- Email: [email protected]
Silver nanoclusters formation in ion-exchanged glasses by thermal annealing, UV-laser and X-ray irradiation
ARTICLE IN PRESS
Journal of Crystal Growth 310 (2008) 234–239 www.elsevier.com/locate/jcrysgro
Silver nanoclusters formation in ion-exchanged glasses by thermal annealing, UV-laser and X-ray irradiation Jian Zhanga, Wen Dongb, Jiawei Shenga,, Jingwu Zhenga, Juan Lia, Liang Qiaoa, Liqiang Jianga a
College of Chemical Engineering and Materials Science, Zhejiang University of Technology, Hangzhou, Zhejiang 310014, China b College of Biological Environmental Engineering, Zhejiang University of Technology, Hangzhou, Zhejiang 310014, China Received 13 July 2007; received in revised form 26 September 2007; accepted 5 October 2007 Communicated by M. Schieber Available online 10 October 2007
Abstract The Ag-exchanged commercial soda-lime silicate glasses were treated by three methods: thermal annealing, UV-laser irradiation, and X-ray irradiation, in order to promote the silver nanoclusters formation. Absorption spectrometry and electron spin resonance measurement results indicated that the silver ions transferred to silver atoms after the above three treatments. The silver atoms diffused and then aggregated to become nanoclusters after thermal annealing in air, or after UV-laser irradiation. However, X-ray irradiation, which induced defects and reduction of Ag0 atoms, would not promote the silver nanocluster formation. After annealing at 600 1C for 45 h, the spherical nanoclusters with a diameter of 3–8 nm were formed. The nanoclusters with a diameter of about 2 nm were formed after 30 min UV-laser irradiation without subsequent heating. The surface plasmon resonance peak position of silver nanoclusters changed from 411 nm after thermal annealing to 425 nm after UV-laser irradiation. The peak position shift was due to the nanoclusters size difference. r 2007 Elsevier B.V. All rights reserved. PACS: 81.05.Kf; 61.80.Cb; 78.20.e; 61.46.+w; 76.30.Mi Keywords: A1. Optical microscopy; B1. Glasses; B1. Nanomaterials
1. Introduction Introducing nanosize metal particles such as silver into glass has been used for changing the color of decorative glass and recently for fabricating optical devices. Nonlinear optical materials, such as composites formed by metal nanoclusters in glass, are potentially important in the field of all optical switching technology [1–4]. During the past decades, the physical and in particular the linear and nonlinear optical properties of metallic nanocluster in glass have been investigated using numerous experimental and theoretical approaches. Detailed investigation has been devoted to both the linear and nonlinear properties Corresponding author. Tel.: +86 571 88320851; fax: +86 571 88320142. E-mail address: [email protected] (J. Sheng).
0022-0248/$ - see front matter r 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.jcrysgro.2007.10.007
introduced by optical absorption due to the surface plasmon resonance (SPR). In the past years, silicate glasses containing silver nanoclusters have been the promising materials for nonlinear optical device fabrication. Ion exchange process with thermal treatment is an important method for obtaining silver nanoclusters in a silicate matrix. On the other hand, it is well known that the formation of colloidal silver clusters in glass causes yellow coloration, due to the SPR in the metal particles. Light-ion irradiation and annealing of Ag-exchanged glasses in hydrogen atmosphere (or in a high-vacuum atmosphere) are now widely accepted methods to prepare glass materials containing silver nanoclusters, mainly because they are easy to operate and commercially viable for technological applications in optoelectronic devices [4–7]. Although a substantial amount of literature is available on properties of silver
ARTICLE IN PRESS J. Zhang et al. / Journal of Crystal Growth 310 (2008) 234–239
235
nanoclusters in glass exhibiting high optical nonlinear response [8–12], only minimal work has been reported on the optical spectroscopic investigations of silver ionexchanged glass thermally annealing in air, or treated by UV-laser and X-ray irradiation. At present, little is known about the aggregation behavior and Ag+ chemical environment. Recently, in different context, we have reported the X-ray irradiation-induced defects in soda-lime silicate glass [13,14], the metallic silver nanocluster formation and dissolution in silver-doped glass by X-ray irradiation followed by thermal annealing [15]. We also reported the UV-laser induced silver nanoclusters in silver ionexchanged glass [16]. In this paper, we studied the silver nanoclusters formation in the ion-exchanged glasses. We described the silver nanoclusters formation in commercial flat soda-lime silicate glasses using optical absorption techniques and TEM measurements in an attempt to understand the formation mechanism of silver nanoclusters during annealing. Three treatments, involving thermal annealing, UV-laser annealing, and X-ray irradiation annealing, have been included in the present work. Once the formation mechanism of the silver nanocluster is understood, the information about the factors affecting the nucleation and growth process of silver nanoclusters will be known, and consequentially the optimum thermally processed can be achieved.
l ¼ 0.061 nm, 50 kV, 50 mA) for 30 min. Optical absorption spectra were recorded using a Lambda 9 Perkin-Elmer spectrophotometer at ambient room temperature and all recorded optical spectra were referenced to air. TEM images for representative samples were observed by a transmittance electron microscope (JEOL-2010 TEM) at electron beam energy of 300 keV. The first-derivative electron spin resonance (ESR) spectra were recorded at room temperature on a BRUKER 300E, operating at 9.7 GHz. First-derivative ESR absorption spectra were recorded at room temperature on a BRUKER 300E, operating at 9.7 GHz. DPPH (1,1-diphenyl-2-piterylhydrozyl) was used to adjust the g-values, in which, the g-value was defined by the relation hv ¼ gbH, where h is the Planck’s constant, n the spectrometer frequency, b the Bohr magneton, and H is the magnitude of the laboratory applied magnetic field at resonance. The g-values and characteristics of ESR spectra associated with the irradiation-induced defects and silver nanoclusters. The Ag0 peaks have two distinct groups in ESR spectra, due to the 107Ag0 and 109Ag0. The formation of Ag++ could be found in the vicinity of g ¼ 2.056. The most fundamental radiationinduced defects in glasses are the nonbridging oxygen hole center, which show a broad derivative peak in the vicinity of g ¼ 1.99–2.00.
2. Experimental procedure
3.1. Silver nanoclusters formation during thermal annealing
Commercially available soda-lime silicate glass substrates composed of (wt%): 74.2 SiO2, 14.3 Na2O, 1.9 Al2O3, 8.1 CaO, and 1.5 MgO were utilized for experimentation. Samples (2 mm thick and of approximately 15 15 mm2), which had previously been scoured and dried, were preheated and subsequently dipped in a molten salt bath formed by a mixture of 98 mol% NaNO3 and 2 mol% AgNO3 in a crucible of Al2O3. The ion exchanges took place at a temperature of 320 1C with a processing time of 10 min. To ensure that the glass structure did not change during testing, the ion-exchange temperature was kept well below the glass-transition temperature of the specimen. After inter-diffusion, samples were removed from the molten bath and washed with several portions of distilled water and acetone to remove any silver nitrate adhering to their surface. To promote the silver nanocluster formation, the Agexchanged samples were treated by three methods: thermal annealing, UV-laser irradiation, and X-ray irradiation. Thermal annealing was performed in air for 2–45 h at temperatures in the range 500–600 1C. Laser irradiation was performed with an ArF UV-laser (193 nm), utilizing at an output energy density of 30 mJ/cm2/pulse, with a repetition frequency of 10 Hz and a 20 ns pulse duration. The UV-laser irradiation time ranged from 5 s to 30 min. The X-ray irradiation was carried out using a X-ray Spectrometer (ThermoARL) at room temperature (Rh Ka,
Prior to the ion-exchange process, the glass samples were colorless and had no measurable absorption in the visible region. The soda-lime silicate glasses did not change color after Ag-exchanging for 10 min. No significant SPR of the silver nanoclusters (400–430 nm) was observed, indicating that the silver aggregate formation did not occur, or the silver nanoclusters were less than 1 nm in size during the 10 min ion-exchange. The formation of silver nanoclusters is limited, primarily, to the mobility of silver and the concentration of reducing elements present in the glass. Under our experimental procedures, the silver mobility was minimal and the concentration of reducing agent was relatively low. Fig. 1 shows the absorption spectra for Ag-exchanged glasses after thermal annealing in air. No obvious absorption band was developed after heating at 550 1C for 2 h. An absorption band at approximately 410 nm was clearly observed after heating at 600 1C for more than 2 h, which was apparently due to the SPR band of silver nanoclusters formed in the glass matrix. When the annealing time increased from 2 to 6 h, a small blue shift of the SPR was observed, indicating that the refractive index property of the matrix or the size of silver nanoclusters was annealing time-dependent (Fig. 1(a)). The intensity of absorption at 411 nm increased significantly while the peak position did not change with the increase of annealing time from 6 to 45 h (Fig. 1(b)).
3. Results and discussion
ARTICLE IN PRESS J. Zhang et al. / Journal of Crystal Growth 310 (2008) 234–239
236
5 Blank As-exchanged 550°C-2h 600°C-2h 600°C-4h 600°C-6h
Absorption, cm-1
4
3
2
1
0 300
350
400 450 500 Wavelength, nm
550
Absorption, cm-1
8
600
650
600°C-8h 600°C-12h 600°C-15h 600°C-35h 600°C-45h
6
Fig. 2. TEM image of Ag-exchanged glass after thermal annealing at 600 1C for 45 h.
4
2
0 300
350
400 450 500 Wavelength, nm
550
600
650
Fig. 1. Optical absorption spectra of silver nanoclusters embedded in Agexchanged glass after thermal annealing.
The width of the absorption band reduced systematically with the increase of annealing time, indicating an increase of volume fraction of silver nanoclusters in the glass. From reported data, the substrate temperature plays a fundamental role in the formation of silver precipitates in Ag-exchanged glass. To form silver nanoclusters in Agexchanged glasses with thermal annealing, the temperature was set much lower in reducting atmospheres than in air. When treated in air, the nanocluster formation is limited due to low mobility of silver and a lower level of reducing elements present in the glass. Therefore a higher annealing temperature is desirable. [17]. The average size of silver nanoclusters, d, can be determined by means of the equation [2,18,19]: d¼
2vf , Do
(1)
where vf=1.39 108 cm/s is the Fermi velocity for bulk silver and Do is the bandwidth at half maximum (FWHM) of the SPR band. The FWHM is determined by assuming
the absorption peak as a normal Gaussian distribution and fitting the flank of the absorption band on the long wavelength side. The value of Do is determined using Do=2pc(1/l1–1/l2), where cis the speed of light in vacuum, l1 and l2 are the wavelength at FWHM. After annealing at 600 1C for 45 h, the calculated average size was around 7 nm. The TEM image (Fig. 2) shows the presence of spherical nanoclusters with a diameter of 3–8 nm. In the Ag-exchanged glass, silver atoms are expected to be bound to nonbridging oxygen (NBO) atoms, after the substitution for the sodium atoms in the glass matrix. The silver introduced by ion-exchange process is mostly comprised of Ag+, with a minor population of Ag0 atoms [5]. When the Ag-exchanged glass was subjected to thermal annealing, a reduction of silver atoms was induced after capturing the electron from glass structure or impurities. During the thermal annealing, more Ag–O bonds are broken and more atoms are formed, thus the neutral silver becomes the dominant state. On the other hand, diffusion of silver towards the surface, with consequential precipitation, are caused by thermal relaxation of the surface tensile stress introduced by the size differential between Ag+ and Na+ during the ion exchange process. Tensile stress introduced by this size difference is still present after cooling [5,20]. The activated Ag+ ions mobilize toward a more relaxed surface, resulting from precipitation or nanocluster formation of silver atoms. Conversely, the dissociation of Ag–O bonds to form Si–O and Ag–Ag bonds result in a net loss in the system free energy [5]. Thus, the silver inclusion of the glass network will move toward the surface and form nanoclusters in order to maintain a minimum energy state within the system. Since the diffusion of silver in the glass matrix is nominal, the
ARTICLE IN PRESS J. Zhang et al. / Journal of Crystal Growth 310 (2008) 234–239
237
above-discussed processes of silver nanocluster formation are commonly carried out through heat treatment. 3.2. UV-laser-induced silver nanoclusters
8
600°C - 45 h UV - laser 30 min
6
4
2
0 300
350
400
450 500 Wavelength, nm
550
600
650
Fig. 5. Optical absorption spectra of Ag-exchanged glasses after 30 min UV-laser irradiation and 600 1C heating for 45 h, respectively.
8 Blank As ion-exchanged 5 sec irradiation 1 min irradiation 10 min irradiation 30 min irradiation
6 Absorption (cm-1)
Fig. 4. TEM image of Ag-exchanged glass after 30 min UV-laser irradiation.
Absorption, cm-1
Effects of laser irradiation were clearly visible with the glass exemplifying a yellow color after laser irradiation. Moreover, the color density increased with increasing irradiation time. Fig. 3 shows the absorption spectra for Ag-exchanged glasses after laser irradiations. An absorption band at approximately 425 nm was clearly observed, which was due to the SPR band of silver nanoclusters. A small blue shift of the SPR is visible in Fig. 3 for the longest irradiation time. The laser irradiation promotes nanocluster formation, and the intensity of the SPR band sharply increases with the irradiation time. For the measured FWHM of Fig. 3, the calculated size, using Eq. (1), were about 1.0 nm after 5 s irradiation and about 2.0 nm after 30 min irradiation, respectively. The average nanocluster size increased with increased irradiation time. The sizes of nanoclusters observed from the sample after 30 min irradiation is shown in Fig. 4, where the spherical individual particles, with an average size of 2 nm, were observed. For comparison, Fig. 5 gives the optical absorption spectra of Ag-exchanged glasses after 30 min UV-laser irradiation and after 600 1C heating for 45 h, respectively. It is clearly found that the sample after 600 1C heating for 45 h had a much narrow FWHM than that of UV-laser irradiation sample. The SPR peak position is 425 nm after UV-laser irradiation, and 411 nm after thermal annealing, respectively. The peak position shift was due to the nanoclusters size difference. Elemental metallic nanoclusters are generally formed in the laser-irradiated area following heat treatment in a reducing atmosphere. Interestingly, our results showed the formation of silver nanoclusters only by UV-laser irradia-
tion without subsequent heating, yielding a yellow color of glass. An electron was driven out from the 2p orbital of a NBO near the silver ions after laser irradiation, while silver ion captured the electron to form an atom. The following equations presented the effects during UV-laser irradiation [16]:
4
2
laser
Glass ! hþ þ e ,
(2)
Agþ þ e ! Ag0 ,
0 300
400
500 600 Wavelength, nm
700
800
Fig. 3. Optical absorption spectra of Ag-exchanged glass after UV-laser irradiation.
+
(3)
where h is a hole and e is the electron. Substrate temperature plays a fundamental role in the formation of silver precipitates in Ag-exchanged glass, with silver nanoclusters usually aggregating only at a high
ARTICLE IN PRESS J. Zhang et al. / Journal of Crystal Growth 310 (2008) 234–239
3.3. The X-ray irradiation-induced silver nanoclusters The ionizing radiation caused the displacement of lattice atoms or cause electron defects that involve changes in the valence state of lattice or impurity atoms in glass. The Ag-exchanged glass showed dark brown after X-ray irradiation, and the induced color was unstable at room temperature, as shown in Fig. 6. Some additional absorption bands were developed as a result of irradiation. Band separation with Gaussian resolution of the induced optical absorption spectra helped to distinguish absorption bands. Three characteristic absorption bands were well modeled, having their peak positions at about 627, 420, and 305 nm, respectively, for the Ag-exchanged glass. When the Ag-exchanged glass was subjected to X-ray, the same process as shown in Eqs. (2) and (3) appeared, subsequently, reduction of Ag0 atoms was induced. The most fundamental radiation-induced defects in glasses are the nonbridging oxygen hole center (NBOHC:Si–O*), where the notation ‘‘’’ represents three bonds with other oxygen in the glass network and ‘‘*’’ denotes an unpaired electron. The induced defects of NBOHCs resulted in above absorption bands of 627 and 420 nm. Another type of defect induced by X-ray irradiation was the trapped electrons, resulting the absorption around 305 nm [21–24]. The formations of Ag0 and NBOHCs were confirmed by ESR measurement, as shown in Fig. 7. Two distinct groups of Ag0 peaks, due to the 107Ag0 and 109Ag0, were clearly observed. However, the formation of Ag++ after X-ray irradiation was also observed in Fig. 7. This might because
3x104
Ag++ g=2.056
2x104 Ag0
Intensity, a.u.
temperature in air. Currently, the mechanisms responsible for the formation of the silver nanoclusters solely by laser irradiation are not fully understood. When Ag-exchanged glass was subjected to laser irradiation, a reduction of silver atoms was induced. We suggest that the neutralized silver promotes nucleation with resultant aggregation during irradiation due to the thermal effect of laser.
1x10
4
0 NBOHCs g=1.99~2.00
-1x104
Ag0
-2x104 -3x104 3000
3200
16
Absorption, cm-1
10
5
0 300
400
500 600 Wavelength, nm
700
800
Fig. 6. Optical absorption of Ag-exchanged glass after X-ray irradiation.
3800
4000
25°C, as irradiated 200°C
14 12
300°C
10
450°C
600°C
8 6 0
1000
2000 3000 4000 Annealing time, s
5000
6000
Fig. 8. Absorption of Ag-exchanged glass at 410 nm with the annealing time after X-ray irradiation.
of the following process: Xray
Blank As exchanged As x-ray irradiated 1 days after irradiation 15 days after irradiation 4 months after irradiation
3600
Fig. 7. ESR spectra of Ag-exchanged glass after X-ray irradiation.
Agþ ! Agþþ þ e . 15
3400
Magnetic induction, Gauss
Absorption at 410 nm, cm-1
238
(4)
Ag0 is an electron trap and Ag++ is a hole trap. The Ag++ led to the absorption band at about 305 nm as well, whose position was similar with trapped electrons. The Ag0 atoms were produced by X-irradiation, however, it is difficult to determine if the aggregation of Ag0 atoms existed or not because of the spectra overlapping at about 420 nm. In order to understand the formation of silver nanoclusters, we detected the absorption at 410 nm from room temperature to 600 1C with a heating rate of 6 1C/min, where the result is given in Fig. 8. The decrease of absorption till about 400 1C was due to the recombination of induced defects. As shown in Fig. 8, the silver nanoclusters did not formed till 600 1C. Unlike the Agdoped glass, which formed silver nanocluster at 400 1C after X-ray irradiation, X-ray irradiation would not promote the silver nanocluster formation in Ag-exchanged
ARTICLE IN PRESS J. Zhang et al. / Journal of Crystal Growth 310 (2008) 234–239
glass. On the other hand, induced defects in Ag-exchanged glass were much stable than in Ag-doped glass, because part of electrons was easily trapped by the Ag+ to produce the stable Ag0 atoms upon irradiating. Such difference might due to the difference of silver surface concentration, subsequently, Ag–O structure between the Ag-doped and Ag-exchanged glasses. For the adopted exchange conditions, the silver concentration at the surface of the glass is about 25% of that of sodium. 4. Conclusions The Ag-exchanged commercial soda-lime silicate glasses were treated by three methods: thermal annealing, UV-laser irradiation, and X-ray irradiation, in order to promote the silver nanoclusters formation. When the Agexchanged glass was subjected to thermal annealing, a reduction of silver atoms was induced after capturing the electron from glass structure or impurities. During thermal annealing in air, the dissociation of Ag–O bonds to form Si–O and Ag–Ag bonds result in a net loss in the system free energy. After annealing at 600 1C for 45 h, the spherical nanoclusters with a diameter of 3–8 nm were formed. Silver nanoclusters in the Ag-exchanged glass were also formed by a single process of UV-laser irradiation, without subsequent heating. Laser irradiation induces the reduction of silver ions and promotes the silver atoms aggregation. The nanoclusters with a diameter of about 2 nm were formed after 30 min laser irradiation. The silver nanoclusters formed by thermal annealing (600 1C, 45 h) had a much narrow FWHM than that of UV-laser irradiation (30 min) sample. The SPR peak position is 411 nm after thermal annealing, and 425 nm after UV-laser irradiation, respectively. The peak position shift was due to the nanoclusters size difference. The Ag-exchanged glass showed dark brown after X-ray irradiation, due to the induced defects and reduction of Ag0 atoms. However, X-ray irradiation would not promote the silver nanocluster formation. X-ray induced defects in Ag-exchanged glass were stable, because part of electrons was easily trapped by the Ag+ to produce the stable Ag0 atoms upon irradiating. Among these three methods, thermal annealing was an easy method to form the large size silver nanoclusters for possible practical applications. Acknowledgments The authors gratefully acknowledge the financial support for this work from Zhejiang Provincial Natural
239
Science Foundation of China (R404108, Y406033), Zhejiang Provincial Special Hired Professor Foundation, and Scientific Research Foundation of Zhejiang Provincial Education Department (20040556).
References [1] D. Ricard, P. Roussignol, C. Flytzanis, Opt. Lett. 10 (1985) 511. [2] W.P. Halperin, Rev. Mod. Phys. 58 (1986) 533. [3] D. Ila, E.K. Williams, R.L. Zimmerman, D.B. Poker, D.K. Hensley, Nucl. Instrum. Methods B 166–167 (2000) 845. [4] A. Miotello, M. Bonelli, G. De Marchi, G. Mattei, P. Mazzoldi, C. Sada, F. Gonella, Appl. Phys. Lett. 79 (2001) 2456. [5] P.W. Wang, Appl. Surf. Sci. 120 (1997) 291. [6] F. Gonella, G. Mattei, P. Mazzoldi, E. Cattaruzza, G.W. Arnold, G. Battaglin, P. Calvelli, R. Polloni, R. Bertoncello Jr., R.F. Haglund, Appl. Phys. Lett. 69 (1996) 3101. [7] M. Kaempfe, H. Hofmeister, S. Hopfe, G. Seifert, H. Graener, J. Phys. Chem. B 104 (2000) 11847. [8] G. De Marchi, F. Gonella, P. Mazzoldi, G. Battaglin, E.J. Knystautas, C. Meneghini, J. Non-cryst. Solids 196 (1996) 79. [9] D. Manikandan, S. Mohan, K.G.M. Nair, Mater. Res. Bull. 38 (2003) 1545. [10] M. Suszynska, L. Krajczyk, R. Capelletti, A. Baraldi, K.J. Berg, J. Non-cryst. Solids 315 (2003) 114. [11] B. Karthikeyan, J. Thomas, R. Philip, Chem. Phys. Lett. 414 (2005) 346. [12] E. Cattaruzza, G. Battaglin, F. Gonella, G. Mattei, P. Mazzoldi, R. Polloni, B.F. Scremin, Appl. Surf. Sci. 247 (2005) 390. [13] J. Sheng, K. Kadono, Y. Utagawa, T. Yazawa, Appl. Radiat. Isot. 56 (2002) 621. [14] J. Sheng, J. Zhang, L. Qiao, J. Non-cryst. Solids 352 (2006) 2914. [15] J. Zhang, W. Dong, L. Qiao, J. Li, J. Zheng, J. Sheng, J. Crystal Growth 305 (2007) 278. [16] J. Sheng, J. Zheng, J. Zhang, C. Zhou, L. Jiang, Physica B 387 (2007) 32. [17] E. Borsella, E. Cattaruzza, G. De Marchi, F. Gonella, G. Mattei, P. Mazzoldi, A. Quaranta, G. Battaglin, R. Polloni, J. Non-cryst. Solids 245 (1999) 122. [18] G. Battaglin, A. Boscolo-Boscoletto, P. Mazzoldi, C. Meneghini, G.W. Arnold, Nucl. Instrum. Methods B 116 (1996) 527. [19] U. Pal, A. Bautista-Hernandez, L. Rodriguez-Fernandez, J.C. Cheang-Wong, J. Non-cryst. Solids 275 (2000) 65. [20] G. Battaglin, E. Cattaruzza, F. Gonella, R. Polloni, F.D. Acapito, S. Colonna, G. Mattei, C. Maurizio, P. Mazzoldi, S. Padovani, C. Sada, A. Quaranta, A. Longo, Nucl. Instrum. Methods B 200 (2003) 185. [21] E.J. Friebele, Radiation effects, in: D.R. Uhlmann, N.J. Kreidl (Eds.), Optical and Properties of Glass, Westerville, 1991, pp. 205–261. [22] C.D. Marshall, J.A. Speth, S.A. Payne, J. Non-cryst. Solids. 212 (1997) 59. [23] P. Ebeling, D. Ehrt, Glastech. Ber. Glass Sci. Technol. 73 (5) (2000) 156. [24] H. Zeng, J. Qiu, X. Jiang, C. Zhu, F. Gan, J. Crystal Growth 262 (2004) 255.
Journal of Crystal Growth 310 (2008) 234–239 www.elsevier.com/locate/jcrysgro
Silver nanoclusters formation in ion-exchanged glasses by thermal annealing, UV-laser and X-ray irradiation Jian Zhanga, Wen Dongb, Jiawei Shenga,, Jingwu Zhenga, Juan Lia, Liang Qiaoa, Liqiang Jianga a
College of Chemical Engineering and Materials Science, Zhejiang University of Technology, Hangzhou, Zhejiang 310014, China b College of Biological Environmental Engineering, Zhejiang University of Technology, Hangzhou, Zhejiang 310014, China Received 13 July 2007; received in revised form 26 September 2007; accepted 5 October 2007 Communicated by M. Schieber Available online 10 October 2007
Abstract The Ag-exchanged commercial soda-lime silicate glasses were treated by three methods: thermal annealing, UV-laser irradiation, and X-ray irradiation, in order to promote the silver nanoclusters formation. Absorption spectrometry and electron spin resonance measurement results indicated that the silver ions transferred to silver atoms after the above three treatments. The silver atoms diffused and then aggregated to become nanoclusters after thermal annealing in air, or after UV-laser irradiation. However, X-ray irradiation, which induced defects and reduction of Ag0 atoms, would not promote the silver nanocluster formation. After annealing at 600 1C for 45 h, the spherical nanoclusters with a diameter of 3–8 nm were formed. The nanoclusters with a diameter of about 2 nm were formed after 30 min UV-laser irradiation without subsequent heating. The surface plasmon resonance peak position of silver nanoclusters changed from 411 nm after thermal annealing to 425 nm after UV-laser irradiation. The peak position shift was due to the nanoclusters size difference. r 2007 Elsevier B.V. All rights reserved. PACS: 81.05.Kf; 61.80.Cb; 78.20.e; 61.46.+w; 76.30.Mi Keywords: A1. Optical microscopy; B1. Glasses; B1. Nanomaterials
1. Introduction Introducing nanosize metal particles such as silver into glass has been used for changing the color of decorative glass and recently for fabricating optical devices. Nonlinear optical materials, such as composites formed by metal nanoclusters in glass, are potentially important in the field of all optical switching technology [1–4]. During the past decades, the physical and in particular the linear and nonlinear optical properties of metallic nanocluster in glass have been investigated using numerous experimental and theoretical approaches. Detailed investigation has been devoted to both the linear and nonlinear properties Corresponding author. Tel.: +86 571 88320851; fax: +86 571 88320142. E-mail address: [email protected] (J. Sheng).
0022-0248/$ - see front matter r 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.jcrysgro.2007.10.007
introduced by optical absorption due to the surface plasmon resonance (SPR). In the past years, silicate glasses containing silver nanoclusters have been the promising materials for nonlinear optical device fabrication. Ion exchange process with thermal treatment is an important method for obtaining silver nanoclusters in a silicate matrix. On the other hand, it is well known that the formation of colloidal silver clusters in glass causes yellow coloration, due to the SPR in the metal particles. Light-ion irradiation and annealing of Ag-exchanged glasses in hydrogen atmosphere (or in a high-vacuum atmosphere) are now widely accepted methods to prepare glass materials containing silver nanoclusters, mainly because they are easy to operate and commercially viable for technological applications in optoelectronic devices [4–7]. Although a substantial amount of literature is available on properties of silver
ARTICLE IN PRESS J. Zhang et al. / Journal of Crystal Growth 310 (2008) 234–239
235
nanoclusters in glass exhibiting high optical nonlinear response [8–12], only minimal work has been reported on the optical spectroscopic investigations of silver ionexchanged glass thermally annealing in air, or treated by UV-laser and X-ray irradiation. At present, little is known about the aggregation behavior and Ag+ chemical environment. Recently, in different context, we have reported the X-ray irradiation-induced defects in soda-lime silicate glass [13,14], the metallic silver nanocluster formation and dissolution in silver-doped glass by X-ray irradiation followed by thermal annealing [15]. We also reported the UV-laser induced silver nanoclusters in silver ionexchanged glass [16]. In this paper, we studied the silver nanoclusters formation in the ion-exchanged glasses. We described the silver nanoclusters formation in commercial flat soda-lime silicate glasses using optical absorption techniques and TEM measurements in an attempt to understand the formation mechanism of silver nanoclusters during annealing. Three treatments, involving thermal annealing, UV-laser annealing, and X-ray irradiation annealing, have been included in the present work. Once the formation mechanism of the silver nanocluster is understood, the information about the factors affecting the nucleation and growth process of silver nanoclusters will be known, and consequentially the optimum thermally processed can be achieved.
l ¼ 0.061 nm, 50 kV, 50 mA) for 30 min. Optical absorption spectra were recorded using a Lambda 9 Perkin-Elmer spectrophotometer at ambient room temperature and all recorded optical spectra were referenced to air. TEM images for representative samples were observed by a transmittance electron microscope (JEOL-2010 TEM) at electron beam energy of 300 keV. The first-derivative electron spin resonance (ESR) spectra were recorded at room temperature on a BRUKER 300E, operating at 9.7 GHz. First-derivative ESR absorption spectra were recorded at room temperature on a BRUKER 300E, operating at 9.7 GHz. DPPH (1,1-diphenyl-2-piterylhydrozyl) was used to adjust the g-values, in which, the g-value was defined by the relation hv ¼ gbH, where h is the Planck’s constant, n the spectrometer frequency, b the Bohr magneton, and H is the magnitude of the laboratory applied magnetic field at resonance. The g-values and characteristics of ESR spectra associated with the irradiation-induced defects and silver nanoclusters. The Ag0 peaks have two distinct groups in ESR spectra, due to the 107Ag0 and 109Ag0. The formation of Ag++ could be found in the vicinity of g ¼ 2.056. The most fundamental radiationinduced defects in glasses are the nonbridging oxygen hole center, which show a broad derivative peak in the vicinity of g ¼ 1.99–2.00.
2. Experimental procedure
3.1. Silver nanoclusters formation during thermal annealing
Commercially available soda-lime silicate glass substrates composed of (wt%): 74.2 SiO2, 14.3 Na2O, 1.9 Al2O3, 8.1 CaO, and 1.5 MgO were utilized for experimentation. Samples (2 mm thick and of approximately 15 15 mm2), which had previously been scoured and dried, were preheated and subsequently dipped in a molten salt bath formed by a mixture of 98 mol% NaNO3 and 2 mol% AgNO3 in a crucible of Al2O3. The ion exchanges took place at a temperature of 320 1C with a processing time of 10 min. To ensure that the glass structure did not change during testing, the ion-exchange temperature was kept well below the glass-transition temperature of the specimen. After inter-diffusion, samples were removed from the molten bath and washed with several portions of distilled water and acetone to remove any silver nitrate adhering to their surface. To promote the silver nanocluster formation, the Agexchanged samples were treated by three methods: thermal annealing, UV-laser irradiation, and X-ray irradiation. Thermal annealing was performed in air for 2–45 h at temperatures in the range 500–600 1C. Laser irradiation was performed with an ArF UV-laser (193 nm), utilizing at an output energy density of 30 mJ/cm2/pulse, with a repetition frequency of 10 Hz and a 20 ns pulse duration. The UV-laser irradiation time ranged from 5 s to 30 min. The X-ray irradiation was carried out using a X-ray Spectrometer (ThermoARL) at room temperature (Rh Ka,
Prior to the ion-exchange process, the glass samples were colorless and had no measurable absorption in the visible region. The soda-lime silicate glasses did not change color after Ag-exchanging for 10 min. No significant SPR of the silver nanoclusters (400–430 nm) was observed, indicating that the silver aggregate formation did not occur, or the silver nanoclusters were less than 1 nm in size during the 10 min ion-exchange. The formation of silver nanoclusters is limited, primarily, to the mobility of silver and the concentration of reducing elements present in the glass. Under our experimental procedures, the silver mobility was minimal and the concentration of reducing agent was relatively low. Fig. 1 shows the absorption spectra for Ag-exchanged glasses after thermal annealing in air. No obvious absorption band was developed after heating at 550 1C for 2 h. An absorption band at approximately 410 nm was clearly observed after heating at 600 1C for more than 2 h, which was apparently due to the SPR band of silver nanoclusters formed in the glass matrix. When the annealing time increased from 2 to 6 h, a small blue shift of the SPR was observed, indicating that the refractive index property of the matrix or the size of silver nanoclusters was annealing time-dependent (Fig. 1(a)). The intensity of absorption at 411 nm increased significantly while the peak position did not change with the increase of annealing time from 6 to 45 h (Fig. 1(b)).
3. Results and discussion
ARTICLE IN PRESS J. Zhang et al. / Journal of Crystal Growth 310 (2008) 234–239
236
5 Blank As-exchanged 550°C-2h 600°C-2h 600°C-4h 600°C-6h
Absorption, cm-1
4
3
2
1
0 300
350
400 450 500 Wavelength, nm
550
Absorption, cm-1
8
600
650
600°C-8h 600°C-12h 600°C-15h 600°C-35h 600°C-45h
6
Fig. 2. TEM image of Ag-exchanged glass after thermal annealing at 600 1C for 45 h.
4
2
0 300
350
400 450 500 Wavelength, nm
550
600
650
Fig. 1. Optical absorption spectra of silver nanoclusters embedded in Agexchanged glass after thermal annealing.
The width of the absorption band reduced systematically with the increase of annealing time, indicating an increase of volume fraction of silver nanoclusters in the glass. From reported data, the substrate temperature plays a fundamental role in the formation of silver precipitates in Ag-exchanged glass. To form silver nanoclusters in Agexchanged glasses with thermal annealing, the temperature was set much lower in reducting atmospheres than in air. When treated in air, the nanocluster formation is limited due to low mobility of silver and a lower level of reducing elements present in the glass. Therefore a higher annealing temperature is desirable. [17]. The average size of silver nanoclusters, d, can be determined by means of the equation [2,18,19]: d¼
2vf , Do
(1)
where vf=1.39 108 cm/s is the Fermi velocity for bulk silver and Do is the bandwidth at half maximum (FWHM) of the SPR band. The FWHM is determined by assuming
the absorption peak as a normal Gaussian distribution and fitting the flank of the absorption band on the long wavelength side. The value of Do is determined using Do=2pc(1/l1–1/l2), where cis the speed of light in vacuum, l1 and l2 are the wavelength at FWHM. After annealing at 600 1C for 45 h, the calculated average size was around 7 nm. The TEM image (Fig. 2) shows the presence of spherical nanoclusters with a diameter of 3–8 nm. In the Ag-exchanged glass, silver atoms are expected to be bound to nonbridging oxygen (NBO) atoms, after the substitution for the sodium atoms in the glass matrix. The silver introduced by ion-exchange process is mostly comprised of Ag+, with a minor population of Ag0 atoms [5]. When the Ag-exchanged glass was subjected to thermal annealing, a reduction of silver atoms was induced after capturing the electron from glass structure or impurities. During the thermal annealing, more Ag–O bonds are broken and more atoms are formed, thus the neutral silver becomes the dominant state. On the other hand, diffusion of silver towards the surface, with consequential precipitation, are caused by thermal relaxation of the surface tensile stress introduced by the size differential between Ag+ and Na+ during the ion exchange process. Tensile stress introduced by this size difference is still present after cooling [5,20]. The activated Ag+ ions mobilize toward a more relaxed surface, resulting from precipitation or nanocluster formation of silver atoms. Conversely, the dissociation of Ag–O bonds to form Si–O and Ag–Ag bonds result in a net loss in the system free energy [5]. Thus, the silver inclusion of the glass network will move toward the surface and form nanoclusters in order to maintain a minimum energy state within the system. Since the diffusion of silver in the glass matrix is nominal, the
ARTICLE IN PRESS J. Zhang et al. / Journal of Crystal Growth 310 (2008) 234–239
237
above-discussed processes of silver nanocluster formation are commonly carried out through heat treatment. 3.2. UV-laser-induced silver nanoclusters
8
600°C - 45 h UV - laser 30 min
6
4
2
0 300
350
400
450 500 Wavelength, nm
550
600
650
Fig. 5. Optical absorption spectra of Ag-exchanged glasses after 30 min UV-laser irradiation and 600 1C heating for 45 h, respectively.
8 Blank As ion-exchanged 5 sec irradiation 1 min irradiation 10 min irradiation 30 min irradiation
6 Absorption (cm-1)
Fig. 4. TEM image of Ag-exchanged glass after 30 min UV-laser irradiation.
Absorption, cm-1
Effects of laser irradiation were clearly visible with the glass exemplifying a yellow color after laser irradiation. Moreover, the color density increased with increasing irradiation time. Fig. 3 shows the absorption spectra for Ag-exchanged glasses after laser irradiations. An absorption band at approximately 425 nm was clearly observed, which was due to the SPR band of silver nanoclusters. A small blue shift of the SPR is visible in Fig. 3 for the longest irradiation time. The laser irradiation promotes nanocluster formation, and the intensity of the SPR band sharply increases with the irradiation time. For the measured FWHM of Fig. 3, the calculated size, using Eq. (1), were about 1.0 nm after 5 s irradiation and about 2.0 nm after 30 min irradiation, respectively. The average nanocluster size increased with increased irradiation time. The sizes of nanoclusters observed from the sample after 30 min irradiation is shown in Fig. 4, where the spherical individual particles, with an average size of 2 nm, were observed. For comparison, Fig. 5 gives the optical absorption spectra of Ag-exchanged glasses after 30 min UV-laser irradiation and after 600 1C heating for 45 h, respectively. It is clearly found that the sample after 600 1C heating for 45 h had a much narrow FWHM than that of UV-laser irradiation sample. The SPR peak position is 425 nm after UV-laser irradiation, and 411 nm after thermal annealing, respectively. The peak position shift was due to the nanoclusters size difference. Elemental metallic nanoclusters are generally formed in the laser-irradiated area following heat treatment in a reducing atmosphere. Interestingly, our results showed the formation of silver nanoclusters only by UV-laser irradia-
tion without subsequent heating, yielding a yellow color of glass. An electron was driven out from the 2p orbital of a NBO near the silver ions after laser irradiation, while silver ion captured the electron to form an atom. The following equations presented the effects during UV-laser irradiation [16]:
4
2
laser
Glass ! hþ þ e ,
(2)
Agþ þ e ! Ag0 ,
0 300
400
500 600 Wavelength, nm
700
800
Fig. 3. Optical absorption spectra of Ag-exchanged glass after UV-laser irradiation.
+
(3)
where h is a hole and e is the electron. Substrate temperature plays a fundamental role in the formation of silver precipitates in Ag-exchanged glass, with silver nanoclusters usually aggregating only at a high
ARTICLE IN PRESS J. Zhang et al. / Journal of Crystal Growth 310 (2008) 234–239
3.3. The X-ray irradiation-induced silver nanoclusters The ionizing radiation caused the displacement of lattice atoms or cause electron defects that involve changes in the valence state of lattice or impurity atoms in glass. The Ag-exchanged glass showed dark brown after X-ray irradiation, and the induced color was unstable at room temperature, as shown in Fig. 6. Some additional absorption bands were developed as a result of irradiation. Band separation with Gaussian resolution of the induced optical absorption spectra helped to distinguish absorption bands. Three characteristic absorption bands were well modeled, having their peak positions at about 627, 420, and 305 nm, respectively, for the Ag-exchanged glass. When the Ag-exchanged glass was subjected to X-ray, the same process as shown in Eqs. (2) and (3) appeared, subsequently, reduction of Ag0 atoms was induced. The most fundamental radiation-induced defects in glasses are the nonbridging oxygen hole center (NBOHC:Si–O*), where the notation ‘‘’’ represents three bonds with other oxygen in the glass network and ‘‘*’’ denotes an unpaired electron. The induced defects of NBOHCs resulted in above absorption bands of 627 and 420 nm. Another type of defect induced by X-ray irradiation was the trapped electrons, resulting the absorption around 305 nm [21–24]. The formations of Ag0 and NBOHCs were confirmed by ESR measurement, as shown in Fig. 7. Two distinct groups of Ag0 peaks, due to the 107Ag0 and 109Ag0, were clearly observed. However, the formation of Ag++ after X-ray irradiation was also observed in Fig. 7. This might because
3x104
Ag++ g=2.056
2x104 Ag0
Intensity, a.u.
temperature in air. Currently, the mechanisms responsible for the formation of the silver nanoclusters solely by laser irradiation are not fully understood. When Ag-exchanged glass was subjected to laser irradiation, a reduction of silver atoms was induced. We suggest that the neutralized silver promotes nucleation with resultant aggregation during irradiation due to the thermal effect of laser.
1x10
4
0 NBOHCs g=1.99~2.00
-1x104
Ag0
-2x104 -3x104 3000
3200
16
Absorption, cm-1
10
5
0 300
400
500 600 Wavelength, nm
700
800
Fig. 6. Optical absorption of Ag-exchanged glass after X-ray irradiation.
3800
4000
25°C, as irradiated 200°C
14 12
300°C
10
450°C
600°C
8 6 0
1000
2000 3000 4000 Annealing time, s
5000
6000
Fig. 8. Absorption of Ag-exchanged glass at 410 nm with the annealing time after X-ray irradiation.
of the following process: Xray
Blank As exchanged As x-ray irradiated 1 days after irradiation 15 days after irradiation 4 months after irradiation
3600
Fig. 7. ESR spectra of Ag-exchanged glass after X-ray irradiation.
Agþ ! Agþþ þ e . 15
3400
Magnetic induction, Gauss
Absorption at 410 nm, cm-1
238
(4)
Ag0 is an electron trap and Ag++ is a hole trap. The Ag++ led to the absorption band at about 305 nm as well, whose position was similar with trapped electrons. The Ag0 atoms were produced by X-irradiation, however, it is difficult to determine if the aggregation of Ag0 atoms existed or not because of the spectra overlapping at about 420 nm. In order to understand the formation of silver nanoclusters, we detected the absorption at 410 nm from room temperature to 600 1C with a heating rate of 6 1C/min, where the result is given in Fig. 8. The decrease of absorption till about 400 1C was due to the recombination of induced defects. As shown in Fig. 8, the silver nanoclusters did not formed till 600 1C. Unlike the Agdoped glass, which formed silver nanocluster at 400 1C after X-ray irradiation, X-ray irradiation would not promote the silver nanocluster formation in Ag-exchanged
ARTICLE IN PRESS J. Zhang et al. / Journal of Crystal Growth 310 (2008) 234–239
glass. On the other hand, induced defects in Ag-exchanged glass were much stable than in Ag-doped glass, because part of electrons was easily trapped by the Ag+ to produce the stable Ag0 atoms upon irradiating. Such difference might due to the difference of silver surface concentration, subsequently, Ag–O structure between the Ag-doped and Ag-exchanged glasses. For the adopted exchange conditions, the silver concentration at the surface of the glass is about 25% of that of sodium. 4. Conclusions The Ag-exchanged commercial soda-lime silicate glasses were treated by three methods: thermal annealing, UV-laser irradiation, and X-ray irradiation, in order to promote the silver nanoclusters formation. When the Agexchanged glass was subjected to thermal annealing, a reduction of silver atoms was induced after capturing the electron from glass structure or impurities. During thermal annealing in air, the dissociation of Ag–O bonds to form Si–O and Ag–Ag bonds result in a net loss in the system free energy. After annealing at 600 1C for 45 h, the spherical nanoclusters with a diameter of 3–8 nm were formed. Silver nanoclusters in the Ag-exchanged glass were also formed by a single process of UV-laser irradiation, without subsequent heating. Laser irradiation induces the reduction of silver ions and promotes the silver atoms aggregation. The nanoclusters with a diameter of about 2 nm were formed after 30 min laser irradiation. The silver nanoclusters formed by thermal annealing (600 1C, 45 h) had a much narrow FWHM than that of UV-laser irradiation (30 min) sample. The SPR peak position is 411 nm after thermal annealing, and 425 nm after UV-laser irradiation, respectively. The peak position shift was due to the nanoclusters size difference. The Ag-exchanged glass showed dark brown after X-ray irradiation, due to the induced defects and reduction of Ag0 atoms. However, X-ray irradiation would not promote the silver nanocluster formation. X-ray induced defects in Ag-exchanged glass were stable, because part of electrons was easily trapped by the Ag+ to produce the stable Ag0 atoms upon irradiating. Among these three methods, thermal annealing was an easy method to form the large size silver nanoclusters for possible practical applications. Acknowledgments The authors gratefully acknowledge the financial support for this work from Zhejiang Provincial Natural
239
Science Foundation of China (R404108, Y406033), Zhejiang Provincial Special Hired Professor Foundation, and Scientific Research Foundation of Zhejiang Provincial Education Department (20040556).
References [1] D. Ricard, P. Roussignol, C. Flytzanis, Opt. Lett. 10 (1985) 511. [2] W.P. Halperin, Rev. Mod. Phys. 58 (1986) 533. [3] D. Ila, E.K. Williams, R.L. Zimmerman, D.B. Poker, D.K. Hensley, Nucl. Instrum. Methods B 166–167 (2000) 845. [4] A. Miotello, M. Bonelli, G. De Marchi, G. Mattei, P. Mazzoldi, C. Sada, F. Gonella, Appl. Phys. Lett. 79 (2001) 2456. [5] P.W. Wang, Appl. Surf. Sci. 120 (1997) 291. [6] F. Gonella, G. Mattei, P. Mazzoldi, E. Cattaruzza, G.W. Arnold, G. Battaglin, P. Calvelli, R. Polloni, R. Bertoncello Jr., R.F. Haglund, Appl. Phys. Lett. 69 (1996) 3101. [7] M. Kaempfe, H. Hofmeister, S. Hopfe, G. Seifert, H. Graener, J. Phys. Chem. B 104 (2000) 11847. [8] G. De Marchi, F. Gonella, P. Mazzoldi, G. Battaglin, E.J. Knystautas, C. Meneghini, J. Non-cryst. Solids 196 (1996) 79. [9] D. Manikandan, S. Mohan, K.G.M. Nair, Mater. Res. Bull. 38 (2003) 1545. [10] M. Suszynska, L. Krajczyk, R. Capelletti, A. Baraldi, K.J. Berg, J. Non-cryst. Solids 315 (2003) 114. [11] B. Karthikeyan, J. Thomas, R. Philip, Chem. Phys. Lett. 414 (2005) 346. [12] E. Cattaruzza, G. Battaglin, F. Gonella, G. Mattei, P. Mazzoldi, R. Polloni, B.F. Scremin, Appl. Surf. Sci. 247 (2005) 390. [13] J. Sheng, K. Kadono, Y. Utagawa, T. Yazawa, Appl. Radiat. Isot. 56 (2002) 621. [14] J. Sheng, J. Zhang, L. Qiao, J. Non-cryst. Solids 352 (2006) 2914. [15] J. Zhang, W. Dong, L. Qiao, J. Li, J. Zheng, J. Sheng, J. Crystal Growth 305 (2007) 278. [16] J. Sheng, J. Zheng, J. Zhang, C. Zhou, L. Jiang, Physica B 387 (2007) 32. [17] E. Borsella, E. Cattaruzza, G. De Marchi, F. Gonella, G. Mattei, P. Mazzoldi, A. Quaranta, G. Battaglin, R. Polloni, J. Non-cryst. Solids 245 (1999) 122. [18] G. Battaglin, A. Boscolo-Boscoletto, P. Mazzoldi, C. Meneghini, G.W. Arnold, Nucl. Instrum. Methods B 116 (1996) 527. [19] U. Pal, A. Bautista-Hernandez, L. Rodriguez-Fernandez, J.C. Cheang-Wong, J. Non-cryst. Solids 275 (2000) 65. [20] G. Battaglin, E. Cattaruzza, F. Gonella, R. Polloni, F.D. Acapito, S. Colonna, G. Mattei, C. Maurizio, P. Mazzoldi, S. Padovani, C. Sada, A. Quaranta, A. Longo, Nucl. Instrum. Methods B 200 (2003) 185. [21] E.J. Friebele, Radiation effects, in: D.R. Uhlmann, N.J. Kreidl (Eds.), Optical and Properties of Glass, Westerville, 1991, pp. 205–261. [22] C.D. Marshall, J.A. Speth, S.A. Payne, J. Non-cryst. Solids. 212 (1997) 59. [23] P. Ebeling, D. Ehrt, Glastech. Ber. Glass Sci. Technol. 73 (5) (2000) 156. [24] H. Zeng, J. Qiu, X. Jiang, C. Zhu, F. Gan, J. Crystal Growth 262 (2004) 255.