Guy, Petra; (1996) A study of distributed Bragg reflector stacks and quantum wells for 1.55 micron Fabry-Perot modulators. Doctoral thesis (Ph.D.), University College London (United Kingdom).

Text A_study_of_distributed_Bragg_r.pdf Download (23MB)

Abstract

The first part of this work was involved with the study of 1.55µm distributed Bragg reflector stacks (DBRs). The operation of the stacks was introduced and perturbation of the spectra was discussed in order to facilitate the analysis of the measured spectra. 1.55µm DBRs reported in the literature were surveyed and compared, and hence the stacks we chose to study in this work were introduced. This work included the study of a range of p-type low reflectance (i.e., having a reflectance of less than 90%) stacks, such as would be likely to be used as the front reflector of a 1.55µm Fabry-Perot modulator. It was concluded that both GaXIn1-xAsyP 1-y/InP and (AlyGa1-y).47In.53 As/Al.48In.52As would provide a very similar refractive index difference of around 0.25, and hence the same reflectance for a given number of periods. The refractive index of GaxIn1-xAs yP1-y having been calculated to be 3.42 at 1.54µm and 1.59µm with 75% arsenic. Whilst we calculated the index of (Al yGa1-y).47In.53As as between 3.42 and 3.46 for 1.62µm to 1.56µm, for around 10% aluminium. However, the quaternary aluminium containing material warrants further study, since we felt that the refractive index may be increased if lower aluminium levels could be achieved. We also looked into the possibility of using lattice matched Ga.47In.53As and utilising the band gap shift which occurs with n-type doping to reduce the absorption of this material at 1.55µm. Transmission measurements carried out on doped Ga.47In.53 As confirmed this band edge shift would occur with the sort of doping levels one would expect to use in the back reflector of the modulator, (1×10 18cm-3 − 1 × 1019cm -3). DBRs using Ga.47In.53As/InP were grown, and refractive indices of between 3.48 to 3.5 were calculated at 1.56µm to 1.58µm for doping levels of mid to high 1018cm -3 respectively. These n-doped stacks were therefore found to provide the highest refractive index difference of all the 1.55µm materials studied, and we have therefore shown that over 99% reflectance can be achieved with only 30 periods using this material. A stack using Ga.47In .53As/Al.48In.52As was also studied, and in this mirror doping levels in the Ga.47In.53As were thought to be low 1018cm-3. This material combination demonstrated the largest refractive index difference, but the peak reflectance was reduced as a result of absorption. Doping levels of around 3-4 × 1018cm-3 were therefore found to be optimum, since the higher the concentration of dopants, the lower the refractive index of the material, whilst if the number of dopants is too low, the ternary layer will be absorbing. Secondly, this work was involved with the study of the multiple quantum well active regions which form the intrinsic cavity of the Fabry-Perot modulator. Several aluminium based systems were studied, and it was shown that high quality materials could be grown at 1.55µm which would provide sharp excitonic features and narrow (32nm, 17meV FWHM room temperature) photoluminescence line widths. Absorption measurements were carried out, and an absorption coefficient of 4.9 × 103cm -1 in the well at the operating voltage was calculated. Based on this figure, we concluded that only 20–30 wells would be sufficient to provide a high contrast low insertion loss modulator, if a back mirror reflectance of around 99.5% was used, and we have already shown that this figure can be easily achieved if n-doped Ga.47In.53As is used as the high index material in the back mirror.

Download activity - last month

Export as CSVXMLJSON

Download activity - last 12 months

Export as XMLCSVJSON

Downloads by country - last 12 months

Export as CSVXMLJSON

Archive Staff Only

View Item