|
A thin layer of INGaAs is the critical photo-absorbing layer in epitaxially grown layers used in the monolithic integration of active photodetectors and a passive mux/demux. This thin InGaAs active layer requires tight thickness control that is beyond the accuracy and reproducibility of conventional thickness measurement techniques such as a step-height profilometer. In order to address this issue, a technique using Field Emission Scanning Electron Microscopy (FESEM) combined with a chemical delineation etch (use to reveal thin layers) was developed. The etch used was dilute H2SO4:H2O2:H20 which will etch all ternaries and quaternaries but does not etch InP. The selectivity of this etch, between the ternary InGaAs material and the InP, was measured to be approximately 500:1. We found that using the FESEM technique provided a minimum measurable thickness of approximately 50 nm, with a precision of approximately 5 nm. The accuracy of our measurement was verified by TEM analysis on selected samples. We also found that SIMS data, obtained from an independent analytical laboratory and originally performed to monitor doping levels in the layers, could be used to monitor the thickness and/or the variations in composition of the thin InGaAs photo-absorbing layer. The key is to make use of the integrated SIMS peaks rather than the full-width-half-maximum (FWHM) of the peaks. For an InGaAs layer, the integrated arsenic SIMS signal varies with the layer thickness and the arsenic concentration. The variations in either of these, in different growth batches, can be detected by SIMS. We found that using integrated SIMS profiles gives a fully independent thickness and/or compositional control monitor to compare with our FESEM measurements and can be used as a trusted control fo the layer quality. This comes at no additional cost since in our case, SIMS profiles are required to monitor doping levels. In this paper, a correlation is shown between the thickness of a thin layer of InGaAs measured by FESEM and the integrated arsenic SIMS peaks from several independent epitaxial growths. Deviations from the norm are explained from small thickness or compositional changes as supported by FESEM and photo-luminescence measurements.
|
