Longitudinal spatial hole burning, its impact on laser operation, and suppression

Abstract

Longitudinal spatial hole burning (LSHB) is a common effect that appears in lasers with standing-wave cavities, especially in edge-emitting semiconductor laser diodes. The standing-wave pattern introduces a non-uniform distribution of the photon density along the laser cavity. Due to the anti-correlationship between the carrier and photon densities (as the stimulated emission depletes carriers where photons exist), the carrier density distribution becomes non-uniform as well, which further causes the nonuniformity of the optical gain of the active region and the refractive index of the waveguide along the cavity. As such, the lasing condition, determined by the gain and refractive index (through the amplitude and phase condition, respectively), is jeopardized. While distributed feedback (DFB) lasers [1, 2] become the only viable solution now in many applications where the single-mode operation is indispensable, LSHB turns out to be a serious issue that deteriorates a number of DFB laser characteristics. It may cause lasing instability [3], linewidth broadening [4], and increased susceptibility to external feedback, to name just a few. Severe LSHB shakes the fundamental lasing condition and even results in coherent collapse [5]. Actually, LSHB is one of the major limiting factors that cause the relative low yield of DFB lasers in their mass productions, especially for the DFB laser having high coupling factor and/or long cavity design for achieving high side-mode-suppression ratio (SMSR) as required. In 100 Gbps and beyond coherent fiber-optic communication systems, lasers with linewidth in several tens of kHz are usually demanded [6] as the light source or local oscillator. It is mainly LSHB that prevents normal DFB lasers from achieving this goal, as otherwise the linewidth reduction can readily be realized by extending the cavity length. For directly modulated DFB lasers, any external feedback (from, e.g., the fiber-pigtail, or splitter/combiner, or filter, or wavelength multiplexer/demultiplexer) may lead to the lasing wavelength shift, or optical power fluctuation, or even coherent collapse, and lasing termination. Although an optical isolator can be co-packaged with the DFB laser to solve this problem, it is certainly not cost-effective. Chen’s group has been working on the sampled gratingbased DFB lasers and has made many significant contributions on multiple wavelength laser sources and tunable laser sources [7, 8]. Their recent work on LSHB suppression by introducing a pair of phase shifts in the basic sampled grating DFB structure [9] has shed light on solving the aforementioned issue through a viable approach. Comparing with the existing methods [10], not only this approach appears to be more effective, but it also bears the unique feature of fabrication readiness and easiness. This work is also comprehensively and coherently presented with the design concept and working principle clearly explained, and with the simulation and experimental results shown consistently. Therefore, it deserves the great attention of researchers in the relevant area.