Project topics and materials
Get Complete Project Material File(s) Now! »
Self-assembly of quantum dots with inorganic capping
In the first class of QDs solid, each particle is stabilized in the matrix thanks to the use of surface ligands (see Figure ??). The organic core allows for precise tuning of the discrete elec-tronic states by adjusting precisely the size of the QDs. This size dependence originates from the strength of the quantum confinement. The ligands capping the core are used to control the growth during synthesis, passivate the surface dangling bonds and stabilize the QDs into the array. As mentioned above, one of the first synthesis of single-component self-assembly was reported on CdSe dots by Murray et al. [?] in 1995.
a) Schematic representation of X- and L-type ligands and their exchange for ligands of the same type. b, e) Schematic of solid-state ligand exhange (b) and solution-phase ligand exchange (e). c, d) TEM images of 6 nm PbSe QD films before (c) and after (d) solid-state exchange of long-chain oleate for compact, thiocyanate ligands. Scale bars, 20 nm. f) SEM image of solution-deposited, thiocyanate-exchanged 3.9 nm CdSe QD films. Scale bar,50 nm. Figures taken and adapted from [?]. g) Carrier mobility as function of ligand size in ambipolar PbSe (6.1 nm) QD field-effect transistors. Extracted from [?]. d) Effect of the various ligand size on the carrier mobility length of a 3.9 nm PbSe QD solid. Extracted from [?].
Unfortunately, the long organic ligands have quickly been problematic for charge transport applications. Because of their length and insulating behavior, the long ligands introduce large interdot distance and prevent strong electronic coupling between the neighboring QDs. In-stead the carriers are forced to move trough interdot tunneling across the film. The transport behavior of these materials is well described by variable range hopping [?]. As a result, very low mobilities have been reported (from 10 12 to 10 4 cm2/(Vs)) [?, ?, ?].
In order to tackle this problem, chemical methods have been introduced to exchange the long native ligand with shorter organic [?, ?, ?, ?] and inorganic ligands [?]. The typical ligands exchange takes place when the native ligands are replaced by shorter ligands with head group providing the same type of binding (see Fig. ??a). The ligand exchange can be performed by a ’solid exchange’ technique. It starts with the deposition of the QDs on a substrate in order to form a QD solid. Then the substrate with the QD film is immersed in a new solution where the exchange happens (Fig ?? b). A second method, called the ’solution exchange’ consists in exchanging the long native ligands with more compact ligands in solution (Fig ?? e). This al-lows the deposition of dense and close-packed QD assemblies. The large area coating enabled by the ’solution exchange’ process is of great interest for the integration in solution-deposited QDs devices.
The effect of the decreasing ligand size on the mobility in QD film is shown in Fig ?? (g) where an increase in the mobility over three decades (from 10 4 to 10 1 cm2/(Vs)) is observed for PbSe QD based-FET. Similarly, Fig ?? (h) shows the increase in mobility over 2 decades for similar devices using the time resolved microwave conductivity technique (TRMC). The highest mobility is peaked at 2 cm2/(Vs). Unfortunately the ligand volume lost in the ’solid exchange’ can compromise the order of the QD film and often causes the formation of voids and cracks [?, ?]. The ’solution exchange’ process damages the geometry of the self-assembly leading to a lack of long range order in these samples [?], as shown in Fig.?? f. Thus, band formation only happens when the coupling energy overcomes the energetic and translational inherent disorder of these systems. The decreasing ligand size coupled with the doping of col-loidal QDs used to control the carrier statistics and Fermi energy in a semiconductor has led to significant increase in the devices performances. But in contrast with bulk semiconductor, doping in colloidal semiconductor QDs is rather complicated due to the difficulty to incorpo-rate foreign atoms of different size inside the QD small volume. These difficulties are often attributed to ’self-purification’, an intrinsic mechanism whereby impurities are expelled from the QDs [?, ?].
Throughout the years many studies on group III-V, II-VI, IV-VI semiconductors have been reported using different types of ligands and processes, in order to increase the mobility in QD film. Examples of the highest mobilities reported in the litterature on self-assembled QD solids prepared with organic ligands are listed in Table 1.1. While most studies report mobilities in the range 0.1 cm2/(V.s), the best examples of chemical engineering of the exchange ligand process and doping of QDs have increased the mobilities up to 5 – 27 cm2/(V.s). An exception-ally high mobility, over 400 cm2/(V.s) has been reported by the Talapin group upon sintering colloidal CdSe QDs. Achieving such mobilities approaching the bulk mobilities makes QD solid film viable for optoelectronics applications.
Epitaxially connected quantum dot superlattices
The intense research on overcoming the limitations stated above have mainly focused on ex-changing the native ligands with shorter organic or inorganic substitutive ligands. These ap-proaches, described in the previous section, also led to the discovery of new methods for the creation of QD superlattices where each QD is epitaxially connected through covalent bonds. High angle annular dark field scanning electron transmission (HAADF-STEM) image of the square (A) and (B) honeycomb structure. The equilateral triangle shows the long-range ordering of the structure. Scale bar is 50 nm. Figure taken from [?, ?]. Square lattice based on the 2D self-assembly of colloidal semiconductor QDs. Figure taken and adapted from [?].
In order to synthesize such highly ordered 2D superlattices, a dispersion of QDs in a non-polar solvent (usually toluene) is dropcasted onto a surface of an immiscible liquid, ethylene glycol in this case. Then the toluene is evaporated in a nitrogen purged glovebox. During this evaporation, the QDs concentrate over time and a mono-layer of a square lattice (visible in Figure ??) is formed at the air-liquid interface. The detachment of the capping molecules from specific facets (here the f100g facets) and their dissolution in ethylene glycol drives the crystal attachment. This structure can later be transformed into 2D CdSe lattices through cation(a) A schematic representation of the experimental method used for the synthesis of QD super-lattices. (b) Model showing two mechanisms to trigger the epitaxial connection of the QDs. (c) PbSe superlattices formed with aniline starting from 7.8 nm QDs with partial ligand shell. (d) HRTEM image showing the high degree of epitaxial connections. Figure taken and adapted from [?]
Due to the epitaxial connection of the QDs, a strong interdot coupling in this type of QD solid is expected. For example, band structure of a honeycomb CdSe superlattice predicted the exis-tence of Dirac cones using atomistic tight binding method calculations [?]. The band structure of such graphene-type honeycomb lattice made of CdSe QDs is shown in Figure ?? (A,B) where (A) shows the conduction band and (B) the valence band. The associated theoretical model of the QD honeycomb structure with the QDs connected through their (110) facets is shown in Figure ?? (C). The conduction band is constituted of two envelops with two and six bands. At the K and K’ points, two Dirac cones with a linear dispersion relation similar to the graphene band structure are predicted. Additionally, flat bands are predicted as a result of the absence of hybridization between the 1S and 1P states of the QDs under the effect of the strong quan-tum confinement. Similar features have been predicted in the case of HgTe QDs with the same honeycomb geometry [?]. even; (G) and (H) : odd) composing the neck between two fused QDs. The model for this square superlattice is shown in (D). Figures taken and adapted from [?].
In the case of square superlattices composed of QDs with atomically bonded facets, the the-oretical calculations predict electronic bands structures composed of successive band [?]. This is due to the strong coupling between the wave functions of the nearest-neighbor QDs. In the case of PbSe superlattices, the position in k space of the conduction and valence band edges depends on whether the number of biplanes of atoms in the QDs is odd (Figure ?? (E) and (F)) or even (Figure ?? (G) and (H)).
The coupling of the wavefunction between QDs, and thus the width of the bands, strongly depends on the necking between QDs, i.e., the number of atoms at the QD bonding plane. This is illustrated in Figure ??. The square superlattices where modeled with 2D lattices of <001>-oriented CdSe QDs attached via perpendicular f100g facets. Each QD has the form of a truncated cube (see Figure ?? (D)) with a truncation factor q. In Figure ?? (a) and (d) the band structure for q = 0 (a homogeneous 2D film) is shown and compared with the corresponding electronic structure of a QD superlattice with q = 0.45 (Figure ?? (b) and (e)) as well as that of individual NCs with the same truncation (Figure ?? (c) and (f)). It is clear that the 2D film presents very dispersive bands (because carriers are free to move in two directions) in com-parison to the truncated superlattice. In the latter case, the truncation creates an opening of the band gap induce by the periodic scattering of the electronic waves. Therefore, the band structures predicted by the theoretical calculations should lead to high charge carrier mobility in the epitaxially connected QDs superlattices.
Table of contents :
Introduction
1 Individual and superlattice of quantum dots semiconductor
1.1 The quantum confinement effect
1.1.1 What are QDs ?
1.1.2 Quantum confinement in semiconductor QDs
1.2 Lead Chalcongenide quantum dots
1.3 Quantum mechanical coupling and QD arrays
1.4 Synthesis of colloidal quantum dots
1.4.1 Colloidal nanocrystal synthesis
1.5 Array of coupled nanocrystals
1.5.1 Self-assembly of quantum dots with inorganic capping
1.5.2 Epitaxially connected quantum dot superlattices
1.6 Conclusion
2 Experimental Techniques
2.1 Scanning Tunneling Microscopy
2.1.1 Introduction
2.1.2 The rectangular potential barrier and the transmission coefficient
2.1.3 The tunnel current
2.2 The microscope
2.2.1 The operating principles
2.2.2 Lateral resolution of the STM
2.2.3 The Omicron LT-STM
2.2.4 Scanning Tunneling Spectroscopy (STS)
2.2.5 The spectroscopy technique applied to semiconductor colloidal quantum dots
2.3 MultiProbe Scanning Tunneling Microscopy
2.3.1 Electrical transport measurements
2.3.2 The Nanoprobe, A multiprobe STM
2.4 Conclusion
3 Scanning tunneling spectroscopy of square PbSe QD superlattice on gold substrate
3.1 Scanning Transmission Electron Microscopy (STEM) images of PbSe QDs superlattice
3.2 STM images of PbSe QDs superlattice
3.3 STS measurements on as-synthesized PbSe QD superlattice
3.4 STS characterization of annealed PbSe QD superlattice
3.5 Conclusion
4 Scanning tunneling spectroscopy characterization of PbSeQDsuperlattice deposited on silica substrate
4.1 Experimental details
4.2 STM images of the PbSe QD superlattice
4.3 STS characterization of the superlattice
5 Transport properties of PbSe QDs square superlattice by Multiprobe STM
5.1 Experimental details
5.2 Surface conductivity of PbSe superlattice
5.3 Conductance anisotropy and influence of the cracks in the transport across the PbSe superlattice
5.4 Influence of bigger cracks on transport
5.5 Estimation of the mobility
Conclusion
Bibliography
