Preparation and characterizations of cadmium sulfide nanoparticles

Optik 126 (2015) 1240–1244

Contents lists available at ScienceDirect

Optik journal homepage: www.elsevier.de/ijleo

Preparation and characterizations of cadmium sulfide nanoparticles Feroz A. Mir a,∗ , Indrajit Chattarjee b , Aijaz A. Dar c , K. Asokan d , G.M. Bhat a a

University Science Instrumentation Centre (USIC), University of Kashmir, Srinagar 190006, J&K, India Indian Institute of Science Education and Research, Kolkatta 741252, India c Department of Chemistry, University of Kashmir, Srinagar 190006, J&K, India d Material Science Division, Inter University Accelerator Centre, New Delhi 110067, India b

a r t i c l e

i n f o

Article history: Received 19 February 2014 Accepted 4 March 2015 Keywords: Nanoparticles Bulk Absorption Emission Dielectric

a b s t r a c t Cadmium sulfide nanoparticles have been synthesized by chemical precipitation method. The X-ray diffraction patterns conform their hexagonal structure. Raman spectra show first, second and third order longitudinal optical modes, and are slightly shifted to lower wavenumber side as compared to its bulk. Optical absorption spectra of cadmium sulfide nanoparticles show that band edge is slightly shifted toward longer wavelength side (red shift) and this shift is due to the quantum confinement effect. Particle size calculated by absorption data shows quantum confinement effect. Photoluminescence shows a broad emission with peak centered at around 460 nm after excitation with 360 nm. The frequency dependent dielectric constant of cadmium sulfide nanoparticles was also investigated and normal behavior with applied field is seen. The values of real part of dielectric constant and imaginary part of dielectric constant show dispersion at low frequencies and become almost saturated at higher frequencies. The above characterizations support the quantum dot formation of cadmium sulfide by this low cost technique. © 2015 Elsevier GmbH. All rights reserved.

1. Introduction Semiconductor nanoparticles (NPs) have shown extra ordinary florescence properties and high quantum efficiency [1]. Mostly II–IV semiconductor NPs of dimensions below Bohr diameter exhibit interesting optoelectronic properties due to quantum size effect and are potential candidates for several of applications. In general with surface-area effects, quantum effects start to dominate the properties of matter as size is decreased to the nanoscale. These can have visible impact on the structural, optical, and electrical properties of materials [2]. In principle; the various properties of semiconductor materials can be varied by playing with their size and shapes. Therefore, the current research interest for condensed matter community is the preparation of semiconductor nanostructured with tuned size and shape. Cadmium sulfide (CdS) belongs to the II–VI group, a direct bang gap (Eg = 2.42 eV) semiconductor and is one of the promising materials for use in photoelectric conversion in solar cell [3], thin film transistor (TFT) [4], nonlinear optics [5], semiconductor laser [6] and flat panel display [7]. CdS with a room-temperature (RT) direct band gap of 2.42 eV is extremely photosensitive throughout the entire spectrum from

∗ Corresponding author. Tel.: +91 9018567472; fax: +91 01942424564. E-mail address: [email protected] (F.A. Mir). http://dx.doi.org/10.1016/j.ijleo.2015.03.022 0030-4026/© 2015 Elsevier GmbH. All rights reserved.

infra-red down to ultraviolet, which makes it a potential and attractive semiconductor in the field of optoelectronics. In general, CdS posses three phases, in bulk form it exists in a hexagonal wurtzite-type (W) crystal structure with a = 0.4160 nanometer (nm) and c = 0.6756 nm but in nano crystalline phase it can have a cubical zinc blende (Z) structure and a high-pressure rock salt phase [8]. Their exist various methods for preparations CdS NPs, the solid phase [9], precipitation in the liquid phase [10] and growth in nano-sized micells [11]. However, many techniques for the synthesis of CdS nano-crystals have been developed but a challenge still exists to synthesize CdS nano-crystals in the quantum confinement range. Therefore, here we discuss a simple chemical precipitation technique to synthesize CdS nano particles.

2. Experimental CdS NPs are prepared by precipitation technique from CdSO4 , using thiourea and NH4 OH. All the chemicals are of AR grade from Merck Ltd. In this work, NPs are prepared by mixing aqueous solution of CdSO4 of 0.1 M concentration with aqueous solution of thiourea 0.02 M concentration. The pH of the mixture solution is kept at 10 by adding NH4 OH. The mixture stirred for 1 h. Then this mixture was sonicated (with 50 W power and 40 kHz operational frequency) at room temperature for 1 h. The color of the solution changed to yellow and precipitation occurs. The precipitated

F.A. Mir et al. / Optik 126 (2015) 1240–1244

1241

CdS NPs are filtered, washed, dried and then used for different characterizations. These CdS NPs were characterized by X-ray diffraction (XRD) ˚ using Rigaku Rotating Anode with Cu K␣ radiation ( = 1.5418 A) (H-3R) diffractometer. The surface morphology of the NPs was carried out using high resolution Field Emission Scanning Electron Microscope (FESEM, S-4300, Hitachi, Japan). The UV–vis spectroscopy of the samples was studied by using Shimadzu UV-1650 spectrometer. Photoluminescence (PL) was done Shimadzu spectroflourometer RF-5301. Fourier Transform Infrared (FT-IR) spectrum was recorded on Perkin-Elmer Paragon-1000 spectrophotometer Esquire 3000 spectrometer by KBr pallet technique. Raman measurements were carried out by using Renishaw InVia Raman microscope. The Argon (Ar) laser (514 nm) was used for excitation and the laser power was kept 1 milli Watt (mW), with 20× objective. The dielectric study of CdS NPs was studied in the 40 Hz–1 MHz frequency range using with an Agilent 4285A precision Inductance Capacitance Resistance (LCR) meter. The dielectric constant and dielectric loss were studied by parallel plate geometry method (the pellets of these NPs were polished by silver paint to form the conducting electrodes). All the above characterizations were done at 300 K. 3. Results and discussions Fig. 1 shows the XRD pattern of CdS NPs. Several peaks of ˚ c = 6.713 A˚ (wurtzite-type hexagonal phase with a = b = 4.136 A, structure), have been obtained due to diffraction from (1 0 0), (0 0 2), (1 1 1), (1 0 2), (1 1 0), (1 0 3), (1 1 2) and (2 0 3) planes of CdS. The peaks are well matched with standard PDF card for hexagonal CdS (JCPDS file card 10-454). It is also noticed that the observed peaks are broad and which indicates nano crystalline formation of the system. The lattice parameter has been computed as 5.29 A˚ ˚ which is very close to the standard value (5.42 A). The particle size (L) was calculated from the width of first peak using Debye Scherrer formula [12]. L=

0.9 ˇ cos 

(1)

where L is the coherence length, ˇ is full width at half maximum (FWHM) in radian of the XRD peaks,  is the diffraction angle and  is the wavelength of X-ray used. In addition, the diameter of crystallite (D) is given by 4 D= L 3

(2)

(002)

Intensity(Arbt. units)

(100) (111)

(110) (112) (102)

(103) (203)

20

30

40

50

2θ (Degree) Fig. 1. XRD pattern of CdS NPs.

60

70

Fig. 2. (a and b) SEM and EDS images of CdS NPs respectively.

The average size calculated for samples were approximately 3.15 nm. One can observe the line width become broader due to the decrease in grain size. Broad peaks also reflect the increase in the concentration of lattice imperfection due to the decrease in the internal micro strain within the system [12]. Smaller grain size maximizes the imperfect regions of the material, which is also supported by the smaller strain and dislocation densities. Information on the strain and the particle size was obtained from the FWHM of the diffraction peaks. After applying the correction for instrumental broadening, the FWHM’s can be expressed as a linear combination of the contributions from the strain and particle size through the following Williamson–Hall equation [12]: 4ε sin  1 ˇ cos  = + L  

(3)

where ε is the effective strain. A plot of (ˇ cos )/ versus (4 sin )/ for this film (figure not shown here) is used to calculate the existing strain. The estimated ε turns out around 0.058 which is small but quite significant. Hence force us to predict that there may be some non-uniform strain and departure from uniform shape along the different crystallographic orientations. The surface morphology of the prepared sample can be studied by FESEM. Fig. 2a shows the FESEM image for CdS NPs The surface of every particle is smooth and looks perfectly round shaped or spherical and are uniformly distributed over the entire scanning area. The particle size of CdS NPs as seen in the FESEM micrographs is of the order of few hundred to tens of nanometres. Fig. 2b shows the energy dispersive X-ray analysis (EDAX) of the CdS NPs. The chemical constituents present in the sample according to the EDAX results are Cd = 49.15% and S = 50.85% for CdS. This conforms the purity of the sample. Fig. 3a shows the FTIR spectrum of CdS NPs, which shows stretching bands at 600–700, 800–950, 1000–1265, 1435 cm−1 and 1625 due to C S, C C, C O CH2 and CO2 , respectively. Hydrogen bonded stretching at 3200–3600 cm−1 due to inter molecular

1242

F.A. Mir et al. / Optik 126 (2015) 1240–1244

Fig. 3. (a and b) The FT-IR and Raman spectra of CdS NPs respectively. Fig. 4. (a and b) UV–visible and PL spectra of of CdS NPs respectively.

hydrogen bonds (due to moisture) [13]. The Cd S stretching vibration should be observed at 408 cm−1 [14], this was not found due to large noise and weak signal. Raman spectroscopy is considered to be vibrant tool to probe the secondary phases (or minute changes in basic structure) in various materials which are beyond the detection limit of basic characterizations tools like XRD or neutron diffraction. Fig. 3b shows the room-temperature Raman spectra of CdS NPs. Raman peaks (with decreasing order of intensity) at 300, 601, and 907 cm−1 , respectively, are due to the first, second, and third order longitudinal optical (LO) phonon vibrational modes of the present system. However, in case of bulk, these observed modes at their respective positions appears few cm−1 above than that of modes observed with lower grain size (in nanostructures) [15]. In our study, we observed shift in peak position as well as in asymmetry of the Raman modes toward low wavenumber side and is due to the change in grain size [16]. The variation in crystallite size affects the vibrational properties. Since the grains are in nanometers, it affects the phonon spectra due to confinement of both, optical phonons and acoustic phonons. For bulk crystals, the photon momentum is quite small (p ≈ 0) at Brillouin zone, so light interacts only with phonons having zero momentum. It provides that the wave vector selection rule for first order Raman scattering is (q ¼ 0), where q is the wave vector of scattered phonon [16]. This (q ¼ 0) selection rule is basically a consequence of infinite periodicity of the crystal lattice (because of large dimensions; as surface to volume ratio in nano scale increases). However, if the periodicity of the crystal is disturbed as in the case of nano crystalline materials, this selection rule is relaxed. The energy bandgap plays a key role in determining the optical properties of materials. Fig. 4a shows UV–visible spectra of CdS NPs. The UV–vis spectra show a broad absorption peak without maximum appears in the range of 300–550 nm. The absorption peak at 350 nm is red shifted compared to that of bulk CdS. The red-shifted

absorption edge is due to the quantum confinement of the excitons present in system. The broadening of absorption spectrum is mainly due to the quantum confinement of the CdS NPs. The effect of the quantum confinement depends on the size of the crystallites. As the size of the particles decreases, the degree of confinement and its effect increases. The confinement in a nanocrystalline size permits the energy transfer efficiently to the atom. In order to calculate the band gap of system the following relations could be used [6,17]; (˛h)

n/2

= B(h − Eg )

(4)

where ˛ is the absorption coefficient, B is a constant called the band tailing parameter, h is the incident photon energy, n is an integer, that defines the sort of transition, for n = 1 (direct transition) and n = 4 (indirect transition). The energy magnitude is determined by extrapolation for the linear range of the experimental curve (˛h)n/2 to x-axis. The usual method for determining the value of Eg involves plotting (˛h)n/2 against h (Tauc plot)[8]. In the present study the value of n = 4 has been used to calculate the band gap energy. By extrapolating the straight part of this relation to the h axis (Tauc plot, shown in inset of Fig. 4a), Eg have been determined for CdS NPs. The Eg calculated for CdS NPs is 2.5 eV, which are higher than that of bulk CdS (2.42 eV) due to quantum size effect (QSE) [2,8,14]. Also using the values of OA the CdS NPs size were also determined by the effective mass model approximation (Brus equation [1]), Eg = EBulk +

2  2 2 er 2

 1 me

+

1 mh

 (5)

where Eg is the band gap of the NPs, Ebulk is the band gap of the bulk material, r is the particle radius, me and mh are effective masses of the electrons and holes, respectively, and m0 is the free electron

F.A. Mir et al. / Optik 126 (2015) 1240–1244

mass. Here, me = 0.21m0 , mh = 0.80m0 and εr = 5.6 is the permittivity of the sample [14]. Further, UV–vis absorption data was used to analyze size of CdS NPs by using Eq. (6). Since, the peaks positions can be related with the mean diameter of particles, smaller diameter meaning lower wavelength. Results obtained from this analysis shows that CdS NPs have size around 3.09 nm, and indicates the effect of quantum confinement or (QSE). The QSE in direct-gap semiconductors nanocrystals determines a shift of the optical absorption edge to higher energies with decreasing size, which can explain the UV–vis red-shift effect [9–11]. The average size of the NPs estimated from the optical data analysis are in good agreement with that calculated from the XRD data. The PL of CdS NPs is shown in Fig. 4b. The material was excited at wavelength of 360 nm. The result indicates that the band to band transition of CdS NPs is absent, although a broad emission peak centered at 515 nm (green emission) is observed. It is a well established fact that in luminescence spectra other peaks are seen and are due to the impurities and crystal defects. Generally, there will be some possibilities of transition or recombination in luminescence emission such as (a) band-edge recombination, (b) free excitonic transition, (c) exciton band to neutral donor, (d) exciton band to neutral acceptor, (e) donor to acceptor recombination, (f) excitation from interstitial sulfur to conduction band (green emission), (g) interstitial cadmium to valence band (yellow emission), (h) sulfur vacancy to the valence band (red emission), and (i) excitation from cadmium vacancy to the valence band [17]. The sulfur vacancy and interstitial cadmium are responsible for donor and the acceptor states arise from cadmium vacancy and interstitial sulfur. Whereas the anion and cation vacancies are reported to exist 0.7–0.9 eV below the conduction band and 1.6–1.9 eV above the valence band, respectively [18,19]. Therefore, the presence of sulfur vacancy in the structure of nanoparticles causes a broad emission at the wavelength of 460 nm (near green region). This also shows blue shifted (lower side) compared to its bulk and hence conforms QSE [20,21]. The capacitance is used to calculate the dielectric constant (ε ) by using the following equation: ε =

Ct ε0 A

(6)

where C is the capacitance (F), ε0 is the free space dielectric constant value (8.854 × 10−12 C/m), A is the capacitor area (m2 ) and t is the thickness (m) of the material. The imaginary part of the dielectric constant (ε ) was calculated by using the following relation [22]: ε = ε tan ı

(7)

where tan ı is the dielectric loss. Fig. 5 shows the variation of ε measured as a function of frequency, while the corresponding ε are depicted in inset of Fig. 5. It is observed from the plots (Fig. 5) at the ε decreases exponentially with increasing frequency and then attains almost a constant value in the high frequency region. The value of ε is 12,742 at 1 kHz and 1232 at 1 MHz respectively. It is observed that Debye type ion mechanism is responsible for higher values of dielectric constant at low frequencies, which decreases as the applied frequency increases [21,23]. At relatively lower frequency, the larger is the value of ε . The contributions of all polarizations such as electronic and space charge are predominant in the low frequency region [24,25]. The same trend is observed in the case of ε versus frequency and which is also a characteristic of Debye relaxation mechanism [20,21]. This could be due to the additional contribution of hopping of polarons in the localized state. The observed behavior is due to the large relaxation time. At this juncture the charge carriers will reorient with respect to field.

1243

Fig. 5. The real part of dielectric constant and insect shows the imaginary part of dielectric constant of CdS NPs.

Once the reorientation completes, relaxation time reduces, hence ε decreases [24]. The origins of the dielectric loss are the dipole losses, vibrational losses and conduction losses. These losses are mostly temperature and field dependent [24,25]. The value of ε is 68,404 at 1 kHz and 1222 at 1 MHz respectively. In addition the grain boundary could be one of the reasons for observed dielectric data. Also the wide grain boundaries are characterized by a high density of nonequilibrium defects: point defects and their complexes, dislocations, nanopores in triple joints, etc. As a result, grain boundaries have a disordered structure and enhanced energy [21]. From XRD analysis of these NPs, the dislocation density increases. Therefore, these dislocations, in turn could be responsible for the low-frequency polarization. It is known [1] that II–VI compounds contain dislocations with unsaturated dangling bonds in the core. These dangling bonds are the main source of the electrical activity of dislocations and their charge [24,25]. 4. Conclusions CdS NPs have been synthesized by chemical precipitation techniques. The XRD of NPs conforms hexagonal phase formation. Raman spectra modes at 1LO, 2LO and 3LO and are slightly shifted to lower wavenumber side. Optical band gap is red shift due to the quantum confinement effect. PL also shows a broad emission at around 460 nm. The frequency-dependent dielectric constant of CdS NPs was also investigated and normal behavior with frequency is observed. The successful synthesis and reduction in particle size are also supported by these characterization techniques carried out on these NPs. Acknowledgements One of the authors (FAM) would like to thank UGC India for providing him Dr. D.S. Kothari UGC-PDF to carry out this work. Mr. Shakeel is acknowledged for helping in FT-IR measurements. References [1] L.E. Brus, Acc. Chem. Res. 23 (1990) 183. [2] C. Guozhong, Nanostructures and Nanomaterials, 1st ed., Imperial College Press, London, 2004. [3] R.S. Singh, V.K. Rangari, S. Sanagapalli, V. Jayaraman, S. Mahendra, V.P. Singh, Sol. Energy Mater. Sol. Cells 82 (2004) 315–330. [4] F.Y. Gan, I. Shih, IEEE Trans. Electron. Dev. 49 (2002) 15–18. [5] M.H.M. Ara, Z. Dehghani, I.E. Saievar, Int. J. Nanotechnol. 6 (2009) 1006–1014. [6] X. Duan, Y. Huang, R. Agarwal, C.M. Lieber, Nature (London) 421 (2003) 241–245. [7] S. Schmitt-Rink, D.S. Chemla, D.A.B. Miller, Adv. Phys. 38 (1989) 89–188. [8] R. Banerjee, R. Jayakrishnan, P. Ayyub, J. Phys.: Condens. Matter 12 (2000) 10647–10654.

1244

F.A. Mir et al. / Optik 126 (2015) 1240–1244

[9] W. Wang, Z. Liu, C. Zheng, C. Xu, Y. Liu, G. Wang, Mater. Lett. 57 (2003) 2755–2760. [10] S. Wagesh, Z.S. Ling, X.X. Rong, J. Cryst. Growth 255 (3–4) (2003) 332–337. [11] B. Liu, G.Q. Xu, L.M. Gan, C.H. Chew, W.S. Li, J. Appl. Phys. 89 (2001) 1059–1063. [12] B.D. Cullity, Elements of X-ray Diffraction, Addison-Wesley, 1978. [13] M. Trchova, I. Sedenkova, E.N. Konyushenko, J. Stejskal, P. Holler, G. CiricMarjanovic, J. Phys. Chem. B 110 (2006) 9461. [14] V. Singh, P. Chauhan, J. Phys. Chem. Solids 70 (2009) 1074; J.R.L. Fernandez, M. De Souza-Parise, P.C. Morais, Surf. Sci. 601 (2007) 3805. [15] D.R. Chuu, C.M. Dai, Phys. Rev. B 45 (1992) 11805. [16] A.K. Arora, M. Rajalakshmi, T.R. Ravindran, V. Subramanian, J. Raman Spectrosc. 38 (2007) 604. [17] J.I. Pankove, Optical Processes in Semiconductors, Prentice-Hall, Englewood Cliffs, NJ, 1971.

[18] R.S. Kane, R.E. Cohen, R. Silbey, Langmuir 15 (1999) 39. [19] L. Qi, H. Colfen, M. Antonietti, Nano Lett. 1 (2001) 65. [20] R. Gangopadhyay, A. De, Chem. Mater. 12 (2000) 608; Y. Yang, H. Chen, X. Bao, J. Cryst. Growth 252 (2003) 251. [21] G. Carrot, S.M. Scholz, C.J.G. Plummer, J.G. Hilborn, J.L. Hedrick, Chem. Mater. 11 (1999) 3571. [22] C. Kittel, Introduction to Solid State Physics, 7th ed., Wiley, New York, 1996. [23] A. Kip, Fundamentals of Electricity and Magnetism, 2nd ed., McGraw-Hill, New York, 1969. [24] K.H. Cole, R.H. Cole, J. Chem. Phys. 9 (1941) 341. [25] A.K. Jonscher, Nature 267 (1977) 673.