7 Basic properties of nanoparticles-VI

Dr. Anchal Srivastava

Contents of this Unit

1.  Optical properties of semiconducting nanoparticles

2.Luminescence

2.1 Photoluminescence

2.2 Surface states

2.3 Thermo-luminescence

3.  Summary

Learning Outcomes

• After studying this module, you shall be able to understand
• Effect of size reduction on optical properties of nanoparticles. Luminescence and its types.
• Surface states causing reduction in luminescence intensity.

1. OPTICAL PROPERTIES OF SEMICONDUCTING NANOPARTICLES

Because of their role in quantum dots, nanoparticles made of the elements, which are normal constituents of semiconductors, have been the subject of much study, with particular emphasis on their electronic properties. The title of this section, “optical properties of semiconducting nanoparticles,” is somewhat misleading. Nanoparticles made of cadmium, germanium, or silicon are not themselves semiconductors. A nanoparticle of Si, can be made by laser evaporation of a Si substrate in the region of a helium gas pulse. The beam of neutral clusters is photolyzed by a UV laser producing ionized clusters who’s mass to charge ratio is then measured in a mass spectrometer. The most striking property of nanoparticles made of semiconducting elements is the pronounced changes in their optical properties compared to those of the bulk material. There is a significant shift in the optical absorption spectra toward the blue (shorter wavelength) as the particle size is reduced.

In a bulk semiconductor a bound electron-hole pair, called an exciton, can be produced by a photon having energy greater than that of the band gap of the material. The band gap is the energy separation between the top filled energy level of the valence band and the nearest unfilled level in the conduction band above it. The photon excites an electron from the filled band to the unfilled band above. The result is a hole in the otherwise filled valence band, which corresponds to an electron with an effective positive charge. Because of the Coulomb attraction between the positive hole and the negative electron, a bound pair, called an exciton, is formed that can move through the lattice. The separation between the hole and the electron is many lattice parameters. The existence of the exciton has a strong influence on the electronic properties of the semiconductor and its optical absorption. The exciton can be modeled as a hydrogen-like atom and has energy levels with relative spacing analogous to the energy levels of the hydrogen atom but with lower actual energies. Light-induced transitions between these hydrogen like energy levels produce a series of optical absorptions that can be labeled by the principal quantum numbers of the hydrogen energy levels. Figure 6.1 presents the optical absorption spectra of cuprous oxide (Cu2O), showing the absorption spectra due to the exciton. We are particularly interested in what happens when the size of the nanoparticle becomes smaller than or comparable to the radius of the orbit of the electron-hole pair. There are two situations, called the weak-confinement and the strong-confinement regimes. In the weak regime the particle radius is larger than the radius of the electron-hole pair, but the range of motion of the exciton is limited, which causes a blue shift of the absorption spectrum. When the radius of the particle is smaller than the orbital radius of the electron-hole pair, the motion of the electron and the hole become independent, and the exciton does not exist. The hole and the electron have their own set of energy levels. Here there is also a blue shift, and the emergence of a new set of absorption lines. Figure 6.2 shows the optical absorption spectra of a CdSe nanoparticle at two different sizes measured at 10 K. One can see that the lowest energy absorption region, referred to as the absorption edge, is shifted to higher energy as the particle size decreases. Since the absorption edge is due to the band gap, this means that the band gap increases as particle size decreases. Notice also that the intensity of the absorption increases as the particle size is reduced. The higher energy peaks are associated with the exciton, and they shift to higher energies with the decrease in particle size. These effects are a result of the confinement of the exciton that was discussed above. Essentially, as the particle size is reduced, the hole and the electron are forced closer together, and the separation between the energy levels changes.

Figure 6.1 Optical absorption spectrum of hydrogen like transition of excitons in Cu2O.

Figure 6.2 Optical absorption spectrum of CdSe for two nanoparticles having sizes 20A and 40A.

2. LUMINESCENCE

2.1    Photoluminescence:

The technique of photoluminescence excitation (PLE) has become a standard one for obtaining information on the nature of nanostructures such as quantum dots. In bulk materials the luminescence spectrum often resembles a standard direct absorption spectrum, so there is little advantage to studying the details of both. High photon excitation energies above the band gap can be the most effective for luminescence studies of bulk materials, but it has been found that for the case of nanoparticles the efficiency of luminescence decreases at high incoming photon energies. Nonradiative relaxation pathways can short-circuit the luminescence at these high energies, and it is of interest to investigate the nature of these pathways. Various aspects of luminescence spectroscopy covered in the review by Chen (2000) are examined here.

Figure 6.3 Spectra taken at 10 K for 5.6 nm diameter CdSe quantum dots: (a) absorption spectrum (solid line) and photoluminescence spectrum (dashed line) obtained with excitation at 2.655 eV (467 nm); (b) photoluminescence spectrum obtained with the emission position marked by the downward arrow on the upper plot.

The photoluminescence excitation technique involves scanning the frequency of the excitation signal, and recording the emission within a very narrow spectral range. Figure 6.3 illustrates the technique for the case of -5.6-nm CdSe quantum dot nanoparticles. The solid line in Fig. 6.3a plots the absorption spectrum in the range from 2.0 to 3.1 eV and the superimposed dashed line shows the photoluminescence response that appears near 2.05 eV. The sample was then irradiated with a range of photon energies of 2.13-3.5eY and the luminescence spectrum emitted at the photon energy of 2.13eV is shown plotted in Fig. 6.3b as a function of the excitation energy. The downward-pointing arrow on Fig. 6.3a indicates the position of the detected luminescence. It is clear from a comparison of the absorption and luminescence spectra of this figure that the photoluminescence (b) is much better resolved.

Figure 6.4 Spectra for CdSe nanoparticles of diameter 3.2 nm, showing absorption spectrum (solid line), excitation spectrum for emission at the 2.175 eV band edge fluorescence maximum (dark dashed line), and excitation spectrum for emission at the 1.65 eV deep-trap level (light dashed line).

The excitation spectra of nanoparticles of CdSe with a diameter of 3.2 nm exhibit the expected band-edge emission at 2.176 eV at the temperature 77 K, and they also exhibit an emission signal at 1.65 eV arising from the presence of deep traps. Figure 6.4 compares the PLE spectra for the band-edge and deep-trap emissions with the corresponding absorption, and we see that the band-edge emission is much better resolved. This is because, as is clear from Figure 6.1, each particle size emits light at a characteristic frequency so the PLE spectrum reflects the emission from only a small fraction of the overall particle size distribution. Shallow traps that can be responsible for band-edge emission have the same particle size dependence in their spectral response. This considerably reduces the inhomogeneous broadening, and the result is a narrowed, nearly homogeneous spectrum. The emission originating from the deep traps does not exhibit this same narrowing, which explains the low resolution of the 1.65 eV-emission spectrum of Fig. 6.4.

Figure 6.5 Normalized photoluminescence excitation spectra for seven CdSe quantum dots ranging in size from -1.5 nm (top spectrum) to -4.3 nm (bottom spectrum).

We mentioned above that there is a blue shift, that is, a shift of spectral line positions to higher energies as the size of a nanoparticle decreases. This is dramatically illustrated by the photoluminescence emission spectra presented in Figure 6.5 arising from seven quantum dot samples ranging in size from – 1.5 nm for the top spectrum to -4.3 nm for the bottom spectrum. We see that the band edge gradually shifts to higher energies, and the distances between the individual lines also gradually increase with the decrease in particle size. Another way to vary spectral parameters is to excite the sample with a series of photon energies and record the fluorescence spectrum over a range of energies, and this produces the series of spectra illustrated in Figure 6.6. On this figure the peak of the fluorescence spectrum shifts to higher energies as the excitation photon energy increases. We also notice from the absorption spectrum, presented at the bottom of the figure for comparison purposes that for all photon excitation energies the fluorescence maximum is at lower energies than the direct absorption maximum.

Figure 6.6 Fluorescence spectra for 3.2 nm diameter CdSe nanocrystals for various indicated excitation energies at 77K: (a) experimental and (b) simulated spectra. For comparison purposes, the experimental absorption spectrum is shown at the bottom left.

2.2 Surface states:

As nanoparticles get smaller and smaller, the percentage of atoms on the surface becomes an appreciable fraction of the total number of atoms. For example, we see from Table 1.1 that a 5.7-nm-diameter FCC nanoparticle formed from atoms with a typical diameter of d = 0.3 nm (shell 10 in the table) has 28% of its 2869 atoms on the surface, and a smaller (2.1-nm) nanoparticle (shell 4) has 63% of its 147 atoms on the surface. Irregularities of the surface topology can provide electron and hole traps during optical excitation. The presence of trapped electron-hole pairs bleaches the exciton absorption, but this absorption recovers when the trapped electron-hole pairs decay away. We will describe how this complex process has been studied by time-resolved laser spectroscopy, which furnishes us with details about how the initial excitation energy passes through various intermediate states before finally being dissipated.

The surface states of two nanoparticles of CdS with dimensions of 3.4 and 4.3 nm, respectively, were studied by fluorescence spectroscopy. We see from the resulting spectra presented in Fig. 6.7 that they each exhibit a sharp fluorescence at 435 nm and 480 nm, respectively, arising from excitons, and a broad fluorescence emission at longer wavelengths. We also see from this figure that the addition of nitromethane (CH3NO2) quenches the fluorescence by bringing about a shift toward longer wavelengths, plus an appreciable decrease in the magnitude of the broad band, and in addition it practically eliminates the sharp exciton emission. The temperature dependence of the excitonic and trapped carrier recombination fluorescence bands from CdS nanoparticles both exhibit a decrease in intensity and a shift toward longer wavelengths when the temperature is raised from 4 to 259K, as illustrated in Fig. 6.8. These spectral data suggest that the hole traps lie much deeper (i.e., have much lower energies) than do electron traps.

Figure 6.7 Fluorescence spectra for two samples of CdSe nanoparticles obtained before (solid lines) and after (dashed lines) the addition of 10-3 M nitromethane CH3NO2

Figure 6.8 Fluorescence spectra of CdS nanoparticles recorded at a series of temperatures from 4 to 259 K, using λ = 360 nm excitation.

To try and elucidate the mechanisms involved in the exciton relaxation and the detrapping of electrons, the time dependencies of the excitonic fluorescence and the trapped fluorescence were determined at a series of temperatures from 4 to 269 K, and the results are displayed in Figs. 6.9 and 6.10 respectively. Both types of decay were found to have a complicated multiexponential behavior, with the rates of decay changing as the processes proceeded. The decay time was shortest for the excitonic emission at intermediate temperatures, requiring the time τ1/2 of less than 10 ns for the decay to reach half of its initial intensity at 121 K. In contrast to this, the trapped fluorescence decayed much more slowly, being particularly slow at intermediate temperatures, with the rate constant τ1/2 = l00 nsec at 70K. It had been determined independently that the trapping of electrons in CdS nanoparticles is extremely fast, in the picosecond (ps) time range requiring 10-13s or less time to complete the trapping, so all the trapped electrons are in place before there is an appreciable onset of the fluorescence.

Figure 6.9 Decay curves for exciton fluorescence of CdS nanoparticles.

Figure 6.10 Decay curves for the trapped fluorescence of CdS nanoparticles.

To probe into spectral changes that take place during the initial extremely short picosecond timescale (1000 ps = 1 ns, or 1 ps = l0-3 ns = 10-15 s), the data from decay curves of the type presented in Figs. 6.9 and 6.10 were used to reconstruct luminescence spectra at various times during the early stages of the emission process, and the results are presented in Fig. 6.11. The four spectra at the top of the figure cover the timespan from 0.05 to 1 ns, and they demonstrate that there is a gradual shift of the ~556nm spectral line peak toward longer wavelengths during the first nanosecond of the emission, with the spectral features remaining stable during the remainder of the decay. The initial extremely fast component of the decay, for times less than 0.05 ns, arises from resonant emission, and the subsequent fast component that underwent the wavelength shift Δλ ~2 nm shown in Fig. 6.11 was attributed to longitudinal optical (LO) phonon vibrations.

Figure 6.11 Time resolved CdS luminescence spectra of the 75 cm-1 shift toward longer wavelength of the 556 nm line during the first nanosecond after the onset of the emission. The excitation was at the wavelength I = 549 nm. The intensities of the spectra were adjusted to facilities lineshape comparison.

The model sketched in Fig. 6.12 has been proposed to explain these results. The initial 400 nm laser excitation produces electron-hole pairs that either form free excitons or become trapped at surface states. Some of the free excitons decay rapidly by the emission of a ~ 1.87-eV photon, and others quickly become trapped and then decay almost as rapidly with the emission of a 1.85-eV photon. The electron-hole pairs trapped at surface states decay much more slowly, either radiatively by the emission of photons in the range from 1.77 to 1.83 eV, or nonradiatively. The rapid decays occur over a picosecond timescale and the slower decays over a nanosecond timescale. This model provides a reasonable explanation of the dynamics of the nanoparticle luminescence that we have been discussing.

Figure 6.12 Sketch of a model to explain the luminescence emission from laser generated electron-hole pairs medium sized CdSe nanocrystals.

2.3 Thermoluminescence

Another spectral technique that can provide information on surface states, detrapping, and other processes involved in light emission from nanoparticles is thermoluminescence, the emission of light brought about by heating. Sometimes electron-hole pairs produced by irradiating a sample do not recombine rapidly, but become trapped in separate metastable states with prolonged lifetimes. The presence of traps is especially pronounced in small nanoparticles where a large percentage of the atoms are at the surface, many with unsatisfied chemical bonds and unpaired electrons. Heating the sample excites lattice vibrations that can transfer kinetic energy to electrons and holes held at traps, and thereby release them, with the accompaniment of emitted optical photons that constitute the thermal luminescence. To measure thermoluminescence, the energy needed to bring about the release of electrons and holes from traps is provided by gradually heating the sample and recording the light emission as a function of temperature, as shown in Fig. 6.13 for CdS residing in the cages of the material zeolite-Y, which will be discussed in the next section. The energy corresponding to the maximum emission, called the glow peak, is the energy needed to bring about the detrapping, and it may be considered as a measure of the depth of the trap. This energy, however, is generally insufficient to excite electrons from their ground states to excited states. For example, at room temperature (300 K) the thermal energy kBT= 25.85 meV is far less than typical gap energies Eg, although it is comparable to the ionization energies of many donors and acceptors in semiconductors (see Table 6.1). It is quite common for trap depths to be in the range of thermal energies.

Figure 6.13 Glow curves of CdSe clusters in zeolite-γ for CdS loadings of 1, 3, 5, and 20% (curves 1-4 respectively). Curves 5 is for bulk CdS and curve 6 is for a mechanical mixture of CdS with zeolite-γ powder.

4. SUMMARY

• In this module you study
• Size dependent optical properties of semiconductors  Luminescence. Photoluminescence.
• Surface states affecting the luminescence properties. Luminescence due to heat.

Learn More:

1.W. Chen, “Fluorescence, Thermoluminescence and Photostimulated Luminescence of nanoparticles,” in

Nalwa (2000), Vol. 4, Chapter 5, p.325.

2.H. S. Nalwa, ed. Handbook of Nanosctructures Materials and Nanotechnology, Vol 2, Spectroscopy and

theory; Vol. 4, Optical Properties, Academic Press, San Diego, 2000.