Nonlinearity difference in the two passbands of a distributed-feedback semiconductor laser amplifier

Abstract

The nonlinearity difference in the two passbands of a distributed-feedback semiconductor laser amplifier was studied experimentally. A theoretical explanation was given by using the transmission matrix approach. The difference of nonlinearity in the two passbands was found to be enhanced greatly by the mechanism of asymmetric facet reflection.Active optical bistability in Febry-Perot (FP) and distributed-feedback (DFB) semiconductor laser amplifiers has the advantages that the power required for switching is approximately three orders of magnitude smaller than that of the passive bistability and, at the same time, the optical gain of the active bistability can provide a sufficient optical power to drive the subsequent optical gates in the system. 1 ' 2 An additional advantage offered by the DFB laser amplifier bistable element is that its filter characteristic can dramatically reduce spontaneous emission 3 in comparison with the FP laser amplifier, and a higher contrast of optical logical gate can be achieved. 4 The two passbands of a DFB laser amplifier, which are determined by the Bragg diffraction, also make it possible for it to be used in systems with two different wavelengths.In this Letter we extend the average field approximation reported in Refs. 1 and 2 to give a unified treatment of optical bistability for FP and DFB laser amplifiers and for a DFB laser amplifier with finite facet reflections. This particular structure is important in determining the single-longitudinal-mode behavior of DFB lasers and the filter characteristic of DFB laser amplifiers. Here we show, experimentally and theoretically, that this additional structure in a DFB laser amplifier will result in a large nonlinearity difference with respect to its two passbands.The experimental setup is as follows. Two identical DFB buried-heterostructure laser diodes were used as a probe and an amplifier; both of them had one facet antireflection coated and the other facet cleaved. The lasing wavelengths were approximately 1554 nm, and the probe was biased at 2.5 times its threshold current. Two diffraction-limited lenses with a N.A. of 0.6 were used for the beam coupling. An optical isolator inserted between the probe and the amplifier provided 40 dB of isolation, and a half-wavelength plate was used to match the polarization of the two lasers. A monochromator and a FP scanning interferometer were used for rough and fine measurement of the spectrum. Frequency matching and adjusting were accomplished by controlling the laser heat-sink temperature, and the change of the coupled optical power caused by the change of temperature was controlled to less than 5%. Figure 1 gives the measured output near the two frequencies corresponding to the lasing band [ Fig. 1(a)] and the nonlasing band [ Fig. 1(b)] of the amplifier biased at 95% of its threshold. The probe power coupled into the amplifier was approximately 0.9 MW. The short-wavelength band (which would lase if biased above threshold) shows a bistability loop, whereas the long-wavelength band shows...

Effect of chirped gratings on reflective optical bistability in DFB semiconductor laser amplifiers

“…This iteration procedure is continued until convergence (we refer the reader to [14] for further details). Although neglected in this paper, finite facet reflectivities may be included using appropriate matrices [17].…”

Switching Behavior of Bistable DFB Semiconductor Laser Amplifiers

“…A pronounced spike in the output transient response of a bistable DFB-SLA with a uniform grating was observed by Adams and Wyatt [4]. Hui and Sapia demonstrated that the nonlinearity difference of the bistability loops in the two pass bands of a uniform DFB-SLA could be enhanced greatly by asymmetric facet reflections [5]. A few years later, in another report, Hui showed that the bistability switching speed could be improved by increasing the bias level from below to above threshold [6].…”

“…The results are obtained for g 0 L D 1:21 for uniform, g 0 L D 0:63 for QWS, and g 0 L D 2:81 for tapered grating DFB-SLAs, which correspond to three equal transmission peaks (G) centered about the normalized detunings of ıL D 7:2, 1.58, and The active layer refractive index 3.553 -n 4 The buffer layer refractive index 3.405 -n 5 The 1.0, respectively. In fact, these results demonstrate that by introducing a phase shift to the uniform grating, the feedback tends to increase.…”

Transfer-matrix analysis of optical bistability in DFB semiconductor laser amplifiers with nonuniform gratings

“…The product of all transfer matrices yields a total transfer matrix which characterizes propagation through the entire amplifier. Although we neglect facet reflections throughout this paper, nonzero facet reflectivities can be included in a straightforward manner [14], [17].…”

“…Optical bistability in DFB SLA's has been theoretically studied for amplifiers with uniform gratings, where an average intensity distribution was (appropriately) used [15]- [17]. Grating nonuniformities, however, cause the intensity distribution to vary significantly within a DFB device.…”