Investigation of interface abruptness and In content in (In,Ga)N/GaN superlattices

Abstract

We investigate designed InN/GaN superlattices (SLs) grown by plasma-assisted molecular beam epitaxy on c-plane GaN templates in situ by line-of-sight quadrupole mass spectroscopy and laser reflectivity, and ex situ by scanning transmission electron microscopy, X-ray diffraction, and photoluminescence (PL). The structural methods reveal concordantly the different interface abruptness of SLs resulting from growth processes with different parameters. Particularly crucial for the formation of abrupt interfaces is the Ga to N ratio that has to be bigger than 1 during the growth of the GaN barriers, as Ga-excess GaN growth aims at preventing the unintentional incorporation of In accumulated on the growth surface after the supply of InN, that extends the (In,Ga)N quantum well (QW) thickness. Essentially, even with GaN barriers grown under Ga-excess yielding to 1 monolayer (ML) thick QWs, there is a real discrepancy between the designed binary InN and the actual ternary (In,Ga)N ML thick QWs revealed by the above methods. The PL emission line of the sample with atomically abrupt interfaces peaks at 366 nm, which is consistent with the In content measured to be less than 10%.

Substitutional synthesis of sub-nanometer InGaN/GaN quantum wells with high indium content

“…However, the mechanism of In x Ga 1−x N growth from an intermetallic adlayer is in contrast to our observations: in both sample series A and B, the amount of indium present in the adlayer under metal-rich conditions at the beginning of the GaN deposition stage is at least 2 MLs. Furthermore, excess indium inside surface droplets is sufficiently mobile at 610 °C to wet the surface, providing an extra indium reservoir 44 . Therefore, it would be expected that growth from the In-Ga intermetallic adlayer would result in QWs thicker than 1–2 MLs and/or with a diffuse upper interface.…”

Influence of Mg doping on In adsorption and In incorporation in (In,Ga)N superlattices

“…In the first set of in-situ experiments, the Mg, In, and N fluxes were supplied in different orders on a GaN(0001) template while line-of-sight QMS 19 was utilized to measure the desorption signal of 69 Ga, 115 In, 24 Mg, 25 Mg, and 26 Mg from the substrate surface. However, for the 69 Ga, 24 Mg, 25 Mg, and 26 Mg signals monitored by QMS, no change was observed due to the relatively low growth temperature, and the shadowing effect by the 28 the N shutter induced a change in the partial pressure and it yielded an offset in the record of the desorption signal of all species that were supplied or not.…”

“…The second set of experiments consisted of (In,Ga)N/GaN SLs formed with 10 repeats of ( 20MLs InN/ 40 MLs GaN) sequences supplied at 610 °C following the approach described in ref. [19], where Mg was inserted at a different time in the sequence for each sample. Note that former studies have shown that the deposition of thick binary InN QWs at such high growth temperature yield actually QWs with a self-limited thickness of 1 ML of (In,Ga)N and a maximum In content of 0.33 [18][19][20][21] .…”

“…[19], where Mg was inserted at a different time in the sequence for each sample. Note that former studies have shown that the deposition of thick binary InN QWs at such high growth temperature yield actually QWs with a self-limited thickness of 1 ML of (In,Ga)N and a maximum In content of 0.33 [18][19][20][21] . For these experiments, the N, In, and Ga growth rates were fixed to 10.0 nm/min, 8.7 nm/min, and 10.8 nm/min in GaN growth rate units and the Mg flux was set to 6.2 × 10 -9 Torr that corresponds to a nominal concentration of 8 × 10 20 cm -3 ML -1 according to prior SIMS measurement on thick (In,Ga)N layers grown at the same temperature.…”

“…[14][15][16] Yet, although the fabrication of InN/GaN SPSLs was demonstrated by Yoshikawa et al 17 , more recent studies have concluded that (In,Ga)N/GaN SPSLs formed, in which the QWs have a self-limited thickness of one monolayer and a maximum In content of 0.3 instead of 1. [18][19][20][21] To date, the fabrication and the properties of (In,Ga)N/GaN SLs doped with Mg have not been studied. In the literature, some work on p-type conductivity of thick InN:Mg and (In,Ga)N:Mg layers has been reported, [22][23][24][25][26] but little is known concerning the effect of Mg on the In incorporation in (In,Ga)N:Mg and on (In,Ga)N / GaN:Mg SLs.…”