Dynamic assembly of CdSe nanoplatelets into superstructures

Get Complete Project Material File(s) Now! »

Shape control

The study of the formation and shape evolution of colloidal semiconducting nanocrystals are of great importance to understand the nucleation and growth process. The nucleation is the first step in the growth of any nanocrystal. Through a fluctuation of the medium in which the nanocrystals are formed, several atoms assemble to a small crystal seed that is thermodynamically stable, and thus does not fall into free atoms or ions.1 Then, growth can be controlled by the kinetics of the reaction or the use of surfactants.

Kinetic control

The control of the growth kinetics of II-VI semiconductors can modify the shape of the particles from a spherical to rod morphology. This also implies that kinetic control can be used to manipulate the average particle size and size distribution. The model of LaMer and Dinegar3 exemplifies the correlation between the solute concentration and different reaction periods. The model includes three periods that start by a supersaturation (I), then a nucleation event (II) followed by a rapid growth (III) and finally by a slower growth by Ostwald ripening (IV) (Fig. 5). Peng et al. further studied the growth of CdSe nanocrystals and observed the shape evolution from dots to rods. They showed that the growth process also develops in two steps, related to the concentration of the monomers: focusing and defocusing.13
The authors reported that chemical potential of elongated nanocrystals is commonly high compared to dot-shaped nanocrystals. This means that the growth of elongated nanocrystals will take place when the chemical potential of the monomers in solution is relatively high, and usually this occurs when the activity of the monomers (monomers concentration) is also high. This means that by simply increasing the monomers concentration in the solution, dot-shaped CdSe nanocrystal were transformed in rod-shaped. When the Cd concentration of the monomer is higher than the solubilities of all the particles, the nanocrystals grow only along a unique axis. And also some variables as ratio and volume of the crystal increase faster which results in a one-dimensional growth stage (1D-growth stage) and the size distribution narrows down. This is the focusing of size distribution, which leads to a close formation of monodisperse colloidal nanocrystals.
But if the Cd monomer concentration is at an intermediate level the crystal grows simultaneously in three dimensions. The ratio stays constant and the crystal volume increases. In addition to the concentration of the monomers, thermodynamic parameters (in addition to temperature) also promote the three-dimensional growth. Indeed, when the monomer concentration drops below the critical point, the length of the rods diminished and their width increased and the size distribution broadens. This means that the anisotropic structure is not thermodynamically favourable as the monomers move away from the rod and grow onto the sides promoting one-dimension to two-dimension intraparticle ripening (1D to 2D ripening). In this process with sufficient time, the quantum rods can progress into dot-shaped. This is the defocusing of size distribution (Ostwald ripening).

Shape control by the surfactants

The performance of surfactants (or surface ligands) has initially been optimized to promote colloidal stability rather than shape and size control. However, surfactants, more than making the nanocrystals dispersible in organic solvents also show extraordinary control over crystal growth at the nanoscale and produce well-defined morphologies.
On the basis of the above facts, to preserve the control of the growth rates, different surfactants need to be used. Peng’s group15,16 further discovered that pure TOPO is not convenient to obtain quantum rods. The presence of pure TOPO induces a faster growth at the high monomer concentration that causes big rod-like particles. The additional components in technical grade TOPO generate slow growth kinetics. To simulate the presence of impurities in technical TOPO (alkyl phosphonic and phosphinic acids – that bind relatively strongly to cadmium), they add a hexyl-phosphonic acids (HPA) that coordinates stronger than TOPO to cadmium. That allows adjusting the growth rate and improves the shape control of the nanocrystals.
Highly hydrophobic surfactants (e.g double tail surfactants) have been demonstrated to be exceptional shape directing agents over crystal growth at nanoscale due to a better surface passivation ability. Double tail surfactants are more hydrophobic and surface active in comparison to the single tail surfactants. When a surfactant is highly hydrophobic as in the case of didodecyl dimethylammonium bromide (DTAB) the formation of cubes with {100} crystal planes is generated. However weaker hydrophobicity (as dimethylene bis-(dodecyldimethylammoniun bromide) preserves hexagons bound with {111} crystal planes.
The parameters cited above allowed different authors to synthesize CdSe QDs,18 Q-rods,19 tetrapods,20 nanoplatelets21 and quantum rings (Fig. 6).22 Among CdSe nanocrystals, we will focus now on nanoplatelets.
Figure 6. TEM micrographs of different CdSe nanocrystals: from left to right, rods, tetrapods and nanoplatelets.

CdSe Nanoplatelets: synthetic aspects

Two-dimensional CdSe semiconductor nanoplatelets or quantum disks have been obtained in two different crystal structures würtzite and zinc blende. The latter has been synthesized by Ithurria and Dubertret in 200821 and is one of the most of interesting kind of nanocrystals thanks to their unique optical properties.
An approach to the synthesis of zinc blende nanoplatelets is in some way the soft template method previously mentioned. The synthesis for the CdSe NPLs relies on fatty acid ligands that bond on the basal planes of the nanoplatelets or quantum disk. To explain, the formation CdSe nanoplatelets, the groups of Peng on the one side and Dubertret, on the other side do not agree on the mechanism of formation of the NPLs and therefore on the best conditions to get them. Dubertret’s group, the initiator of nanoplatelets, established that using cadmium acetate (CdOAc2) or short carboxylates as precursor is necessary to trigger the lateral extension of the NPLs. On the contrary, Li and Peng23 claimed that any fatty acid can be used to get NPLs such as cadmium butanoate (CdBu2), and cadmium octanoate (CdOc2). Nevertheless, both groups concluded that fatty acids are anyway necessary in the synthesis of 2D CdSe nanostructures, either by using these fatty acids as precursors or by adding it during the reaction. In fact, the temperature range of the synthesis is determined by the hydrocarbon chain length of the fatty acids which must be between 140 and 250°C in order to get platelets (Fig.7)..
Among fatty acids that can be used for the synthesis of NPLs, Peng and co-workers tried two fatty acids, stearic acid and decanoic acid (Fig. 8). The results revealed that the up temperature limit increased as the chain length of the fatty acids raised. Therefore, the temperature of apparition of quantum dots occurred early for the dodecanoic acid. The figure 8 show the maximum absorbance of the CdSe nanoplatelets obtained at 180°C for the dodecanoic acid and 240°C for the stearic acid. Li and Peng attributed this effect to the influence of the hydrocarbon chain on the thermal stability of the NPLs, one of the key parameters on the soft-template growth mechanism.
Figure 8. UV-vis of CdSe quantum disks grown at different temperatures and obtained with stearic acid (left) and decanoic acid (rigth). The inset shows the absorbance of the excitonic absorption versus the temperature of the reaction.
It should be pointed out that the lateral dimensions rely upon the concentration of fatty acids, the chain length of fatty acids, and the reaction temperature and the lateral size can be tuned from some nanometers to a few hundred nanometers. It was found that the lower the concentration of the fatty acids the larger the lateral dimension was obtaining. But also increasing the length of fatty acids can lead to the same result. Regarding the effect of the temperature, low temperature resulted in quantum disk of small lateral dimensions.23 The continuous injection of precursors also induces the growth of lateral extension in different nanoparticles such as CdTe and CdSe NPLs. Another variable to control on these NCs is the thickness.
Further studies21 demonstrated that the latter the acetates are added to the synthesis the thicker the NPLs would be formed. In the case of CdS, it was shown that low temperature and shorter alipathic chains induce thinner nanoplatelets. The NPLs formation appears between 130 and 140°C in a period of few minutes, the low temperature limit is possibly determined by the activation of elemental Se in the reaction system. The growth of these ultrathin nanocrystals is achieved at higher temperature ranging from 130°C to 250°C whereas at lower temperature the formation of würzite NPLs take place.24 Higher temperature could eliminate the packing of ligands that is critical for the formation of soft-template process.
Li and Peng also studied the monomers concentration effect on the formation of nanoplatelets. We mentioned in paragraph I.4.1 that in order to get II-IV quantum rods, the concentration of the monomers should be sufficiently high to induce the 1D- growth. However, they didn’t found a relevant difference by varying the precursor concentration. One possible explanation is that the Se powder used for the synthesis was no activated, owed to the low temperature reaction used for analyzing.
They also observed that the Cd and Se precursors ratio should be necessary higher than 1:1 for the growth of CdSe disks. This result is consistent with to the fact that in two-dimensional nanocrystals both the top and the bottom are Cd atoms terminated.
Dubertret’s group24 demonstrated that CdSe nanoplatelets growth begins with the nucleation of ̴ 2 nm diameter nanocrystals seeds. These small nanocrystals would then immediately associate, to form NPLs that gradually expand their lateral dimension (Fig. 9). The formation of nanoplatelets could start by the self-assembly of this well-defined seed to extend laterally (path 1). The self-organization of small cluster that assemble in patches has been observed in ultrathin PbS nanoplatelets with a rock salt crystal structure (path 3). These seeds have two cation-rich facets that are bound to ligands and also can extend laterally thanks to in situ continuous reaction of Cd and Se precursors on the NPLs edges (path 2).
Figure 9. Evolution of the CdSe NPLs lateral extension. It begins with the formation of small seeds. The seeds self-organize to assemble and form the lateral extension (path 1). Lateral dimensions extend by continuous reaction of precursors (path 2).

Optical properties of CdSe nanoplatelets

In recent years, colloidal quasi-two dimensional (2D) II-VI semiconductor nanocrystals (NPLs) such as CdS, CdTe and CdSe has gained growing interest due their original optical properties that empower advanced optoelectronic devices. These flat nanocrystals have electronic lateral dimensions that are much larger than the exciton Bohr radius. But their thickness is similar to those of the ultrathin semiconducting quantum wells (QWs).
Cadmium-based nanoparticles offer26 advantageous features such as a strong oscillator strength and direct bandgaps in the visible region. Thanks to these, many properties of these NPLs such as size, shape, composition, crystal structure or surface ligands can be analyzed by using spectroscopic studies as fluorescence, absorption, photoluminescence excitation and fluorescence lifetime.
Contrary to the würtzite structure the zinc blende NPLs present their thickness along the short axis direction (001). The others two long axis directions correspond to their lateral NPLs planes. The thickness control is quantized to an integer number of monolayers (MLs). Recently, it has been synthesized 3 different populations of nanoplatelets with thickness of 0.9, 1.2 and 1.5 nn. This thickness is characterized by monolayers of CdSe with corresponding numbers of Se layers, which we term 3ML, 4ML, 5ML; maxima of exciton absorption can be found at 463, 513, and 550 nm with lateral dimensions of 60±5 nm by 40±7 nm for 3ML, 27±3 nm by 7±2 nm for 4 ML, 25±3 nm by 10±1 nm for 5ML.
The emission quantum yield (QY) has been reported around 50% for the 5ML, between 55 and 34 % for 4ML an 10% for 3ML. The PL lifetime is much faster in NPLs than in 0D QDs, it can go from few nanoseconds at room temperature to 300 ps at 4 K, which makes the NPls the fastest colloidal emitters until now.
For each population of NPLs we can identify on the absorbance spectrum two transitions: a sharp peak that corresponds to the first excitonic transition electron/heavy hole (lowest energy), and a broader signal corresponding to the electron/light hole transition (highest energy). These two transitions are at 513 nm and 480 nm for the 4 ML (Fig. 10).
Figure 10. Absorption (solid lines) and photoluminescence spectra (dotted lines) emitting at 462nm (3ML), 512nm (4ML), 550 nm (5ML) CdSe nanoplatelets.


Effect of the NPLs thickness and lateral dimensions

The strong quantum confinement in the NPLs is only in the vertical direction, meaning that the only parameter that defines the exciton energy is the thickness of the nanoplatelets.25 The strong quantum confinement, at the same time, is reflected in a extremely narrow intense bands on both absorption and PL spectra with full width at half maximum (fwhm) close to kT tipically between 7 and 10 nm at room temperature and a small Stokes shift.
On the other hand, it was reported that increasing the lateral size of NPLS does not produce any spectral shift on PL. However, growing the lateral dimensions has been shown to decrease the Photoluminescence efficiency (PL-QE) and to greatly accelerate the photoluminescence decay rate (Fig. 11). To explain this observation it was found that a broad part of the population of NPLs present hole traps and the probability to find these defects such as Cd vacancy in the CdSe NPLs increases as the lateral area increases. Therefore, this suggests that in ensemble NPLs where the mean lateral size is larger the nonradiative trap channel is introduced by the defected NPls subpopulation.
Figure 11. Evolution of the photoluminescence quantum efficiency with increasing lateral size.
Dubretret and co-workers concluded that the zinc blende NPLs presents an atomically flat surface.26 And being atomically flat enhances the interaction between ligands decorating their surface compared with highly curved surfaces.

Effect of the ligand

Ligands play an important role in the colloidal nanocrystal semiconductors properties. For exemple, ligands molecules can be employed as precursors in the synthesis of NCs; they can influence the nucleation and growth kinetic as well as the morphology, size control and crystalline structure of the NCs as we have studied before. They also contribute to the colloidal, chemical and photo-stability in several media by modifying the functionality and reactivity of the NCs. Moreover, it determines significantly the physical properties of the NCs. Among these features, studying the interactions between the ligand and the interphase in the 2D semiconductors nanoplatelets is becoming of great interest.
A key advantage of the NPLs compared to the QDs and nanorods is the surface chemistry, while for QDs and nanorods their highly curved surface introduce disorder in the ligand layer, NPLs present a relatively large and atomically flat surface.
Recently the group of Antanovich32, evidenced the impact of surface passivating ligands on the optical and structural properties of zinc blende CdSe nanoplatelets. They observed that upon ligand-exchange of native oleic acid (OA) with hexadecanethiol (HDT) and n-hexadecylphosphonic acid (HDPA) on the surface with different monolayers CdSe nanoplatelets the optical spectra become significantly red-shifted (Fig. 12).
It has been reported for QDs that some capping ligands as phenylchalcogenols Ph-X (X= SH, SeH, TeH) and non-innocent ligands (e.g. phenyldithiocarbamate) can induce large red-shift in PL around 10 -40 nm. Nevertheless, HDPA and HDT are ligands that modify the confinement and exciton transition energy by altering the NCs band gap. They demonstrated that the exciton energy shift was related to structural changes. They exhibited by XRD that the functionalization of the NPLs with organic ligands induces an anisotropic distortion of the unit cell by ligands comprising a contraction of the lateral direction and an expansion in the thickness direction of the ZB unit cell (Fig. 13). This means that cubic symmetry of the ZB CdSe lattice is diminished into a tetragonal symmetry. Since the lattice distortion can be attributed to several elements, the observed lattice strain magnitude was attributed to the head group interaction with the surface Cd atoms of the NPLs. This assumption is in line with several reports that indicate that an increase of ligand coverage could generate the augmentation of repulsive interaction of headgroups, leading to tensile strain. Therefore, ligand induced strain alters the well thickness and hence the exciton confinement.32
Figure 13. Proposed scheme of anisotropic lattice distortion by ligand exchange.

Self-assembly of CdSe nanoplatelets

In 2013, Dubertret and coworkers33 reported the self-organization of CdSe nanoplatelets into stacks to form 1D-superlattice through the addition of an antisolvent in the colloidal nanoplatelets solution. They noted that addition of ethanol could trigger the formation of anisotropic supracrystals. Hassinen and coworkers34 exposed that short-chain alcohols remove and replace the carboxylate ligands at the surface of QDs nanocrystals. Therefore, the addition of ethanol could restrain the steric repulsion imposed by the oleic acid brush making possible the nanoparticles stacking. Another explanation comes from the fact that ethanol is a bad solvent for the aliphatic chain of the carboxylate ligands. In other terms, the contact between aliphatic chains of the oleic acids is energetically favoured when the amount of ethanol is increased. In these supraparticles, the NPL building blocks are oriented with their lateral planes that are perpendicular to the long axis of the column-like assemblies, whose length can contain about 106 individual nanoparticles (Fig. 14).
Interestingly, the light emitted by these microneedles is steadily polarized in the direction perpendicular of the principal direction of the microneedles. This indicate the polarization in the plane of an individual nanoplatelet (Fig 15).
Figure 15. Epifluorescence measurements of stacked NPLs versus the polarization direction.
Another approach to promoting the self-organization of CdSe NPLs is by the slow evaporation of colloidal solution. Abécassis and coworkers used oleic acid followed by a slow drying and then a redispersion to induce the self-assembly into micrometre-long threads. They detected by fluorescence microscopy that these threads can periodically break and restore dynamically. Moreover, free NPL join to both terminations of the existing threads with no formation of new chains. In other terms, the length of threads increases through the addition of NPLs (Fig. 16). For this reason the 1-dimensional superstructures was named living polymers. Two different forces may favour the self-assembly of the system: Van der Waals and depletion interactions. The first one destabilize colloidal solution within time and the second depends linearly on the oleic acid concentration.
Figure 16. Dependence of the threads average length on the number of NPL added.
The gradual process of formation of stack of nanoplatelets was studied by Guzelturc et al.36 They observed that the photoluminescence intensity decreases as the NPLs are formed into stacks upon addition of ethanol (Fig. 17).
They showed that the quenching of the photoluminescence was linked to the existence of exciton migration. According to their explanation, the excitons migrate from one platelet to another within the stacked NPLs assemblies until they come to a non emissive well (Fig. 18). It has been reported that CdSe NCs bearing poorly passivated surfaces sites, crystal and surface defects can outcome in non-emmisive nanocrystals. For instance, hole trapping was associated to Cd vacancies and poor surface passivation in CdSe NCs leading to a non-radiative recombination of the exciton. Non-radiative recombination is when an electron in the conduction band recombines with a hole in the valence band and the excess of energy is emitted in the form of heat.37 As a consequence, the existence of non-emissive platelets in stacked quantum wells can strongly decrease the photoluminescence intensity.

Table of contents :

Chapter I: Semiconducting nanocrystals
I.1 Definition
I. 2 Optical properties of CdSe semiconductor
I. 3 Synthesis
I. 3. 1 Physical methods
I. 3. 2 Chemical methods
I. 4 Shape control
I. 4. 1 Kinetic control
I. 4. 2 Shape control by the surfactants
I. 5 CdSe Nanoplatelets: synthetic aspects
I. 6 Optical properties of CdSe nanoplatelets
I. 6. 1 Effect of the NPLs thickness and lateral dimensions
I. 6. 2 Effect of the ligand
I. 7 Self-assembly of CdSe nanoplatelets
I. 8 References
Chapter II: Dynamic assembly of CdSe nanoplatelets into superstructures
II. 1 Self-assembly
II. 1. 1 Concept
II. 1. 2 Types of assembly
II. 1. 3 Assembly of nanoparticles
II. 1. 4 Motivation of this work
II. 2 Azobenzenes
II. 3 Synthetic strategy
II. 4 First generation of azobenzene ligands: C3, C11 and C18
II. 4. 1 CdSe nanoplatelets purification
II. 4. 2 Ligand exchange
II. 4. 3 Results and discussion
II. 5 Second generation of azobenzene ligands: tBuC3, tBuC11, tBuC18
II. 5. 1 NPLs treatment
II. 5. 2 Ligand exchange
II. 5. 3 Results and discussion
Flores Arias Yesica – doctoral Thesis – 2018
II. 6 Conclusion and outlooks
II. 7 Methods
II. 8 References
Chapter III: Composite materials made of CdSe nanoplatelets and metallophthalocyanines 
III. 1 Phthalocyanines
III. 1. 1 Structures
III. 1. 2 Solubility issues
III. 1. 3 Electronic properties of cobalt phtalocyanine
III. 2 Composite material made of CoPc and CdSe NPLs by addition of a bridging ligand 65
III. 2. 1 Synthetic strategy
III. 2. 2 Results
III. 2. 3 Discussion
III. 3 Composite material made of CoPc and CdSe NPLs by weak interactions
III. 3. 1 Synthetic strategy
III. 3. 2 Results and discussion
III. 4 Conclusion and perspectives
III. 5 Experimental section
III. 6 References


Related Posts