Open Access
Add to BibSonomy
Abstract
Semiconductor lasers can be athermalised, which allows them to maintain their lasing wavelength with varying ambient temperature. Athermalisation can be achieved with careful control of injection currents into the laser diode. This allows the removal of a thermoelectric cooler and hence enables athermal wavelength division multiplexing. In this study, a high-order surface grating laser is operated athermally on eight, 100 GHz spaced channels from the ITU dense wavelength division multiplexing (DWDM) grid, achieving a wavelength stability of ±0.003nm/±0.4 GHz from 20 to 72–108°C depending on the channel. High side mode suppression ratios(>40 dB) and output powers(>+10 dBm) are also observed for a majority of the tuning ranges. Additionally, operation on six, 12.5 GHz spaced channels is also demonstrated with similar performance.
Published by The Optical Society under the terms of the Creative Commons Attribution 4.0 License. Further distribution of this work must maintain attribution to the author(s) and the published article's title, journal citation, and DOI.
1. Introduction
The annual rise in internet traffic has demanded continual industrial innovation which has led to the development of dense wavelength division multiplexing (DWDM). Lasers for DWDM applications generally use thermoelectric coolers (TECs) to maintain their temperature near 20°C and keep their wavelengths within the required channels, however these TECs increase energy usage and reduce package lifetime which is undesirable. A key challenge in upcoming DWDM systems is the reduced channel spacing, requiring wavelength stability of $\pm$3.125, $\pm$12.5 and $\pm$25.0 GHz for 12.5, 50 and 100 GHz spaced grids respectively [1]. This is challenging to achieve as laser diodes are susceptible to large wavelength variations (0.1 nm/ °C or 12.5 GHz/ °C) with respect to ambient temperature changes.
Coarse WDM (CWDM) uses lasers that operate without a TEC, but this is because any temperature induced wavelength drift is accommodated by using 20 nm wide channels. Due to growth in demand for increased bandwidth and requirements for improved energy efficiency right down to the device level increase, smaller channel spacings will be required. In what we term athermal WDM (AWDM), the laser wavelength is kept constant within a channel, while the currents injected into the laser are varied to maintain this wavelength without the need for a TEC. This requires precise monitoring of the laser diode temperature and fine control of injection currents, but as a result allows drastic shrinking of CWDM channel spacing. Previous work on athermal operation has demonstrated impressive performance over the entire C-band using Digital Super-mode Distributed Bragg Reflector (DS-DBR) lasers [2,3], however to achieve this, fine current control over 5 laser sections was required. These schemes have been simplified with auto-tuned algorithms to three current regimes [4], however they still use complicated devices which is undesirable. Another solution that has been proposed is to use hybrid structures for passive athermalisation [5,6], however these significantly increase device fabrication which is unsuitable for current DWDM systems. Furthermore, current control schemes could potentially be deployed to already existing transmitters which is not possible for passive athermalisation. As a simpler alternative, high-order surface grating lasers have been previously demonstrated as an effective solution [7]. Nevertheless, in all the aforementioned cases, large, continuous athermalisation ranges could not be realised without making use of discontinuous tuning, which would require down time in the transmission link while the retuning is performed, which is not optimal.
In this paper, we use a monolithic high-order, surface grating laser operating in the C-band. This device is operated athermally over eight, 100 GHz spaced channels. Additionally six, 12.5 GHz spaced channels are demonstrated, with their position being in-between two of the 100 GHz spaced channels. Both of these schemes use only continuous tuning. The achieved wavelength variation is only $\pm$0.003 nm/$\pm$0.4 GHz for every channel over a temperature range of 20 to 72–108°C depending on the channel. This work demonstrates variations of device parameters depending on the wavelength channel used, in particular the differing range of temperatures over which a device can achieve athermalisation. However, with this performance, the aforementioned device may prove to be a promising candidate for tightly spaced AWDM networks.
2. Device structure and athermalisation theory
The device used in these measurements is a monolithic, 700 μm long, three-section all active DBR laser, with an integrated semiconductor optical amplifier (SOA). A sketch of a sample device is shown in Fig. 1. It should be noted that this illustration only demonstrates one device on a 12-laser array that covers the C-band [8].
Fig. 1. An example surface grating laser. It is a multi-section device with an active grating section. Note the SOA is curved at 7°.
This device is an all-active, ridge waveguide laser, with the active medium consisting of 5 AlGaInAs quantum wells (QWs) which have the material gain peak at around 1545 nm at room temperature and with slots of the grating section etched into the 2 μm wide ridge. The lasers were fabricated on a commercially purchased wafer, which consists of 5 AlGaInAs QWs, above which there is a 1.6 μm p-doped InP layer (referred to as cladding here), 50-nm thick, p-doped InGaAsP layer, and a 200 nm InGaAs contact layer. Below the quantum wells there is 120 μm layer of InP. The slots and the ridge of the laser are fabricated using two inductively coupled plasma (ICP) etch steps using Cl2 and N2 gas. This results in a 1.35 μm high ridge. The laser is subsequently contacted, cleaved and the facets are coated with high reflection (HR) and anti-reflection (AR) films, before being eutectic bonded onto an AlN carrier. As the slots were etched into the ridge, the need for regrowth steps was removed. The three sections of the ridge are the gain section, the grating section and the SOA section, with lengths of 470, 230 and 200 μm respectively, forming a 700 μm long laser. The gain is used to control the position of the Fabry-Perot modes, the grating is used to control the position of the Bragg peak and the SOA can be used to control the output power. Note that the SOA section is curved at an angle of 7° to reduce back reflections from the front facet. This SOA section is outside of the lasing cavity, and thus has minimal effect on the laser wavelength tuning [9], only affecting the series resistance and heating of the substrate, however these effects are small when compared to the tuning capability of the gain/grating sections. As the goal of these devices is to simplify manufacturing, high-order gratings which result in large feature size were employed, however these come at the expense of a low free spectral range (FSR), which results in several reflection peaks falling within the gain bandwidth, deteriorating device performance. Previously, it had been shown that employing more than a single period within the grating structure significantly reduces this effect [10], therefore such designs were chosen for this study. The specific grating parameters are shown in Table 1. These grating parameters have been modelled, optimised and characterised in Refs. [7,8,10]. The overall result of the optimisation is a compromise between output power and side mode suppression ratio (SMSR).
Table 1. Parameters of the triple period grating.
The elimination of the regrowth step results in there being no passive regions in the device, and as such, the carrier density clamps for both wavelength tuning sections (gain and grating sections) above threshold, causing thermal tuning to dominate. Therefore, to achieve continuous athermalisation, changes in ambient temperature can be offset via changes in device self-heating due to variance of injection currents across the device. Hence, the continuous athermalisation condition is maintained by sustaining the local ridge temperature over the entire athermal range. Thus, to achieve continuous athermalisation over a large ambient temperature range, it is necessary to induce a large amount of self-heating via a large amount of current injection at low ambient temperatures, effectively operating at higher device temperature. This is why relatively short laser sections were chosen. Previously, it has been demonstrated that laser diodes with a large red-shift of their Bragg peak, with respect to the material gain peak at 20°C, exhibit improved performance in terms of output power and slope efficiency at higher temperatures [11,12]. Furthermore, this redshift also produces an increased stability in output power and SMSR, while thermally tuning, which has been recently demonstrated by our group, in addition to having added benefits of increasing athermal range [13]. This is due to the material gain peak and the Bragg peak drifting at different rates with respect to ambient temperature (0.5nm/°C and 0.1nm/°C respectively), resulting in more favourable overlap at higher ambient temperatures and thus, improved lasing performance. The next section will make use of these facts to achieve wide range, mode-hop free athermalisation over several WDM channels.
3. Results
Measurements for this study were carried out on the setup in Fig. 2.
The array of devices was placed onto a copper heatsink which had an integrated TEC and thermistor for monitoring ambient temperature. The gain, grating and SOA sections of a single device were individually contacted to allow independent control of currents to each section. For wavelength and SMSR measurements, output light was coupled using a lensed optical fibre and passed through a 90:10 optical splitter, with 90$\%$ of the light going to a spectroscopic measurement (HP 86120B wavemeter for wavelength, Agilent 86140B optical spectrum analyser for SMSR) and 10$\%$ of the light going to a PIN photodiode to assist with coupling of light into the fibre more quickly. For output power measurements, a bare photodiode was used. Changes in ambient temperature were then simulated using a TEC, and the device temperature was maintained at a constant value by controlling the injection current. The TEC temperature in these measurements was adjusted in 10°C increments until a mode-hop was approached, at which point finer increments were used to obtain the maximum range of temperatures over which athermal operation is maintained. This is better known as the athermal range. As mentioned earlier, the gain and grating currents were varied accordingly to maintain the athermalisation condition, while the SOA current was kept at a constant 30 mA, which is roughly the midway point between transparency and saturation of the SOA.
The tuning map of the device at 20°C is plotted in Fig. 3, with the shape of the map being a result of the asymmetric shift of the cavity modes with respect to the reflective peak. The currents used to achieve athermalisation for 100 GHz channels are plotted on the map.
Fig. 4. Wavelength stability of 100 GHz spaced channels. Note the increasing athermal range with increasing wavelength (from lower right to upper left).
The squares are used to denote the start of the athermal path at 20°C. We use this map to guide the athermal operation, by helping us choose a starting current and to avoid mode hops. Note, that the tuning map changes with temperature, this is why some of the paths appear to cross mode boundaries further along the path. Figure 4 shows the wavelengths for the 100 GHz spaced channels. It is observed that the wavelength remains constant to within $\pm$0.003 nm/$\pm$ 0.4 GHz of the central wavelength. The athermal range can also be seen from these figures, but is better illustrated in Fig. 5.
Fig. 5. Athermal range of 100 GHz channels versus injection current at 20°C. Note how the range saturates at high total currents.
Initially, there appears to be a linear rise in athermal range with increased current at 20°C, however this effect seems to saturate. This behaviour can be explained by taking a closer look at injection currents into each device section given in Fig. 6(a-c).
Fig. 6. Variation of (a) gain, (b) grating and (c) total injection currents versus temperature over the athermal paths for 100 GHz spaced channels.
From this figure, it is clear that there is not a large variation in gain current at 20°C, however there still is a small increase in gain current with increasing wavelength. The main difference between channels comes in the form of the grating current, as this varies by a significant amount between 100 GHz spaced channels and is responsible for the differences in athermal range. However, once a grating current of $\approx$200 mA is reached, this effect saturates and small differences to athermal range can be attributed to variations in gain current. This is because past 200 mA, the local temperature of the grating becomes significantly larger than that of the gain section and therefore, gain current becomes the limiting factor in the athermal range. This conclusion is obtained using thermal impedance values recorded in Ref. [14], and from voltage measurements at a given current for each section. This then can be used to calculate input power which can be used to obtain average section temperature. Additionally, the increased athermal range seems to come with a reduced output power as shown in Fig. 7(a). This can be compensated to some degree with the short SOA section, by varying the injection current into the SOA, but has not been studied in this set of experiments. The reduced power is due to shorter wavelength devices operating at lower local temperatures and hence, not suffering as severely from temperature sensitive effects such as leakage current and Auger recombination. Additionally, all channels exhibit a decrease in output power with increasing ambient temperature, which is due to a reduction in carrier injection, reducing the gain provided to the main lasing peak and thus decreasing output power. Similarly, the SMSR is observed to decrease with temperature as shown in Fig. 7(b). A faster decrease is observed for shorter wavelengths which is due to them initially operating at lower injection currents, thus these wavelengths are closer to threshold than longer wavelengths resulting in a more rapidly decreasing SMSR with respect to temperature.
Fig. 7. Variation of (a) output power and (b) SMSR and versus temperature over the athermal paths for 100 GHz spaced channels.
Finally, wall-plug efficiency, which is the ratio of output power versus electrical input power into the laser, was recorded for each channel and this result is shown in Fig. 8. The input power into the TEC is excluded for this measurement as the TEC is only used to simulate changes in ambient temperature. It is observed that wall-plug efficiency decreases with decreasing channel number, which is due to increased device temperature, furthering temperature sensitive inefficiencies such as carrier leakage. The observed efficiency is seen to be relatively low, at around 1$\%$, which is due to the devices operating at very high local temperatures. This is a trade off made to achieve a very high athermal range. If a smaller athermal range is required, then the amount of device self-heating could be reduced, and thus a reduction in device temperature could be achieved, improving the wall-plug efficiency.
Additionally, as the wavelength stability achieved was better than that required for 12.5 GHz spaced channels on the DWDM grid, which allow a maximum spectral excursion of $\pm$ 3.125 GHz, these lower spaced channels were also studied for athermal operation. The chosen channels were between channels 9 and 10, with their specific wavelengths and stability illustrated in Fig. 9. As in the case of 100 GHz spaced channels, these more tightly spaced channels also exhibit a wavelength variation of $\pm$0.003 nm/$\pm$0.4 GHz which is well within the ITU requirement. This level of stability could lead to further downscaling of channel bandwidth, below 12.5 GHz, however as the requirements in terms of wavelength stability are not yet defined, closer spacings were not studied.
Fig. 9. Wavelength stability of the 12.5 GHz spaced channels. Maximum drift of $\pm$0.003 nm/$\pm$0.4 GHz from the channel centre is demonstrated.
The athermal range is then illustrated in Fig. 10. It is observed to increase with decreasing channel number, however some of these channels seem to have a lower athermal range when compared to the 100 GHz channels they are surrounded by (channel 9 and 10). This is because channels 9.875 to 9 were operated within a different mode than channel 10, requiring lower total current for shorter wavelengths within the mode, and hence not allowing each channel to achieve the same amount of self-heating, leading to variations in athermal range.
The currents used for each channel are plotted in Fig. 11(a-c). It is observed that while the grating current for channel 10 is the lowest, the gain current is not. As previously mentioned, for longer wavelengths, gain current becomes the dominant factor in deciding athermal range and thus, this results in channel 10 having longer athermal range than some of the other channels.
Fig. 11. Variation of (a) gain, (b) grating and (c) total injection currents versus temperature over the athermal paths for 12.5 GHz spaced channels.
Output power and SMSR are also illustrated in Fig. 12 and are observed to be mostly consistent with trends demonstrated for 100 GHz spaced channels, which is expected.
Fig. 12. Variation of (a) output power and (b) SMSR and versus temperature over the athermal paths for 12.5 GHz spaced channels.
Wall plug efficiency similarly is observed to follow the same trend, with shorter wavelengths achieving improved efficiency, as illustrated in Fig. 13. The athermalisation results shown here are the first to be performed on multiple channels, particularly of 12.5 GHz spacing on simple to fabricate devices, over such a wide temperature range. It is clearly observed that currents finely tuned over the two tuning sections of our device can achieve comparable, or even improved results over more complex, DS-DBR devices in terms of wavelength stability and athermal range. Thus, an array of ten of our devices could be a promising competitor to these more complex transmitters.
Fig. 13. Wallplug efficiency versus temperature over the athermal paths for 12.5 GHz spaced channels.
4. Conclusion
A simple to fabricate, high-order surface grating laser is operated athermally, achieving mode-hop free performance over eight, 100 GHz spaced channels with a wavelength variation of $\pm$0.003 nm/$\pm$0.4 GHz, with an athermal range of 72–108°C depending on the channel. This performance has also been shown on a more tightly spaced, 12.5 GHz grid. With high output power, low spectral excursion and high SMSR, this device is a promising candidate for uncooled or athermal WDM networks. Much work has to be done to look at device lifetimes under athermal operation. However, similar testing for CWDM was required given the high temperature environments in which this particular technology operates.
Funding
Science Foundation Ireland (15/IA/2854).
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. ITU-T, “ITU-T Recommendation G.989.1 40-Gigabit-capable passive optical networks (NG-PON2): General requirements,” Itu-T G-Series Recommendations (2013).
2. S. H. Lee, A. Wonfor, R. V. Penty, I. H. White, G. Busico, R. Cush, and M. Wale, “Self-configuring athermal tunable DS-DBR laser for passive optical networks,” in CLEO/QELS: 2010 Laser Science to Photonic Applications, (2010), pp. 1–2.
3. J. Zhu, A. Wonfor, S. H. Lee, S. Pachnicke, M. Lawin, R. V. Penty, J.-P. Elbers, R. Cush, M. J. Wale, and I. H. White, “Athermal Colorless C-Band Optical Transmitter System for Passive Optical Networks,” J. Lightwave Technol. 32(22), 4253–4260 (2014). [CrossRef]
4. L. Ponnampalam, C. Renaud, M. Fice, R. Cush, R. Turner, P. Firth, M. Wale, and A. Seeds, “High temperature operation of athermal widely tuneable laser with simplified wavelength control for WDM-PON systems,” Opt. Express 22(20), 24405–24410 (2014). [CrossRef]
5. J. Bovington, S. Srinivasan, and J. E. Bowers, “Athermal laser design,” Opt. Express 22(16), 19357–19364 (2014). [CrossRef]
6. B. R. Koch, E. J. Norberg, B. Kim, J. Hutchinson, J.-H. Shin, G. Fish, and A. Fang, “Integrated Silicon Photonic Laser Sources for Telecom and Datacom,” in 2013 Optical Fiber Communication Conference and Exposition and the National Fiber Optic Engineers Conference (OFC/NFOEC), (2013), pp. 1–3.
7. M. J. Wallace, R. O. Meehan, R. Enright, F. Bello, D. McCloskey, B. Barabadi, E. N. Wang, and J. F. Donegan, “Athermal operation of multi-section slotted tunable lasers,” Opt. Express 25(13), 14414–14426 (2017). [CrossRef]
8. A. Abdullaev, Q. Lu, W. Guo, M. J. Wallace, M. Nawrocka, F. Bello, A. Benson, J. O’Callaghan, and J. F. Donegan, “Improved performance of tunable single-mode laser array based on high-order slotted surface grating,” Opt. Express 23(9), 12072–12078 (2015). [CrossRef]
9. M. McDermott, R. McKenna, C. Murphy, D. Mickus, H.-Z. Weng, S. Naimi, Q. Lu, W.-H. Guo, M. Wallace, N. Abadía, and J. F. Donegan, “1.3 µm wavelength tunable single-mode laser arrays based on slots,” Opt. Express 29(10), 15802–15812 (2021). [CrossRef]
10. G. Jain, M. J. Wallace, R. McKenna, K. Brazel, F. Bello, Q. Lu, W. Guo, and J. F. Donegan, “Design Optimization for Semiconductor Lasers With High-Order Surface Gratings Having Multiple Periods,” J. Lightwave Technol. 36(22), 5121–5129 (2018). [CrossRef]
11. F. Xie, C. Caneau, H. P. Leblanc, S. Coleman, M.-T. Ho, C. A. Page, L. C. Hughes, and C.-E. Zah, “Impact of Wavelength Detuning on the Performance of Mid-IR Distributed Feedback Quantum Cascade Lasers,” IEEE J. Sel. Top. Quantum Electron. 19(4), 1200508 (2013). [CrossRef]
12. Y. Nishimoto, H. Yagi, K. Miura, D. Plumwongrot, K. Ohira, T. Maruyama, and S. Arai, “High T0 Operation of 1590 nm GaInAsP/InP Quantum-Wire Distributed Feedback Lasers by Bragg Wavelength Detuning,” Jpn. J. Appl. Phys. 46(No. 17), L411–L413 (2007). [CrossRef]
13. D. Mickus, R. Mckenna, C. Murphy, and J. F. Donegan, “Wide range thermal and athermal operation of slotted surface grating lasers,” Opt. Express 29(11), 16893 (2021). [CrossRef]
14. R. McKenna, D. Mickus, S. Naimi, C. Murphy, M. McDermott, S. Corbett, D. McCloskey, and J. F. Donegan, “Spatially resolved self-heating and thermal impedance of laser diodes using CCD-TR imaging,” OSA Continuum 4(4), 1271–1281 (2021). [CrossRef]
