Birendra Dutt

Roadmap to an Efficient Germanium-on-Silicon Laser: Strain vs. n-Type Doping

We provide a theoretical analysis of the relative merits of tensile strain and n-type doping as approaches to realizing an efficient low-power germanium laser. Ultimately, tensile strain offers threshold reductions of over 200x, and significant improvements in slope efficiency compared with the recently demonstrated 0.25% strained electrically pumped germanium laser. In contrast, doping offers fundamentally limited benefits, and too much doping is harmful. Moreover, we predict that tensile strain reduces the optimal doping value and that experimentally demonstrated doping has already reached its fundamental limit. We therefore theoretically show large (>; 1%) tensile strain to be the most viable path to a practical germanium-on-silicon laser.

Theoretical Analysis of GeSn Alloys as a Gain Medium for a Si-Compatible Laser

In this paper, a theoretical analysis of unstrained GeSn alloys as a laser gain medium was performed. Using the empirical pseudopotential method, the band structure of GeSn alloys was simulated and verified against experimental data. This model shows that GeSn becomes direct bandgap with 6.55% Sn concentration. The optical gain of GeSn alloys with 0-10% Sn concentration was calculated with different n-type doping concentrations and injection levels. It is shown theoretically that adding Sn greatly increases the differential gain owing to the reduction of energy between the direct and indirect conduction bands. For a double-heterostructure laser, the model shows that at a cavity loss of 50 cm-1, the minimum threshold current density drops 60 times from Ge to Ge0.9Sn0.1, and the corresponding optimum n-doping concentration of the active layer drops by almost two orders of magnitude. These results indicate that GeSn alloys are good candidates for a Si-compatible laser.

Impact of minority carrier lifetime on the performance of strained germanium light sources

Abstract We theoretically investigate the impact of the defect-limited carrier lifetime on the performance of germanium (Ge) light sources. For Ge LEDs, we show that improving the material quality can offer even greater enhancements than techniques such as tensile strain, the leading approach for enhancing Ge light emission. For Ge lasers, we show that the defect-limited lifetime becomes increasing important as tensile strain is introduced, and that defect-limited lifetime must be improved if the full benefits of strain are to be realized. We conversely show that improving the material quality supersedes much of the utility of n-type doping for Ge lasers.

Ultimate limits of biaxial tensile strain and n-type doping for realizing an efficient low-threshold Ge laser

We theoretically investigate the methodology involved in the minimization of the threshold of a Ge-on-Si laser and maximization of the slope efficiency in the presence of both biaxial tensile strain and n-type doping. Our findings suggest that there exist ultimate limits beyond which no further benefit can be realized through increased tensile strain or n-type doping. In this study, we quantify these limits, showing that the optimal design for minimizing threshold involves approximately 3.7% biaxial tensile strain and 2 × 1018 cm−3 n-type doping, whereas the optimal design for maximum slope efficiency involves approximately 2.3% biaxial tensile strain with 1 × 1019 cm−3 n-type doping. Increasing the strain and doping beyond these limits will degrade the threshold and slope efficiency, respectively.

Theoretical Modeling for the Interaction of Tin Alloying With N-Type Doping and Tensile Strain for GeSn Lasers

We investigate the interaction of tin alloying with tensile strain and n-type doping for improving the performance of a Ge-based laser for on-chip optical interconnects. Using a modified tight-binding formalism that incorporates the effect of tin alloying on conduction band changes, we calculate how threshold current density and slope efficiency are affected by tin alloying in the presence of tensile strain and n-type doping. Our results show that while there exists a negative interaction between tin alloying and n-type doping, tensile strain can be effectively combined with tin alloying to dramatically improve the Ge gain medium in terms of both reducing the threshold and increasing the expected slope efficiency. Through quantitative modeling, we find that the best design is to include large amounts of both tin alloying and tensile strain but only moderate amounts of n-type doping, if researchers seek to achieve the best possible performance in a Ge-based laser.

Approaches for a viable Germanium laser: Tensile strain, GeSn alloys, and n-type doping

We model the impact of tensile strain, GeSn alloys and n-type doping on the performance of germanium-based lasers. Ultimately, doping offers limited benefits, whereas GeSn and strain can reduce the lasing threshold by >100x.

Recent development in GeSn alloys for silicon-compatible laser

We present experimental results for photoluminescence from GeSn alloys designed for silicon-compatible laser and compare with our earlier published theoretical predictions. The results demonstrate clear improvement over the strained and highly doped Ge approach. This is an encouraging result on the path toward demonstration of the laser compatible with the silicon CMOS process.