Surface-grating distributed Bragg reflector (DBR) lasers have been explored and developed for fiber communications since last decade. These single-growth-step lasers have the advantage of no crystal re-growth as required in buried-grating lasers, making fabrication simpler. As to the device characterization, it has been known that Bragg reflectivity strongly affects threshold current, side-mode suppression ratio, and slope efficiency of a DBR laser. Unfortunately, precisely measuring the Bragg reflectivity is usually not feasible, since it is difficult to couple sufficient power under subthreshold condition for a stop-band measurement. On the other hand, either Bragg reflectivity or coupling coefficient obtained from theoretical calculations may deviate from the actual situation substantially, mainly because estimations of the coupling coefficient and the grating losses are not easy to achieve.
Instead, we have characterized DBR lasers differently by using a novel self-consistent method. In the approach, the measurement for mode spacing at the Bragg peak was carried out, the coupling coefficient and the grating losses were then localized through the relation between effective and real grating lengths of the lasers. These precisely estimated device parameters were further put into the calculations of Bragg reflectivity, side-mode suppression ratio, threshold current density, and slope efficiency for the lasers. It was found that the optimal coupling strength, i.e. , for the highest side-mode suppression falls in the region between 1.0 and 1.2 with a laser geometry of La = 200 μm and Lg = 500 μm. Exact value of the above strength can be obtained, if a precise grating loss is given.
Because the mode-spacing under the subthreshold would reflect the mechanism of gain-loss competition inherent in the laser cavity, and the estimations were repeatedly checked by experimental results during computation, we believe our results would be relatively more accurate than those obtained by using other methods. Detailed comparison among similar researches will be demonstrated on the conference.
From our analyses, it can be found that operating at the optimal coupling strength not only provides the highest SMSR, but also gives acceptable slope efficiency and, however, would also come along with inevitable penalty of the threshold current, which is resulting from the higher DBR mirror loss with a shorter effective cavity length under the weaker Bragg reflection. An acceptable compromise can be made, if a slight increase, beyond the optimal coupling strength, is appropriately selected.