveguide  albeit with different cleaves to the longer chip, which had a slightly angled cleave one end thereby reducing the reflectivity at that end. Since the signal gain beyond 171mW pump could not be measured directly due to the lasing threshold, and the gain curve in Fig. 3(c) is still increasing, it can be concluded that the maximum gain of the chip exceeded 15dB or 3dB/cm approaching an estimated 4dB/cm for the full pump power available. This is the highest gain per length in any Tellurite device according to the authors’ knowledge. Extrapolation of the gain vs pump power curve (Fig. 3(c)) leads to a saturated gain greater than 20dB if sufficient pump power is available.
The gain spectrum also indicates a significant signal excited state absorption (ESA) cross-section in the region from 1640nm to beyond 1700nm (limit of the optical spectrum analyzer). High signal gain, corresponding to a high 4I13/2 population, allows absorption of the light around 1650nm promoting electrons to the 4I9/2 state as a result of the absorption process. It is clear that the ESA is increasing with the strength of the pump and signal gain. Using the ESA calculation detailed in  the ESA cross section was calculated and is displayed together with ground sate absorption cross section on Fig. 3(d). The values of these cross sections are in good agreement with previously reported values for Tellurite fibers .
When the supercontinuum source was turned off, lasing occurred around 1532nm with random mode hopping as there was no clearly defined feedback peak. Due to the low reflectivities of the end facets, the threshold of the laser cavity was relatively high. The peak laser output power measured on the OSA at the final fiber connector was up to −12dBm at ~300mW coupled pump power, corresponding to >0.1mW exiting each chip facet.
To reduce the laser threshold, a gold coated cleaved fiber was butt-coupled to one end of the waveguide to provide an enhanced reflection. In this case, the gold coated mirror and one facet of the waveguide created a laser cavity. This increased the reflectivity of one end from around 10% to an estimated >80%. The exact increase in the reflectivity of the combination of the end facet and the coated mirror was difficult to determine because the mirror was not in complete contact with the end of the waveguide but at a distance due to the slightly angled cleaved face of the waveguide. The total round trip cavity loss of the new configuration was now 16dB as opposed to 28dB for the bare facets. The pump power level required for a round trip gain of 16dB was 100mW as shown on Fig. 3(c). In this configuration it was only possible to pump the laser from one end. The gold coated mirror however reflected/recycled any unabsorbed pump power resulting from the pump induced bleaching of the 980nm absorption, partly compensating for the loss of the second pump. These improvements allowed the laser to operate at lower pumping power than in the first case. The lower pump power required to reach the lasing threshold as a result of the increased reflectivity from the gold mirror resulted in a gain spectrum peaked around 1550nm (see Fig. 3(a)) and consequently a lasing signal at around 1555-1560nm was observed with no wavelength stabilization. Lasing occurred above 100mW of pump power and a maximum of −3dBm or 0.5mW was observed at the output connector after the pump WDM, corresponding to an estimated ~1mW of laser emission from the facet. The laser output was power stable over long periods but tended to mode hop randomly with thermal fluctuations in the alignment of the fibers. Upon stopping the lasing action after long periods and revaluating the waveguide losses, no changes were visible from photosensitive effects from either the pump, the circulating intracavity 1550nm power, or any of the upconversion products.
In order to produce a wavelength stabilized laser, a Bragg fiber grating was inserted between the WDM and the chip as shown in Fig. 2(b). One of the Bragg gratings used had center wavelength at 1550.27nm, FWHM of 0.24 with reflectivity of 95.2% and the other grating had center wavelength at 1535.06nm, FWHM of 0.216 and reflectivity of 94.3%. In both cases, laser action was achieved. Figure 4 was the result of using the 1550.39nm Bragg grating. The lasing occurred at 1550.27nm in a narrow line limited by the OSA resolution bandwidth of 1 GHz, with output power beyond the WDM being −17.73dBm. Allowing for the lens, coupling, and FBG transmission losses this equates to about 1mW exiting at the facet (note this does not account for the higher threshold caused by the ~6dB intracavity loss to the grating and back caused by the input coupling loss). Emission was again stable in time with no signs of photosensitive effects from either the 980nm pump, the circulating 1550nm power, or the upconversion products.
As noted above, parasitic process such as upconversion and pump excited state absorption were also clearly occurring. To compare the relative upconversion from 980nm vs 1480nm pumping, the visible emission was measured. A 60µm-core multimode fiber was used as a probe at a distance <100µm from the one end of the waveguide. The signal was sent to a high resolution spectrometer (Model HR2000 from Ocean Optics Inc.), spectra at maximum power for 980nm and 1480nm pumps being plotted in Fig. 5(a). Note that the integration time setting for the 980nm pump was 20x shorter than for the 1480nm case. The 546nm emission is therefore 20x stronger for 980nm pumping compared with 1480nm. The trend of the emissions against the power is similar with power trend of xn with n = 1.485 and 1.532 for 1480nm and 980nm, respectively.
The green emission comes from the transitions from the 4S3/2 and 2H11/2 levels to the ground level 4I15/2. With a 980nm pump, the 4I11/2 level is directly excited and the 4I13/2 1550nm transition band is filled by the decay from 4I11/2 via radiative and non-radiative transitions. However, the life time of the 4I11/2 level is usually quite long at around 200µs due to the low phonon energy of the tellurite host . A population therefore builds up at the 4I11/2 level leading to excited state absorption of an additional 980nm pump photon directly pumping the 4F7/2 level. Multiphonon mediated decay between the closely spaced 4F7/2 and 4S3/2/2H11/2 bands then populates these latter states that lead to green emissions (and indeed an additional weak blue emission not visible on these spectra).
With 1480nm pumping, the ions are directly excited to the 4I13/2 level. However, phonon assisted transitions, co-operative upconversion and excited state absorption processes due to high population of the 4I13/2 can populate the 4F7/2, 4S3/2, and 4H11/2 levels at high pump intensities. However at least three photons/ions are involved in such processes in total compared to two for the 980nm pump, giving rise to the lower efficiencies observed.
Another significant difference between pumping at 980nm and 1490nm is the relative intensity at 659nm compared with 546nm. With 980nm pumping, the ratio 659nm to 546nm emission is only around 8%, whereas it is ~70% for 1490nm pumping. The difference is caused by the fact that when 980nm pump is used, the green emissions are directly and efficiently pumped by two pump photons, therefore dominating the spectrum.
The power fitting lines to the 546nm emission in Fig. 5(a) and 5(b) show very similar dependence of the power of ~1.5 on the pump power for both 980nm and 1480nm pump. This shows that the higher level excitation process due to both 980nm and 1480nm are complicated and not the same as reported in other materials such as Al2O3 where the power dependence was reported to be 2.04 . Whilst the exact mechanisms at play here are not clear, it may be that in the high pumping regime where these measurements were made that similar effects to the “quasi ground state” lasing observed in low phonon energy fluoride hosts  are occurring. Further investigation is needed to determine the exact upconversion processes in this particular host.
Low threshold operation can be achieved via a number of routes. The laser thresholds demonstrated in these experiments are high due to large intra-cavity loss/low reflectivity facets. Incorporating suitable Bragg gratings directly on the waveguide (by etching for example) would leave only propagation loss (~2dB) to overcome, and the threshold would be reduced significantly. Further improvement could be achieved by reducing the loss to the levels shown in passive devices, <0.2dB/cm or 1dB for 5cm length with appropriate fabrication . In this situation, the estimated threshold would be below 50mW for the cavity of Fig. 2(b), and much lower for integrated waveguide gratings. Further improvements then must come from improved pumping efficiency via reduction of the 980nm pump ESA.
In order to achieve better pump efficiency, improvements in the lifetime and reduction in upconversion is required. The OH contamination known to exist in the current material reduces the lifetime significantly as shown previously . In addition, co-doping of Er3+ and Ce3+ can be utilized to achieve higher 980nm pump efficiency through shortening the 4I11/2 lifetime by resonant energy exchange [20, 23, 24]. Shen  showed that it is possible to achieve close to 100% inversion with 980nm pump in Er3+/Ce3+ doped tellurite fiber.
In order to obtain some estimation of the improvements possible, a commercial package from Optiwave System Inc. was used to model the Er-doped waveguide performance. The model is based on the solution of the propagation equations using, directly, the solutions of the involved electromagnetic fields and the exact transverse Er distribution. The modal and propagation equations are solved using the finite-element method and the Runge-Kutta algorithm, respectively. The parameters used for this model (some from [24,25]) are listed in Table 1, where A32 is the decay rate from level 4I11/2 to 4I13/2, A43 is the decay rate from 4I9/2 to 4I11/2, A54 is the effective decay rate from level 4F7/2 to 4I9/2. C2 and C3 are the co-operative upconversion coefficients for the 4I13/2 level and 4I11/2 levels; C14 is the cross-relaxation coefficients for level 4I15/2 and 4I9/2; and σ24 is the peak excited state absorption cross-section. The various interactions incorporated in the model are shown in Fig. 6(a). The model predicts a maximum gain as high as 27dB over 5cm length as shown in Fig. 6(b). Importantly Fig. 6c shows that gains of 10dB are available across much of the spectrum at only ~50mW pump power, indicating low threshold lasing should be possible with high rear reflections and low propagation losses. Thus with only slight improvements, an amplifier with gain per unit length of 5.4dB/cm is possible and low threshold/high power output lasing will become attainable.
Although considerable scope for improvement exists, a waveguide with >15dB gain itself has applications to a number of active devices, for instance integrated on chip CW, Q-switched, or mode locked lasers. There have been reports of rare earth waveguide lasers [5 ,6, 8, 27–29], CW devices with both linear cavities with fibre Bragg gratings [5, 8, 28] and a ring cavity with couplers . Mode locked operation of an Er-doped glass based waveguide laser has been achieved [6, 8, 30] though not yet in a fully planar integrated form. Lastly, the coupling losses need to be addressed as these contribute directly to the noise figure of a line or preamplifier device. The coupling losses can be reduced very substantially by making use of the vertical tapering method based on aperture masking for a PVD deposited film demonstrated in , as the film here is also deposited by PVD. With such techniques coupling losses to SMF-28 fibre of well below 1dB are expected.
A high internal gain 980nm pumped single-mode Tellurium Dioxide Erbium doped planar rib waveguide amplifier was demonstrated. A planar waveguide Tellurite laser was also realized for the first time. Modelling indicates that higher gains and low threshold/high output power lasing will be attainable with minor improvements in a planar device.
This work was funded by the Australian Research Council mainly via Discovery Project DP0987056 and partly through the Centre for Ultrahigh bandwidth Devices for Optical Systems (CUDOS) CE110001018. The fabrication and characterization were possible in part due to access to the Australian National Fabrication Facilities (ANFF) and Australian Microscopy and Microanalysis Research Facility (AMMRF).
References and links
1. S. Berneschi, G. Nunzi Conti, and G. C. Righini, “Planar waveguide ampflifiers,” Ceramist: J. Korean Ceram. Soc. 10, 75–85 (2007).
2. G. N. van den Hoven, R. J. I. M. Koper, A. Polman, C. van Dam, J. W. M. van Uffelen, and M. K. Smit, “Net optical gain at 1.53 µm in Er-doped Al2O3 waveguides on silicon,” Appl. Phys. Lett. 68(14), 1886–1888 (1996). [CrossRef]
3. J. E. Román, M. Hempstead, C. Ye, S. Nouh, P. Camy, P. Laborde, and C. Lerminiaux, “1.7 μm excited state absorption measurement in erbium‐doped glasses,” Appl. Phys. Lett. 67(4), 470–472 (1995). [CrossRef]
4. Y. C. Yan, A. J. Faber, H. de Waal, P. G. Kik, and A. Polman, “Erbium-doped phosphate glass waveguide on silicon with 4.1 dB/cm gain at 1.535 m,” Appl. Phys. Lett. 71(20), 2922–2924 (1997). [CrossRef]
5. G. Della Valle, A. Festa, G. Sorbello, K. Ennser, C. Cassagnetes, D. Barbier, and S. Taccheo, “Single-mode and high power waveguide lasers fabricated by ion-exchange,” Opt. Express 16(16), 12334–12341 (2008). [CrossRef] [PubMed]
6. G. Della Valle, R. Osellame, G. Galzerano, N. Chiodo, G. Cerullo, P. Laporta, O. Svelto, U. Morgner, A. G. Rozhin, V. Scardaci, and A. C. Ferrari, “Passive mode locking by carbon nanotubes in a femtosecond laser written waveguide laser,” Appl. Phys. Lett. 89(23), 231115 (2006). [CrossRef]
7. G. Della Valle, S. Taccheo, P. Laporta, G. Sorbello, E. Cianci, and V. Foglietti, “Compact high gain Erbium-ytterbium doped waveguide amplifier fabricated by Ag-Na ion exchange,” Electron. Lett. 42(11), 632–633 (2006). [CrossRef]
8. R. Osellame, G. Della Valle, N. Chiodo, S. Taccheo, P. Laporta, O. Svelto, and G. Cerullo, “Lasing in femtosecond laser written optical waveguides,” Appl. Phys., A Mater. Sci. Process. 93(1), 17–26 (2008). [CrossRef]
9. R. A. H. El-Mallawany, Tellurite glasses handbook: physical properties and data (CRC Press, US, 2002).
10. C. E. Chryssou, F. Di Pasquale, and C. W. Pitt, “Er3+-doped channel optical waveguide amplifiers for WDM systems: A comparison of tellurite, alumina and Al/P silicate materials,” IEEE J. Sel. Top. Quantum Electron. 6(1), 114–121 (2000). [CrossRef]
11. S. Dai, C. Yu, G. Zhou, J. Zhang, G. Wang, and L. Hu, “Concentration quenching in Erbium-doped tellurite glasses,” J. Lumin. 117(1), 39–45 (2006). [CrossRef]
12. A. Mori, “Tellurite-based fibers and their applications to optical communication networks,” Nippon Seramikkusu Kyokai Gakujutsu Ronbunshi/J. Ceram. Soc. Jpn. 116(1358), 1040–1051 (2008). [CrossRef]
13. S. Shen, A. Jha, X. Liu, M. Naftaly, K. Bindra, H. J. Bookey, and A. K. Kar, “Tellurite glasses for broadband amplifiers and integrated optics,” J. Am. Ceram. Soc. 85(6), 1391–1395 (2002). [CrossRef]
14. T. T. Fernandez, G. Della Valle, R. Osellame, G. Jose, N. Chiodo, A. Jha, and P. Laporta, “Active waveguides written by femtosecond laser irradiation in an Erbium-doped phospho-tellurite glass,” Opt. Express 16(19), 15198–15205 (2008). [CrossRef] [PubMed]
15. T. T. Fernandez, S. M. Eaton, G. Della Valle, R. M. Vazquez, M. Irannejad, G. Jose, A. Jha, G. Cerullo, R. Osellame, and P. Laporta, “Femtosecond laser written optical waveguide amplifier in phospho-tellurite glass,” Opt. Express 18(19), 20289–20297 (2010). [CrossRef] [PubMed]
17. S. J. Madden and K. T. Vu, “Very low loss reactively ion etched Tellurium Dioxide planar rib waveguides for linear and non-linear optics,” Opt. Express 17(20), 17645–17651 (2009). [CrossRef] [PubMed]
18. K. T. Vu and S. J. Madden, “Reactive ion etching of tellurite and chalcogenide waveguides using hydrogen, methane, and argon,” J. Vac. Sci. Technol. A 29(1), 011023 (2011). [CrossRef]
19. G. Tittelbach, B. Richter, and W. Karthe, “Comparison of three transmission methods for integrated optical waveguide propagation loss measurement,” Pure Appl. Opt. 2(6), 683–700 (1993). [CrossRef]
20. Y. G. Choi, K. H. Kim, S. H. Park, and J. Heo, “Comparative study of energy transfers from Er3+ to Ce3+ in tellurite and sulfide glasses under 980 nm excitation,” J. Appl. Phys. 88(7), 3832–3839 (2000). [CrossRef]
21. G. N. van den Hoven, E. Snoeks, A. Polman, C. vanDam, J. W. M. vanUffelen, and M. K. Smit, “Upconversion in Er-implanted Al2O3 waveguides,” J. Appl. Phys. 79(3), 1258–1266 (1996). [CrossRef]
23. S. Shen, L. Huang, P. Joshi, and A. Jha, “Gain characteristics of Er3+/Ce3+ codoped tellurite short fibre amplifier pumped at 980 nm,” Electron. Lett. 39(25), 1797–1799 (2003). [CrossRef]
24. S. Shen, B. Richards, and A. Jha, “Enhancement in pump inversion efficiency at 980 nm in Er3+, Er3+/Eu3+ and Er3+/Ce3+ doped tellurite glass fibers,” Opt. Express 14(12), 5050–5054 (2006). [CrossRef] [PubMed]
25. Y. Hu, S. Jiang, G. Sorbello, T. Luo, Y. Ding, B. C. Hwang, J. H. Kim, H. J. Seo, and N. Peyghambarian, “Numerical analyses of the population dynamics and determination of the upconversion coefficients in a new high Erbium-doped tellurite glass,” J. Opt. Soc. Am. B 18(12), 1928–1934 (2001). [CrossRef]
26. A. P. López-Barbero, W. A. Arellano-Espinoza, H. L. Fragnito, and H. E. Hernández-Figueroa, “Tellurite-based optical fiber amplifier analysis using the finite-element method,” Microw. Opt. Technol. Lett. 25, 103–107 (2000). [CrossRef]
27. G. Della Valle, R. Osellame, S. Taccheo, N. Chiodo, G. Galzerano, G. Cerullo, P. Laporta, R. Ramponi, U. Morgner, and A. C. Ferrari, “Waveguide lasers in Er:Yb-doped phosphate glass fabricated by femtosecond laser writing,” in 9th International Conference on Transparent Optical Networks(2007), pp. 289–292.
28. G. Della Valle, S. Taccheo, R. Osellame, A. Festa, G. Cerullo, and P. Laporta, “1.5 μm single longitudinal mode waveguide laser fabricated by femtosecond laser writing,” Opt. Express 15(6), 3190–3194 (2007). [CrossRef] [PubMed]
29. J. D. B. Bradley, R. Stoffer, L. Agazzi, F. Ay, K. Worhoff, and M. Pollnau, “Widely wavelength-selective integrated ring laser in Al2O3:Er3+,” in Lasers and Electro-Optics/Quantum Electronics and Laser Science Conference: 2010 Laser Science to Photonic Applications, CLEO/QELS 2010(2010).
30. H. Byun, D. Pudo, S. Frolov, A. Hanjani, J. Shmulovich, I. Ippen, and F. X. Kartner, “Integrated Low-Jitter 400-MHz Femtosecond Waveguide Laser,” IEEE Photon. Technol. Lett. 21, 763–765 (2009). [CrossRef]
31. S. Madden, Z. Jin, D. Choi, S. Debbarma, D. Bulla, and B. Luther-Davies, “Low loss coupling to sub-micron thick rib and nanowire waveguides by vertical tapering,” Opt. Express 21(3), 3582–3594 (2013). [CrossRef] [PubMed]