Suppressed phase separation in thick GaInAsSb layers across the compositional range grown by molecular beam epitaxy for 1.7–4.9 μm infrared materials
(Color online) Predicted thermodynamic stability of Ga1-xInxAsySb1-y (with the y = 0.91x condition for lattice matching to GaSb) grown at 500 C,10 as well as bandgaps as a function of indium concentration.11 White regions: binodal growth; dark gray: metastable growth; light gray: spinodal decomposition. The figure also shows the dependence of cut-off wavelength of the lattice-matched alloy on indium concentration.
(Color online) Calibration curves for obtaining target arsenic or antimony concentrations. Point symbols are experimental and represent different indium concentrations; colors represent growth at different temperatures. Solid and dashed lines represent fits to the data according to Eq. (1) with a temperature dependent, adjustable incorporation parameter. Because incorporation rates of Group V elements depend on temperature, different color solid curves are plotted for different growth temperatures as suggested by the data. For arsenic (antimony) concentrations below 40% (60%), antimony (arsenic) flux was held constant at a level slightly above stoichiometry for GaSb (InAs) growth, while the arsenic flux was ramped up as shown in panel (a) [panel (b)].
(Color online) Schematic of the structure of samples grown. The Ga1-xInxAsySb1-y layers were nearly lattice-matched to GaSb for x = 0 to 1.
(Color online) Schematic of energy bands, gaps, and offsets of samples grown in this study at selected indium concentrations in Ga1-xInxAsy Sb1-y, with band parameters from Ref. 10: (a) Ga0.10In0.90As0.82Sb0.18, (b) Ga0.40In0.60As0.56Sb0.44, (c) Ga0.52In0.48As0.42Sb0.58, (d) GaSb. (a)–(b) For higher indium concentrations (0.6 ≤ x ≤ 0.9), Al0.20In0.80As0.73Sb0.17 barriers were necessary to obtain hole confinement. Good hole confinement for 0.9 < x was not possible using these clads. (c)–(d) For lower indium concentrations (x < 0.60), AlAs0.08Sb0.92 barriers gave sufficient hole confinement.
(Color online) Photoluminescence from Ga1-xInxAsySb1-y samples with x varying from 0 to 0.9 continuously spanning the wavelength range 1.7 to 4.9 ums: (a) Peak PL normalized to 1 (b) Relative PL intensity, with detector and FTIR response normalized out.
(Color online) Summary of measured room temp and 77 K PL and HRXRD FWHM of samples.
(Color online) Comparison of measured room temperature PL peak emission as a function of indium concentration compared to analytical expression Eq. (2).
(Color online) HRXRD for two of the samples (a) Ga0.52In0.48 As0.42Sb0.58 and (b) Ga0.10In0.90As0.82Sb0.18. Beyond the single alloy peak and substrate peak, no other significant peaks appeared over a wide range of angles, indicating only one alloy phase is present. Note peaks for binary phases InAs, GaAs, and InSb would appear at angles of 0.44, 5.6, and −4.25 degrees, respectively.
(Color online) Evolution of surface morphology with increasing indium concentration for selected samples as measured by AFM:(a) Ga0.95In0.05As0.04Sb0.96, TG=450 C (b) Ga0.80In0.20As0.17Sb0.83, TG = 450 C (c) Ga0.70In0.30As0.26Sb0.74, TG = 450 C (d) Ga0.70In0.30As0.26Sb0.74, TG = 410 C (e) Ga0.52In0.48As0.42Sb0.58, TG = 410 C (f) Ga0.40In0.60 As0.56Sb0.44, TG = 410 C (g) Ga0.10In0.90As0.82Sb0.18, TG = 410 C (h) InAs0.91Sb0.09, TG = 410 C.
(a)–(c) XTEM images for three selected samples, which shows a high degree of homogeneity of the material. Note some shadowing and mill marks occurred on the cross-sections, and was present in the GaSb substrate/buffer as well. EDS confirmed, however, there is no compositional difference between the shading.
(Color online) Comparison of room temperature PL of an unoptimized x = 0.70 Ga1-xInxAsySb1-y and a similar optimized superlattice PL sample.
Strain and RMS roughness corresponding to samples shown in Fig. 9.
Left column shows the target alloy, and the right four columns show the elemental concentration averaged over four equally spaced points distributed across the 2 um layer as determined by EDS in cross-sectioned samples.
Article metrics loading...
Full text loading...