Type of presentation: Poster

MS-1-P-2977 qHRTEM analysis of the (211)B In(Ga)As QDs/GaAs heterostructure

Florini N.¹, Kioseoglou J.¹, Dimitrakopulos G. P.¹, Walther T.², Hatzopoulos Z.³, Pelekanos N. T.³, Kehagias T.¹

¹Physics Department, Aristotle University of Thessaloniki, GR-54124 Thessaloniki, Greece, ²Department of Electronic and Electrical Engineering, University of Sheffield, Mappin St, Sheffield S1 3JD, UK, ³Materials Science & Technology and Physics Departments, University of Crete and IESL/FORTH, GR-71003 Heraklion, Greece

kehagias@auth.gr

The nanostructural properties and strain state of In(Ga)As quantum dots (QDs), embedded in GaAs (211)B by plasma-assisted molecular beam epitaxy (PAMBE), such as shape and dimensions, existence of associated dislocations, thickness of the wetting layer and the possibility of interdiffusion or segregation phenomena in the QDs, were investigated by high-resolution and scanning transmission electron microscopy (HRTEM-STEM) methods. HRTEM imaging showed that the wetting layer thickness does not exceed 2 monolayers. Moreover, the embedded QDs are not associated with any linear defects, suggesting fully strained and optically active nanostructures. However, the shape and dimensions of QDs cannot be precisely extracted, due to the dark strain contrast surrounding the QDs. Conversely, STEM annular bright-field (ABF) imaging revealed that In(Ga)As QDs are elongated along the [-111] direction [Fig. 1(a)].

Quantitative measurements of the local strain in the QDs from HRTEM images have been performed, by the geometrical phase analysis (GPA). GPA is used to determine the strain field in a HRTEM image with respect to a reference region. The GPA lattice strain e_g = (α – α_GaAs)/α_GaAs, α being the in-plane or the out-of-plane strained values of In(Ga)As QDs, is defined relative to the unstrained GaAs underlayer corresponding values, which was taken as reference. GPA measurements using a g/2 mask, showed that the in-plane strain approximates 0, implying a fully registered heterostructure at the interface, as also illustrated in the HRTEM images of two individual QDs in Figs. 1(b) and (c). Assuming a biaxial strain state of the QDs, the corresponding GPA strain surface plots of two QDs along the growth direction are also shown. The quantitative analysis of the In_xGa_(1-x)As QDs on GaAs resulted in chemical composition maps of the investigated QDs. The In content was found to vary from 52% at the base of the QDs to almost 100% at the apex area [Figs. 1(b) and (c)], implying possible Ga segregation in the initial stages of QD growth and formation of an InGaAs alloy. However, since the QDs are entirely embedded in GaAs, possible influence from the matrix cannot be excluded.

Moreover, the samples were analyzed by energy dispersive X-ray (EDX) spectroscopy in order to estimate the chemical composition on the InAs QDs in comparison to the quantitative GPA measurements.

Research co-financed by the European Union (European Social Fund–ESF) and Greek national funds through the Operational Program "Education and Lifelong Learning" of the National Strategic Reference Framework (NSRF)–Research Funding Program: THALES, project NANOPHOS.

Fig. 1: (a) ABF STEM image depicting the embedded InAs QDs. (b) & (c) HRTEM images of embedded InAs QD along the [0-11] zon eaxis and the corresponding GPA strain surface plots along the growth direction.