G. Grenet

Instabilities in crystal growth by atomic or molecular beams

Abstract When growing a crystal, a planar front is desired for most of the applications. This plane shape is often destroyed by instabilities of various types. In the case of growth from a condensed phase, the most frequent instabilities are diffusion instabilities , which have been studied in detail by many authors but will be briefly discussed in simple terms in Section 2. The present review is mainly devoted to instabilities which arise in ballistic growth, especially molecular beam epitaxy (MBE). The reasons of the instabilities can be geometric, but they are mostly kinetic (when the desired state cannot be reached because of a lack of time) or thermodynamic (when the desired state is unstable). The kinetic instabilities which will be studied in detail in Sections 4 and 5 result from the fact that adatoms diffusing on a surface do not easily cross steps (Ehrlich–Schwoebel or ES effect). When the growth front is a high symmetry surface, the ES effect produces mounds which often coarsen in time according to power laws. When the growth front is a stepped surface, the ES effect initially produces a meandering of the steps, which eventually may also give rise to mounds. Kinetic instabilities can usually be avoided by raising the temperature, but this favours thermodynamic instabilities of the thermodynamically unstable materials (quantum wells, multilayers …) which are usually prepared by MBE or similar techniques. The attention will be focussed on thermodynamic instabilities which result from slightly different lattice constants a and a +δ a of the substrate and the adsorbate. They can take the following forms. (i) Formation of misfit dislocations, whose geometry, mechanics and kinetics are analysed in detail in Section 8. (ii) Formation of isolated epitaxial clusters which, at least in their earliest form, are ‘coherent’ with the substrate, i.e. dislocation-free (Section 10). (iii) Wavy deformation of the surface, which is presumably the incipient stage of (ii) (Section 9). The theories and the experiments are critically reviewed and their comparison is qualitatively satisfactory although some important questions have not yet received a complete answer. Short chapters are devoted to shadowing instabilities, twinning and stacking faults, as well as the effect of surfactants.

Role of buffer surface morphology and alloying effects on the properties of InAs nanostructures grown on InP(001)

We show the role played by the buffer surface morphology and by alloying effects on the size, shape and lateral distribution of InAs nanostructures grown on InP(001) substrates by molecular beam epitaxy. Three buffers, viz., In0.53Ga0.47As, In0.52Al0.48As, and InP lattice matched on InP have been studied. Differences in nanostructure morphology and in carrier confinement have been evaluated by atomic force microscopy and by low-temperature photoluminescence measurements, respectively. Alongside the classical relaxation mode through two-dimensional/three-dimensional surface morphology change, a chemical relaxation mode has to be introduced as a competitive mode of relaxation of strained layers. This chemical relaxation mode, due to alloying between the InAs deposit and the buffer, is thought to be responsible for most of the observed differences in the InAs nanostructure properties.

Surface effects on shape, self-organization and photoluminescence of InAs islands grown on InAlAs/InP(001)

InAs nanostructures were grown on In0.52Al0.48As alloy lattice matched on InP(001) substrates by molecular beam epitaxy using specific growth parameters in order to improve island self-organization. We show how the change in InAs surface reconstruction via growth temperature from (2×4) to (2×1) and/or the use of InAlAs initial buffer surface treatments improve the island shape homogeneity (either as quantum wires or as quantum dots). Differences in island shape and in carrier confinement are shown by atomic force microscopy and by photoluminescence measurements, respectively. We point out that such shape amendments induce drastic improvements to island size distribution and discernible changes in photoluminescence properties, in particular concerning polarization.

Staggered vertical self-organization of stacked InAs/InAlAs quantum wires on InP(001)

Abstract Using atomic force microscopy (AFM) imaging, transmission electron microscopy (TEM) and photoluminescence (PL), we have studied InAs stacked islands on InP(001) versus the InAlAs spacer layer thickness (SLT). We have found that first wire-like island shape is strongly favored by such a stacking process and second in the 10–25 nm SLT range, the wire size and height are dependent on the SLT. TEM images show off a new surprising staggered vertical island organization that can be explained by the phase separation appearing in the InAlAs spacer layers.

In situ XPS investigation of indium surface segregation for Ga1−xInxAs and Al1−xInxAs alloys grown by MBE on InP(001)

We present an in situ X-Ray photoemission spectroscopy (XPS) investigation of indium surface segregation for As-stabilized Ga1−xInxAs and Al1−xInxAs ternary bulk alloys grown by molecular beam epitaxy on InP( the following conditions: (i) lattice-matched Ga0.47In0.53As and Al0.48In0.52As grown at 300 and 525°C, and (ii) compressive strained and relaxed Ga0.25In0.75As grown at 525°C. Our results show, first, that a chemically shifted component indicative of an InAs surface layer appears within the In 4d core level spectrum. Second, below this InAs top layer, we found a composition gradient which increases with growth temperature and is greater for AlInAs than for GaInAs compounds. Finally, we show that strain seems to have no direct effect on the indium surface segregation at least for Ga0.25In0.75As at 525°C. We analyse and interprete the indium segregation phenomenon as induced by differences in In, Al and Ga surface mobilities during the layer by layer epitaxial growth process.

Surface spinodal decomposition in low temperature Al0.48In0.52As grown on InP(001) by molecular beam epitaxy

Abstract The clustering development for lattice-matched Al0.48In0.52As grown on (001) oriented InP substrates by molecular beam epitaxy (MBE) has been investigated by ex-situ transmission electron microscopy (TEM) and in-situ scanning tunnelling microscopy (STM). For a growth temperature of 450°C, a V/III beam equivalent pressure (BEP) ratio equal to 20 and a growth rate of 1 μm h−1, the clusters are strongly anisotropic: typically, 2 nm along the [110] direction, 30–50 nm along the [ 1 1 ¯ 0 ] direction and 20 nm along the [001] direction. We show theoretically that such a spinodal decomposition would be forbidden if the surface of the deposited film is perfectly flat. We also demonstrate that this decomposition appears to be possible if the surface roughness is sufficient to allow a partial elastic relaxation.

Alloying effects in self-assembled InAs/InP dots

Abstract We have studied the influence of alloying induced chemical exchange reactions on the formation of self-assembled InAs dots prepared on InP and In 0.52 Al 0.48 As buffers lattice-matched to InP(001). Atomic force microscopy and low-temperature photoluminescence measurements have been used to characterize the dot properties. Strong differences in the islanding process are observed depending on the growth conditions and on the nature of the buffer layer. They are associated to variation in intermixing between the InAs deposit and the buffer layer.

Feasibility of strain relaxed InAsP and InGaAs compliant substrates

With a view to investigating the feasibility of using ultrathin films as compliant substrates, we present some preliminary results concerning InAs/sub 0.25/P/sub 0.75/ (0.8% compressively-stressed on InP) film stuck onto a Si host substrate via borophosphorosilicate glass (BPSG). In an attempt to study relaxation mechanisms without any limitation on material viscosity, we also report on how a pseudomorphic In/sub 0.65/Ga/sub 0.35/As layers (0.8% compressively-stressed on InP) elastically relaxes when stuck onto a Si host substrate via a thick Apiezon wax layer. It appears that uniform and flat in-plane elastic relaxation is actually possible only for samples of small areas. For larger samples, there is a competition between undulating and sliding as stress relaxation processes. However, we think that the resulting morphology in terms of undulation periodicity and amplitude is compatible with the use of such layers as seed layers for subsequent epitaxial overgrowths.

Epitaxial growth of SrO on Si(001): Chemical and thermal stability

Heteroepitaxial SrO films grown on Si(001) are characterized by reflection high energy electron diffraction and x-ray photoelectron spectroscopy. Special emphasis is put on the interface chemical, structural, and thermal stability because SrO films can be used as template layers for growing crystalline high-k oxides on Si(001). Ultrathin SrO layers of good crystalline quality with sharp interface with Si(001) can be grown at low temperature (50°C) and low partial oxygen pressure (<10−7Torr). In this case, plastic strain relaxation occurs rapidly at about one-monolayer SrO coverage. At higher temperature (500°C), both strontium and oxygen react with silicon to form a crystalline silicate with a composition close to Sr2SiO4. This silicate is thermodynamically unstable and, when annealed, transforms into a different silicate close to SrSiO3.

Probing deeper by hard x-ray photoelectron spectroscopy

We report an hard x-ray photoelectron spectroscopy method combining high excitation energy (15 keV) and improved modelling of the core-level energy loss features. It provides depth distribution of deeply buried layers with very high sensitivity. We show that a conventional approach relying on intensities of the core-level peaks is unreliable due to intense plasmon losses. We reliably determine the depth distribution of 1 ML La in a high-κ/metal gate stack capped with 50 nm a-Si. The method extends the sensitivity of photoelectron spectroscopy to depths beyond 50 nm.