Journal of Non-Crystalline Solids 353 (2007) 550–554 www.elsevier.com/locate/jnoncrysol
Photoluminescence characterization of sol–gel prepared low density silica samples C.M. Carbonaro *, D. Chiriu, R. Corpino, P.C. Ricci, A. Anedda Dipartimento di Fisica, Universita` di Cagliari, Cittadella Universitaria, SP no. 8 Km 0.700, I-09042 Monserrato (CA), Italy Available online 9 February 2007
Abstract By exciting with synchrotron radiation in the 4–10 eV range the emission properties of sol–gel synthesized porous silica at room temperature and at 8 K are investigated. The modifications of the ultraviolet emission in the range 3.7–4.2 eV excited at different energies (5.64, 6.20, 6.53, and 6.89 eV) support the hypothesis of the contribution of two emitting centers. In particular the two emissions are spectrally resolved when exciting at 5.64 eV and a blue shift of the emission peak is reported as the excitation energy increases. The analysis of the decay times indicates that the two centers are temporally resolved with the 5.64 eV excitation both at room temperature and at 8 K and allows to estimate lifetimes of about 3 and 18 ns. When exciting at 6.53 eV, the larger relative contribution of the 4.0 eV band does not allow to efficiently resolve the two decay times and the estimated life time of the 3.8 eV band is shortened. 2007 Elsevier B.V. All rights reserved. PACS: 78.30.Ly; 78.47.+p; 78.55. m; 78.55.Mb Keywords: Optical spectroscopy; Defects; Luminescence; Silica
1. Introduction The investigation of the optical properties of mesoporous silica in the visible and ultraviolet (UV) energy range has recently gathered attention because of the similarities with oxidized porous silicon and silicon nanostructure and because of the possible technological applications in the optoelectronic field for visible and UV emitting silicon/silica based devices [1–3]. Three main different emission bands were reported: a green band (around 2.2–2.4 eV) [2,4–6], a blue band (at about 2.8 eV) and an UV band (in the 3.7–4.2 eV range) [7–9]. Due to the huge surface-to-volume ratio, different models of surface centers have been hypothesized, including H- and C-related centers [2,6,10,11]. Recently, the UV emission has been associated to OH interacting silanol species [8,9] and the contribution *
Corresponding author. Tel.: +39 0706754823; fax: +39 070510171. E-mail address:
[email protected] (C.M. Carbonaro).
0022-3093/$ - see front matter 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.jnoncrysol.2006.10.022
of two different kinds of these surface centers has been suggested in order to explain the reported composite nature of the emission [12,13]. The aim of this work is to proceed further in the analysis of the UV band by investigating the UV emission under different excitation energies in the UV range: the observed modifications of the photoluminescence (PL) spectra support the hypothesis of the contribution of two different centers to the examined emission. Indeed the PL spectrum shows the presence of two spectrally resolved emission bands when the excitation is set at 5.64 eV and displays a blue shift of the peak emission from 3.8 to 4.1 eV as the excitation energy increases. The time decay curves collected at 3.8 and 4.0 eV at different excitation energies also indicate the presence of two contributions. Since the contribution of the two emissions depends on the excitation energy [13], the estimated decay times display a dependence of the resolved life times on the excitation energies.
C.M. Carbonaro et al. / Journal of Non-Crystalline Solids 353 (2007) 550–554
2. Experimental setup
3. Results The PL spectra excited in the UV range (5.64, 6.20, 6.53 and 6.89 eV) are reported in Figs. 1 and 2 at room temperature and at 8 K, respectively. The recorded data show the presence of two main emissions around 2.7 and in the 3.7–4.2 eV range. The analysis of the blue band (2.7 eV), whose relative contribution with respect to the UV band increases at low temperature, has been reported elsewhere and will not be discussed here for the sake of brevity [16,17]. As reported in Table 1, the peak position and the full width at half maximum of the UV band depend on the excitation energy. It is worth to note that both the spectra at RT and LT excited at 5.64 eV display the pres-
600 PL amplitude (arb. units) PL amplitude (arb. units) PL amplitude (arb. units) PL amplitude (arb. units)
Measurements were performed on sol–gel synthesized commercial porous silica monoliths (diameter 5 mm, thickness 2 mm) produced by Geltech Inc. (US) [14]. Investigated samples have pore diameter of 5.5 nm (BET analysis, 5% of standard deviation), specific surface area of 540 m2/g and density of 0.9 g/cm3. Samples were analyzed as prepared, none chemical and/or physical treatment was carried out to remove the water contained within the porous structure. Photoluminescence (PL) and excitation of PL (PLE) measurements were carried out at the SUPERLUMI experimental station on the I beamline of the HASYLAB synchrotron laboratories at Desy (Hamburg) by using the pulsed excitation light of the synchrotron radiation (SR). The PLE measurements were performed in the 4–10 eV energy range with 0.3 nm of bandwidth. Excitation spectra were corrected for the spectral efficiency of the excitation source by using a sodium salicylate standard. The PL signal was dispersed by a Spectra-Pro 300i monochromator and detected in the 1.5–5.0 eV energy range with a 1024 · 1024 LN-CCD (Princeton Instruments) for the PL measurements, ad with a photomultiplier (Hamamatsu 206358P) for the PLE measurements. The spectral emission bandwidth was 11 nm. PL and PLE spectra were recorded under multi-bunch operation and detected with an integral time window of 192 ns correlated to the SR pulses. Decay time in the ns domain were gathered under single-bunch operation, using 1024 channels to scan the 192 ns interval time between adjacent pulses (pulse width of 0.2 ns [15]). A continuous-flow liquid helium cryostat was used to set the temperature of the sample chamber at 8 K (LT). The estimated errors on the luminescence amplitude is ±1%. The uncertainty on the spectral features of the PL (±1% for the peak position and ±2% for the line width) was deduced from the characteristics of the monochromator and the spectral bandwidth. The uncertainty on the life times was estimated by means of a best-fit procedure of the time decay curves (linear correlation factor R > 0.99).
551
500
E
exc
= 5.64 eV
400 300 200 100 0 700
E
exc
= 6.20 eV
600 500 400 300 200 100 0 1200
E
exc
= 6.53 eV
1000 800 600 400 200 0 800
Eexc = 6.89 eV
600 400 200 0 1.5
2
2.5
3 3.5 Energy (eV)
4
4.5
5
Fig. 1. Room temperature PL spectra excited at 5.64, 6.20, 6.53 and 6.89 eV. Solid lines are guides for the eyes.
ence of two spectrally resolved contributions at about 3.8 and 4.1 eV. In addition the PL amplitudes indicate that the 4.1 eV band is better excited at 6.53 eV both at room temperature and 8 K. Figs. 3 and 4 report the time decay curves of the 3.8 and 4.0 eV emission energies excited at 5.64 and 6.53 eV both at room temperature and 8 K. The semi-logarithmic plot evidences the non-single exponential decay of the recorded data. Indeed the experimental decay curves can successfully be fitted with a theoretical curve given by the superimposition of two exponential decays plus a constant background (linear correlation factor R > 0.99). Fitting results are reported in Table 2; the fitting curves are reported in the figures. One can also note that the estimated life times depend on the monitored emission and on the excitation energy. As expected, the whole set of estimated life times increases at 8 K.
552
C.M. Carbonaro et al. / Journal of Non-Crystalline Solids 353 (2007) 550–554
E
400
exc
em
= 5.64 eV
300 200 100 0 280 240 200 160 120 80 40 0
E
exc
= 3.8 eV @ RT
8
PL Amplitude (arb.units)
E
= 6.20 eV
10
E
= 6.53 eV
E
= 5.64 eV
exc
7
10
exc
6
10
5
10
6 4
10
E
em
= 4.0 eV @ RT
5
10
400
E
exc
PL Amplitude (arb.units)
PL amplitude (arb. units) PL amplitude (arb. units) PL amplitude (arb. units) PL amplitude (arb. units)
500
= 6.53 eV
300 200 100
4
E
= 6.53 eV
E
= 5.64 eV
exc
10
exc
1000 100 10
0 E
300
exc
1
= 6.89 eV
0
250 200
20
40
60 80 100 Time (ns)
120
140
Fig. 3. Time decays of the emissions at 3.8 and 4.0 eV excited at 6.53 and 5.64 eV at room temperature. Solid lines are fitting results.
150 100 50 0 1.5
2
2.5
3 3.5 Energy (eV)
4
4.5
5
Fig. 2. Low temperature (8 K) PL spectra excited at 5.64, 6.20, 6.53 and 6.89 eV. Solid lines are guides for the eyes.
Table 1 Peak position (Ep) and full width at half maximum (FWHM) of the UV band reported in Figs. 1 and 2 Eexc (eV)
RT Ep (eV)
FWHM (eV)
Ep (eV)
FWHM (eV)
5.64 6.20 6.53 6.89
3.8–4.0 ± 0.1 4.0 ± 0.1 4.1 ± 0.1 4.1 ± 0.1
0.8 ± 0.1 0.6 ± 0.1 0.6 ± 0.1 0.7 ± 0.1
3.8–4.1 ± 0.1 3.9 ± 0.1 4.1 ± 0.1 4.0 ± 0.1
0.7 ± 0.1 0.7 ± 0.1 0.7 ± 0.1 0.8 ± 0.1
LT
4. Discussion Porous silica properties are related to the surface chemistry of the samples. The surface of sol–gel synthesized PS is generally terminated with OH groups bounded to a silicon atom, SiOH units called silanols [18,19]. The concentration of OH groups at the surface is about 4–5 OH/nm2
and it is found to be almost independent on the synthesis conditions of porous silica [18]. Depending on the hydrogen bonding interaction, different forms of silanol units can be expected: isolated, geminals and vicinals [18,19]. The analysis of the effects of the chemical and physical structure of the porous silica surface on the PL properties suggests the attribution of the reported bands to emitting surface centers [7]. Indeed experimental and theoretical works have indicated OH and H related centers as possible candidates of the PL observed in porous silica and porous silicon [4–6,11,20]. Recently, we have investigated the UV emission recorded in mesoporous silica samples under UV excitation energy (at 5.64 eV). The reported analysis indicated the presence of two different surface centers with peak position at about 3.7 and 3.9 eV and decay times of about 20 and 2 ns, respectively [12,13]. Due to the relation between surface hydration conditions and the intensity of the UV band [8,9], we proposed to associate the two emissions to two different kinds of interacting silanol species. In order to further proceed in the characterization of the emission properties of porous silica samples, we examined the modifications of the UV PL band under different excitation energies in the UV range (at 5.64, 6.20, 6.53 and 6.89 eV). The peak position of the UV band depends on the excitation energy: it decreases from 4.1 to 3.8–3.9 eV
C.M. Carbonaro et al. / Journal of Non-Crystalline Solids 353 (2007) 550–554
10
5
E PL Amplitude (arb.units)
em
10
= 3.8 eV @ LT
4
E
= 6.53 eV
E
= 5.64 eV
exc
1000
exc
100
10
1 E PL Amplitude (arb.units)
em
10
8
10
7
= 4.0 eV @ LT
E
= 5.64 eV
E
= 6.53 eV
exc exc
10
6
10
5
10
4
1000 0
20
40
60 80 Time (ns)
100
120
140
Fig. 4. Time decays of the emissions at 3.8 and 4.0 eV excited at 6.53 and 5.64 eV at low temperature (8 K). Solid lines are fitting results.
Table 2 Time decays of 3.8 and 4.0 eV PL bands excited at 5.64 and 6.53 eV as deduced from best-fit procedure (R > 0.99, see text and Fig. 3) Eexc (eV) 5.64 6.53 5.64 6.53
(RT) (RT) (LT) (LT)
Eem = 3.8 eV
Eem = 4.0 eV
s1 (ns)
s2 (ns)
s1 (ns)
s2 (ns)
3.5 ± 0.1 2.3 ± 0.1 3.5 ± 0.1 2.7 ± 0.1
18.4 ± 0.5 6.3 ± 0.5 20.9 ± 0.5 9.3 ± 0.5
2.2 ± 0.1 3.2 ± 0.1 2.6 ± 0.1 2.9 ± 0.1
6.5 ± 0.5 7.2 ± 0.5 7.8 ± 0.5 8.2 ± 0.5
as the excitation energy decreases. Moreover, it is worth to note that the PL spectra recorded under the 5.64 eV excitation at RT and LT show the presence of two resolved emission peaks. Thus, the shift of the peak position under different excitation energies can be interpreted as due to a different relative contribution of two emissions whose superimposition originates the UV band. Moreover, we can also deduce from the analysis of the peak position that the contribution at higher energy, let us say around 4.0 eV, is more efficiently excited than the contribution at lower energy, around 3.8 eV, when the excitation energy is 6.5 eV or larger, while the vice versa is observed at smaller excitation energies. We point out that the slight differences
553
observed in the peak position of the UV band with respect to previously reported results can be accounted for by a different relative contribution of the two emitting centers due to the slightly different characteristics of the investigated porous samples [12,13]. The analysis of the time decays of the 3.8 and 4.0 eV excited at 5.64 and 6.53 eV supports the previous interpretation: indeed both the experimental data can be successfully fitted with the superimposition of two exponential decays plus a constant background. We note that the fitting results of the 3.8 eV emission excited at 5.64 eV, both at room temperature and 8 K, agree quite well with the previously estimated decay times [12,13], giving two life times of about 3 and 18 ns. These life times were previously attributed to the 4.0 and 3.8 eV emission bands, respectively [12,13]. However, the decay time of the 3.8 eV band (s2 in Table 2) decreases when estimated from the amplitude decay of the 4.0 eV emission, whatever the excitation energy is. The estimated life time decreases from 18 ns to about 7 ns. A comparable decrease is also estimated when monitoring the emission at 3.8 eV excited at 6.53 eV. Different hypothesis can be suggested: (i) the decay time of the 3.8 eV band depends on the excitation energy, calling for a stretched distribution of the emitting centers [21]; (ii) the estimated life time pertains to a third emitting center preferentially excited above 6 eV; (iii) when monitoring the 4.0 eV emission, the relative contribution of the 3.8 eV band is low, even when exciting at 5.64 eV, and the estimated decay time is strongly affected by the larger contribution of the faster 4.0 eV component; in a similar way when monitoring the 3.8 eV excited at 6.53 eV, the relative contribution of the 4.0 eV band is large and strongly affects the estimation of the life time. One should consider that the estimated PL features are averaged values of a distribution of centers. If Gaussian profiles are assumed for the two emissions at 3.8 and 4.0 eV [13], one can observe that the two distribution largely overlap. Indeed, when collecting the amplitude decay at 4.0 eV, also the emission of the centers in the tail of the distribution of the 3.8 eV band is recorded. These centers are expected to be surrounded by different physical environment as compared to the peak of the distribution, in terms of bond length and angles of the centers itself. In this picture, these centers can experience different non-radiative de-excitation pathways and display different life times. Further measurements as a function of the excitation energy and the temperature in a larger range of emission energies all over the UV emission band needed to elucidate this aspect. 5. Conclusions The optical properties of mesoporous silica in the visible and ultraviolet range were investigated. The PL spectra excited at different excitation energies in the UV range support the previously reported hypothesis of the contribution of two different emitting centers. Indeed the PL spectrum shows the presence of two spectrally resolved emission
554
C.M. Carbonaro et al. / Journal of Non-Crystalline Solids 353 (2007) 550–554
bands when the excitation is set at 5.64 eV and displays a blue shift of the peak emission from 3.8 to 4.1 eV as the excitation energy increases. The decay time spectra collected at 3.8 and 4.0 eV at different excitation energies also indicate the presence of two contributions. The reported data confirm the presence of two emitting centers characterized by different emission energies (3.8 and 4.0 eV), different excitation spectra (the 4.0 eV band is better excited above 6.5 eV) and different decay times, of about 18 and 3 ns, even though the observed dependence on the excitation energy of the life time of the 3.8 eV band deserves to be further investigated. Acknowledgements We thank M. Kirm of the G. Zimmerer group and A. Paleari for the SR experimental time at DESY (Hamburg). This work was supported by the European Community – Research Infrastructure Action under the FP6 ‘Structuring the European Research Area’ Programme (through the Integrated Infrastructure Initiative) ‘Integrating Activity on Synchrotron and Free Electron Laser Science’. This study was also supported by a national research project (PRIN2002) of MIUR (Ministero dell’Istruzione, dell’Universita` e della Ricerca). References [1] G.G. Qin, J. Lin, J.Q. Duan, G.Q. Yao, Appl. Phys. Lett. 69 (1996) 1689.
[2] Y.D. Glinka, S.-H. Lin, L.P. Hwang, Y.-T. Chen, Appl. Phys. Lett. 77 (2000) 3968. [3] H.Z. Song, X.M. Bao, N.S. Li, X.L. Wu, Appl. Phys. Lett. 72 (1998) 356. [4] Y.D. Glinka, S.N. Naumenko, V.M. Ogenko, A.A. Chuiko, Opt. Spectrosc. (USSR) 71 (1992) 250. [5] Y.D. Glinka, S.-H. Lin, Y.-T. Chen, Phys. Rev. B 62 (2000) 4733. [6] Y.D. Glinka, S.-H. Lin, Y.-T. Chen, Appl. Phys. Lett. 75 (1999) 778. [7] N. Chiodini, F. Meinardi, F. Morazzoni, A. Paleari, R. Scotti, D. Di Martino, Appl. Phys. Lett. 76 (2000) 3209. [8] B. Yao, H. Shi, X. Zhang, L. Zhang, Appl. Phys. Lett. 78 (2001) 174. [9] A. Anedda, C.M. Carbonaro, F. Clemente, F. Corpino, S. Grandi, P. Mustarelli, A. Magistris, J. Non-Cryst. Solids 322 (2003) 68. [10] A.G. Cullis, L.T. Canham, P.D. Calcott, J. Appl. Phys. 82 (1997) 909. [11] J.L. Gole, D.A. Dixon, Phys. Rev. B 57 (1998) 12002. [12] A. Anedda, C.M. Carbonaro, F. Clemente, R. Corpino, P.C. Ricci, J. Phys. Chem. B 109 (2005) 1239. [13] C.M. Carbonaro, F. Clemente, R. Corpino, P.C. Ricci, A. Anedda, J. Phys. Chem. B 109 (2005) 14441. [14] United State Patent, Patent number 5076980, 1991. [15] G. Zimmerer, Nucl. Instrum. Methods Phys. Res., Sect. A 308 (1991) 178. [16] A. Anedda, C.M. Carbonaro, F. Clemente, R. Corpino, P.C. Ricci, Opt. Mater. 27 (2005) 958. [17] A. Anedda, C.M. Carbonaro, F. Clemente, R. Corpino, P.C. Ricci, Mater. Sci. Eng. C 25 (2005) 631. [18] J. Brinker, W.G. Sherer, Sol–Gel Science: The Physics and Chemistry of Sol–Gel Processing, Academic, San Diego, 1990. [19] L. Klein, Sol–Gel Technology for Thin Films, Fibers, Preforms, Electronics and Specialty Shapes, Noyes, Park Ridge, NJ, 1988. [20] A.S. Zyubin, A.M. Mebel, S.H. Lin, Y.D. Glinka, J. Chem. Phys. 116 (2002) 9889. [21] S. Angello, R. Boscaino, M. Cannas, A. Cannizzo, F.M. Gelardi, S. Grandi, M. Leone, Phys. Rev. B 68 (2003) 165201.
