User:LPCNO/Droplet epitaxy

=Introduction=

Electronic properties in semiconductor are well known and the spin physics in ferromagnetic metal had already lead to some applications as head for hard disk, memories. Mixing of these two fields to achieve the spin degree of freedom control in semiconductor near the charge control in semiconductor is a promising way for future “electronic” device. Two other ideas for electronic devices are optoelectronic: increase the integration of optic devices in electronic circuits. Or Valleytronics: the control of an other degree of freedom as the valley index of the carrier as in transition metal dichalcogenide.

Up to now, the spintronic devices (real or only theorical) are based of control of the spin of free carriers. It is easier to transport information with free carriers. But playing with the spin of the nuclei is promising for memories. Indeed the life time of nuclear spin is very long in certain structure as quantum dots (QD).

On the goal of reducing size of semiconductors for increasing the electronic confinement, different steps appear. First produces a 1D confine material as a quantum well after produces a 2D confine as quantum wire then produces a 3D confine as quantum dot.

This article focuses on quantum dots and particularly quantum dots grown by droplet epitaxy (DE)

= Growth=

The growth methods start to be very challenging for controlling size, strain, chemical composition, density of dots, symmetry, shape,….these methods can be separated in two groups chemical ones as chemical vapor deposition or Colloidal synthesis and physical ones as MBE or lithography technique. The lithography techniques are based on patterned dots with ion or electron beam on a thin layer. It leads to big dots (around 100 nm). We focus on MBE techniques. The main method to produce QD is Stranski–Krastanov growth based on strain release after few atomic layers. Unfortunately this technique does not work on substrate with crystallographic axis. Indeed in the axis the strain released is perform by dislocations instead of island formations. Another drawback is the formation of an unavoidable wetting layer which decreased the confinement and prevents fully isolation of dots. A method invented by N. Koguchi and his team around 1990, called droplet epitaxy permits to produce dots without wetting layer and without strain. It is also a MBE method at low temperature (around 300°C) which is compatible with a microelectronic industry line.

This technique can also produce dots with different shape as ring ,hexagon or pyramid.

And dots are promising due their high symmetry on pyramidal shape (C3v in Koster notation) with very small fine structure splitting (FSS) and no heavy-light hole mixing. This properties lead to good emitter of single photons and also an efficient entangled photon source for quantum cryptography or quantum computing.

= Structural Characterization=

Structural characterization can be realized by direct observation of the atomic position with techniques as atomic force microscopy (AFM) or scanning tunneling microscope (STM). This permits a precise determination of shape, size and distribution of these properties on the sample.

With AFM, growers can control the size and the height of dots. We can also determine the density of dots. It is a non destructive technique.

With STM on ultra-vacuum or with high resolution transmission electronic microscope, the resolution is atomic so the crystalline structure, chemical composition and symmetry of dots can be checked. These techniques are destructive.

((Others techniques, more indirect can lead to an roof estimations of the structure of dots as photoluminescence REF or C(V). ))?

= Optical Properties=

The Optical properties presented here are not exclusive of dots grown by DE but more an illustration of what is possible with this kind of dots. Experiments are photoluminescence ones, i.e. excitation and detection are done with light.

Efficient emitters in the visible
The GaAs/AlGaAs quantum dots emit in the visible spectra. The photoluminescence of an ensemble of dots shows an emission between 700 and 800 nm.

Single photon emitters on silicon substrate
Quantum dots grown by DE on GaAs substrate are good single photon emitters and it is not possible to directly deposite GaAs on silicon (Si).

To integrate them on an electronic circuit is needed to be deposited on Si substrate, in order to be compatible with a microelectronic industry line. Historically, it was first deposited on germanium (Ge), more precisely GaAs was deposited on Ge and quantum dots are still good single photon emitters. After, the same group deposited GaAs on Si-on-Ge. The germanium absorbs the lattice mismatch between silicium and germanium. So the bottom part of Ge is full of dislocations but the top part of Ge is clean. The quality of dots is the same as deposited on GaAs.

Entangled photon pairs source
Due to their high symmetry, the FSS of the neutral exciton is very low for some dots (below 20 µeV). If the FSS becomes smaller than the radiative broadening, entangled photon pair can be emitted. In strain free dots this can happened without external tuning. Indeed it is possible to tune for other system, FSS with magnetic field or even electric field.

For future applications no external tuning is easier to implemented but it is less versatile: Only the dots with high symmetry works or when tunning is possible, all dots are candidate for emmision of entengled photons. Some works are in progress for electrical tuning on strain free dots systems.

Another interesting way is the isolation of neutral exciton and biexciton in a charge tuneable device to increase the brightness of the entangled photon source.

Anomalous Hanle effect
Due to the absence of strain in the dot, the nuclear spin is free to move. This leads to anomalous Hanle effect. The polarization of the light not monotonously decreases in function of transverse magnetic field. It exhibits an increase when the field compensated the in-build nuclear magnetic field. Applying at field in a random direction (between transverse and longitudinal can lead to very asymmetric behavior as described in Optical orientation.

Optical nuclear magnetic resonance (NMR)
Quadripolar effects originated from the strain, this is a supplementary complication for performing optically detected nuclear magnetic resonance.

In strain free dots, NMR will theoretically be easier to realize but have not be demonstrate yet.

=References=