Std. Quantum Well
Keep this page updated with the most recent SQW data!
The first structure/device created during system bring-up is the standard quantum well (SQW). The SQW is a metric for the status of source calibration and cleanliness of the system. Two SQW structures are fabricated:
InGaAs/GaAs
AlGaAs/GaAs
The InGaAs/GaAs SQW acts as a first-order sanity check that all sources EXCEPT Al are operating and responding correctly. If the InGaAs/GaAs SQW is functioning similarly to previous growth campaigns, the Al can be melted. Once the Al is melted, it is not refrozen until the end of the campaign because it wets to the pBN crucible and, when cooled, will contract and can crack its crucible -- Al is only cooled when the system is brought down. Thus, the AlGaAs/GaAs SQW is grown second to check the status of the Al source material. After the AlGaAs/GaAs SQW is performing correctly, the system is ready for the all growths in the new growth campaign.
This page should be updated at the start of new growth campaigns with the updated SQW data! Comparison between growth campaigns is vital to understanding if the system is ready for other growths and shows the evolution of the system over months/years of operation.
Measurement and Comparison Metrics
How to Measure PL Correctly
Follow the Photoluminescence Procedures when measuring SQWs.
PL Intensity Degradation and Mitigation
Laser quickly degrades sample, over only a few minutes. Measure previous PL for the most recent (2 or 3) campaigns during the same PL measurement session. Measure each sample twice with approximately the same time between the 1st and 2nd measurement. For example, make the first measurement almost immediately after placing the sample on the sample chuck. Then, make the second measurement shortly after the first one finishes, say 5 minutes later. It is important to compare the relative peak size between the different 1st measurements for each sample and the 2nd measurements for each sample.
The degradation in intensity is explained by the breaking of bonds near the surface of the samples due to the high energy density of the laser light. Breaking bonds creates surface defects and lowers the optical response of the material by creating non-radiative recombination sites. One way to mitigate the defect states is to place the quantum well layer farther from the surface. The degradation has been studied by LASE for InGaAs/GaAs SQWs but not for AlGaAs/GaAs SQWs, however, similar behavior is expected.
PL Peak Shifting
Peak shifting indicates there is a change in the carrier confinement of the QW. Recall that increasing the width ("thickness" in growth terminology) causes (1) the separation between quantized energy levels decreases and (2) the position of the quantized levels is closer to the bottom of the conduction band. This change in the position of the energy levels shows up as a shift in the PL emission peak.
Another explanation for the peak shifting is the In or Al calibration are slightly off. This leads to stoichiometrically different material than what is expected, thus changing the bandgap of he material (since it is a different material). This is easily seen in a plot of bandgap energy v. lattice constant.
For example: using this plot, say the In is incorrectly calibrated and results in extra In deposition. This leads to a smaller bandgap or longer wavelength emission.
Most Recent InGaAs/GaAs
Recipe
InGaAs/GaAs SQW recipe -- used during B140929B growth
Plots
The following is a comparison of recent scans made 2014-09-30
Comments and Analysis
These plots compare the previous two campaigns -- B130626B and B140307A -- and the current campaign -- B140929B. All scans shown are the 1st of the two measurements taken for each sample. B140307A and B140929B appear to line up very well with peak intensities near 1000 nm. Although B140929B luminesces less than B140307A, but this is not a problem since the relative difference is only ~25%. In general, the current and previous campaign have almost identical In calibration. Notice that B130626B has a peak intensity at a shorter wavelength, 980 nm, which indicates a larger bandgap. This suggests there is less In in the InGaAs layer (see Energy v. Bandgap diagram).
Most Recent AlGaAs/GaAs
Recipe
AlGaAs/GaAs SQW recipe -- used during B141020A growth
Plots
The following is a comparison of recent scans made 2014-10-20
Comments and Analysis
This PL intensity plot is more telling for the AlGaAs/GaAs SQW than the InGaAs/GaAs SQW. The PL intensity from a recent bring up (B141002B) is significantly (~5x) lower than the previous campaign (B140409A). The Al cell was not sufficiently cleaned. Al is more susceptible to oxygen contamination due to gettering, thus high-temperature outgassing is required to "clean" the Al source material. Following a junk growth and high-temperature outgassing, another attempt at the AlGaAs/GaAs SQW was performed (B141020A), and the PL intensity recovered to similar levels as the previous campaign. Multiple junkgrowth/high-temperature outgassings followed by SQW growth sessions might be needed to recover the PL to appropriate levels.