Femtosecond Transient Absorption Studies in Cadmium Selenide Nanocrystal Thin Films Prepared by Chemical Bath Deposition Method

Dynamics of photo-excited carrier relaxation processes in cadmium selenide nanocrystal thin films prepared by chemical bath deposition method have been studied by nondegenerate femtosecond transient pump-probe spectroscopy. The carriers were generated by exciting at 400 nm laser light and monitored by several other wavelengths. The induced absorption followed by a fast bleach recovery observed near and above the bandgap indicates that the photo-excited carriers (electrons) are first trapped by the available traps and then the trapped electrons absorb the probe light to show a delayed absorption process. The transient decay kinetics was found to be multiexponential in nature. The short time constant, <1 picosecond, was attributed to the trapping of electrons by the surface and/or deep traps and the long time constant, ≥20 picoseconds, was due to the recombination of the trapped carriers. A very little difference in the relaxation processes was observed in the samples prepared at bath temperatures from 25◦C to 60◦C.


INTRODUCTION
Cadmium selenide (CdSe) nanocrystals (NCs) are important materials for optoelectronics, solar cells, and several biological applications [1][2][3][4][5]. Their optical tunability and stability make them ideal candidate for the future optoelectronics devices such as all optical switches and photovoltaics [6][7][8][9]. CdSe NCs are often grown over different substrates in the form of thin films. Several methods are employed for growing these thin films such as chemical deposition (CD) [10][11][12][13], electrochemical deposition (ECD) [14], photochemical deposition (PCD) [15], chemical vapour deposition (CVD) [16,17], and molecular beam epitaxy (MBE) [9] methods. Apart from the form of thin films, nanoparticles forms are also prepared and studied in depth [18][19][20]. The chemical deposition (CD) is the easiest, the cheapest, and the most convenient method, which can produce such materials in a large scale. By changing the deposition parameters, such as concentrations of the individual reagents, bath temperature, and the duration of the deposition time, the NC size and the film thickness can be controlled.
The study of their optical properties is an essential element in the development of above-mentioned devices. Photo-excited carrier relaxation dynamics in these materials have been reported by a number of groups in the recent years. Maly et al. have studied the effect of surface states on the carrier relaxation dynamics in the CdSe NC films in a picosecond time-scale resolution [21]. Simurda et al. have reported the effect of surface passivation by ammonia in the carrier relaxation dynamics in the CdSe NC thin films [22]. Ai et al. have studied charge carrier dynamics in CdSe NC films using a femtosecond degenerate pump-probe technique [23]. Several other groups have reported the carrier dynamics in CdSe quantum structures in various conditions [24][25][26][27][28][29]. However, the carrier relaxation dynamics in the CdSe NC films prepared by the CD method are still to be investigated in a thoroughly manner. In this study, we report the carrier relaxation dynamics in CdSe NC films prepared by the CD method at different temperatures by femtosecond transient nonresonant nondegenerate pump-probe spectroscopy. We probed the carrier relaxation processes at various probe wavelengths by exciting at photon energy of 3.1 eV (400 nm).

2
Journal of Nanomaterials

Sample preparation
CdSe NC films were grown on clean glass substrates by the chemical deposition (CD) method [10][11][12][13]. Briefly, 10 mL 0.5 mol dm −3 CdSO 4 solution was taken in a 40 mL beaker. Ammonia solution was added slowly with a constant stirring until the solution became clearly transparent. A freshly prepared 10 mL Na 2 SeSO 3 solution was added slowly to it. Clean glass slides were inserted into the beaker and inclined at about 20 • angles to its wall. The bath temperature was kept at four different temperatures: 25 • C, 40 • C, 50 • C, and 60 • C. The deposition was carried out for 30 minutes to 3 hours depending on the bath temperatures. One side of the slide was cleaned by a cotton plug dipped in a dilute HCl. The films were air-dried and stored for further study. The orange-coloured as-grown films were characterized by XRD, SEM, EDAX, and optical absorption spectroscopy. The samples were characterized by atomic force microscope (AFM) at ambient conditions.

Femtosecond transient absorption spectrometer
Ultrafast carrier relaxation dynamics study was performed by using a multipass amplified femtosecond Ti:sapphire laser system from Avesta, Russia (1 kHz repetition rate at 800 nm, 50 femtoseconds, 800 μJ/pulse). The details of the setup are described elsewhere [30]. Typically, the 800 nm output pulses from the multipass amplifier are split into two parts to generate pump and probe pulses. One part, with 200 μJ/pulse, is frequency doubled in a BBO crystal to generate pump pulse at 400 nm. The other part is focused onto a 1.5 mm thick sapphire window to generate a white-light continuum for the probe pulses (470 nm to 1000 nm). The pump and probe energies at the sample position were 18 μJ/pulse and 200 nJ/pulse, respectively. The angle between the pump and probe pulses is about 8 • . The pump pulses are set at magic angle (54.7 • ) with respect to the probe pulses, which are horizontally polarised. The pump and probe spot sizes at the sample position were about 400 μm and 200 μm, respectively. The pump-probe setup works in transmission mode. The wavelength of the white light (probe pulses) was tuned by the interference filter just before the photodiodes. The signal collected by the photodiodes is fed to the boxcar, and the data acquisition is done by using the Labview programme in a picosecond. The typical noise in the measured absorbance change is less than about 0.4%.

RESULTS AND DISCUSSION
The optical absorption spectra (Figure 1) of the films exhibited distinct excitonic absorption peaks (E 1 ) at around 540 nm (2.29 eV), 550 nm (2.25 eV), 562 nm (2.21 eV), and 574 nm (2.16 eV) for the 25 • C, 40 • C, 50 • C, and 60 • C samples, respectively. The optical absorption spectra extend beyond 700 nm. The extended absorption beyond 700 nm that arises in the chemical method of preparation can be attributed to the surface states. The absorption peak corresponds to the lowest allowed transition 1S e -1S h [27,[31][32][33]. There is a systematic shift in the excitonic absorption peak towards the higher-wavelength side with the increase in the reaction of bath temperature. This indicates that there is an increase in the grain size of the nanocrystals with temperature. The room temperature bandgap of the bulk CdSe is 1.74 eV. The NC grain size was determined from the energy of quantum confinement (i.e., ΔE = E 1 − E g ) using modified Brus equation, which ranges from 1.9 nm to 2.5 nm for the 25 • C to 60 • C samples, respectively. This indicates that the NC consists of quantum size particles (de Broglie wavelength of the charge carriers = 4.8 nm at room temperature). The grain sizes of the NC films were also calculated from the XRD data using Scherrer's formula and listed in Table 1. The slight deviation in the grain size obtained from two different results could be due to a broad nature of the absorption spectra. The AFM picture of one of the CdSe samples prepared at 40 • C is shown in Figure 2. The picture indicates a nice uniform growth of the nanocrystals of CdSe over the glass substrates. The formation of CdSe nanocrystals prepared by the chemical bath deposition method has been reported by Kale et al. [34] where they have observed a direct evidence of the formation of CdSe nanocrystals from the AFM and SEM data. The thickness of the samples determined from the weight difference method was found to be between 300 and 500 nm. The thickness of the samples as determined from the AFM measurements fairly matches with the previous ones.
The emission from these samples was too weak to be detected by normal steady-state fluorescence spectrometer at room temperature; however, a very week emission could be observed only at higher wavelengths >600 nm for all M. C. Rath et al.  There was no change in the kinetic profiles with the reduction in the pump energy from 18 μJ/pulse to 8 μJ/pulse except that there was a reduction in the signal intensity. The kinetic profiles at various probe wavelengths were fitted with multiexponential fitting parameters by using the Labview fitting programme, and the best-fit results are listed in Table 1.
In this case, the pump pulses photo-excite the valence band (VB) electrons to the conduction band (CB) creating a hot electron density in the CB and holes in the VB. The probe pulses monitor these photo-excited electrons in the CB by a band-to-band absorption process. However, it is true that on photo-excitation, charge carriers are generated which is due to hole in the valence band and electron in the conduction band. So it is quite obvious to think that both the electron and the hole can have transient absorption in the visible region. However, the charge carriers can have different cross-section of transient absorption in different probing regions. Burda et al. [35] carried out ultrafast transient spectroscopy on CdSe nanoparticles monitoring at visible, near IR, and mid-IR regions separately. They have shown clearly that cross-section of electron is more in the visible and mid-IR regions, and on the other hand, the hole has more crosssection in the near-IR region. They have observed that excited dynamics in the visible and mid-IR regions are solely  The pump-probe kinetic profiles obtained at the bandgap probe wavelengths can be modeled to an empirical equation as written below: where k 1 , k 2 , and k 3 are the rate constants for the three processes such as carrier trapping, induced absorption, and carrier recombination, as indicated in Scheme 1. In (1), n is the number density of the photo-excited electrons at the conduction band (CB), m is the inherent free-electron density available at the trap centers, and α is the overall probe absorption in the materials. A model kinetic profile is shown the bottom of Scheme 1. The initial bleach signal is due to the band filling effect for the bandgap probe wavelengths. A bleach signal is not expected for the probe wavelength 900 nm. The time constants for the bleach recovery and the formation of an induced absorption process are found to be less than 1 picosecond and slightly higher than 2 picoseconds, respectively. The fast time constant (<1 picosecond) is due to the trapping of the electrons at the surface traps and/or deep traps. The second component (>2 picoseconds) is due to the absorption of the trapped electrons at the trap centers as represented in Scheme 1. The relative percentages of the observed time constants were calculated from the pre-exponential factors, and the overall percentage was fixed at 100. The relative percentage of the first time constant is found to be always higher than that of the second one. Therefore, the trapping of the electrons and the formation of the absorbing trapped electrons for the further absorption to a higher energy state in the CB are very much comprehensive to each other. The long decay time constants obtained at these probe wavelengths indicate a slow recombination process of the trapped carriers. Similar slow recombination dynamics has been observed by Benjamin et al. [36] in GaAs semiconductor materials, which have attributed to the recombination dynamics due to midgap trap states. The amplitude of the bleach signal was higher than that of the induced absorption, and the residual part of the absorption decay signal is very less as compared to the previous two. Therefore, the samples could be classified as one of the fast optical response materials. There was a slight difference in the time constants obtained in the four different samples. However, the observed time constants follow a systematic pattern following the model given in Scheme 1. The carrier-relaxation pattern in the 40 • C and 50 • C samples seems to have a similar trend, whereas the 25 • C and 60 • C samples found to have a slightly different pattern as compared to the previous two samples.
The results obtained at 900 nm were slightly different from the above. The probe pulse monitors the initially freely available electrons presented in the deep traps and then the trapped electrons populated there. This is because the CdSe nanostructured materials prepared by the CD method have a lot of defect states. As the energy of this probe photon is very less than that of the pump photons as well as that of the excitonic absorption peak position, so there was no bleach signal for this probe wavelength. The decay time constants obtained from kinetic profiles at this probe wavelength indicate the recombination time for the electrons and the holes at the trap centers as well as at the two band structures. The decay time constants match well with the absorption decay time constants obtained for other probe wavelengths. Similarly, there was a slight difference in the time constants obtained in the four different samples. This could be due to the fact that the change in the grain size of around 1 nm within the quantum confinement regime does not make any significant deviations in the photo-excited carrier relaxation processes among these four sets of samples. Hence, it is predicted that these kinds of materials could be useful for the devices based on a fast optical response.

CONCLUSION
The photo-excited carrier relaxation processes in cadmium selenide nanocrystal thin films prepared by chemical deposition method have been studied by nondegenerate femtosecond pump-probe spectroscopic method. The induced absorption followed by a fast bleach recovery within a less than 3-picosecond time scale indicates that the photo-excited electrons are first trapped by the trap sites and then the trapped electrons absorb the probe light to a higher-energy level in the conduction band which exhibits such a delayed absorption process. The short time constant, <1 picosecond, was attributed to the trapping of electrons by the surface and/or deep traps, and the long-time constant, >100 picoseconds, was due to the recombination of the trapped carriers. The amplitude of the bleach signal is very high as compared to that of the residual part of the induced absorption decay, which indicates that the carrier trapping process is a major phenomenon in these materials. The samples obtained at 40 • C and 50 • C were found to have similar carrier relaxation pattern which is different from the other two samples.