Tunable lasers provide the telecommunications industry with an important tool for the design of new network architectures. Unlike lasers manufactured for just a single wavelength, a tunable laser can be rapidly retuned to support several different wavelengths in the International Telecommunication Union (ITU) grid for wavelength-division multiplexing (WDM). As the number of wavelengths in dense WDM systems escalates toward the 100-plus range, tunable lasers become increasingly important in containing design and manufacturing complexity by reducing the number of different lasers that must be produced and deployed.
The flexibility afforded by tunable lasers could support, for the first time, wavelength-routing capabilities in the optical domain-a key element in the development of all-optical networks. With tunable lasers, network operators will be able to route traffic over fixed paths to different points in the network by changing the laser emission wavelength. Other potential applications might include network protection and the use of a single wavelength per large business customer to create future virtual private networks (see photo here and on cover).
Tunable lasers also can be used as spares so that network operators do not have to maintain large and costly inventories of lasers, each dedicated to a single wavelength. Operators would also gain flexibility in reconfiguring the wavelengths used in their systems if they could retune the lasers already there instead of ordering and installing new ones.
Range of solutions
Tunable lasers can be produced with a variety of different laser structures, each with advantages and disadvantages. Two of the basic structures are distributed Bragg reflector (DBR) and distributed feedback (DFB) lasers. In a DBR laser, the active region, which provides gain, and the grating, which provides wavelength selectivity, are separated. In a DFB laser, the two functions are combined.
Distributed Bragg reflector lasers are the simplest and oldest of the structures. These lasers have a reflector at each end of the laser cavity and use current tuning in the passive reflector region. The typical continuous tuning range is about 6 nm, and it is limited by current injection. The main drawback in using DBR lasers as tunable lasers is the difficulty of controlling the length of the optical path between the reflectors at the ends of the cavity.
null
Distributed feedback lasers have a reflector embedded in the cavity gain region. They are commonly tuned to different wavelengths by changing the temperature, either through drive-current changes or a controlled heat sink. The temperature can be precisely controlled-to within a tenth of a degree or better-to produce a well-defined output wavelength.
The DFB lasers are well behaved and characterized, and extensive studies have verified their reliability. While DFB lasers offer manufacturing, performance, and operational advantages, their effective tuning range tends to be limited to about 5 nm because the efficiency and output power of the device goes down as the tuning temperature goes up.
To extend the tuning range, these devices can be integrated into arrays. Ensemble or side-by-side laser arrays integrate several lasers on one chip and couple them into a single output. One laser at a time is driven to select a wavelength. However, this solution is not continuously tunable, the combining mechanism is optically inefficient, and chip size leads to yield issues. Cascade or in-line laser arrays overcome the coupling loss by allowing light to pass transparently through other laser sections on the device. The main challenge in this case is achieving mode stability for each of the laser sections.
More-complex structures
Other laser structures exist that offer a wider tuning range than DFB lasers, although they are more complex to manufacture and control. Sampled-grating tunable lasers have grating reflectors at either end of the cavity. The "sampling" of the grating produces a spectral comb response. The back and front sampled gratings have slightly different pitch so that the resulting spectral combs have slightly different mode spacing.
Tuning to a specific wavelength is achieved by controlling the current in the two grating sections so as to align the two combs at the chosen wavelength. The laser, therefore, "hops" between wavelengths. An additional contact is normally required to adjust the phase so that an integral number of half-wavelengths exist. If this phase adjustment is not included, then mode stability can suffer and noise increases.
While these types of lasers offer a wider tuning range-typically around 70 nm-they have complex drive requirements when compared to DFB lasers. They have more electrodes, and accurate wavelength selection requires matching of numerous input currents with the appropriate electrodes, although disk- or CD-ROM-based lookup tables can be used to automate the process.
Sampled gratings with grating-assisted codirectional coupler filters add a filter to select one of the sampled grating peaks. This modification makes it easier and cleaner to select a wavelength but at the cost of additional manufacturing complexity.
External-cavity lasers are another configuration in which the wavelength selection and tuning functions are external to the semiconductor material. Changes in the mechanical size or geometry of the cavity are used to tune the laser, rather than temperature tuning or current changes applied to the semiconductor material. Basically, these lasers require just a simple Fabry-Perot chip to drive a much-longer cavity that normally consists of passive components. A separate tunable bulk grating filter allows the external cavity to operate only at the frequency set by the filter.
Commonly used in laboratory instruments that operate in a stable environment, external-cavity lasers are bulky and sensitive to mechanical shock and movement, although some manufacturers are making good progress in addressing these drawbacks. They are easy to manufacture, but their high cost is an issue for telecommunications.
FIGURE 1. Three-section tunable DFB laser combines an active region to provide gain and a grating to provide wavelength selectivity.
All the structures described above are semiconductor-based. Another possible solution is fiber-based structures. This approach uses erbium-doped fiber amplifiers (EDFAs) as the optical gain medium. However, work in this area is still in its infancy.
A 15-nm DFB solution - Last year, Nortel Networks reported a frequency-stabilized DFB laser that can be temperature-tuned over seven
100-GHz channels (5 nm). The laser is mounted on a double-stage thermoelectric cooler for frequency tuning. The submount temperature range is -10°C to +50°C. The +50°C limit was established to minimize laser degradation over its life.
null
A more recent development is a continuous-wave, three-section DFB laser (see Fig. 1). This cascade-laser-array type of device operates over a 15-nm tuning range, supporting 19 channels at 100-GHz spacing on a single chip. The device is grown using a standard two-step metal-organic copper-vapor-deposition system. The laser active region consists of compressively strained indium gallium arsenide phosphide multiple-quantum-well layers.
The laser chip contains three separate electrode sections. For each laser section, a different grating pitch is produced using an electron-beam writer, enabling a tuning range of 5 nm per electrode. The laser device is continuously tuned to different wavelengths by varying the temperature between 0°C and 50°C and biasing one laser segment at a time with approximately 100 mA of input current.
null
The pitch under each electrode is offset so that the Bragg center wavelength of the individual segment increases with the number of sections. This design allows light to pass transparently through the next section in the cascade without being reflected by the corresponding grating reflection band (see Fig. 2). Thus the 15-nm tunable three-section DFB laser combines three laser sections on a single device.
This technology offers a very simple, controllable, and potentially easy-to-manufacture device. Unlike other options that have to juggle with multiple electrodes and input currents to select the correct wavelength, each electrode addresses only one region of the spectrum, and a temperature scan is sufficient to produce the desired wavelength. The technology is also extendable to a 30-nm tuning range-doubling the number of wavelengths-by coupling two of the devices together into the same fiber.
The major technical challenge was to achieve good mode stability for each laser section of the device throughout the entire temperature range. Without mode stability, data transmission over fiber would be severely degraded. The device exhibits continuous single-mode operation over the 15-nm range, with a side-mode suppression ratio in excess of 50 dB. Fiber-coupled power in excess of 0 dBm was measured for all channels (see Fig. 3)
The 15-nm tunable laser is packaged with an internal wavelength stabilization device. Once a wavelength is selected, the etalon-based stabilization device locks it precisely so that it does not drift out of its acceptable range and interfere with adjacent wavelengths in the narrowly spaced DWDM grid. Ensuring the accuracy of these wavelengths over the life of the laser is a key to maintaining the reliability of DWDM systems.
Tunable laser devices such as these, as well as those being developed by various suppliers around the world, provide a range of structural options. With technological advances, tunable lasers will indeed play an increasingly important role in emerging optical networks. o
CHRIS F. CLARKE is senior product manager at Nortel Networks, Optoelectronics Division, Brixham Road, Paignton, Devon, UK TQ4 7BE; JIN HONG and MIK SVILANS are senior device designers and CLAUDE ROLLAND is senior manager of device R&D at Nortel Networks, Box 3511, Station C, Ottawa, Ontario, Canada K2B 5T2; e-mail: [email protected].