Nickel-induced enhancement of photoluminescence in nc-Si–SiOx nanostructures

The effect of nickel silicide interlayer on the intensity of photoluminescence (PL) from Si nanoclusters (nc) in normally deposited and obliquely deposited in vacuum SiOx/Ni/Si structures have been studied using spectral and time-resolved PL measurements. It has been shown that the intensity of PL b...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Datum:2014
Hauptverfasser: Michailovska, K.V., Indutnyi, I.Z., Shepeliavyi, P.E., Dan’ko, V.A.
Format: Artikel
Sprache:English
Veröffentlicht: Інститут фізики напівпровідників імені В.Є. Лашкарьова НАН України 2014
Schriftenreihe:Semiconductor Physics Quantum Electronics & Optoelectronics
Online Zugang:http://dspace.nbuv.gov.ua/handle/123456789/118413
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Назва журналу:Digital Library of Periodicals of National Academy of Sciences of Ukraine
Zitieren:Nickel-induced enhancement of photoluminescence in nc-Si–SiOx nanostructures / K.V. Michailovska, I.Z. Indutnyi, P.E. Shepeliavyi, V.A. Dan'ko // Semiconductor Physics Quantum Electronics & Optoelectronics. — 2014. — Т. 17, № 4. — С. 336-340. — Бібліогр.: 24 назв. — англ.

Institution

Digital Library of Periodicals of National Academy of Sciences of Ukraine
id irk-123456789-118413
record_format dspace
spelling irk-123456789-1184132017-05-31T03:03:15Z Nickel-induced enhancement of photoluminescence in nc-Si–SiOx nanostructures Michailovska, K.V. Indutnyi, I.Z. Shepeliavyi, P.E. Dan’ko, V.A. The effect of nickel silicide interlayer on the intensity of photoluminescence (PL) from Si nanoclusters (nc) in normally deposited and obliquely deposited in vacuum SiOx/Ni/Si structures have been studied using spectral and time-resolved PL measurements. It has been shown that the intensity of PL band in SiOx/Ni/Si samples is essentially higher than that in reference SiOx/Si samples (without the nickel interlayer) with the same characteristics and treatment. The PL intensity enhancement factor is equal to 5.77 for normally deposited samples and 18 for obliquely deposited samples. The unchanged spectral shape of PL bands and similar position of PL maximum in samples with and without nickel silicide interlayer indicates that in the SiOx/Ni/Si structures after annealing no additional emitting centers are introduced to compare with reference one. Time-resolved measurements showed that PL decay rate was decreased from 8.2*10⁴ s⁻¹ for SiOx/Si specimens to 2.86*10⁴ s⁻¹ for SiOx/Ni/Si one. The emission decay rate distribution was determined by fitting the experimental decay curves to the stretchedexponential model. The observed narrow decay rate distribution, decrease of the PL decay rate and enhancement of the PL intensity in SiOx/Ni/Si samples can be assigned to the processes of nickel silicide passivation of the dangling bonds at the interface of Si nanoparticles and the silicon oxide matrix, which is more effective in porous samples. 2014 Article Nickel-induced enhancement of photoluminescence in nc-Si–SiOx nanostructures / K.V. Michailovska, I.Z. Indutnyi, P.E. Shepeliavyi, V.A. Dan'ko // Semiconductor Physics Quantum Electronics & Optoelectronics. — 2014. — Т. 17, № 4. — С. 336-340. — Бібліогр.: 24 назв. — англ. 1560-8034 PACS 78.67.Bf, 78.55.Ap http://dspace.nbuv.gov.ua/handle/123456789/118413 en Semiconductor Physics Quantum Electronics & Optoelectronics Інститут фізики напівпровідників імені В.Є. Лашкарьова НАН України
institution Digital Library of Periodicals of National Academy of Sciences of Ukraine
collection DSpace DC
language English
description The effect of nickel silicide interlayer on the intensity of photoluminescence (PL) from Si nanoclusters (nc) in normally deposited and obliquely deposited in vacuum SiOx/Ni/Si structures have been studied using spectral and time-resolved PL measurements. It has been shown that the intensity of PL band in SiOx/Ni/Si samples is essentially higher than that in reference SiOx/Si samples (without the nickel interlayer) with the same characteristics and treatment. The PL intensity enhancement factor is equal to 5.77 for normally deposited samples and 18 for obliquely deposited samples. The unchanged spectral shape of PL bands and similar position of PL maximum in samples with and without nickel silicide interlayer indicates that in the SiOx/Ni/Si structures after annealing no additional emitting centers are introduced to compare with reference one. Time-resolved measurements showed that PL decay rate was decreased from 8.2*10⁴ s⁻¹ for SiOx/Si specimens to 2.86*10⁴ s⁻¹ for SiOx/Ni/Si one. The emission decay rate distribution was determined by fitting the experimental decay curves to the stretchedexponential model. The observed narrow decay rate distribution, decrease of the PL decay rate and enhancement of the PL intensity in SiOx/Ni/Si samples can be assigned to the processes of nickel silicide passivation of the dangling bonds at the interface of Si nanoparticles and the silicon oxide matrix, which is more effective in porous samples.
format Article
author Michailovska, K.V.
Indutnyi, I.Z.
Shepeliavyi, P.E.
Dan’ko, V.A.
spellingShingle Michailovska, K.V.
Indutnyi, I.Z.
Shepeliavyi, P.E.
Dan’ko, V.A.
Nickel-induced enhancement of photoluminescence in nc-Si–SiOx nanostructures
Semiconductor Physics Quantum Electronics & Optoelectronics
author_facet Michailovska, K.V.
Indutnyi, I.Z.
Shepeliavyi, P.E.
Dan’ko, V.A.
author_sort Michailovska, K.V.
title Nickel-induced enhancement of photoluminescence in nc-Si–SiOx nanostructures
title_short Nickel-induced enhancement of photoluminescence in nc-Si–SiOx nanostructures
title_full Nickel-induced enhancement of photoluminescence in nc-Si–SiOx nanostructures
title_fullStr Nickel-induced enhancement of photoluminescence in nc-Si–SiOx nanostructures
title_full_unstemmed Nickel-induced enhancement of photoluminescence in nc-Si–SiOx nanostructures
title_sort nickel-induced enhancement of photoluminescence in nc-si–siox nanostructures
publisher Інститут фізики напівпровідників імені В.Є. Лашкарьова НАН України
publishDate 2014
url http://dspace.nbuv.gov.ua/handle/123456789/118413
citation_txt Nickel-induced enhancement of photoluminescence in nc-Si–SiOx nanostructures / K.V. Michailovska, I.Z. Indutnyi, P.E. Shepeliavyi, V.A. Dan'ko // Semiconductor Physics Quantum Electronics & Optoelectronics. — 2014. — Т. 17, № 4. — С. 336-340. — Бібліогр.: 24 назв. — англ.
series Semiconductor Physics Quantum Electronics & Optoelectronics
work_keys_str_mv AT michailovskakv nickelinducedenhancementofphotoluminescenceinncsisioxnanostructures
AT indutnyiiz nickelinducedenhancementofphotoluminescenceinncsisioxnanostructures
AT shepeliavyipe nickelinducedenhancementofphotoluminescenceinncsisioxnanostructures
AT dankova nickelinducedenhancementofphotoluminescenceinncsisioxnanostructures
first_indexed 2025-07-08T13:55:58Z
last_indexed 2025-07-08T13:55:58Z
_version_ 1837087287280664576
fulltext Semiconductor Physics, Quantum Electronics & Optoelectronics, 2014. V. 17, N 4. P. 336-340. © 2014, V. Lashkaryov Institute of Semiconductor Physics, National Academy of Sciences of Ukraine 336 PACS 78.67.Bf, 78.55.Ap Nickel-induced enhancement of photoluminescence in nc-Si–SiOx nanostructures K.V. Michailovska, I.Z. Indutnyi, P.E. Shepeliavyi, V.A. Dan’ko V. Lashkaryov Institute of Semiconductor Physics, NAS of Ukraine 41, prospect Nauky, 03028 Kyiv, Ukraine E-mail: indutnyy@isp. kiev.ua Abstract. The effect of nickel silicide interlayer on the intensity of photoluminescence (PL) from Si nanoclusters (nc) in normally deposited and obliquely deposited in vacuum SiOx/Ni/Si structures have been studied using spectral and time-resolved PL measurements. It has been shown that the intensity of PL band in SiOx/Ni/Si samples is essentially higher than that in reference SiOx/Si samples (without the nickel interlayer) with the same characteristics and treatment. The PL intensity enhancement factor is equal to 5.77 for normally deposited samples and 18 for obliquely deposited samples. The unchanged spectral shape of PL bands and similar position of PL maximum in samples with and without nickel silicide interlayer indicates that in the SiOx/Ni/Si structures after annealing no additional emitting centers are introduced to compare with reference one. Time-resolved measurements showed that PL decay rate was decreased from 8.2104 s–1 for SiOx/Si specimens to 2.86104 s–1 for SiOx/Ni/Si one. The emission decay rate distribution was determined by fitting the experimental decay curves to the stretched- exponential model. The observed narrow decay rate distribution, decrease of the PL decay rate and enhancement of the PL intensity in SiOx/Ni/Si samples can be assigned to the processes of nickel silicide passivation of the dangling bonds at the interface of Si nanoparticles and the silicon oxide matrix, which is more effective in porous samples. Keywords: time-resolved photoluminescence, thin film, silicon oxide, nanoparticle, nickel silicide. Manuscript received 22.04.14; revised version received 16.07.14; accepted for publication 29.10.14; published online 10.11.14. 1. Introduction Thin-film structures containing Si nanoparticles embedded into SiOx matrix attract attention of many researchers because of their promising applications in advanced electronic and optoelectronic devices [1-5]. But in spite of intensive researches over the past years, the reported efficiencies of nc-Si–SiOx light-emitting structures are still low and not high enough for practical application. The most important factors influencing the characteristics of PL are nanoparticle size and state of nanoparticle–SiOx interface. The passivation of nonradiative states and defects at this interface is an essential requirement in order to increase the intensity of PL. Hydrogen passivation through standard forming gas annealing is impractical for device application because hydrogen is easy-to-dissociate at elevated temperatures, which leads to the invalidation of hydrogen passivation. The Si–SiOx interface can be modified by chemical compounds of necessary composition. Such treating is the most efficient in porous structures. Recently [6, 7], we have proposed the method of porous nc-Si–SiOx Semiconductor Physics, Quantum Electronics & Optoelectronics, 2014. V. 17, N 4. P. 336-340. © 2014, V. Lashkaryov Institute of Semiconductor Physics, National Academy of Sciences of Ukraine 337 light-emitting structure formation using oblique deposition of Si monoxide (SiO) in vacuum. The electron microscopy studies show that, during this deposition, SiOx films with a porous (column-like) structure are formed. During high-temperature annealing of these films, the thermally stimulated formation of Si nanoinclusions occurs in a restricted volume of the SiOx column. Because of free space (cavities) between the oxide column, the structures is more susceptible to chemical treatments, e.g., to treatment with HF solution or vapor [8]. As a result of HF vapor treatment, approximately 200-fold increase in the PL intensity is observed. It was reported also [3] that introduction of thin nickel interlayer between SiOx films and Si substrates results in enhancing of the PL intensity of SiOx/Ni/Si structures by a factor of 4 in comparison with samples without any Ni interlayer. Using the examples of light- emitting diodes based on SiO1.56/Ni/Si structures, it was shown that NiSi2 distribution in SiO1.56 film could improve turn-on voltage and give a benefit to the electroluminescence efficiency [3]. In this paper, we report the results of studying the nickel induced enhancement of PL emission in porous and solid nc-Si– SiOx light-emitting nanostructures. 2. Experimental A thin (~12 nm) layer of Ni film was deposited by thermal evaporation of nickel powder onto (100)- oriented Si wafer. After deposition, the samples were annealed in vacuum at 450 °C for 10 min. The annealing provides the required energy for the system to overcome the energy barriers for solid state reaction of formation of intermediate nickel silicides [5]. Then, silicon oxide (SiOx) film was deposited onto the nickel silicide film by thermal evaporation of Cerac. Inc. SiO with 99.9% purity in the vacuum chamber (the residual pressure   Pa102...1 3 ). Before SiOx deposition, the substrates were oriented at the angles (α) of 60° or 0° (normal deposition) between the normal to the substrate surface and direction to the evaporator. Because of the additional oxidation by residual gases during evapo- ration of SiO the compositionally nonstoichiometric SiOx (x ~1.25 for normal deposition and x ~1.54 for 60°) films were obtained. The thickness of the SiOx films were monitored in situ by the quartz-crystal-oscillator monitor system (KIT-1) and measured after deposition by a microinterferometer (MII-4). The thickness of normally deposited SiOx films was equal to 400 nm, deposited at 60° – 700…800 nm. The samples without Ni interlayer were used as the reference ones in these measurements. As-deposited samples with and without Ni film were annealed simultaneously in vacuum at the temperature 975 °C for different duration ranging from 4 to 15 min. This high-temperature annealing leads to decomposition of silicon oxide (where x changed from 1.25 to 1.95 for samples obtained under normal deposition and from 1.54 to 2.0 for 60° ones) and formation of Si nanoparticles embedded into such oxide matrix [9, 10]. (The composition of the oxide matrix in as-deposited and annealed samples (parameter x) were determined using compositional dependence of the position of the main IR band in spectra of SiOx layers within the range of 1000 to 1100 cm–1, as it was ascer- tained in [11]. This band corresponds to the Si–O–Si stretching mode. FTIR measurements were carried out with Perkin-Elmer Spectrum BXII spectrometer.) The structure of obliquely deposited SiOx films was studied in our previous paper [12] by SEM apparatus ZEISS EVO 50XVP. Such SiOx films have a porous inclined column-like structure with the column diameters of 10 to 100 nm. The porosity of films depends on the angle of deposition and equals to ~40% for α = 60°. High-temperature annealing of these films does not change the porosity and column-like structure of the samples [7]. The PL spectra of obtained SiOx/Ni/Si and SiOx/Si samples were recorded at room temperature within the wavelength range 440 to 900 nm by using a system based on ZMR-3 monochromator equipped with a photomultiplier and detection system. The PL spectra were normalized to the spectral sensitivity of the experimental setup. The 337-nm line of a nitrogen laser with a spot size of about 2 mm in diameter was employed to excite the PL. Decay curve measurements were performed using the same N2 laser with the pulse duration 9 ns, which was short as compared to PL average lifetimes of our samples (approximately tens of microseconds). The PL lifetime was measured at different emission wavelengths. The time trace was recorded with a resolution of 0.5 μs. 3. Results and discussion Fig. 1 presents the PL spectra of normally deposited SiOx/Ni/Si samples after annealing at 975 °C for 15 and 4 min (curve 1 and 2, accordingly). Fig. 1 shows also PL spectra of the reference normally deposited SiOx/Si sample (curve 3) under annealing at the same temperature for 15 min. At room temperature, all the samples exhibited a broad band with a maximum of emission centered within 760…780 nm. This strong near-infrared luminescence band within the wavelength range 600…900 nm is associated with quantum confinement effects, that is with electron-hole pairs (or exciton) recombination in nc-Si. The intensity of emission from SiOx/Ni/Si samples IPL(Ni) is enhanced significantly in comparison with the reference sample (IPL) and depends on the annealing time. So, the PL enhancement factor η =   PLPL II Ni measured at 780 nm increases from 2.45 to 5.77 for 4- and 15-min annealing time, accordingly. Fig. 2 shows the PL spectra for the SiOx/Ni/Si (curve 1) and SiOx/Si (curve 2) samples deposited at 60° and annealed at 975 °C for 15 min in vacuum. The Semiconductor Physics, Quantum Electronics & Optoelectronics, 2014. V. 17, N 4. P. 336-340. © 2014, V. Lashkaryov Institute of Semiconductor Physics, National Academy of Sciences of Ukraine 338 emission spectrum of both samples exhibit a broad band with the peak position at 630 nm. A spectral shift of PL peaks to shorter wavelengths in obliquely deposited samples in comparison with normally deposited ones can be caused by decrease in nc-Si dimensions. The size of nc-Si in these systems depends on the silicon content in deposited SiOx layers [13]: decrease in the Si content (i.e., increase of x) results in decreased nc-Si sizes and blue shift of PL spectrum. Besides, during high- temperature annealing of porous films with column-like structure the thermally stimulated formation of nc-Si occurs in a restricted volume of the SiOx columns. It can also cause the reduction of the nc-Si size and blue shift of the PL peak. As seen from Figs 1 and 2, addition of Ni interlayer results in enhancement of the PL intensity more prominent in porous samples. For porous, obliquely deposited samples, the enhancement factor η measured at 630 nm is equal to 18. 600 650 700 750 800 850 900 0.0 0.1 0.2 0.3 0.4 0.5 0.6 I P L , a rb . u ni ts nm 3 2 1 Fig. 1. PL spectra of normally deposited samples: SiOx/Ni/Si after annealing in vacuum at 975 °C for 15 (1) and 4 min (2) and the reference SiOx/Si sample after annealing at 975 °C for 15 min (3). 500 550 600 650 700 750 800 0 2 4 6 8 10 I P L , a rb . u ni ts ,nm 1 2 Fig. 2. PL spectra of SiOx/Ni/Si (1) and reference of SiOx/Si (2) samples obliquely deposited at the angle 60° after annealing in vacuum at 975 °C for 15 min. It is known that, in a low excitation regime, the PL intensity is generally given by the expression IPL ~ (τPL/τR)σφN, where τPL and τR are the photo- luminescence and radiative lifetimes, respectively, φ is the photon flux of the laser pump (constant throughout our experiment), N is the density of nanocrystals in the film, and σ is their excited cross-section [17]. The unchanged spectral shape of PL bands and similar position of PL maximum in samples with and without Ni interlayer mean that in the SiOx/Ni/Si structures after annealing no additional emitting centers are introduced to compare with the reference one. In the quantum confinement scheme, the light emission from nc-Si is caused by radiative recombination of electron- hole pairs (or excitons) confined within nanoparticle [15]. It is deduced that the radiative lifetimes τR would be close to each other in SiOx/Ni/Si and SiOx/Si samples due to the same luminescence scheme and similar size of nc-Si (similar position of the PL maximum). One can suppose that nc-Si density (N) and PL lifetime (τPL) are the main factors changing the PL intensity in SiOx/Ni/Si samples. More direct demonstration of enhanced electron-hole pair recombination involved comparative measurements of the PL decay rate in the investigated structures. Time-resolved PL measurements were performed using the normally deposited SiOx/Ni/Si and SiOx/Si samples. Fig. 3 shows PL decay curve for SiOx/Ni/Si (curve 1) and SiOx/Si (curve 2) samples at 780 nm. One can see that the PL intensity for the SiOx/Ni/Si samples decayed slower than for the reference samples. The obtained decay curves of the PL intensity may be described well by a stretched exponential function:                     0 1 exp t tCtIPL , (1) where C, τ0 and β are some constant, decay time and stretched parameter (0 < β ≤ 1), respectively. It is known that the stretched exponential function is widely used for the decay curve analysis of Si nanocrystals [16]. The least-squares fit of Eq. (1) to experimental data brings values of τ0 and β. The obtained decay times τ0 were equal to 35 and 12.2 μs for SiOx/Ni/Si and SiOx/Si samples, respectively. This range of PL lifetimes is comparable with those reported in the literature for silicon nanocrystals at room temperature [17]. The PL decay rate k  1 0 k at 780 nm is decreased from 8.2104 s–1 for SiOx/Si to 2.86104 s–1 for SiOx/Ni/Si samples. It was also determined that the dispersion parameter β is 0.83 for SiOx/Ni/Si and 0.74 for SiOx/Si samples. In general, the parameter β is related to the stretching of the decay process and is a direct measure of the width of the decay rate distribution. In the case of stretched exponential relaxation function, the PL decay may be analyzed more thoroughly by recovering the distribution of recombination rates [16]. Semiconductor Physics, Quantum Electronics & Optoelectronics, 2014. V. 17, N 4. P. 336-340. © 2014, V. Lashkaryov Institute of Semiconductor Physics, National Academy of Sciences of Ukraine 339 0 10 20 30 40 0.0 0.2 0.4 0.6 0.8 1.0 I P L , a rb . u ni ts t, s 1 2  ex = 337 nm  em = 780 nm T= 295 K Fig. 3. PL decay curves measured at 780 nm for the normally deposited SiOx/Ni/Si (1) and SiOx/Si (2) samples after annealing at 975 °C for 15 min. 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.0 0.2 0.4 0.6 0.8 1.0 D is tr ib ut io n Ф (k ) Decay rate, s-1 without Ni 2 1 with Ni Fig. 4. Decay rate distributions of SiOx/Ni/Si (1) and SiOx/Si (2) normally deposited samples obtained from the stretched- exponential decay model. The decay rate distributions are normalized by the peak value of Ф(k) for both samples. Using the values of τ0 and β measured at λ = 780 nm, we calculated the asymptotic form of the decay rates probability density function Ф(k) that might be obtained by the saddle-point method [18]:     aa kk a kФ      exp 2 )( 2/1 , (2) where   11 a and    1/1 0 1   a . Fig. 4 shows the Ф(k) distributions calculated from Eq. (2) for SiOx/Ni/Si (curve 1) and SiOx/Si (curve 2) samples. The obtained distribution for the reference SiOx/Si sample is very broad with a long tail directed towards shorter lifetimes, which demonstrates the strongly non-single exponential character of decay curves. In SiOx/Ni/Si samples, the decay rate distribution Ф(k) is more narrow as compared with SiOx/Si sample and shifts towards lower decay rates. To explain the observed feature, it should be mentioned that the obtained Ф(k) function provides information about both the radiative (kR) and nonradiative (kNR) relaxation rates [19]. A very low quantum efficiency of nc-Si emission suggests that nonradiative processes should be predominant (kNR >> kR). It allows us to relate the changes observed in the decay rate distribution Ф(k) to the different quantity of defect states (dangling silicon bonds) in the matrix containing nc-Si. We believe that the interface between nc-Si and the SiOx matrix plays a crucial role in distributing decay rates [20]. SiO2 is an ideal matrix for nc-Si, as it can passivate a large fraction of the dangling Si bonds. There is a remarkable content of broken silicon bonds in SiOx films, which may act as nonradiative recombination paths for the excited carriers and cause the quenching of PL [21]. The observed narrow decay rate distribution and the increase of the PL decay time in SiOx/Ni/Si samples can be related to the processes of nickel silicide passivation of nonradiative recombination centers. It has been reported previously that nickel silicide thin films were formed by vacuum thermal processing (350-750 °C) of Ni thin films deposited onto (100) p-type Si substrates [4]. Ni as the mobile species moves through NiSi during the transition from NiSi to the NiSi2 phase at 750…1000 °C [22]. NiSi2 distributed in SiOx film passivates the broken bonds on the nc-Si surface [23]. The model of NiSi2 passivation was proved by thermodynamic analysis and Fourier transform infrared spectroscopy in SiO1.56/Ni/Si systems [3]. In the porous SiOx/Ni/Si samples, the nickel diffusion processes significantly accelerated due to the presence of voids and the surface of nanocolumns. This leads to more effective passivation of the nonradiative recombination centers and more significant increase in the PL emission intensity. In conclusion, we can note that the PL intensity enhancement factor calculated from the experimental results for SiOx/Ni/Si and SiOx/Si samples annealed at 975 °C is η = 5.77 for normal deposited samples. Enhancement of PL lifetimes in these samples by the factor ~2.87 due to the processes of nickel silicide passivation is not enough for explanation of the increase in PL intensity. Therefore, we proposed that the PL intensity enhancement can be caused both by increasing the PL lifetime and also by increasing the nc-Si density. The presence of Ni gives an additional driving force to the separation process of SiOx which was discussed in the previous paper [24]. 5. Conclusion In summary, we have presented the effect of Ni on the PL emission from Si nanoparticles embedded in the silicon oxide matrix. It has been shown that the intensity of near-infrared emission band in SiOx/Ni/Si samples was significantly higher than that in the samples without the Ni interlayer. It was assumed that nickel in SiOx film may passivate the residual dangling bonds at the interface of nc-Si and SiOx matrix. Time-resolved PL measurements showed the decrease in the PL decay rate in SiOx/Ni/Si samples as compared with the SiOx/Si one. Decrease in the PL decay rate and increase in the density Semiconductor Physics, Quantum Electronics & Optoelectronics, 2014. V. 17, N 4. P. 336-340. © 2014, V. Lashkaryov Institute of Semiconductor Physics, National Academy of Sciences of Ukraine 340 of nc-Si could be the main factors enhancing the PL intensity in nanostructures with nickel silicide interlayer. References 1. K.S. Min, K.V. Shcheglov, S.M. Yang, H.A. Atwater et al., Defect related versus excitonic visible light emission from ion beam synthesized Si nanocrystals // Appl. Phys. Lett. 69, p. 2033-2035 (1996). 2. M. Fujii, A. Mimura, S. Hayashi, K. Yamammoto et al., Improvement in photoluminescence efficiency of SiO2 films contaning Si nanocrystals by P doping // J. Appl. Phys. 87, p. 1855-1857 (2000). 3. D.X. Li, Y. He, J.Y. Feng, The study of nickel- induced enhancement of near-infrared luminescence in Si-rich silicon oxide films // Physica E, 41, p. 812-816 (2009). 4. M. Tinani, A. Mueller, Y. Gao et. al., In situ real- time studies of nickel silicide phase formation // J. Vac. Sci. Technol. B, 19(2), p. 376-383 (2001). 5. F.F. Zhao, J.Z. Zheng, Z.X. Shen et al., Thermal stability of NiSi and NiSi2 thin films // Microelectron. Eng. 71(1), p. 104-111 (2004). 6. I.Z. Indutnyy, I.Yu. Maidanchuk, V.I. Min’ko, Visible photoluminescence from annealed porous SiOx films // J. Optoelectron. and Adv. Mater. 7, p. 1231-1236 (2005). 7. V.A. Dan’ko, I.Z. Indutnyy, I.Y. Maidanchuk et al., Formation of the photoluminescence structure based on SiOx porous films // Optoelektronika i poluprovodnikovaya tekhnika, 39, p. 65-72 (2004) (in Ukrainian). 8. I.Z. Indutnyi, E.V. Michailovskaya, P.E. Shepeliavyi and V.A. Dan’ko, Visible photoluminescence of selectively etched porous nc-Si–SiOx structures // Fizika tekhnika poluprovodnikov 44(2), p. 218-222 (2010) (in Russian) [Semiconductors, 44(2), p. 206- 210 (2010)]. 9. V.A. Dan’ko, V.Ya. Bratus’, I.Z. Indutnyi, I.P. Lisovskyy, S.O. Zlobin, K.V. Michailovska, P.E. Shepeliavyi, Controlling the photo- luminescence spectra of porous nc-Si–SiOx structures by vapor treatment // Semiconductor Physics, Quantum Electronics & Optoelectronics, 13(4). p. 413-417 (2010). 10. V.Ya. Bratus’, V.A. Yukhimchuk, L.V. Berezhinsky et al., Structural transformation and silicon nanocrystallite formation in SiOx films // Semiconductors, 35(7), p. 821-826 (2001). 11. M. Nakamura, Y. Mochizuki, K. Usami et al., Infrared absorption spectra and compositions of evaporated silicon oxide (SiOx) // Solid State Communs. 50, p. 1079-1081 (1984). 12. I.Z. Indutnyi, K.V. Michailovska, V.I. Min’ko, P.E. Shepeliavyi, Effect of acetone vapor treatment on photoluminescence of porous nc-Si–SiOx nanostructures // Semiconductor Physics, Quantum Electronics & Optoelectronics, 12(2), p. 105-109 (2009). 13. D. Nesheva, C. Raptis, A. Perakis et al., Raman scattering and photoluminescence from Si nanoparticles in annealed SiOx thin films // J. Appl. Phys. 92, p. 4678-4683 (2002). 14. C. Garcia, B. Garrido, P. Pellegrino et al., Size dependence of lifetime and absorption cross section of Si nanocrysrals embedded in SiO2 // Appl. Phys. Lett. 82, p. 1595-1597 (2003). 15. C. Delerue, G. Allan, M. Lannoo, Theoretical aspects of the luminescence of porous silicon // Phys. Rev. B, 48, p. 11024-11036 (1993). 16. G. Zatrub, A. Podhorodecki, J. Misiewicz et al., On the nature of the stretched exponential photoluminescence decay for silicon nanocrystals // Nanoscale Res. Lett. 6, p. 106 (2011). 17. M. Dovrat, Y. Goshen, J. Jedrzejewski, I. Balberg, A. Sa’ar, Radiative versus nonradiative decay process in silicon nanocrystals probed by time- resolved photoluminescence spectroscopy // Phys. Rev. B, 69, 155311 (2004). 18. R. Sato, K. Murayama, A universal distribution function of relaxation in amorphous materials // Solid State Communs. 63, p. 625-627 (1987). 19. A.F. van Driel, I.S. Nikolaev, P. Vergeer et al., Statistical analysis of time-resolved emission from ensembles of semiconductor quantum dots: interpretation of exponential decay models // Phys. Rev. B, 75, 035329 (2007). 20. G. Hadjisavvas, P.C. Kelires, Structure and energetics of Si nanocrystals embedded in α-SiO2 // Phys. Rev. Lett. 93, p. 226104 (2004). 21. I. Mihalcescu, J.C. Vial, R. Romestain, Carrier localization in porous silicon investigated by time- resolved luminescence analysis // J. Appl. Phys. 80, p. 2404 (1996). 22. M. Bhaskaran, S. Sriram, T.S. Perova et al., In situ micro-Raman analysis and X-ray diffraction of nickel silicide thin films on silica // Micron, 40(1), p. 89-93 (2009). 23. H.F. Yan, Y.J. Xing, Q.L. Hang, D.P. Yu, Y.P. Wang et al., Growth of amorphous silicon nanowires via a solid-liquid-solid mechanism // Chem. Phys. Lett. 323, p. 224-228 (2000). 24. Y. He, K. Ma, I. Bi, J.Y. Feng, Z.J. Zhang, Nickel- induced enhancement of photoluminescence from Si-rich silica films // Appl. Phys. Lett. 88, 031905 (2006). Semiconductor Physics, Quantum Electronics & Optoelectronics, 2014. V. 17, N 4. P. 336-340. PACS 78.67.Bf, 78.55.Ap Nickel-induced enhancement of photoluminescence in nc-Si–SiOx nanostructures K.V. Michailovska, I.Z. Indutnyi, P.E. Shepeliavyi, V.A. Dan’ko V. Lashkaryov Institute of Semiconductor Physics, NAS of Ukraine 41, prospect Nauky, 03028 Kyiv, Ukraine E-mail: indutnyy@isp. kiev.ua Abstract. The effect of nickel silicide interlayer on the intensity of photoluminescence (PL) from Si nanoclusters (nc) in normally deposited and obliquely deposited in vacuum SiOx/Ni/Si structures have been studied using spectral and time-resolved PL measurements. It has been shown that the intensity of PL band in SiOx/Ni/Si samples is essentially higher than that in reference SiOx/Si samples (without the nickel interlayer) with the same characteristics and treatment. The PL intensity enhancement factor is equal to 5.77 for normally deposited samples and 18 for obliquely deposited samples. The unchanged spectral shape of PL bands and similar position of PL maximum in samples with and without nickel silicide interlayer indicates that in the SiOx/Ni/Si structures after annealing no additional emitting centers are introduced to compare with reference one. Time-resolved measurements showed that PL decay rate was decreased from 8.2(104 s–1 for SiOx/Si specimens to 2.86(104 s–1 for SiOx/Ni/Si one. The emission decay rate distribution was determined by fitting the experimental decay curves to the stretched-exponential model. The observed narrow decay rate distribution, decrease of the PL decay rate and enhancement of the PL intensity in SiOx/Ni/Si samples can be assigned to the processes of nickel silicide passivation of the dangling bonds at the interface of Si nanoparticles and the silicon oxide matrix, which is more effective in porous samples. Keywords: time-resolved photoluminescence, thin film, silicon oxide, nanoparticle, nickel silicide. Manuscript received 22.04.14; revised version received 16.07.14; accepted for publication 29.10.14; published online 10.11.14. 1. Introduction Thin-film structures containing Si nanoparticles embedded into SiOx matrix attract attention of many researchers because of their promising applications in advanced electronic and optoelectronic devices [1-5]. But in spite of intensive researches over the past years, the reported efficiencies of nc-Si–SiOx light-emitting structures are still low and not high enough for practical application. The most important factors influencing the characteristics of PL are nanoparticle size and state of nanoparticle–SiOx interface. The passivation of nonradiative states and defects at this interface is an essential requirement in order to increase the intensity of PL. Hydrogen passivation through standard forming gas annealing is impractical for device application because hydrogen is easy-to-dissociate at elevated temperatures, which leads to the invalidation of hydrogen passivation. The Si–SiOx interface can be modified by chemical compounds of necessary composition. Such treating is the most efficient in porous structures. Recently [6, 7], we have proposed the method of porous nc-Si–SiOx light-emitting structure formation using oblique deposition of Si monoxide (SiO) in vacuum. The electron microscopy studies show that, during this deposition, SiOx films with a porous (column-like) structure are formed. During high-temperature annealing of these films, the thermally stimulated formation of Si nanoinclusions occurs in a restricted volume of the SiOx column. Because of free space (cavities) between the oxide column, the structures is more susceptible to chemical treatments, e.g., to treatment with HF solution or vapor [8]. As a result of HF vapor treatment, approximately 200-fold increase in the PL intensity is observed. It was reported also [3] that introduction of thin nickel interlayer between SiOx films and Si substrates results in enhancing of the PL intensity of SiOx/Ni/Si structures by a factor of 4 in comparison with samples without any Ni interlayer. Using the examples of light-emitting diodes based on SiO1.56/Ni/Si structures, it was shown that NiSi2 distribution in SiO1.56 film could improve turn-on voltage and give a benefit to the electroluminescence efficiency [3]. In this paper, we report the results of studying the nickel induced enhancement of PL emission in porous and solid nc-Si–SiOx light-emitting nanostructures. 2. Experimental A thin (~12 nm) layer of Ni film was deposited by thermal evaporation of nickel powder onto (100)-oriented Si wafer. After deposition, the samples were annealed in vacuum at 450 °C for 10 min. The annealing provides the required energy for the system to overcome the energy barriers for solid state reaction of formation of intermediate nickel silicides [5]. Then, silicon oxide (SiOx) film was deposited onto the nickel silicide film by thermal evaporation of Cerac. Inc. SiO with 99.9% purity in the vacuum chamber (the residual pressure ( ) Pa 10 2 ... 1 3 - × ). Before SiOx deposition, the substrates were oriented at the angles (α) of 60° or 0° (normal deposition) between the normal to the substrate surface and direction to the evaporator. Because of the additional oxidation by residual gases during evapo-ration of SiO the compositionally nonstoichiometric SiOx (x ~1.25 for normal deposition and x ~1.54 for 60°) films were obtained. The thickness of the SiOx films were monitored in situ by the quartz-crystal-oscillator monitor system (KIT-1) and measured after deposition by a microinterferometer (MII-4). The thickness of normally deposited SiOx films was equal to 400 nm, deposited at 60° – 700…800 nm. The samples without Ni interlayer were used as the reference ones in these measurements. As-deposited samples with and without Ni film were annealed simultaneously in vacuum at the temperature 975 °C for different duration ranging from 4 to 15 min. This high-temperature annealing leads to decomposition of silicon oxide (where x changed from 1.25 to 1.95 for samples obtained under normal deposition and from 1.54 to 2.0 for 60° ones) and formation of Si nanoparticles embedded into such oxide matrix [9, 10]. (The composition of the oxide matrix in as-deposited and annealed samples (parameter x) were determined using compositional dependence of the position of the main IR band in spectra of SiOx layers within the range of 1000 to 1100 cm–1, as it was ascer-tained in [11]. This band corresponds to the Si–O–Si stretching mode. FTIR measurements were carried out with Perkin-Elmer Spectrum BXII spectrometer.) The structure of obliquely deposited SiOx films was studied in our previous paper [12] by SEM apparatus ZEISS EVO 50XVP. Such SiOx films have a porous inclined column-like structure with the column diameters of 10 to 100 nm. The porosity of films depends on the angle of deposition and equals to ~40% for α = 60°. High-temperature annealing of these films does not change the porosity and column-like structure of the samples [7]. The PL spectra of obtained SiOx/Ni/Si and SiOx/Si samples were recorded at room temperature within the wavelength range 440 to 900 nm by using a system based on ZMR-3 monochromator equipped with a photomultiplier and detection system. The PL spectra were normalized to the spectral sensitivity of the experimental setup. The 337-nm line of a nitrogen laser with a spot size of about 2 mm in diameter was employed to excite the PL. Decay curve measurements were performed using the same N2 laser with the pulse duration 9 ns, which was short as compared to PL average lifetimes of our samples (approximately tens of microseconds). The PL lifetime was measured at different emission wavelengths. The time trace was recorded with a resolution of 0.5 μs. 3. Results and discussion Fig. 1 presents the PL spectra of normally deposited SiOx/Ni/Si samples after annealing at 975 °C for 15 and 4 min (curve 1 and 2, accordingly). Fig. 1 shows also PL spectra of the reference normally deposited SiOx/Si sample (curve 3) under annealing at the same temperature for 15 min. At room temperature, all the samples exhibited a broad band with a maximum of emission centered within 760…780 nm. This strong near-infrared luminescence band within the wavelength range 600…900 nm is associated with quantum confinement effects, that is with electron-hole pairs (or exciton) recombination in nc-Si. The intensity of emission from SiOx/Ni/Si samples IPL(Ni) is enhanced significantly in comparison with the reference sample (IPL) and depends on the annealing time. So, the PL enhancement factor η = ( ) PL PL I I Ni measured at 780 nm increases from 2.45 to 5.77 for 4- and 15-min annealing time, accordingly. Fig. 2 shows the PL spectra for the SiOx/Ni/Si (curve 1) and SiOx/Si (curve 2) samples deposited at 60° and annealed at 975 °C for 15 min in vacuum. The emission spectrum of both samples exhibit a broad band with the peak position at 630 nm. A spectral shift of PL peaks to shorter wavelengths in obliquely deposited samples in comparison with normally deposited ones can be caused by decrease in nc-Si dimensions. The size of nc-Si in these systems depends on the silicon content in deposited SiOx layers [13]: decrease in the Si content (i.e., increase of x) results in decreased nc-Si sizes and blue shift of PL spectrum. Besides, during high-temperature annealing of porous films with column-like structure the thermally stimulated formation of nc-Si occurs in a restricted volume of the SiOx columns. It can also cause the reduction of the nc-Si size and blue shift of the PL peak. As seen from Figs 1 and 2, addition of Ni interlayer results in enhancement of the PL intensity more prominent in porous samples. For porous, obliquely deposited samples, the enhancement factor η measured at 630 nm is equal to 18. 600650700750800850900 0.0 0.1 0.2 0.3 0.4 0.5 0.6 I PL , arb. units nm 3 2 1 – Fig. 1. PL spectra of normally deposited samples: SiOx/Ni/Si after annealing in vacuum at 975 °C for 15 (1) and 4 min (2) and the reference SiOx/Si sample after annealing at 975 °C for 15 min (3). 500550600650700750800 0 2 4 6 8 10 I PL , arb. units ,nm 1 2 Fig. 2. PL spectra of SiOx/Ni/Si (1) and reference of SiOx/Si (2) samples obliquely deposited at the angle 60° after annealing in vacuum at 975 °C for 15 min. It is known that, in a low excitation regime, the PL intensity is generally given by the expression IPL ~ (τPL/τR)σφN, where τPL and τR are the photo-luminescence and radiative lifetimes, respectively, φ is the photon flux of the laser pump (constant throughout our experiment), N is the density of nanocrystals in the film, and σ is their excited cross-section [17]. The unchanged spectral shape of PL bands and similar position of PL maximum in samples with and without Ni interlayer mean that in the SiOx/Ni/Si structures after annealing no additional emitting centers are introduced to compare with the reference one. In the quantum confinement scheme, the light emission from nc-Si is caused by radiative recombination of electron-hole pairs (or excitons) confined within nanoparticle [15]. It is deduced that the radiative lifetimes τR would be close to each other in SiOx/Ni/Si and SiOx/Si samples due to the same luminescence scheme and similar size of nc-Si (similar position of the PL maximum). One can suppose that nc-Si density (N) and PL lifetime (τPL) are the main factors changing the PL intensity in SiOx/Ni/Si samples. More direct demonstration of enhanced electron-hole pair recombination involved comparative measurements of the PL decay rate in the investigated structures. Time-resolved PL measurements were performed using the normally deposited SiOx/Ni/Si and SiOx/Si samples. Fig. 3 shows PL decay curve for SiOx/Ni/Si (curve 1) and SiOx/Si (curve 2) samples at 780 nm. One can see that the PL intensity for the SiOx/Ni/Si samples decayed slower than for the reference samples. The obtained decay curves of the PL intensity may be described well by a stretched exponential function: ( ) ú ú û ù ê ê ë é ÷ ÷ ø ö ç ç è æ t - = b - b 0 1 exp t t C t I PL , (1) where C, τ0 and β are some constant, decay time and stretched parameter (0 < β ≤ 1), respectively. It is known that the stretched exponential function is widely used for the decay curve analysis of Si nanocrystals [16]. The least-squares fit of Eq. (1) to experimental data brings values of τ0 and β. The obtained decay times τ0 were equal to 35 and 12.2 μs for SiOx/Ni/Si and SiOx/Si samples, respectively. This range of PL lifetimes is comparable with those reported in the literature for silicon nanocrystals at room temperature [17]. The PL decay rate k ( ) 1 0 - t = k at 780 nm is decreased from 8.2(104 s–1 for SiOx/Si to 2.86(104 s–1 for SiOx/Ni/Si samples. It was also determined that the dispersion parameter β is 0.83 for SiOx/Ni/Si and 0.74 for SiOx/Si samples. In general, the parameter β is related to the stretching of the decay process and is a direct measure of the width of the decay rate distribution. In the case of stretched exponential relaxation function, the PL decay may be analyzed more thoroughly by recovering the distribution of recombination rates [16]. 0 10 20 30 40 0.0 0.2 0.4 0.6 0.8 1.0 I PL , arb. units t, s 1 2  ex = 337 nm  em = 780 nm T= 295 K Fig. 3. PL decay curves measured at 780 nm for the normally deposited SiOx/Ni/Si (1) and SiOx/Si (2) samples after annealing at 975 °C for 15 min. 0.000.050.100.150.200.250.300.350.40 0.0 0.2 0.4 0.6 0.8 1.0 Distribution Ф(k) Decay rate, s -1 without Ni 2 1 with Ni Fig. 4. Decay rate distributions of SiOx/Ni/Si (1) and SiOx/Si (2) normally deposited samples obtained from the stretched-exponential decay model. The decay rate distributions are normalized by the peak value of Ф(k) for both samples. Using the values of τ0 and β measured at λ = 780 nm, we calculated the asymptotic form of the decay rates probability density function Ф(k) that might be obtained by the saddle-point method [18]: ( ) ( ) [ ] a a k k a k Ф - - - t - × t × pb t = exp 2 ) ( 2 / 1 , (2) where ( ) 1 1 - b - b = a and ( ) [ ] 1 / 1 0 1 - b - b t = t a . Fig. 4 shows the Ф(k) distributions calculated from Eq. (2) for SiOx/Ni/Si (curve 1) and SiOx/Si (curve 2) samples. The obtained distribution for the reference SiOx/Si sample is very broad with a long tail directed towards shorter lifetimes, which demonstrates the strongly non-single exponential character of decay curves. In SiOx/Ni/Si samples, the decay rate distribution Ф(k) is more narrow as compared with SiOx/Si sample and shifts towards lower decay rates. To explain the observed feature, it should be mentioned that the obtained Ф(k) function provides information about both the radiative (kR) and nonradiative (kNR) relaxation rates [19]. A very low quantum efficiency of nc-Si emission suggests that nonradiative processes should be predominant (kNR >> kR). It allows us to relate the changes observed in the decay rate distribution Ф(k) to the different quantity of defect states (dangling silicon bonds) in the matrix containing nc-Si. We believe that the interface between nc-Si and the SiOx matrix plays a crucial role in distributing decay rates [20]. SiO2 is an ideal matrix for nc-Si, as it can passivate a large fraction of the dangling Si bonds. There is a remarkable content of broken silicon bonds in SiOx films, which may act as nonradiative recombination paths for the excited carriers and cause the quenching of PL [21]. The observed narrow decay rate distribution and the increase of the PL decay time in SiOx/Ni/Si samples can be related to the processes of nickel silicide passivation of nonradiative recombination centers. It has been reported previously that nickel silicide thin films were formed by vacuum thermal processing (350-750 °C) of Ni thin films deposited onto (100) p-type Si substrates [4]. Ni as the mobile species moves through NiSi during the transition from NiSi to the NiSi2 phase at 750…1000 °C [22]. NiSi2 distributed in SiOx film passivates the broken bonds on the nc-Si surface [23]. The model of NiSi2 passivation was proved by thermodynamic analysis and Fourier transform infrared spectroscopy in SiO1.56/Ni/Si systems [3]. In the porous SiOx/Ni/Si samples, the nickel diffusion processes significantly accelerated due to the presence of voids and the surface of nanocolumns. This leads to more effective passivation of the nonradiative recombination centers and more significant increase in the PL emission intensity. In conclusion, we can note that the PL intensity enhancement factor calculated from the experimental results for SiOx/Ni/Si and SiOx/Si samples annealed at 975 °C is η = 5.77 for normal deposited samples. Enhancement of PL lifetimes in these samples by the factor ~2.87 due to the processes of nickel silicide passivation is not enough for explanation of the increase in PL intensity. Therefore, we proposed that the PL intensity enhancement can be caused both by increasing the PL lifetime and also by increasing the nc-Si density. The presence of Ni gives an additional driving force to the separation process of SiOx which was discussed in the previous paper [24]. 5. Conclusion In summary, we have presented the effect of Ni on the PL emission from Si nanoparticles embedded in the silicon oxide matrix. It has been shown that the intensity of near-infrared emission band in SiOx/Ni/Si samples was significantly higher than that in the samples without the Ni interlayer. It was assumed that nickel in SiOx film may passivate the residual dangling bonds at the interface of nc-Si and SiOx matrix. Time-resolved PL measurements showed the decrease in the PL decay rate in SiOx/Ni/Si samples as compared with the SiOx/Si one. Decrease in the PL decay rate and increase in the density of nc-Si could be the main factors enhancing the PL intensity in nanostructures with nickel silicide interlayer. References 1. K.S. Min, K.V. Shcheglov, S.M. Yang, H.A. Atwater et al., Defect related versus excitonic visible light emission from ion beam synthesized Si nanocrystals // Appl. Phys. Lett. 69, p. 2033-2035 (1996). 2. M. Fujii, A. Mimura, S. Hayashi, K. Yamammoto et al., Improvement in photoluminescence efficiency of SiO2 films contaning Si nanocrystals by P doping // J. Appl. Phys. 87, p. 1855-1857 (2000). 3. D.X. Li, Y. He, J.Y. Feng, The study of nickel-induced enhancement of near-infrared luminescence in Si-rich silicon oxide films // Physica E, 41, p. 812-816 (2009). 4. M. Tinani, A. Mueller, Y. Gao et. al., In situ real-time studies of nickel silicide phase formation // J. Vac. Sci. Technol. B, 19(2), p. 376-383 (2001). 5. F.F. Zhao, J.Z. Zheng, Z.X. Shen et al., Thermal stability of NiSi and NiSi2 thin films // Microelectron. Eng. 71(1), p. 104-111 (2004). 6. I.Z. Indutnyy, I.Yu. Maidanchuk, V.I. Min’ko, Visible photoluminescence from annealed porous SiOx films // J. Optoelectron. and Adv. Mater. 7, p. 1231-1236 (2005). 7. V.A. Dan’ko, I.Z. Indutnyy, I.Y. Maidanchuk et al., Formation of the photoluminescence structure based on SiOx porous films // Optoelektronika i poluprovodnikovaya tekhnika, 39, p. 65-72 (2004) (in Ukrainian). 8. I.Z. Indutnyi, E.V. Michailovskaya, P.E. Shepeliavyi and V.A. Dan’ko, Visible photoluminescence of selectively etched porous nc-Si–SiOx structures // Fizika tekhnika poluprovodnikov 44(2), p. 218-222 (2010) (in Russian) [Semiconductors, 44(2), p. 206-210 (2010)]. 9. V.A. Dan’ko, V.Ya. Bratus’, I.Z. Indutnyi, I.P. Lisovskyy, S.O. Zlobin, K.V. Michailovska, P.E. Shepeliavyi, Controlling the photo-luminescence spectra of porous nc-Si–SiOx structures by vapor treatment // Semiconductor Physics, Quantum Electronics & Optoelectronics, 13(4). p. 413-417 (2010). 10. V.Ya. Bratus’, V.A. Yukhimchuk, L.V. Berezhinsky et al., Structural transformation and silicon nanocrystallite formation in SiOx films // Semiconductors, 35(7), p. 821-826 (2001). 11. M. Nakamura, Y. Mochizuki, K. Usami et al., Infrared absorption spectra and compositions of evaporated silicon oxide (SiOx) // Solid State Communs. 50, p. 1079-1081 (1984). 12. I.Z. Indutnyi, K.V. Michailovska, V.I. Min’ko, P.E. Shepeliavyi, Effect of acetone vapor treatment on photoluminescence of porous nc-Si–SiOx nanostructures // Semiconductor Physics, Quantum Electronics & Optoelectronics, 12(2), p. 105-109 (2009). 13. D. Nesheva, C. Raptis, A. Perakis et al., Raman scattering and photoluminescence from Si nanoparticles in annealed SiOx thin films // J. Appl. Phys. 92, p. 4678-4683 (2002). 14. C. Garcia, B. Garrido, P. Pellegrino et al., Size dependence of lifetime and absorption cross section of Si nanocrysrals embedded in SiO2 // Appl. Phys. Lett. 82, p. 1595-1597 (2003). 15. C. Delerue, G. Allan, M. Lannoo, Theoretical aspects of the luminescence of porous silicon // Phys. Rev. B, 48, p. 11024-11036 (1993). 16. G. Zatrub, A. Podhorodecki, J. Misiewicz et al., On the nature of the stretched exponential photoluminescence decay for silicon nanocrystals // Nanoscale Res. Lett. 6, p. 106 (2011). 17. M. Dovrat, Y. Goshen, J. Jedrzejewski, I. Balberg, A. Sa’ar, Radiative versus nonradiative decay process in silicon nanocrystals probed by time-resolved photoluminescence spectroscopy // Phys. Rev. B, 69, 155311 (2004). 18. R. Sato, K. Murayama, A universal distribution function of relaxation in amorphous materials // Solid State Communs. 63, p. 625-627 (1987). 19. A.F. van Driel, I.S. Nikolaev, P. Vergeer et al., Statistical analysis of time-resolved emission from ensembles of semiconductor quantum dots: interpretation of exponential decay models // Phys. Rev. B, 75, 035329 (2007). 20. G. Hadjisavvas, P.C. Kelires, Structure and energetics of Si nanocrystals embedded in α-SiO2 // Phys. Rev. Lett. 93, p. 226104 (2004). 21. I. Mihalcescu, J.C. Vial, R. Romestain, Carrier localization in porous silicon investigated by time-resolved luminescence analysis // J. Appl. Phys. 80, p. 2404 (1996). 22. M. Bhaskaran, S. Sriram, T.S. Perova et al., In situ micro-Raman analysis and X-ray diffraction of nickel silicide thin films on silica // Micron, 40(1), p. 89-93 (2009). 23. H.F. Yan, Y.J. Xing, Q.L. Hang, D.P. Yu, Y.P. Wang et al., Growth of amorphous silicon nanowires via a solid-liquid-solid mechanism // Chem. Phys. Lett. 323, p. 224-228 (2000). 24. Y. He, K. Ma, I. Bi, J.Y. Feng, Z.J. Zhang, Nickel-induced enhancement of photoluminescence from Si-rich silica films // Appl. Phys. Lett. 88, 031905 (2006). © 2014, V. Lashkaryov Institute of Semiconductor Physics, National Academy of Sciences of Ukraine 336 _1477916050.unknown _1477916190.unknown _1475215692.unknown _1477915497.unknown _1475213648.unknown _1475214571.unknown _1475213639.unknown