Solid State Communications 143 (2007) 213–216 www.elsevier.com/locate/ssc
Charge retention and optical properties of Ge nanocrystals embedded in GeO2 matrix Y. Batra ∗ , D. Kabiraj, D. Kanjilal Inter University Accelerator Centre, Aruna Asaf Ali Marg, New Delhi-110 067, India Received 2 March 2007; accepted 22 May 2007 by C. Tejedor Available online 27 May 2007
Abstract Germanium (Ge) nanocrystals (NCs) have attracted a lot of attention due to their excellent optical properties. In this paper we report on the formation of Ge nanoparticles embedded in GeO2 matrix by electron beam evaporation and subsequent annealing. Charge retention properties of Ge NCs thus synthesized are also investigated. Fourier transform infrared (FTIR) spectroscopic studies are carried out to verify the evolution of the NCs. Micro-Raman analysis also confirms the formation of Ge nanoparticles in the annealed films. Development of Ge nanoparticles is established by photoluminescence (PL) analysis. The memory effect of Ge NCs is revealed by the hysteresis in the capacitance–voltage (C–V ) curves of the fabricated metal-oxide-semiconductor (MOS) structure containing Ge NCs. c 2007 Elsevier Ltd. All rights reserved.
PACS: 72.80.Cw; 73.21.La; 73.40.Qv Keywords: A. Semiconductors; A. Nanocrystals; D. Optical properties; D. Charge retention properties
1. Introduction Group IV semiconductors, silicon (Si) and germanium (Ge), are inefficient emitters of light because of their indirect band gap. But their efficiency can be increased by changing their structure. In nanoparticles the momentum selection rule doesn’t hold good. Indirect transitions in Si and Ge nanoparticles are quite prominent, giving rise to higher luminescence efficiency. Recently, Si and Ge nanoparticles have attracted a lot of research interest. This is because of their excellent optical and electronic properties, which can be utilized for applications in the field of optoelectronic devices [1]. Si and Ge nanoparticles have been extensively investigated as promising candidates for charge storage in nanocrystal gate memories [2–8]. However, Ge is a rather attractive candidate due to its smaller band gap than that of Si and thus faster write and erase speeds [4] . Also, Ge nanoparticles have a larger Bohr exciton radius (∼25.3 nm) than Si nanoparticles (∼4 nm); because of this quantum confinement effects for Ge nanoparticles can be seen
∗ Corresponding author. Tel.: +91 11 26892601; fax: +91 11 26893666.
E-mail address:
[email protected] (Y. Batra). c 2007 Elsevier Ltd. All rights reserved. 0038-1098/$ - see front matter doi:10.1016/j.ssc.2007.05.026
at a much larger size [9]. In fact enhancement in the radiative transition yield for nanoparticles is the result of confinement of carriers in structures with sizes less than the Bohr exciton radius. Moreover the direct band gap (0.88 eV) of Ge is close to the indirect band gap (0.75 eV). So the Ge nanostructures are expected to have a direct band gap semiconductor nature. Embedded Ge nanoparticles have been mainly synthesized by various techniques such as ion implantation [10,11], oxidation of SiGe alloys [12] and a solution synthesis process [13]. Photoluminescence (PL) from Ge quantum dots dispersed in amorphous GeO2 was reported by Gorokhov et al. [14]. The Ge:GeO2 layers were synthesized by chemical vapor deposition of GeO from a germanium monoxide vapor flow followed by decomposition of the deposited material into Ge and GeO2 . It is to be emphasized that various interesting features exhibited by Ge:GeO2 layers in previous studies make them promising objects for study of their optical and electronic properties. But the emphasis so far has been on structural and optical properties of the NCs thus synthesized and electronic properties of the Ge NCs in GeO2 are as yet unexplored.
214
Y. Batra et al. / Solid State Communications 143 (2007) 213–216
Fig. 1. Schematic of the synthesis process of a MOS structure containing Ge NCs.
We present here the memory effect of a metal-insulatorsemiconductor (MIS) structure with Ge NCs embedded in a GeO2 matrix. GeO2 acts as an insulator in this MIS structure. Micro-Raman characterization is done to verify the presence of Ge NCs, and supports the observed memory effect of Ge NCs. 2. Experimental details Prior to deposition, Si substrates were thoroughly cleaned by standard cleaning procedures. GeO2 was deposited on Si substrates by electron beam evaporation under high vacuum conditions. The base pressure of the deposition chamber was ∼10−8 mbar. A pallet formed from GeO2 powder was used as the evaporation material. The films were deposited at room temperature. The thickness of the films was 30 nm and 100 nm, measured by a calibrated quartz crystal monitor during deposition. The 30 nm films were used for fabrication of the MOS structure and the 100 nm films were used for optical characterization. The deposition rate was kept at 0.06 nm/s. After deposition, films were annealed at various temperatures in an Ar atmosphere in a tubular furnace with a ramp rate of 20 ◦ C/min. Aluminum contacts of diameter ∼2 mm were deposited on the trilayer structure through a shadow mask by resistive heating in a high vacuum chamber. Fig. 1 shows the schematic of the fabricated MOS structure containing Ge NCs. All the capacitors thus fabricated were characterized by a HP 4285A C–V analyzer. All measurements were done at room temperature. Fourier transform infrared (FTIR), micro-Raman and photoluminescence (PL) spectroscopy studies were carried out to verify the presence of Ge NCs in the fabricated MIS structure. FTIR measurements were taken with a resolution of 4 cm−1 with a Nexus 670 FTIR system. Micro-Raman analysis was carried out with a Renishaw Raman setup using the 514.5 nm excitation wavelength of an argon ion laser. Measurements were taken with 1 cm−1 resolution. Raman results confirm the presence of Ge nanoparticles in the GeO2 matrix. The excitation wavelength for PL analysis was 325 nm from a He–Cd laser. A Michelle 900 monochromator system with a Si charge coupled detector was used for the measurements.
Fig. 2. FTIR spectra of as-prepared and annealed films of GeO2 .
3. Results and discussion FTIR spectroscopy is a powerful tool for probing the oxygen content present and thus gives information about the stoichiometry of the system under investigation. In a recent report on a FTIR study of GeOx thin films [15], a shift in the frequency of Ge–O–Ge stretching vibration mode was observed [16]. This indicates the increase in oxygen composition of the films. Fig. 2 shows the FTIR spectra of as-prepared and annealed films. As-deposited film shows an intense band at 821 cm−1 which is attributed to the stretching vibration mode of Ge–O–Ge [8]. With annealing, the band shifts towards a higher wavenumber. The intensity of the band increases with increase in annealing temperature. This indicates an increase in the oxygen content of the films. During the evaporation process the electron beam may decompose the GeO2 resulting in deposition of substoichiometric GeO2 thin films. After annealing, the substoichiometric films undergo phase separation resulting in the formation of Ge nanoparticles in GeO2 [15], which is clear from
Y. Batra et al. / Solid State Communications 143 (2007) 213–216
215
Fig. 3. Raman spectra of as-prepared and annealed films of GeO2 .
Fig. 4. PL spectra of as-prepared and annealed films of GeO2 .
the shift of the 821 cm−1 peak towards a higher frequency in the FTIR spectra. GeO2 films were also characterized by micro-Raman spectroscopy. Fig. 3 shows the Raman spectra of the films. It is clear from the spectrum that films annealed at 500 ◦ C show a sharp peak appears at 300 cm−1 which is asymmetrically broadened at lower frequency. This is the signature of the presence of a nanocrystalline Ge phase. For semiconductor nanoparticles, the spatial confinement of phonons in a finite volume is expected to partially relax the wave vector selection rules. This leads to the modification of the bulk Raman peak such as peak frequency shifting, peak broadening and emergence in peak asymmetry. Serincan et al. [17] carried out Raman analysis of the Ge nanoparticles formed by ion implantation of Ge+ ions in a SiO2 matrix and reported that the width and spectral shape of the peak in the case of Ge nanoparticles is different from those of bulk Ge but similar to nanocrystalline Ge. The peak for Ge nanocrystals was down shifted and asymmetrically broadened towards the lower frequency with respect to the bulk Ge peak which was observed around 300 cm−1 . In another report on Raman characterization of GeOx thin films [15], it was seen that after annealing at 500 ◦ C the spectrum shows a broad band at 270 cm−1 , which shows the presence of an amorphous phase of Ge. But for films annealed at 600 ◦ C the amorphous peak disappears and a sharp peak appears at 300 cm−1 , which is asymmetrically broadened at lower frequency. This corresponds to the formation of Ge NCs. However, we also adopt the phonon confinement model [18,19] to estimate the average size of the Ge nanocrystallites. For spherical nanocrystals, the first order Raman spectrum I (w) is R exp(−q 2 d 2 /4a 2 )d3 q I (w) ∝ , (1) {[w0 − w(q)]2 + (Γ /2)2 }
the natural line width (∼3.5 cm−1 ) and w(q) is the dispersion relation for optical phonons in crystalline Ge. For a crystallite size larger than the lattice constant aGe , we may take
where w0 = 300.5 cm−1 ; q is expressed in units of 2π/aGe , with aGe = 0.565 nm being the lattice constant of Ge. Γ is
w(q) = [A + B cos(πq/2)]1/2
(2)
where A = 1.578 × 105 cm−2 and B = 1.000 × 105 cm−2 , the values obtained experimentally from neutron scattering [20]. We can calculate the mean crystallite size of the Ge nanoparticles from Eqs. (1) and (2). For films annealed at 500 ◦ C, it is ∼9 nm. The PL from Si and Ge nanoparticles has been studied extensively in the last two decades. Kanemitsu et al. [21] reported the visible luminescence at around 2.2 eV from Ge nanocrystals in a SiO2 matrix grown by a rf co-sputtering technique. This was explained on the basis of quantum confinement of the charge carriers in the system. Fig. 4 presents PL spectra of the films after annealing at various temperatures, which shows that only the annealed films give an appreciable luminescence signal at around 575 nm (2.2 eV). This is assigned to the evolution of Ge nanoparticles from a sub-stoichiometric oxide of Ge after annealing, which is supported by FTIR as well as the Raman results discussed above. The charge storage effect in the Ge NCs is demonstrated by hysteresis behavior in the high frequency C–V characteristics (Fig. 5). C–V curves were obtained by sweeping the voltage from the accumulation to the inversion region of the fabricated MOS capacitor and then scanning back. When voltage is swept to a large positive value, a number of electrons will be stored in Ge NCs by a direct tunneling process through oxide, resulting in the shift in capacitance. A flat band voltage shift of 0.8 V is observed for the annealed samples. The stored charge density corresponding to a flatband voltage shift is calculated by integration of the C–V curve, and is 4.7 × 1011 cm−2 . As-prepared samples don’t show any appreciable voltage shift. This can be explained on the basis of the absence of Ge NCs in the unannealed samples and evolution of Ge nanoparticles
216
Y. Batra et al. / Solid State Communications 143 (2007) 213–216
financial support in the form of a fellowship. Help received from Mr F. Singh is gratefully acknowledged. References
Fig. 5. High frequency (1 MHz) C–V characteristics of the annealed MOS structure.
from sub-stoichiometric oxide of Ge after annealing, which is supported by FTIR as well as the Raman results discussed above. 4. Conclusions Fabrication of a MIS structure with Ge NCs embedded in a GeO2 matrix was achieved by electron beam evaporation. Formation of Ge NCs was accomplished by annealing of GeO2 thin films deposited on Si substrates. The presence of Ge NCs is verified by FTIR spectroscopy and micro-Raman studies. PL analysis also confirms the evolution of Ge nanoparticles after annealing. With annealing, a phase separation is induced in the films leading to Ge aggregates in the GeO2 matrix. Acknowledgments One of the authors (YB) is thankful to the Council for Scientific and Industrial Research (CSIR), India for providing
[1] K.D. Hirschman, L. Tsybeskov, S.P. Dattagupta, P.M. Fauchet, Nature (London) 384 (1996) 338. [2] S. Tiwari, F. Rana, H. Hanafi, A. Hartstein, E.F. Crabbe, K. Chan, Appl. Phys. Lett. 68 (1996) 1377. [3] Y. Kim, K.H. Park, T.H. Chung, H.J. Bark, J.Y. Yi, W.C. Choi, E.K. Kim, J.W. Lee, J.Y. Lee, Appl. Phys. Lett. 78 (2001) 934. [4] A. Kanjilal, J.L. Hansen, P. Gaiduk, A.N. Larsen, N. Cherkashin, A. Claverie, P. Normand, E. Kapelanakis, D. Skarlatos, D. Tsoukalas, Appl. Phys. Lett. 82 (2003) 1212. [5] S. Tiwari, F. Rana, K. Chan, L. Shi, H. Hanafi, Appl. Phys. Lett. 69 (1996) 1232. [6] T.Z. Lu, M. Alexe, R. Scholz, V. Talalaev, R.J. Zhang, M. Zacharias, J. Appl. Phys. 100 (2006) 014310. [7] C.J. Park, K.H. Cho, W.C. Yang, H.Y. Cho, Suk-Ho Choi, R.G. Elliman, J.H. Han, C. Kim, Appl. Phys. Lett. 88 (2006) 71916. [8] P.F. Lee, X.B. Lu, H.L.W. Chan, E. Jelenkovic, K.Y. Tong, Nanotechnology 17 (2006) 1202. [9] Yoshioto Maeda, Nobuo Tsuamoto, Yoshiaki Yazawa, Yoshihiko Kanemitsu, Yasuaki Masumoto, Appl. Phys. Lett. 64 (1991) 3168. [10] C. Bonafos, M. Carrada, N. Cherkashin, H. Coffin, D. Chassaing, G. Ben Assayag, T. Muller, K.H. Heinig, M. Perego, M. Fanciulli, P. Dimitrakis, P. Normand, J. Appl. Phys. 95 (2005) 5696. [11] I.D. Sharp, Q. Xu, C.Y. Liao, D.O. Yi, J.W. Beeman, Z. Liliental-Weber, K.M. Yu, D.N. Zakharov, J.W. Ager III, D.C. Chrzan, E.E. Hallre, J. Appl. Phys. 97 (2005) 124316. [12] M.I. Ortiz, A. Rodriguez, J. Sangrador, T. Rodriguez, M. Avella, J. Jimenez, C. Ballesteros, Nanotechnology 16 (2005) S197. [13] D. Gerion, N. Zaitseva, C. Saw, M.F. Casula, S. Fakra, T.V. Buuren, G. Galli, Nano Lett. 4 (2004) 597. [14] Eugene B. Gorokhov, Vladimir A. Volodin, Denis V. Marin, Vasily A. Shvets, Andrew G. Borisov, 5th International Siberian Workshop and Tutorial EDM(2004), Session 1, p. 37. [15] M. Ardyanian, H. Rinnert, M. Vergnat, Appl. Phys. Lett. 89 (2006) 11902. [16] H. Rinnert, M. Vergnat, A. Burneau, J. Appl. Phys. 89 (2001) 237. [17] U. Serincan, G. Kartopu, A. Guennes, T.G. Finstad, R. Turan, Y. Ekinci, S.C. Bayliss, Semicond. Sci. Technol. 19 (2004) 247. [18] H. Richter, Z.P. Wang, L. Ley, Solid State Commun. 39 (1981) 625. [19] I.H. Campbell, P.M. Fauchet, Solid State Commun. 58 (1986) 739. [20] R. Tubino, L. Piseri, G. Zerbi, J. Chem. Phys. 56 (1986) 739. [21] Y. Kanemitsu, U. Hiroshi, Y. Masumoto, Y. Maeda, Appl. Phys. Lett. 61 (1992) 2187.