A novel technique for doping silicate glasses with transition metals and rare-earth for waveguides applications

A novel technique for doping silicate glasses with transition metals and rare-earth for waveguides

applications

S. Alia,*, F. Gonellab, E. Cattaruzzab, A. Quarantac a. Department of Physics, Abdul Wali Khan University, 23200 Mardan, Pakistan

b. Physical Chemistry Dept., Università Ca' Foscari, Dorsoduro 2137, Venezia I-30123, Italy c. Materials Engineering Dept., Università di Trento, via Mesiano 77, 38050 Povo (TN), Italy

*Correspondence author: Dr. Shahid Ali, [email protected]

Abstract - Field-Assisted Solid-State Ion-Exchange (FASSIE) technique for doping silicate glasses with either transition metals or rare earths has been attracting much attention for its potential application in light waveguides, luminescent materials and for the possibility to realize systems in which formation of metal nanoclusters is controlled by suitable post-exchange techniques. In this framework, metallic films of either silver or gold are deposited onto soda-lime and borosilicate glasses by the rf-sputtering technique. Owing to an external electric field, metal ions diffuse into the glass replacing its alkali content at different values of processing temperature and electric field. The nanocomposites are then characterized by secondary ion mass and m-line spectrometries, optical absorption and transmission electron microscopy, indicating that the migration not only depends on the experimental conditions but also on the matrix and the chemical phenomena occurring at the metal/glass interface. For both transition metals, dark m-lines detection suggests that the samples may actually support guided modes, although the quality of the surface creates some problems for the prism coupling used in the measurements. In particular, the yet broaden lines seem to indicate two modes in the visible and one in the IR. Keywords - Field-assisted ion-exchange, glass nanocomposites, optical waveguides

I. INTRODUCTION

The doping of glasses with transition metals or rare-earths for altering the physical or electronic structure of a glass, then yielding to a change in the refractive index and in several optical properties, is exploited for its potential applications in optical materials and functional photonic devices [1-4]. These doped glasses, possibly containing dopant nanoclusters, are as vital to the future development of photonic systems as integrated circuits are to the electronic systems. Number of techniques have been employed to achieve an in-depth and homogeneous diffusion, but all come with some limitations and constraints [5-7]. Due to advancements in this field and the need to further explore the glass modification, a specific suitable technique, field-assisted solid-state ion exchange (FASSIE) has been recently developed [8]. The technique

was successfully realized for doping silicate glasses not only with monovalent but also with multivalent ions, with homogeneous in-depth concentration profiles [9-10]. Nevertheless, an exhaustive description and a complete understanding of the diffusion are still lacking, because of the complexity of the phenomena involved in the process; indeed, it is a non-equilibrium technique, where the dopant ions diffuse to either replace alkali ions of a glass or to fill the gaps/defects in the glass matrix. This less explored technique is exploited for the diffusion of transition metals, namely, cobalt, chromium, silver and gold, as well as rare-earth erbium in various matrices like soda-lime, borosilicate and pure silica glasses. The dopants diffusion into a given matrix depends not only on the nature of the dopant and on the experimental conditions, but also on the local structure and the chemical phenomena occurring at the metal-glass interface. The direct doping of surface layers of soda-lime glass with erbium at high concentration explores new ways for erbium-doped glasses towards numerous applications [11-12]. Rutherford backscattering spectrometry confirms the existence of relatively high concentration of erbium in the silicate matrices, while photoluminescence reveals that the incorporated rare-earth ions are in an optically active configuration. The structural, optical, and compositional analyses of the doped erbium assess the effectiveness of the technique, making it a suitable route for the controlled preparation of erbium-doped materials. In general, the presented preliminary experimental findings well address the novelty of the technique, indicating that high concentration of dopants may be made to diffuse into the glass matrix, which are suitable for a subsequent post-exchange treatment and can be utilized for advanced application in the field of photonics materials and thin film integrated optoelectronic technology. Several characterization techniques are used during the experimental work, namely, secondary ion mass and Rutherford backscattering spectrometries, optical absorption spectroscopy, extended x-ray absorption fine structure, ellipsometry, m-line technique and photoluminescence, each of which bolsters the versatility of the technique.

ICP2010-10

II. EXPERIMENTAL

Thin films of transition metals, 200-500 nm thick,

are deposited onto 1 mm thick 25×75 mm2 soda-lime

glass (SLG) slides (atomic % composition: 59.6 O, 23.9 Si, 10.1 Na, 2.6 Mg, 2.4 Ca, 0.7 Mg, 0.5 K, 0.2 S, and traces) and onto 1 mm thick 25×75 mm2 BK7 borosilicate glass slides (atomic % composition: 60.2 O, 22.4 Si, 11.0 B, 3.8 Na, 1.8 K, 0.8 Ba) in a radiofrequency magnetron sputtering deposition apparatus. A metallic gold film of 200 nm thickness is also deposited on the backside of the slides to achieve a good ohmic contact with the iron plate acting as a negative electrode. Deposition on both sides of the samples are performed by means of a 13.56 MHz radiofrequency source in a pure Ar atmosphere, at a pressure of 50×10-2 Pa. As for the deposition of Er, a metallic erbium film of 100 nm thickness is deposited on the top side of the slide by the same radiofrequency magnetron sputtering deposition apparatus. The deposited sample is then heated at 350°C for 2 hours in ambient air, with the aim to oxidize the deposited metallic erbium film. The thermal treatment also redistributes the sodium possibly removed from the surfaces during the cleaning process. Two other depositions are then performed, covering both sides of the sample with a metallic gold film, 1-2 hundreds nm thick. Each deposited sample is then cut into five pieces (15×25 mm2) and the edges are then cleaned in order to prevent direct current flow (short-circuiting). Each sample is then ion exchanged in air in an oven by the FASSIE technique, whose setup is shown in figure 1, with applied electric fields E ranging from 100 to 600 V/mm and temperature T ranging from 200 to 400 °C. The process duration varies from 1 to 4 hrs. After the field-assisted diffusion,the residual Au at the surface is removed mechanically. The current density through the sample is monitored during all the process with an interfaced computer.

The in-depth compositional profiles of the dopant and glass constituents are determined by secondary ion mass spectrometry (SIMS) using an IMS-4f CAMECA spectrometer, equipped with a normal incidence electron gun to compensate the surface charge build-up while profiling insulating samples. A 14.5 keV Cs+ primary beam and negative secondary ion detection (rastered area: 0.125×0.125 mm2) are used. Rutherford backscattering spectrometry (RBS) measurements are performed by using a 4He+ beam at the energy of 2.0 MeV with a detector placed at 160°. Optical absorption (OA) spectra in the 200–900 nm range are recorded using a UV–VIS-NIR Jasco V-570 dual-beam spectrophotometer. Bright-field transmission electron microscopy (TEM) analysis is performed with a FEI CM30T microscope operating at 300 kV and a field-emission gun (FEG) microscope (FEI Tecnai F20 Super Twin) operating at 200 kV, equipped with an energy-dispersive X-ray (EDX) spectrometer and a Gatan controller for performing scanning transmission electron microscopy (STEM).

+ –i

t

oven

isolation (silica)metal contact

substrate with film

metal contact

Fig. 1. Field-Assisted Solid-State Ion exchange (FASSIE) setup. Isolation is made by a pure silica slide, while a general purpose sourcemeter supplies voltage and measures current, and the current vs. time response is plotted on the interfaced computer.

III. RESULTS AND DISCUSSION

The work presented here is not intended to deeply investigate the doped glasses and their respective compositional analyses. Rather, it is concentrated on the versatility of the technique for doping silicate glasses with monovalent and multivalent transition metals as well as optically important rare-earth erbium. The mechanism of diffusion and the shape of the in-depth compositional profile come to depend strongly on the nature of dopant, on the experimental parameters as well as on the host matrix. As gold is diffused in silicate glasses for the first time with this technique [8-9], the novelty of technique is described mainly by reviewing this specific case study. Diffusion into the matrix depends critically on the two processing parameters, i.e. applied electric field and temperature. These parameters have a combined effect on the oxidation of the dopant layer and its diffusion into the glass. As current through the electrolyte depends mainly on space-charge occurrence, thickness of the oxidized layer due to oxygen migration and the number of free dopant ions available for diffusion in the matrix, an increase in a current is observed at strong experimental parameters. The actual behavior of the drift current can be observed by applying electric field on a cold sample, while temperature is then allowed to rise, the current behavior revealing peculiar information about the mechanism of diffusion. As the temperature starts rising, two maxima (peaks) are observed, showing substantial variation in the magnitude of current with the processing conditions. A very sharp and high peaked response is followed by a relatively small and well-separated increase, with a continuous fall afterwards. Such a peculiar response is reported in figure 2, for a series of samples treated at the same temperature of 400 °C and electric field ranging from 400 to 600 V/mm. The two peaks observed in all profiles are approximately of the same intensity, making it difficult to draw conclusions about the nature of species that take part in the current flow across the glass matrix. However, the resulting two distinct and well-separated peaks indicate that two different mechanisms occur in the matrix that contributes to the drift current. The current was observed to increase with the applied field across the matrix, and both peaks linearly follow the change in the

ICP2010-10

applied field. However, the height of the second peak results to be almost unaffected by the processing conditions, especially the applied electric field. It can be concluded that the first peak is due to the migration of highly mobile species of the matrix that have a strong response to a change in the applied electric field, while the second peak is due to less mobile and tightly bound species, such as the incoming dopant ions, for which the effectivness of the applied field is however limited by their lower mobility. At present, no evidence for important formation of metallic or oxide nanoclusters can be supported by our experimental data: the shape of diffusion profiles indicates diffusion mechanisms that do not involve the formation of clusters, a situation similar to those already reported for Cu and Ag migration [13-15], for which more mobile ion species (Cu+ and Ag+) do not however give rise to clusterization, at even higher temperatures in the case of Cu. Besides Na, SIMS and RBS analyses evidenced migration also of the other (less mobile) alkali species of the glass matrix, namely, Ca and K. As observed in figure 3, which shows the Au, Na and K profiles after the ion exchange, a strongly depleted region is formed between gold and alkali species that take place in the process, and this can play a significant role in the electro-assisted migration mechanism, due to local charge unbalancing. Besides the migration of alkali towards the cathode, one shoudl take into account also a migration of oxygen towards the surface of the glass, occurring in the first stage of the process during the oxidation of the deposited layer at the interface. In the case of BK7 glass, its compact local structure, high transition temperature, less content of alkali, and a good adhesion of Au deposited layer, makes it a proper candidate for doping with Au nanoparticles by FASSIE technique. The diffusion of Au cations is rather homogeneous and very small nanoclusters are observed well beneath the glass surface. The very different structures of SLG and BK7 cause different structural modifications during the ion exchange process, and hence at higher electric field the compositional modifications of these glass matrices also affect the in-depth distribution of other alkali and alkaline-earth elements, such as potassium and calcium. In figure 4a, the SIMS profiles of O, Si, Au, Na, K and Ca are shown for a SLG sample treated at 400°C and 400 V/mm for 90 min. The total gold amount, diffused up to about 1.2 μm, is (4±1)×1016 atoms/cm2 as from RBS data [9]. Na is observed to to be completely depleted in the region where Au is diffused, being its yield is negligible throughout the whole examined layer. In addition, the depth distribution of K and Ca are also affected by the field-assisted thermal treatment, showing a (less extended) depletion region below the surface. Figure 4b shows the SIMS depth profiles for a BK7 sample treated at the same experimental conditions as for the SLG sample.

0

1

2

3

4

5

6

0 500 1000 1500 2000

400°C 400V

400°C 500V

400°C 600V

Time (sec)

Cur

rent

(mA

)

Fig. 2. At sufficiently high temperature, current through the sample is driven according to the magnitude of applied electric field. A sizeable increase in the ionic transport occurs as the strength of electric field is raised. In this case, the total amount of gold diffused into the glass matrix is estimated by RBS to be (0.8±0.2)×1016 atoms/cm2, i.e. approximately one fifth of that for the SLG sample. This difference is probably related to the lower alkali content in the BK7 glass that leave their sites, preventing the achievement of local gold concentration as high as in the SLG case. The thickness of the gold-doped region is about 1.1 μm, comparable to that found for the SLG sample, and this suggests that the values of the gold diffusion coefficient in the two different glasses are close. The depth distribution of Na and K exhibit a remarkable depletion in the gold-doped region; the depletion is anyway limited to the region in which the gold diffused, confirming that the BK7 glass is a more stable matrix for the field-assisted diffusion.

0

1 104

2 104

3 104

4 104

5 104

6 104

0 100 200 300 400 500 600

O

Si

Au

Na

K

Depth (nm)

Yiel

d (c

ount

s/se

c)

T=200°C E=200V/mm

Fig. 3. SIMS concentration profiles of O, Si, Au, Na and K for an ion exchanged soda-lime glass treated at T=200 °C under E=200 V/mm of electric field.

ICP2010-10

100

101

102

103

104

105

106

0 500 1000 1500

O Si Au Na K Ca

Depth (nm)

Yiel

d (c

ount

s/se

c)

T=400°C , E=400 V/mm 90 min(a)

100

101

102

103

104

105

106

500 1000 1500

O Si Au Na K

T=400°C , E=400 V/mm 90 min(b)

Yiel

d (c

ount

s/se

c)

Depth (nm) Fig. 4. a) SIMS in-depth profiles of O, Si, Au, Na, K and Ca for SLG sample treated at T=400°C and E=400V/mm for 90 min. b) SIMS in-depth profiles of O, Si, Au, Na and K for BK7 sample treated at the same conditions of temperature and applied electric field as well as same duration of ion exchange. The normalized RBS yield for SLG samples are reported in figure 5, treated at the same temperatures and various field strength. The glass network seems to be heavily affected by the intense processing conditions. Due to high content of Na, the structure is deeply modified due to its complete depletion in the range of a couple of microns. The Aui RBS yield increases as the applied field increases from 100 to 400 V/mm at a fixed temperature of 300 °C. However, the yield decreses beyond this field. The migration of other alkali (like K) and alkali earth elements (like Ca and Mg) are also observed to take place at very intense experimental conditions. An accumulation can be noted at lower applied fields, which is smoothed at higher field strength. Thus, RBS confirms the SIMS analyses of different samples discussed above. However, in the case of BK7, RBS (not reported here) shows that the overall structure of glass remains stable compared to the very perturbed structure of SLG. The optical absorption of the doped layer of BK7 sample exhibits a surface plasmon resonance (SPR) peak for samples treated at less intense conditions. The occurrence of such a high absorption can be attributed to an accumulation of Au cations occurring at low processing temperature and high applied electric field, owing to the different mobilities of the involved species.

0

5

10

15

100 150 200 250 300 350 400 450 500

0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

T=300°C V=100V

T=300°C V=200V

T=300°C V=300V

T=300°C V=400V

T=300°C V=500V

T=300°C V=600V

Au

K , Ca

Si

Na , Mg

O

Channel

Nor

mal

ized

Yie

ld

Energy (MeV)

Fig. 5. The normalized RBS yield for soda-lime glass samples treated at the same temperature and different values of applied electric field. Au yield increases linearly with an increase in the applied voltage, nevertheless, decreasing in the concentration beyond 400 V.

0.25

0.3

0.35

0.4

0.45

0.5

0.55

200 300 400 500 600 700 800 900

T=200°C E=400V/mm

T=200°C E=500V/mm

T=300°C E=300V/mm

500 nm

Wavelength (nm)

Opt

ical

Den

sity

Fig. 6. Optical density profiles for BK7 samples show a considerable SPR for samples treated at low values of temperature and moderately high field strengths. The SPR peak shifts to lower wavelength of 500 nm than the typical value of 530 nm for Au nanoparticles. This shift can be attributed to the size distribution of the nanoparticles. The doped layer close to the surface, enriched in Au cations plasma, causes this substantial absorption. A very broad and prominent SPR is observed for the sample ion exchanged at 200 °C and 500 V/mm, as reported in figure 6. However, as the applied field is decreased by keeping the processing temperature constant, the intensity of the SPR lowers considerably, as reported for the sample treated at 200 °C and 400 V/mm. An increase in the treatment temperature and a reduction in the applied field gives rise to a further distribution into the matrix and thus an SPR of comparatively less intensity is observed for the sample treated at 300 °C and 300 V/mm. Anyways, the relation between the two experimental parameters is very crucial, and if balanced properly, no build-up of dopants near the surface occur. Indeed, the optical absorption depends on the overall concentration of the dopant and/or on the local build-up of cations. Moreover, a possible shifting of the plasmon wavelength to lower values can be due to nanoparticles which are very small in size and well dispersed into the matrix. A detailed study regarding the shape and size of SPR with respect to the size and morphology of the nanoparticles can be found in [16-21]. From a structural point of view, the FASSIE technique is more promising in the case of BK7 glass. The preparation conditions (500 °C and 150 V/mm for 4 hours) are chosen to allow gold ions to diffuse into a much dense matrix, having less content of Na. The structure of the exchanged region of the sample appears not completely homogenous but having a uniform size distribution of the dopants. The diffusion of gold nanoparticles is evident, taking place up to a depth of about half a micron. On the other hand, gold nanoparticles are not present in a region (about 40 nm thick) below the surface. All the experimental evidences (as from SIMS and optical absorption analyses) are confirmed by the scanning TEM microscopy measurements, reported in figure 7, in which the contrast of the image can be related to the atomic number of the elements. A brighter contrast indicates the presence of a heavy element, Au in this case.

ICP2010-10

200 nm

Glass Doped Layer Surface

1 2 3 40

5

10

15

20

25

Frac

tion

(%)

Diameter (nm)

<D>=2.2 nm σ =0.4 nm

Fig. 7. TEM Dark-Field image of BK7 sample. TEM Bright-Field images of the Au clusters in the same sample at different magnifications of 20 nm and 10 nm. The histogram plots the size distribution of these doped clusters. The Gaussian fitting shows that the clusters have a mean diameter of approximately 2 nm. Thus, the white spots indicate the formation of Au clusters, and it can be observed that their concentration is not uniform. Au nanoparticles exhibit a symmetric size distribution centred at 2.2 nm, with a mean diameter <D>=2.2 nm and a standard deviation σ=0.4 nm, as reported in figure. The homogeneity of the matrix indicates that almost all sites left by alkali are replaced by the small sized Au nanoparticles. Glass systems prepared by FASSIE are expected to exhibit different features depending on their future use in industrial devices, which in turn depend on the nature of the dopant element. In particular, some of the investigated systems could be interesting in optical waveguiding techniques, for example, in all-optical switches based on the third-order optical non-linearity exhibited by glass containing Au nanoclusters. In this respect, m-lines measurements (in the dark-line configuration) were preliminary performed on several Au-containing samples. Besides the coupling problems due to the sometimes bad optical quality of the sample surface, m-lines performed at the He-Ne 633 nm wavelength gives two guided modes in Au-doped soda-lime sample (T=400 °C, E=500 V/mm, t=5 h), for a calculated thickness of more than 2 μm. The effective refractive indices structure suggests that significant refractive index depletion may occur in the substrate region below the doped one, where alkali content is strongly depleted without a complete substitution by the gold atoms. Work is in progress to determine the optical waveguiding behavior of the obtained systems, with the aim at controlling their optical performances. As concerns the experiments with rare-earths doping, an RBS spectrum is reported in figure 8 for a soda-lime sample treated at 400 °C and 400 V/mm for 4 hours, after the removal of the residual Er oxide film. The broad intense response in the high energy region (approximately 1.5 MeV) is due to erbium atoms, where the narrow peak at the highest energy is due to the small fraction of the residual Er oxide film. It is evident that the direct diffusion of Er with FASSIE is successful and a significant amount of Er cations diffuse in the SLG matrix. The depletion of alkali and alkaline-earth elements in the doped region of the glass also occurs, as can be observed from the difference between the two spectra at lower energies. The SIMS in-depth profile (not

200 300 400 500Channel

0

5

10

15

Nor

mal

ized

Yie

ld

1.0 1.5Energy (MeV)

Pure SLGEr-doped SLG

Er

K , Ca

Si

MgNa

O

1500 1520 1540 1560 1580 16000

5

10

15

20

0 10 20 30

λexc= 488 nmPL

inte

nsity

(arb

. uni

ts)

Wavelength (nm)

1

0.1

Nor

m. P

L in

t. at

154

0 nm

Time (ms)

Fig. 8. RBS spectra of the soda-lime glass before and after the erbium doping. Energy edges for the different elements are indicated. The inset shows PL spectrum at 1540 nm for the doped SLG obtained by 25 mW Ar laser excitation at 488 nm. reported) indicates that the erbium diffusion took place up to 1 μm into the glass,whereas the combined SIMS and RBS analyses gives a total amount of erbium atoms diffused into the soda-lime glass of about 2.2×1016 Er/cm2 [11-12], with an uncertainty of approximately 10%. The inset in the figure shows the near-IR photoluminescence (PL) spectrum recorded for the sample at issue. The characteristic rare-earth PL at 1.54 μm [22-24], due to resonant excitation (at 488 nm), is detected without further thermal annealing of the sample. This pre-optically active configuration of the rare-earth can be attributed to the processing temperature, which is not so far from the SLG softening point. As concerns the dynamics involved in the fluorescence process, shown in the sub-inset, the decay behavior for the emission signal at 1540 nm has a typical lifetime of 13 milliseconds. In fact, optical activity of the Er3+ diffused into SLG by FASSIE technique is less intense than the Er-doped glasses synthesized with other techniques like ion implantation [25-26], conventional thermal ion exchange [27-28], sol-gel routes [29-30], and laser irradiation [31] or sputtering. But FASSIE technique has an edge over all these techniques in the sense that the doped glass does not need further thermal annealing, which is a mandatory subsequent requirement for Er3+ activation when treated with all these conventional techniques.

IV. CONCLUSIONS

The solid-state field-assisted diffusion technique is extensively used for doping various glasses with transition metals and rare-earth elements. The emphasis is given to the versatility of the technique and its ease of use. Only three external parameters, namely, temperature, electric field and duration of the treatment are to be optimized for getting a considerable, homogeneous and in-depth diffusion. The technique is successful so far, since it has been utilized for doping various glasses with monovalent as well as multivalent ions. Some of the diffusions, for example; Au3+ in silicate glasses and the direct diffusion of Er3+ in soda-lime glass have been recently realized for the first time. The in-depth diffused

ICP2010-10

composites are characterized with various techniques for exploring the possibility of light guiding and transmission applications. It is shown that this technique is an appropriate choice for doping various glasses beyond the solubility limit and subsequent post-exchange treatments can be manoeuvred for the type of application.

ACKNOWLEDGEMENT

Many thanks to Prof. P. Mazzoldi for his fruitful discussions throughout this work. Prof. G. Battaglin, Dr. C. Sada, Dr. V. Bello and M. Ferrari are acknowledged for the RBS, SIMS, TEM and m-line measurements, respectively.

REFERENCES

[1] F. Gonella and P. Mazzoldi, in: Handbook of Nanostructured Materials and Nanotechnology, H. S. Nalwa, ed., Vol. 4, pp. 81-158, Academic Press, S. Diego 2000.

[2] M. Martin, J. J. Videau, L. Canioni, F. Adamietz, L. Sarger, G. L. Flem, App. Opt., Vol. 39, 3, pp. 435-440, 2000.

[3] M. Weisser, F. Thoma, B. Menges, U. Langbein, S. M. Neher, Opt. Comm., Vol. 153, 1-3, pp. 27-31, 1998.

[4] D. Culemann, A. Knuettel, E. Voges, IEEE J. Sel. Top. in Quantum Electronics, Vol. 6, 5, pp. 730-734, 2000.

[5] H. Portales, M. Mattarelli, M. Montagna, A. Chiasera, M. Ferrari, A. Martucci, P. Mazzoldi, S. Pelli, G. C. Righini, J. Non-Cryst. Solids, Vol. 351, 21-23, pp. 1738-1742, 2005.

[6] E. Trave, G. Mattei, C. Maurizio, P. Mazzoldi, G. Pellegrini, C. Scian, G. Battaglin, App. Phy. Let., Vol. 89, 15, pp. 151121, 2006.

[7] V. Ranganathan and L. C. Klein, J. Non-Cryst. Solids, Vol. 354, 30, pp. 3567-3571, 2008.

[8] F. Gonella, E. Cattaruzza, A. Quaranta, S. Ali, N. Argiolas, C. Sada, Solid State Ionics, Vol. 177, 35-36, pp. 3151-3155, 2006.

[9] E. Cattaruzza, G. Battaglin, F. Gonella, S. Ali, C. Sada, A. Quaranta, Mat. Sc. and Engg: B, Vol. 149, 2, pp. 195-199, 2008.

[10] E. Cattaruzza, F. Gonella, S. Ali, V. Bello, T. Cesca, Solid State Phen., Vol. 151, pp 252-258, 2009.

[11] E. Cattaruzza, F. Gonella, G. Peruzzo, A. Quaranta, C. Sada, E. Trave, Mat. Sc. and Engg: B, Vol. 146, 1-3, pp. 163-166, 2008.

[12] E. Cattaruzza, G. Battaglin, F. Gonella, C. Maurizio, S. Ali, E. Trave, J. Phy. D: App. Phy., Vol. 42, pp. 045301-5, 2009.

[13] F. Gonella, F. Caccavale, L. D. Bogomolova, F. D'Acapito, A. Quaranta, J. App. Phy., Vol. 83, 3, pp. 1200-1206, 1998.

[14] J. Spirkova, P. T. Nebolova, M. Mika, Opt. Mat., Vol. 25, 2, pp. 101-107, 2004.

[15] F. Gonella, P. Canton, E. Cattaruzza, A. Quaranta, C. Sada, A. Vomiero, Mat. Sc. and Engg C: Biom. and Supramolecular Sys., Vol. 26, 5-7, pp. 1087-1091, 2006.

[16] K. L. Kelly, E. Coronado, L. L. Zhao, G. C. Schatz, J. Phy. Chem. B, Vol. 107, 3, pp. 668-677, 2003.

[17] I. O. Sosa, C. Noguez, R. G. Barrera, J. Phy. Chem. B, Vol. 107, 26, pp. 6269-6275, 2003.

[18] E. Hao, G. C. Schatz, J. T. Hupp, J. Fluoresc., Vol. 14, 4, pp. 331-341, 2004.

[19] B. J. Wiley, S. H. Im, Z. Y. Li, J. McLellan, A. Siekkinen, Y. Xia, J. Phy. Chem. B, Vol. 110, 32, pp. 15666-15675, 2006.

[20] S. Malynych and G. Chumano, J. Quant. Spect. and Rad. Trans., Vol. 106, 1-3, pp. 297-303, 2007.

[21] A. L. Gonzalez, and Cecilia Noguez, J. Comp. and Theoret. Nanoscience, Vol. 4, 2, pp. 231-238, 2007.

[22] Y. K. Sharma, S. S. L. Surana, R. K. Singh, R. P. Dubedi, Opt. Mat., Vol. 29, 6, pp. 598-604, 2007.

[23] S. Liu, G. Zhao, Y. Li, H. Ying, J. Wang, G. Han, Opt. Mat., Vol. 30, 9, pp. 1393-1398, 2008.

[24] A. Al Choueiry, A. M. Jurdyc, B. Jacquier, F. Gourbilleau, R. Rizk, Mat. Sc. and Engg: B, Vol. 146, 1-3, pp. 157-159, 2008.

[25] S. Berneschi, G. N. Conti, I. Banyasz, A. Watterich, N. Q. Khanh, M. Fried, F. Paszti, M. Brenci, S. Pelli, G. C. Righini, App. Phy. Let., Vol. 90, 12, pp. 121136-3, 2007.

[26] G. G. Bentini, M. Chiarini, M. Bianconi, F. Bergamini, D. Castaldini, G. B. Montanari, A. Bogoni, L. Poti, S. Sugliani, A. Nubile, P. De Nicola, L. Gallerani, G. Pennestri, S. Petrini, Nuc. Inst. and Methods in Phy. Res. B: Beam Int. with Mat. and Atoms, Vol. 266, 12-13, pp. 3120-3124, 2008.

[27] S. Janakova, B. Svecova, L. Salavcova, J. Spirkova, M. Mika, J. Oswald, J. Phy. and Chem. of Solids, Vol. 68, 5-6, pp. 1263-1267, 2007.

[28] B. Svecova, J. Spirkova, S. Janakova, M. Mika, Mat. Sc. and Engg: B, Vol. 149, 2, pp. 177-180, 2008.

[29] P. Mazzoldi, S. Padovani, F. Enrichi, G. Mattei, C. Sada, E. Trave, M. Guglielmi, A. Martucci, G. Battaglin, E. Cattaruzza, F. Gonella, C. Maurizio, Integ. Opt. and Phot. Integ. Circ: Proc. SPIE, Vol. 5451, pp. 311-326, 2004.

[30] R. R. Thomson, H. T. Bookey, N. Suyal, A. K. Kar, App. Phy. Let., Vol. 89, 10, pp. 101108-3, 2006.

[31] Y. Li, Z. He, H. Tang, L. Liu, L. Xu, W. Wang, J. Non-Cryst. Solids, Vol. 354, 12-13, pp. 1216-1220, 2008.

ICP2010-10