applied sciences Article
Luminescent Properties of Silicon Nanocrystals:Spin on Glass Hybrid Materials Marco Antonio Vásquez-Agustín 1 , Orlando Cortazar-Martínez 1 , Alfredo Abelardo González-Fernández 1 , José Alberto Andraca-Adame 2 , Alfredo Morales-Sánchez 3, * and Mariano Aceves-Mijares 1, * 1
2 3
*
Departamento de Electronica, Instituto Nacional de Astrofísica, Óptica y Electrónica (INAOE), Tonantzintla, 72840 Cholula, Puebla, Mexico;
[email protected] (M.A.V.-A.);
[email protected] (O.C.-M.);
[email protected] (A.A.G.-F.) Instituto Politécnico Nacional, CNMN, Avenida Luis Enrique Erro S/N, Zacatenco, Delegación Gustavo Adolfo Madero, C.P. 07738 Ciudad de México, Mexico;
[email protected] Centro de Investigación en Materiales Avanzados SC, Unidad Monterrey-PIIT, 66628 Apodaca, Nuevo León, Mexico Correspondence:
[email protected] (A.M.-S.);
[email protected] (M.A.-M.); Tel.: +52-81-1156-0842 (A.M.-S.); +52-222-266-3100 (M.A.-M.)
Academic Editor: Paolo Minzioni Received: 15 October 2016; Accepted: 8 December 2016; Published: 13 January 2017
Abstract: The photoluminescence characteristics of films consisting of Si nanocrystals either coated with or embedded into Spin on Glass (SOG) were studied. Si nanocrystals showing red or blue luminescence when suspended in alcohol solution were obtained from porous silicon films. These were then either deposited in Si substrates and coated with SOG, or mixed in an SOG solution that was later spun on Si substrates. Both types of films were thermally annealed at 1100 ◦ C for three hours in N2 atmosphere. Transmission electron microscopy measurements showed a mean diameter of 2.5 nm for the Si nanocrystals, as well as the presence of polycrystalline Si nanoagglomerates. These results were confirmed by X-ray diffraction studies, which revealed the (111), (220) and (311) Bragg peaks in Si nanocrystals. Fourier transform infrared spectroscopy studies showed that the coated films present higher chemical reactivity, promoting the formation of non-stoichiometric SiO2 , while the embedded films behave as a stoichiometric SiO2 after the thermal annealing. The PL (photoluminescence) characterization showed that both embedded and coated films present emission dominated by the Quantum Confinement Effect before undergoing any thermal treatment. After annealing, the spectra were found to be modified only in the case of the coated films, due to the formation of defects in the nanocrystals/SiO2 interface. Keywords: Si-NCs (silicon nanocrystals); SOG; HRTEM (high resolution transmission electron microscopy); XRD (X-ray diffraction); FTIR (Fourier transform infrared spectroscopy); PL
1. Introduction In the past, many efforts have been devoted to the fabrication of light emitting films that are compatible with silicon technology [1]. One of the proposed approaches is to use dielectric matrices with silicon nanocrystals (Si-NCs) or nano-agglomerates as active media. Among several other techniques, chemical methods have been used to fabricate such materials [2–10]. The origin of the photoluminescence (PL) of Si-NCs embedded into oxide films is still controversial [11], but there are two principal models commonly used to explain the emission mechanism. One is known as the Quantum Confinement Effect (QCE), in which the radiative process is carried out by electron-hole recombination in the Si-NCs; silicon nanocrystals below 5 nm show an effect of a quasi-direct bandgap due to the boundary conditions [6,8,12,13]. The second model proposes the formation of defect-states Appl. Sci. 2017, 7, 72; doi:10.3390/app7010072
www.mdpi.com/journal/applsci
Appl. Sci. 2017, 7, 72
2 of 13
at the Si-NCs/SiO2 interface where radiative recombination occurs [10,14,15]. In the literature, there is enough evidence of SiO2 defects with specific emissions that simultaneous experiments of luminescence and detection of defects have been done during the years to link them. Just to mention a few, electron paramagnetic resonance (EPR), time resolved photoluminescence, and tunable laser excitation are some of the techniques used [16–18]. Two main different emissive films of this kind have been proposed: those that produce centers with luminescent characteristics as they are synthetized; and films where Si-NCs are obtained separately from the dielectric matrix. In the first method, temperatures of around 1100 ◦ C are needed to promote the nucleation and activation of radiative Si nanopoints, crystalline or not [3,4,19,20]. Examples of these materials are the silicon rich oxides obtained by chemical vapor deposition (CVD), Ionic implantation of Si in a dielectric matrix, and thermally treated chemical solutions [4,21–23]. These kind of silicon rich oxides show photoluminescence mainly in a range from 650 to 950 nm. However, if high energy is used to stimulate them, the range emission can be extended to the blue side of the electromagnetic spectrum. However, the high temperature annealing usually needed to promote the silicon-emissive center formation [3,4,19,20] complicates their potential incorporation in CMOS (complementary metal-oxide-semiconductor) integrated circuits. On the other hand, Si-NCs obtained separately by a variety of techniques and mixed with SiO2 solutions, such as Spin on Glass (SOG), have been of particular research interest because they are usually easy and low cost methods that do not need to be thermal treated to produce emissive films. In most of the cases, it has been reported that such films contain embedded Si-NCs embedded in the dielectric matrix. Their characteristic photoluminescence is rather on the blue side of the spectrum, varying with the diameter of the nanocrystals. Nevertheless, it has been reported that the emission depends on many factors, including the density of Si-NCs in the SOG, the way they are obtained and mixed, and the type and temperature used if they are annealed [24–27]. Then, despite the existence of several studies, much research is still to be done in order to fully understand the luminescence mechanisms, and the optimal technological conditions to obtain reliable highly emissive films. In this work, in order to corroborate the chemical interaction between the Si-NCs and the SOG film, two different films are proposed: Si-NCs coated with a thin layer of SOG and Si-NCs embedded into the SOG film—hereafter, respectively referred to as “coated” and “embedded” films. Moreover, the coated and embedded films were studied with and without thermal annealing at 1100 ◦ C to look deeply into its effects. 2. Experimental Details The Porous Silicon (pSi) samples were prepared by anodic etching of epitaxial, , n-type mono crystalline-silicon with an electrical resistivity between 2 Ω-cm and 0.025 Ω-cm. A solution mixture of hydrofluoric acid (44%), hydrogen peroxide (44%) and methanol (99.9%) with a volume ratio of 4:3:3 was used with a constant current of 40 mA. The pSi samples were obtained using a standard configuration of lateral anodizing in which samples were introduced at a rate of 4 mm/h into the solution [28]. Then, Si-NCs were obtained either scratching or milling the porous silicon layer. Scratching produces Si-NCs with an intense red emission, and milling produces blue PL emission [29]. The Si-NCs were then dispersed into isopropanol [(CH3 )2 CHOH)] to obtain colloidal solutions. The colloidal solutions were used to produce the coated or embedded test structures. Coated samples were fabricated by spraying the colloidal Si-NC solution onto Si substrates heated to 90 ◦ C producing a layer of Si-NCs (~70 nm-thick). Then, these Si-NCs films were coated with a pure SOG solution (synthetized from a 99.99% Sigma Aldrich Tetraethyl orthosilicate, St. Louis, MO, USA). The SOG solution was spun at 3000 rpm for 50 s. The coated samples are schematized in Figure 1a. The embedded films were obtained by mixing the colloidal solution with the SOG solution with a volume ratio of 3:1 (Si-NCs:SOG). Then, the SOG with embedded Si-NC solution was spun onto the silicon substrates at 3000 rpm for 50 s. These films are schematized in Figure 1b.
Appl. Sci. 2017, 7, 72 Appl. Sci. 2016, 6, 434
3 of 13 3 of 13
Figure 1. Schemes for: (a) coated; and (b) embedded films with red or blue Si-NCs (silicon Figure 1. Schemes for: (a) coated; and (b) embedded films with red or blue Si‐NCs (silicon nanocrystals). nanocrystals).
Both coated and embedded films were heated at 150 ◦ C to solidify the SOG and some of them were Both coated and embedded films were heated at 150 °C to solidify the SOG and some of them ◦ C in N atmosphere for 3 h. The thicknesses of all films were measured thermally annealed at 1100at were thermally annealed 1100 °C 2 in N2 atmosphere for 3 h. The thicknesses of all films were with a Gaertner L117 null ellipsometer (Gaertner Scientific Corporation, Skokie, IL, USA) using a measured with a Gaertner L117 null ellipsometer (Gaertner Scientific Corporation, Skokie, IL, USA) He-Ne incident laser of 632.8 nm wavelength and 2 mW of power. The coated and embedded films using a He‐Ne incident laser of 632.8 nm wavelength and 2 mW of power. The coated and embedded resulted with a thickness of ~190 nm The micro-structural properties of films resulted with a thickness of and ~190 ~260 nm nm, and respectively. ~260 nm, respectively. The micro‐structural both coated and embedded films were studied by transmission electron microscopy (TEM; Tecnai F30, properties of both coated and embedded films were studied by transmission electron microscopy FEI Company, OR, USA). Hillsboro, Grazing incidence X-ray diffraction data were obtained with an X (TEM; Tecnai Hillsboro, F30, FEI Company, OR, USA). Grazing incidence X‐ray diffraction data Pert MRD diffractometer The Netherlands). The measurements were PRO obtained with an X from Pert PANalytical PRO MRD (Almelo, diffractometer from PANalytical (Almelo, were The carried out with a Cu Kα (λ = 1.5418 Å) radiation as source. Theα (λ = 1.5418 Å) radiation as source. grazing angle of the incident beam Netherlands). The measurements were carried out with a Cu K ◦ was fixed at 0.5 and the data acquisition angle was varied between 15◦ and 60◦ with a resolution of The grazing angle of the incident beam was fixed at 0.5° and the data acquisition angle was varied ◦ 0.05 , and with an acquisition time per angular step of 800 s. A computer controlled spectrometer between 15° and 60° with a resolution of 0.05°, and with an acquisition time per angular step of 800 s. Bruker Fourier transform infrared (FTIR) model V22 (Bruker Optics Inc.,(FTIR) model V22 Billerica, MA, USA) was A computer controlled spectrometer Bruker Fourier transform infrared (Bruker used toInc., obtain the FTIR spectra theused films.to The absorption spectra were acquired fromabsorption 400 cm−1 Optics Billerica, MA, USA) of was obtain the FTIR spectra of the films. The −1 . The photoluminescence to 4000 cm werephotoluminescence carried out at roommeasurements temperature for all spectra were acquired from 400 cm−1 measurements to 4000 cm−1. The were samples using computer controlled Flouromax-3 (Horiba, Kyoto, Japan). carried out at aroom temperature for spectroflourometer all samples using model a computer controlled spectroflourometer The samples were excited withKyoto, a 300 nm wavelength and the emission recorded in the model Flouromax‐3 (Horiba, Japan). The samples were excited spectra with a were 300 nm wavelength range of 370 nm to 1000 nm. A 300 nm band pass filter was used in the excitation monochromatic and the emission spectra were recorded in the range of 370 nm to 1000 nm. A 300 nm band pass filter source and a 370 nm high pass filter was used for the PL detector. was used in the excitation monochromatic source and a 370 nm high pass filter was used for the PL detector. 3. Results and Discussion
3. Results and Discussion 3.1. Structural Properties 3.1. Structural Properties 3.1.1. Transmission Electron Microscopy Figure 2 shows the cross section TEM images of coated films with red light-emitting Si-NCs before 3.1.1. Transmission Electron Microscopy (A) and after (B) thermal annealing. Before annealing, a clear frontier between the crystalline Si-NCs Figure 2 shows the cross section TEM images of coated films with red light‐emitting Si‐NCs and amorphous SOG layers can be observed. Conversely, the frontier between the amorphous and the before (A) and after (B) thermal annealing. Before annealing, a clear frontier between the crystalline crystalline regions is practically lost after thermal annealing. Furthermore, it is then possible to observe Si‐NCs and amorphous SOG layers can be observed. Conversely, the frontier between the some amorphous barriers between the Si-NCs (white circle in the Figure 2B). This is probably because amorphous and the crystalline regions is practically lost after thermal annealing. Furthermore, it is the thermal treatment promotes the flow of the SOG through the layer of the Si-NCs. The insets of then possible to observe some amorphous barriers between the Si‐NCs (white circle in the Figure 2B). Figure 2B(b) shows an FFT (Fast Fourier Transform) performed to the TEM images (Figure 2B(a)) in This is probably because the thermal treatment promotes the flow of the SOG through the layer of order to determine the distance between atomic layers of Si-NCs. The nanocrystal lattice spacing in the Si‐NCs. The insets of Figure 2B(b) shows an FFT (Fast Fourier Transform) performed to the TEM high resolution transmission electron microscopy (HRTEM) images was estimated using the digital images (Figure 2B(a)) in order to determine the distance between atomic layers of Si‐NCs. The micrograph software (version 3.1.1, Gatan Inc., Pleasanton, CA, USA) to identify the orientation of the nanocrystal lattice spacing in high resolution transmission electron microscopy (HRTEM) images crystalline planes. The distances obtained correspond to the face-centered cubic lattice of crystalline was estimated using the digital micrograph software (version 3.1.1, Gatan Inc., Pleasanton, CA, USA) silicon (joint committee on powder diffraction standards (JCPDS) #27-1402) (see inset Figure 2B(d)). to identify the orientation of the crystalline planes. The distances obtained correspond to the
Appl. Sci. 2016, 6, 434 Appl. Sci. 2016, 6, 434
4 of 13 4 of 13
face‐centered cubic lattice of crystalline silicon (joint committee on powder diffraction standards Appl. Sci. 2017, 7, 72 4 of 13 face‐centered cubic lattice of crystalline silicon (joint committee on powder diffraction standards (JCPDS) #27‐1402) (see inset Figure 2B(d)). (JCPDS) #27‐1402) (see inset Figure 2B(d)).
(A) (A)
(B) (B)
Figure 2. 2. Cross‐section red Figure Cross-section TEM TEM (transmission (transmission electron electron microscopy) microscopy) images images of of coated coated films films with with red Figure 2. Cross‐section TEM (transmission electron microscopy) images of coated films with red 2 atmosphere. In both images, Si‐NCs before (A) and after (B) thermal annealing at 1100 °C for 3 h in N ◦ Si-NCs before (A) and after (B) thermal annealing at 1100 C for 3 h in N2 atmosphere. In both 2 atmosphere. In both images, Si‐NCs before (A) and after (B) thermal annealing at 1100 °C for 3 h in N the white marks the area analysis where where there there are Si‐NCs. The insets of (B) (a) images, thesquare white square marks theof area of analysis are Si-NCs. The insets of are: (B) are: the white square marks the area of analysis where there are Si‐NCs. The insets of (B) are: (a) cross‐section TEM images; (b) (b) FFT (Fast Fourier white (a) cross-section TEM images; FFT (Fast FourierTransform); Transform);(c) (c)the thespecific specificpattern pattern of of the the white cross‐section TEM images; (b) FFT (Fast Fourier Transform); (c) the specific pattern of the white square area; and (d) the electrons diffraction planes of the white square area. The white circle in the square area; and (d) the electrons diffraction planes of the white square area. The white circle in the square area; and (d) the electrons diffraction planes of the white square area. The white circle in the Figure 2B shows a zone where the SOG (Spin on Glass) flowed around the Si‐NC after annealing. Figure 2B shows a zone where the SOG (Spin on Glass) flowed around the Si-NC after annealing. Figure 2B shows a zone where the SOG (Spin on Glass) flowed around the Si‐NC after annealing.
Due to the nature of the coated films, it is not possible to observe the shape of Si‐NCs. However, Due to the nature of the coated films, it is not possible to observe the shape of Si-NCs. However, in Due to the nature of the coated films, it is not possible to observe the shape of Si‐NCs. However, in the case of films with embedded red light‐emitting Si‐NCs (Figure 3A), it is possible to observe the case of films with embedded red light-emitting Si-NCs (Figure 3A), it is possible to observe nearly in the case of films with embedded red light‐emitting Si‐NCs (Figure 3A), it is possible to observe nearly spherical particles, as typically produced by the electrochemical technique [28]. The spherical particles,particles, as typically produced by the electrochemical technique [28]. The nanocrystals’ nearly spherical as from typically produced by with the a electrochemical technique [28]. The nanocrystals’ diameter ranges 1.5 nm to 4.5 nm mean size of 2.5 nm. In addition, a diameter ranges from 1.5 nm to 4.5 nm with a mean size of 2.5 nm. In addition, a variety of nanocrystals’ diameter ranges from 1.5 nm to 4.5 nm with a mean size of 2.5 nm. In addition, a variety of crystalline planes corresponding to polycrystalline agglomerates was discovered, as crystalline corresponding to polycrystalline agglomerates was discovered, observed as in variety of planes crystalline planes corresponding to polycrystalline agglomerates was as discovered, observed in the inset Figure 3A(d). the inset Figure 3A(d). observed in the inset Figure 3A(d). In the case of embedded films made with blue light‐emitting Si‐NCs, these cannot be clearly In the these cannot cannot be be clearly clearly In the case case of of embedded embedded films films made made with with blue blue light-emitting light‐emitting Si-NCs, Si‐NCs, these observed with the naked eye in the TEM images (Figure 3B(a)), making it difficult to select the area observed with the naked eye in the TEM images (Figure 3B(a)), making it difficult to select the area to observed with the naked eye in the TEM images (Figure 3B(a)), making it difficult to select the area to perform the FFT analysis. However, some diffraction patterns can be observed after processing perform the FFT analysis. However, some diffraction patterns can be observed after processing the to perform the FFT analysis. However, some diffraction patterns can be observed after processing the TEM images, as shown in the inset Figure 3B(d). These diffraction patterns indicate crystalline TEM images, as shown in the inset Figure 3B(d). These diffraction patterns indicate crystalline silicon the TEM images, as shown in the inset Figure 3B(d). These diffraction patterns indicate crystalline silicon with cubic structure. This was confirmed with the X‐ray measurements. with cubic structure. This was confirmed with the X-ray measurements. silicon with cubic structure. This was confirmed with the X‐ray measurements.
(A) (A)
(B) (B)
Figure 3. Cross‐section TEM images of embedded films with red (A) and blue (B) Si‐NCs, after Figure 3. Cross-section Cross‐section TEM images of embedded red (A) and (B) blue (B) Si‐NCs, after 2 atmosphere. In both images, the white square marks the thermal annealing at 1100 °C for 3 h in N Figure 3. TEM images of embedded filmsfilms withwith red (A) and blue Si-NCs, after thermal 2 atmosphere. In both images, the white square marks the thermal annealing at 1100 °C for 3 h in N ◦ annealing at 1100 C for 3 h in N2 atmosphere. In both images, the white square marks the area of area of analysis where there is a Si‐NCs. The insets are: (a) the cross‐section TEM images; (b) FFT; (c) area of analysis where there is a Si‐NCs. The insets are: (a) the cross‐section TEM images; (b) FFT; (c) analysis where there is a Si-NCs. The insets are: (a) the cross-section TEM images; (b) FFT; (c) the the specific pattern of the white square area; and (d) the electrons’ diffraction planes of the white the specific pattern of the white square area; and (d) the electrons’ diffraction planes of the white specific pattern of the white square area; and (d) the electrons’ diffraction planes of the white square square areas, for each image. square areas, for each image. areas, for each image.
Appl. Sci. 2017, 7, 72 Appl. Sci. 2016, 6, 434
5 of 13 5 of 13
3.1.2. Grazing Incidence X‐Ray Diffraction 3.1.2. Grazing Incidence X-Ray Diffraction Figure 4 shows the X‐ray diffraction patterns of coated (a) and embedded films (b) before and Figure 4 shows the X-ray diffraction patterns of coated (a) and embedded films (b) before and after annealing. AA well‐defined main peak is observed 28.4°, well as peaks lower ◦ , as after annealing. well-defined main peak is observed at 28.4at well as as peaks with lowerwith intensities intensities at 47.3° and 56.1°, which respectively correspond to (111), (220), and (311) Bragg peaks for at 47.3◦ and 56.1◦ , which respectively correspond to (111), (220), and (311) Bragg peaks for silicon. silicon. These peaks match very well the PDF JCPDS #27‐1402 values for cubic crystalline structure These peaks match very well the PDF JCPDS #27-1402 values for cubic crystalline structure of silicon. of silicon. The diffraction plane corresponds to a face centered cubic lattice as the obtained by the The diffraction plane corresponds to a face centered cubic lattice as the obtained by the TEM analysis TEM analysis (Figures 2 and 3). These peaks indicate that the coated and embedded films contain (Figures 2 and 3). These peaks indicate that the coated and embedded films contain Si-NCs with a face Si‐NCs with a face centered cubic lattice structure. centered cubic lattice structure.
~21.5° ~23°
28°
15
20
25
30
47°
35
40
45
50
28°
60
15
20
25
30
(311)
(220)
(111)
56°
55
Embedded with red Si-NCs Annealed Embedded with blue Si-NCs Annealed
(b)
Intensity (a. u.)
(220)
(111)
Intensity (a. u.)
(311)
Coated with red Si-NC Annealed Coated with blue Si-NC Annealed
(a)
47°
35
40
45
50
56°
55
60
65
Figure 4. Grazing incidence XRD (X‐ray diffraction) (0.5°) patterns from coated (a) and embedded Figure 4. Grazing incidence XRD (X-ray diffraction) (0.5◦ ) patterns from coated (a) and embedded films; (b) with red and blue light‐emitting Si‐NCs, before and after annealing at 1100 °C for 3 h in N films; (b) with red and blue light-emitting Si-NCs, before and after annealing at 1100 ◦ C for 3 h in2 atmosphere. N2 atmosphere.
As observed in Figure 4a, a wide peak centered at ~23° can be seen in as‐deposited samples, As observed in Figure 4a, a wide peak centered at ~23◦ can be seen in as-deposited samples, which shifts to a lower angle of 21.5° after thermal annealing for coated films with both red and blue which shifts to a lower angle of 21.5◦ after thermal annealing for coated films with both red and blue Si‐NCs. This shift has been associated with the formation of amorphous silicon oxide (a‐SiO2) due to Si-NCs. This shift has been associated with the formation of amorphous silicon oxide (a-SiO2 ) due spinoidal decomposition [30]. Therefore, the structural units of the SOG, which can be described by to spinoidal decomposition [30]. Therefore, the structural units of the SOG, which can be described the different tetrahedral units SiO4‐xSix with n = 0−4, gradually transform into well‐defined SiO4 and by the different tetrahedral units SiO4-x Six with n = 0−4, gradually transform into well-defined SiO4 Si‐Si4 tetrahedral structures [23,24]. This transformation, only observed in the coated films, agrees and Si-Si4 tetrahedral structures [23,24]. This transformation, only observed in the coated films, agrees with the TEM observation of the SOG flowing into the Si‐NCs layer after thermal annealing. with the TEM observation of the SOG flowing into the Si-NCs layer after thermal annealing. 3.1.3. Fourier Transform Infrared Spectroscopy 3.1.3. Fourier Transform Infrared Spectroscopy Figure Figure 5 5 shows shows the the typical typical FTIR FTIR spectra spectra before before and and after after annealing annealing of of a a pure pure SOG SOG film film (a), (a), embedded films (b), and coated films (c) containing red light‐emitting Si‐NCs. The same behavior embedded films (b), and coated films (c) containing red light-emitting Si-NCs. The same behavior was was observed in coated and embedded films with blue light‐emitting Si‐NCs. observed in coated and embedded films with blue light-emitting Si-NCs. −1 The typical IR (infrared) vibration modes from silicon oxide films [31,32] at 1078 cm The typical IR (infrared) vibration modes from silicon oxide films [31,32] at 1078 cm−1 ( Si‐O‐Si ( Si-O-Si −1 −1 stretching), 800 cm cm−1 (O‐Si‐O bending), cm−1 (Si‐O‐Si rocking) can be observed in all of the stretching), 800 (O-Si-O bending), and and 457 457 cm (Si-O-Si rocking) can be observed in all of the −1 cases. All the absorption spectra show a peak at 1400 cm 2 stretching bonds, as well −1 due to the Si-CH cases. All the absorption spectra show a peak at 1400 cm due to the Si‐CH 2 stretching bonds, as −1 (Si‐OH as some other peaks related to vibration modes associated with hydrogen bonds at 935 cm well as some other peaks related to vibration modes associated with hydrogen bonds at 935 cm−1 stretching) and >3200 (H 2O) cm−1. The latter decreases in intensity or disappears after the thermal − 1 (Si-OH stretching) and >3200 (H2 O) cm . The latter decreases in intensity or disappears after the annealing [31,32]. It is noticeable that both before and after annealing, the IR spectra of the pure SOG thermal annealing [31,32]. It is noticeable that both before and after annealing, the IR spectra of the film are similar to the IR spectra obtained from the coated and embedded films with Si‐NCs (Figure pure SOG film are similar to the IR spectra obtained from the coated and embedded films with Si-NCs 5b,c). 5b,c). (Figure Figure 6 shows the IR peak related to the Si‐O‐Si asymmetric stretching mode deconvoluted in Figure 6 shows the IR peak related to the Si-O-Si asymmetric stretching mode deconvoluted in their longitudinal (LO) and transversal (TO) modes of the different samples, in order to qualitatively their longitudinal (LO) and transversal (TO) modes of the different samples, in order to qualitatively assess the structural changes produced by the thermal annealing [7,33–35]. assess the structural changes produced by the thermal annealing [7,33–35].
Appl. Sci. 2017, 7, 72 Appl. Sci. 2016, 6, 434
Si-O(S)
OH
SiHx
Si-CH2 Si-O(S)
Embedded / red Si-NCs Annealed (b) OH
Si-O(B) Si-OH
Si-O(R)
SOG Film Annealed
(a)
CO2
500
Si-O(S)
Coated / red Si-NCs Annealed
SiHx
OH
(c) Si-CH2
CO2
Si-O(B) Si-OH
Si-O(R)
Absorbance (a. u.)
CO2
Si-O(B) Si-OH
Si-O(R)
6 of 13 6 of 13
1000 1500 2000 2500 3000 3500 4000 -1
Wavenumber (cm )
Figure 5. (Color (Color Online) Online) FTIR (Fourier transform infrared) spectra and annealing after annealing Figure 5. FTIR (Fourier transform infrared) spectra beforebefore and after at 1100 ◦at C 2 atmosphere of pure SOG film (a); embedded (b); and coated (c) films fabricated 1100 °C for 3 h in N for 3 h in N2 atmosphere of pure SOG film (a); embedded (b); and coated (c) films fabricated with with red Si‐NCs. red Si-NCs. −1 In all of the spectra in Figure 6, it is possible to see the stretching peak (at ~1078 cm In all of the spectra in Figure 6, it is possible to see the stretching peak (at ~1078 cm−1) and the ) and the −1 −1 asymmetric stretching peak (from 1100 cm to 1200 cm ). Notice that these spectra show a shift to − 1 − 1 asymmetric stretching peak (from 1100 cm to 1200 cm ). Notice that these spectra show a shift to lower or higher wavenumbers depending on its structure or composition after annealing. lower or higher wavenumbers depending on its structure or composition after annealing. The stretching infrared (IR) peak observed in coated samples with red and blue Si‐NCs shifts The stretching infrared (IR) peak observed in coated samples with red and blue Si-NCs shifts towards towards a a lower lower wavenumber wavenumber after after thermal thermal annealing, annealing, as as observed observed in in Figure Figure 6a,b, 6a,b, respectively. respectively. However, this IR band shifts towards the opposite direction for the embedded red and blue Si‐NCs However, this IR band shifts towards the opposite direction for the embedded red and blue Si-NCs films, as observed in Figure 6c,d, respectively. A deconvolution of the IR stretching peak (symmetric films, as observed in Figure 6c,d, respectively. A deconvolution of the IR stretching peak (symmetric and and asymmetric) asymmetric) in in the the LO LO and and TO TO modes modes shows shows a a change change in in the the full full width width at at a a half half maximum maximum −1 (FWHM) associated with the formation of amorphous phases of SiO 2 (peak at ~1025 cm− (FWHM) associated with the formation of amorphous phases of SiO2 (peak at ~1025 cm 1) ) in in the the Si‐NCs/SOG interface after thermal annealing. This is an indication of a possible combination Si-NCs/SOG interface after thermal annealing. This is an indication of a possible combination of of non‐stoichiometric and stoichiometric SiO non-stoichiometric and stoichiometric SiO22 [7,36]. The Si‐NCs could react with the oxygen atoms of [7,36]. The Si-NCs could react with the oxygen atoms of the SOG due due toto the the silicon atoms presenting a higher probability of having oxygen neighboring the SOG silicon atoms presenting a higher probability of having oxygen neighboring atoms, −1 and 13.7 atoms, forming Si=O related bonds [14]. This is confirmed by the shift of about 16.4 cm forming Si=O related bonds [14]. This is confirmed by the shift of about 16.4 cm−1 and 13.7 cm−1 to −1 to lower wavenumbers of the stretching Si‐O‐Si peak for each coated film, as observed in Figure cm lower wavenumbers of the stretching Si-O-Si peak for each coated film, as observed in Figure 6a,b. 6a,b. In addition, the observed broadening IR stretching peak be could be to related to between a strain In addition, the observed broadening of the of IR the stretching peak could related a strain between the Si‐NCs and the SiO 2‐based SOG film, producing the formation of a high defect density the Si-NCs and the SiO2 -based SOG film, producing the formation of a high defect density after the after the thermal annealing in N 2 atmosphere [37]. thermal annealing in N2 atmosphere [37]. On the other hand, the shift of the IR stretching peak found in the embedded Si‐NCs towards On the other hand, the shift of the IR stretching peak found in the embedded Si-NCs towards higher wavenumbers after thermal annealing could be related to the formation of a stoichiometric higher wavenumbers after thermal annealing could be related to the formation of a stoichiometric and −1 and 16.2 cm−1 for red and blue embedded Si‐NCs film, and film (shifts cm16.2 −1 and densedense film (shifts of 15.9 of cm15.9 cm−1 for red and blue embedded Si-NCs film, respectively). respectively). Therefore, a structural rearrangement occurs these films annealing, due to the thermal Therefore, a structural rearrangement occurs in these films duein to the thermal improving annealing, improving the stoichiometry because of the phase separation between the Si and the stoichiometry because of the phase separation between the Si and SiO2 [34,38]. In addition,SiO the2 [34,38]. In addition, the decrease in the intensity from the IR asymmetric stretching peak (found decrease in the intensity from the IR asymmetric stretching peak (found between 1100 cm−1 and −1 between cm−1 can and be 1200 can of be the seen in the decrease of the longitudinal and 1 ), which 1200 cm−1100 seencm in ), thewhich decrease longitudinal and transversal optical modes, transversal optical modes, denotes stress relief in the Si‐NCs/SOG interface. denotes stress relief in the Si-NCs/SOG interface.
Appl. Sci. 2017, 7, 72 Appl. Sci. 2016, 6, 434
7 of 13 7 of 13
Mode (SiOx)
Absorbance (a. u.)
0.9 0.8 0.7 0.6
TO1
0.5
LO2
Shift=16.4
0.4
1.1
Coated/red Si-NC Annealed
(a)
TO2
1.0
LO1
TO1
(b)
0.9
Absorbance (a. u.)
1.0
0.8 0.7 0.6 0.5 0.4
0.3
0.3
0.2
0.2
0.1
0.1
LO2
900
950
900
1000 1050 1100 1150 1200 1250 1300 1350
950
1000 1050 1100 1150 1200 1250 1300 1350 -1
-1
Wavenumber (cm )
Wavenumber (cm ) 0.55
0.55
Embedded /red Si-NC Annealed
(c)
0.45
0.50
(d)
Embedded/blue Si-NC Annealed
TO1
0.45
TO1
0.40
Absorbance (a. u.)
Absorbance (a. u.)
TO2 LO1
Shift = 13.7
0.0
0.0
0.50
Coated/blueSi-NC Annealed
Cavidad (SiOx)
1.1
0.35 0.30 0.25
LO2
0.20
TO2
LO1
0.40 0.35
LO2 TO2
0.30
LO1
0.25 0.20 0.15
0.15 0.10
0.10
Shift=15.9
Shift=16.2
0.05
0.05
0.00
0.00 1000
1050
1100
1150
1200
Wavenumber (cm-1)
1250
1300
1000
1050
1100
1150
1200
1250
1300
-1
Wavenumber (cm )
Figure 6. (Color Online) FTIR spectra of coated and embedded films before and after annealing at Figure 6. (Color Online) FTIR spectra of coated and embedded films before and after annealing at ◦ 1100 °C for 3 h in N 1100 C for 3 h in N22 atmosphere. (a,b) are the IR (infrared) spectra of coated films with red and blue atmosphere. (a,b) are the IR (infrared) spectra of coated films with red and blue Si‐NCs, respectively; (c,d) the spectra IR spectra of embedded films with blue red Si-NCs, and blue Si‐NCs, Si-NCs, respectively; (c,d) areare the IR of embedded films with red and respectively. respectively.
3.2. Photoluminescence Properties 3.2. Photoluminescence Properties Si-NCs obtained through porous silicon samples exhibit a typical strong red PL emission with a Si‐NCs obtained through porous silicon samples exhibit a typical strong red PL emission with a quasi-Gaussian curve [31], which is usually attributed to the quantum confinement effect [1,6]. The PL quasi‐Gaussian curve [31], which is usually attributed to the quantum confinement effect [1,6]. The spectra obtained from the Si-NCs-based colloidal solutions for this work are shown in Figure 7a. The PL PL spectra obtained from the Si‐NCs‐based colloidal solutions for this work are shown in Figure 7a. spectrum of the red Si-NC colloidal solution has a main peak at around 639 nm (which is similar to The PL spectrum of the red Si‐NC colloidal solution has a main peak at around 639 nm (which is that of pSi layers; not shown), and a lower intensity secondary PL peak at 400 nm. As observed in similar to that of spectrum pSi layers; and solution a lower intensity secondary peak peaks at 400 atnm. As Figure 7a, the PL ofnot the shown), isopropanol without Si-NCs also PL exhibits about observed in Figure 7a, the PL spectrum of the isopropanol solution without Si‐NCs also exhibits 415 nm, but its intensity is very low as compared to the PL from samples with them. peaks at about 415 nm, but its intensity is very low as compared to the PL from samples with them. The blue Si-NC colloidal solution exhibits an intense PL band at ~400 nm and an additional PL bandThe blue Si‐NC colloidal solution exhibits an intense PL band at ~400 nm and an additional PL of lower intensity around 753 nm. The intense PL band centered at 400 nm is mainly associated band of lower intensity around 753 nm. The intense PL band centered at 400 nm is mainly associated with the blue Si-NCs, since the isopropanol luminescence is known to be too low. These PL emission with the blue Si‐NCs, since the isopropanol luminescence is known to be too low. These PL emission bands have been reported before [8,24,25]. bands have been reported before [8,24,25]. Figure 7b shows the PL spectra of the coated red and blue Si-NC films before thermal annealing. Figure 7b shows the PL spectra of the coated red and blue Si‐NC films before thermal annealing. Coated red Si-NCs show a broad PL band with a maximum peak at ~650 nm and a less intense band at Coated red Si‐NCs show a broad PL band with a maximum peak at ~650 nm and a less intense band ~415 nm, while the coated blue Si-NC films emit an intense PL band with the main peak at ~410 nm at ~415 nm, while the coated blue Si‐NC films emit an intense PL band with the main peak at ~410 and a very small secondary peak at ~800 nm. nm and a very small secondary peak at ~800 nm. to their diameter distribution and their energy The PL emission of the red Si-NCs was related The PL emission of the red Si‐NCs was related to their diameter distribution and their energy bandgap. Assuming a nearly spherical configuration of the nanocrystals, calculations of the energy bandgap. Assuming a nearly spherical configuration of the nanocrystals, calculations of the energy gap were made using the mathematical expression [39]: gap were made using the mathematical expression [39]:
Appl. Sci. 2017, 7, 72 Appl. Sci. 2016, 6, 434
88 of 13 of 13 2 2 EehEeh = 1.1eV 8∗m +hh22//(8* (8 ∗mmh R 1.1eV+hh22//((8* me R R2 )) h R) ,) ,
(1) (1)
where mee and m and mhh are the effective masses of the electron and hole, respectively; R is the radius of the are the effective masses of the electron and hole, respectively; R is the radius of the where m nanocrystals; and h is Planck’s constant. According to Equation (1), considering an me e = 1.08m = 1.08m00,, and a and a nanocrystals; and h is Planck’s constant. According to Equation (1), considering an m −31 kg is the free electron rest mass [40], the estimated diameter −31 kg is the free electron rest mass [40], the estimated diameter of the mhh = 0.56m = 0.56m0, where m 0 = 9.11 × 10 0 , where m 0 = 9.11 × 10 of the Si-NCs producing PL emission at 650 is 2.3 nm.This Thisis isconsistent consistentwith with the the diameter Si‐NCs producing a PL aemission at 650 nm nm is 2.3 nm. diameter distribution obtained from HRTEM images for red Si-NCs, which is between 1.5 nm to 4.5 nm, with a distribution obtained from HRTEM images for red Si‐NCs, which is between 1.5 nm to 4.5 nm, with mean of 2.5 nm. This supports the conclusion stating that the emission of red Si-NCs is mainly due a mean of 2.5 nm. This supports the conclusion stating that the emission of red Si‐NCs is mainly due to QCE. A similar result is obtained from blue colloidal solution and coated blue Si-NCs films before to QCE. A similar result is obtained from blue colloidal solution and coated blue Si‐NCs films before the thermal annealing. Using Equation (1) and the PL spectrum of the blue Si-NCs, the nanocrystals’ the thermal annealing. Using Equation (1) and the PL spectrum of the blue Si‐NCs, the nanocrystals’ diameter ranges from nm to Unfortunately, the observation diameter ranges from 1.3 1.3 nm to 1.9 1.9 nm nm with with aa mean mean of of 1.4 1.4 nm. nm. Unfortunately, the observation of of Si-NCs of less than 1.5 nm is not possible in the available HRTEM. Nevertheless, the presence of the Si‐NCs of less than 1.5 nm is not possible in the available HRTEM. Nevertheless, the presence of the blue Si-NCs was confirmed by XRD. Moreover, the PL emission from the colloidal solution is similar blue Si‐NCs was confirmed by XRD. Moreover, the PL emission from the colloidal solution is similar to inin the coated films. Therefore, the the lightlight emission mechanism in these coatedcoated films to the the one one observed observed the coated films. Therefore, emission mechanism in these is likely related to QCE as well. films is likely related to QCE as well. Isopropanol blue Si-NCs coloidal solution red Si-NCs coloidal solution
Intensity (a. u.)
(a) 6
5x10
6
3x10
6
2x10
0 5 1x100
x84 Coated with blue Si-NC´s Coated with red Si-NC´s
Intensity (a. u.)
(b) 4
9x10
4
6x10
4
3x10
0 400
500
600
700
Wavelength (nm)
800
900
Figure 7.7. (Color (Color Online) (Photoluminescence) spectra obtained at temperature room temperature from Figure Online) PL PL (Photoluminescence) spectra obtained at room from colloidal colloidal solution (a); and from coated films with red and blue Si‐NCs without annealing (b). solution (a); and from coated films with red and blue Si-NCs without annealing (b).
As observed in Figure 8, after thermal annealing, the PL emission spectra from both red and As observed in Figure 8, after thermal annealing, the PL emission spectra from both red and blue blue Si‐NCs coated films are broadened. It now presents a wavelength range from ~400 nm to 850 Si-NCs coated films are broadened. It now presents a wavelength range from ~400 nm to 850 nm. nm. This behavior indicates that, regardless of having different Si‐NCs, similar emission This behavior indicates that, regardless of having different Si-NCs, similar emission mechanisms take mechanisms take place in both samples after the thermal annealing, as discussed in a previous place in both samples after the thermal annealing, as discussed in a previous report [29]. report [29]. The thermal annealing produces important effects in the response of the material. In particular, The thermal annealing produces important effects in the response of the material. In particular, it results in an increase of the radiative surface states density at the Si-NCs/SiO2 interface [41,42], it results in an increase of the radiative surface states density at the Si‐NCs/SiO2 interface [41,42], as as discussed in the previous sections. The shift of the Si-O-Si stretching mode to a lower wavenumber discussed in the previous sections. The shift of the Si‐O‐Si stretching mode to a lower wavenumber indicates a phase change towards an amorphous (non-stoichiometric silicon oxide) and porous indicates a phase change towards an amorphous (non‐stoichiometric silicon oxide) and porous structure [33–37]. This could be related to the formation of a SiOx shell around silicon nanocrystals structure [33–37]. This could be related to the formation of a SiOx shell around silicon nanocrystals formed by a wide variety of Si-O bond defects. In a recent publication, Vaccaro et al. have corroborated formed by a wide variety of Si‐O bond defects. In a recent publication, Vaccaro et al. have the formation of a core shell structure with a crystalline core surrounded by an interface mainly corroborated the formation of a core shell structure with a crystalline core surrounded by an composed by Si3 O defects [43]. interface mainly composed by Si3O defects [43].
Appl. Sci. 2017, 7, 72 Appl. Sci. 2016, 6, 434
99 of 13 of 13
Wavelength (nm) 400 4
8x10
500
600
Coated/red Si-NCs Annealed
(a)
Wavelength (nm)
400
700 800 900
500
600
700 800 900
Coated/blue Si-NCs Annealed
(b) 4
4
PL intensity (a. u.)
PL intensity (a. u.)
8x10 E'
6x10
NOV NBOHC
4
4x10
WOB 4
2x10
E' 4
6x10
NOV WOB 4
4x10
NBOHC 4
2x10
DB
DB
0
0 3.2 3.0 2.8 2.6 2.4 2.2 2.0 1.8 1.6 1.4
3.2 3.0 2.8 2.6 2.4 2.2 2.0 1.8 1.6 1.4
Energy (eV)
Energy (eV)
Figure 8. (Color Online) PL emission of coated films with red (a) and blue (b) Si‐NCs after annealing Figure 8. (Color Online) PL emission of coated films with red (a) and blue (b) Si-NCs after annealing ◦ C°C 2. Compared to Figure 7, spectra these spectra are wider, and towards both shift the 1100 in 2 .NCompared atat 1100 in N to Figure 7, these are wider, and both shift thetowards blue-green blue‐green region of the visible spectrum. region of the visible spectrum.
The thermal annealing also appears to cause the loss of the H‐passivated characteristics of the The thermal annealing also appears to cause the loss of the H-passivated characteristics of the Si‐NCs, modifying its configuration into an oxygen‐passivated surface. This is corroborated by the Si-NCs, modifying its configuration into an oxygen-passivated surface. This is corroborated by the FTIR results (Figure 5c). Therefore, it is likely that part of the emission remains due to recombination FTIR results (Figure 5c). Therefore, it is likely that part of the emission remains due to recombination inside the nanocrystals, oxygen defects also contribute the broadening of the range of inside the nanocrystals, butbut oxygen defects also contribute to the to broadening of the range of emission emission in both types of coated films. in both types of coated films. The broad PL spectra were decomposed in Gaussian curves centered in the position of reported The broad PL spectra were decomposed in Gaussian curves centered in the position of reported Si‐O defect emissions, as shown in Figure 8a,b. As a result, PL bands at about 3.02 eV, 2.73 eV, 2.37 eV, Si-O defect emissions, as shown in Figure 8a,b. As a result, PL bands at about 3.02 eV, 2.73 eV, 1.99 eV, and 1.55 eV, were identified, which are respectively related to: weak‐oxygen bond (WOB) 2.37 eV,1.99 eV, and 1.55 eV, were identified, which are respectively related to: weak-oxygen bond [44], Neutral Oxygen Vacancy (NOV) [45], E’ center [46], non‐bridging oxygen hole center (NBOHC) (WOB) [44], Neutral Oxygen Vacancy (NOV)δ[45], E’δ center [46], non-bridging oxygen hole center [18,47,48] and dangling bonds (DB) [49]. It was found that NBOHC and E’ δ centers are the dominant (NBOHC) [18,47,48] and dangling bonds (DB) [49]. It was found that NBOHC and E’δ centers are the radiative defects for each of the coated films. The band around 2.7 eV has related to a dominant radiative defects for each of the coated films. The band around 2.7 eValso has been also been related twofold coordinated Si [16,17]. However, recently, the red emission around 1.9 eV has been related to a twofold coordinated Si [16,17]. However, recently, the red emission around 1.9 eV has been related to Si O, and so the main contribution can be also ascribed to this defect [43]. These results agree with to Si33O, and so the main contribution can be also ascribed to this defect [43]. These results agree with a a previous report in which the main emission mechanism was associated to Si‐O defects [29]. previous report in which the main emission mechanism was associated to Si-O defects [29]. Figure 99 shows shows the the PL PL emission emission spectra spectra of of embedded embedded films films with with red red (a) (a) and and blue blue (b) (b) Si-NCs Si‐NCs Figure before and after thermal annealing. It is worth mentioning that any PL emission is observed when a before and after thermal annealing. It is worth mentioning that any PL emission is observed when a Si‐NCs:SOG volume ratio is lower than 3:1. As can be seen, the PL spectra are different than those Si-NCs:SOG volume ratio is lower than 3:1. As can be seen, the PL spectra are different than those observed in coated Si‐NCs. Even though coated and embedded films were fabricated using the same observed in coated Si-NCs. Even though coated and embedded films were fabricated using the same Si‐NCs, their their respective spectrum has noticeable alterations. Before annealing, thermal the annealing, the Si-NCs, respective spectrum has noticeable alterations. Before thermal wavelength wavelength range of PL emission of both coated and embedded Si‐NCs films are similar. However, range of PL emission of both coated and embedded Si-NCs films are similar. However, after annealing, after annealing, the emission of the embedded films shows practically no variation, contrary to the the emission of the embedded films shows practically no variation, contrary to the coated films. coated films. The PL spectrum obtained from the embedded films with red Si-NCs presents a peak at ~715 nm, and aThe PL spectrum obtained from the embedded films with red Si‐NCs presents a peak at ~715 more intense one at ~410 nm, which is related to the SOG PL emission, as observed in Figure 9a. nm, and a more intense one at ~410 nm, which is related to the SOG PL emission, as observed in The embedded films with blue Si-NCs exhibit a maximum PL peak at ~430 nm, and a less intense Figure 9a. The embedded films with blue Si‐NCs exhibit a maximum PL peak at ~430 nm, and a less one at ~850 nm, as shown in Figure 9b. The PL emission observed in embedded films with both intense one at ~850 nm, as shown in Figure 9b. The PL emission observed in embedded films with red and blue Si-NCs before thermal annealing is similar to that observed in the colloidal solutions both red blue before the thermal is similar to that observed in the colloidal (Figure 7a).and Since theSi‐NCs SOG around Si-NCsannealing is not chemically bonded to them, the embedded film solutions (Figure 7a). Since the SOG around the Si‐NCs is not chemically bonded to for them, the emission is ascribed to the quantum confinement mechanism, as proposed by other authors similar embedded film emission is ascribed to the quantum confinement mechanism, as proposed by other films [25,50]. authors for similar films [25,50].
Appl. Sci. 2017, 7, 72 Appl. Sci. 2016, 6, 434
10 of 13 10 of 13 Embedded / red Si-NCs Annealed at 1100°C in N2
(a)
4
4x10
4
4x10
Embedded / blue Si-NCs Annealed at 1100°C in N2
(b)
SOG Film
3x10
PL Intensity (a. u.)
PL Intensity (a. u.)
SOG Film 4
4
2x10
4
1x10
4
3x10
4
2x10
4
1x10
0
0 400
500
600
700
800
Wavelength (nm)
900
1000
400
500
600
700
800
Wavelength (nm)
900
1000
Figure 9. 9. (Color (Color Online) PL emission Si‐NCs (b) (b) Figure Online) PL emission spectra spectra of of embedded embedded films films with with red red (a) (a) and blue and blue Si-NCs before and after annealing at 1100 °C in N 2 atmosphere. PL spectra from pure SOG films are ◦ before and after annealing at 1100 C in N2 atmosphere. PL spectra from pure SOG films are included. included.
When thermal annealing is applied to these embedded films, it is clear that the PL emission When thermal annealing is applied to these embedded films, it is clear that the PL emission changes in intensity, but not in shape, as shown in Figure 9a. Because of this, the PL of the embedded changes in intensity, but not in shape, as shown in Figure 9a. Because of this, the PL of the embedded films (with both red and blue Si-NCs) can be attributed to the quantum confinement effect before films (with both red and blue Si‐NCs) can be attributed to the quantum confinement effect before and after thermal annealing. Then, we can assume that there is no nitridation effect or any radiative and after thermal annealing. Then, we can assume that there is no nitridation effect or any radiative defect formation in the Si-NC/SiO2 interface for embedded films. This is confirmed by the absence of defect formation in the Si‐NC/SiO2 interface for embedded films. This is confirmed by the absence of emission bands related to Si-O defects [23]. It is possible that thermal annealing in nitrogen passivates emission bands related to Si‐O defects [23]. It is possible that thermal annealing in nitrogen the non-radiative centers around the Si-NCs, showing only the PL emission from Si-NCs [4,19]. passivates the non‐radiative centers around the Si‐NCs, showing only the PL emission from Si‐NCs This passivation effect increases the PL intensity from the red Si-NCs, but reduces that from the blue [4,19]. This passivation effect increases the PL intensity from the red Si‐NCs, but reduces that from ones. This is also corroborated by the shift of the IR stretching peak towards higher wavenumbers the blue ones. This is also corroborated by the shift of the IR stretching peak towards higher and the reduced intensity of the peak at 1200 cm−1 after thermal annealing, forming a well-defined wavenumbers and the reduced intensity of the peak at 1200 cm−1 after thermal annealing, forming a shoulder in the asymmetric stretching peak (Figure 6c,d). This behavior allows for arguing that the PL well‐defined shoulder in the asymmetric stretching peak (Figure 6c,d). This behavior allows for emission of the embedded films is mainly due to the contribution of Si-NCs. arguing that the PL emission of the embedded films is mainly due to the contribution of Si‐NCs. It is well known that the films with Si-NCs obtained externally to the dielectric film are highly It is well known that the films with Si‐NCs obtained externally to the dielectric film are highly dependent on the way they were mixed with the SOG solution. The experimental evidence presented dependent on the way they were mixed with the SOG solution. The experimental evidence here shows that when the Si-NCs are embedded in the dielectric matrix, the interfaces between the presented here shows that when the Si‐NCs are embedded in the dielectric matrix, the interfaces Si and the SiO2 are well defined, and even after high temperature treatments, there are no chemical between the Si and the SiO2 are well defined, and even after high temperature treatments, there are interactions between the silicon and its surroundings. This is corroborated by other authors who no chemical interactions between the silicon and its surroundings. This is corroborated by other observed a quenching of PL after thermal annealing [27]. authors who observed a quenching of PL after thermal annealing [27]. However, However, when when the the Si-NCs Si‐NCs are are coated coated (as (as opposed opposed to to embedded) embedded) by by the the SOG, SOG, and and then then annealed, the SiO flows around the Si-NCs, promoting the formation of silicon-oxide bonds. Then, annealed, the SiO22 flows around the Si‐NCs, promoting the formation of silicon‐oxide bonds. Then, a asintered type of core elemental silicon surrounded by a shell of non‐stoichiometric oxide is obtained. sintered type of core elemental silicon surrounded by a shell of non-stoichiometric oxide is obtained. The off-stoichiometric oxide contains silicon defects, which are luminescent under excitation. In this The off‐stoichiometric oxide contains silicon defects, which are luminescent under excitation. In this paper, after 1100 ◦ C annealing, the samples show an emission shift toward a range around 500 nm paper, after 1100 °C annealing, the samples show an emission shift toward a range around 500 nm regardless of the Si-NC size. However, other emission ranges can be obtained when the same type of regardless of the Si‐NC size. However, other emission ranges can be obtained when the same type of samples are subjected to different annealing conditions [29]. Therefore, in order to obtain efficient Si samples are subjected to different annealing conditions [29]. Therefore, in order to obtain efficient Si light emitters externally acquired, thethe results indicate that that it is better to cover them light emitters using using nanocrystals nanocrystals externally acquired, results indicate it is better to cover with the dielectric matrix and apply a thermal treatment. them with the dielectric matrix and apply a thermal treatment. 4. Conclusions 4. Conclusions Porous silicon obtained by electrochemical methods has been used to produce colloidal solutions Porous silicon obtained by electrochemical methods has been used to produce colloidal with silicon nanocrystals that show red or blue PL emission. These colloidal solutions were used to solutions with silicon nanocrystals that show red or blue PL emission. These colloidal solutions were fabricate SOG-coated and embedded films. In both solutions, Si-NCs were studied using TEM and used to fabricate SOG‐coated and embedded films. In both solutions, Si‐NCs were studied using XRD techniques. The size of the red light-emitting Si-NCs as observed through TEM agrees with PL TEM and XRD techniques. The size of the red light‐emitting Si‐NCs as observed through TEM results. From FTIR, a notable difference between coated and embedded films was found, mainly in agrees with PL results. From FTIR, a notable difference between coated and embedded films was found, mainly in the Si‐O stretching peak, which shifts to show non‐stoichiometric characteristics for coated films and stoichiometric SiO2 for embedded films.
Appl. Sci. 2017, 7, 72
11 of 13
the Si-O stretching peak, which shifts to show non-stoichiometric characteristics for coated films and stoichiometric SiO2 for embedded films. From PL measurements, it was found that the colloidal solutions, and as-deposited coated and embedded films, show an intense PL emission, which was related to the nanocrystal sizes (QCE). When a thermal annealing process is carried out, the PL of embedded films does not change, and the quantum confinement effect remains as the emission mechanism. However, coated films with both red and blue Si-NCs do change their PL spectrum towards the blue-green region after annealing, demonstrating that a chemical activity occurs only in the covered films after sintering at 1100 ◦ C. A correlation between photoluminescence and the FTIR results shows that the Si-NCs interface evolves, developing Si-O defects for such types of films. The new emission band is ascribed to a wide variety of Si-O related defects formed at the Si-NC/SiO2 interface. The analysis of the emission indicated that a high density of the radiative NBOHC and E’δ center dominates the PL emission from coated films after annealing. Acknowledgments: The authors are thankful to Consejo Nacional de Ciencia y Tecnología (CONACYT) for the economic support and Pablo Alarcon for his help during the sample preparation. Author Contributions: This paper is the result of O.C.-M.’s PhD work, who was under the supervision of M.A.-M. M.A.V.-A. and A.A.G.-F. were mostly involved in the setup of the experiments, also in the redaction of the paper. J.A.A.-A. was responsible for the X-ray measurements and interpretation of the data. A.M.-S. performed the TEM studies and reviewed the manuscript." Conflicts of Interest: The authors declare no conflict of interest.
References 1. 2.
3. 4.
5.
6. 7.
8. 9. 10.
11.
Lehmann, V.; Gösele, U. Porous silicon formation: A quantum wire effect. Appl. Phys. Lett. 1991, 58, 856–858. [CrossRef] ˙ Kramer, N.J.; Westermann, R.H.; Dohnalová, K.; Smets, A.H.; Verheijen, M.A.; Gregorkiewicz, T.; Dogan, ˘ I.; van de Sanden, M.C. Ultrahigh throughput plasma processing of free standing silicon nanocrystals with lognormal size distribution. J. Appl. Phys. 2013, 113. [CrossRef] Bolduc, M.; Genard, G.; Yedji, M.; Barba, D.; Martin, F.; Terwagne, G.; Ross, G.G. Influence of nitrogen on the growth and luminescence of silicon nanocrystals embedded in silica. J. Appl. Phys. 2009, 105. [CrossRef] Quiroga-González, E.; Bensch, W.; Aceves-Mijares, M.; Yu, Z.; López-Estopier, R.; Monfil-Leyva, K. On the photoluminescence of multilayer arrays of silicon rich oxide with high silicon content prepared by low pressure chemical vapor deposition. Thin Solid Films 2011, 519, 8030–8036. [CrossRef] Fauchet, P.M.; Ruan, J.; Chen, H.; Pavesi, L.; Dal Negro, L.; Cazzaneli, M.; Elliman, R.G.; Smith, N.; Samoc, M.; Luther-Davies, B. Optical gain in different silicon nanocrystal systems. Opt. Mater. 2005, 27, 745–749. [CrossRef] Ledoux, G.; Gong, J.; Huisken, F.; Guillois, O.; Reynaud, C. Photoluminescence of size-separated silicon nanocrystals: Confirmation of quantum confinement. Appl. Phys. Lett. 2002, 80, 4834–4836. [CrossRef] Ray, M.; Hossain, S.M.; Klie, R.F.; Banerjee, K.; Ghosh, S. Free standing luminescent silicon quantum dots: evidence of quantum confinement and defect related transitions. Nanotechnology 2010, 21. [CrossRef] [PubMed] Valenta, J.; Janda, P.; Dohnalová, K.; Nižnansky, ˇ D.; Vácha, F.; Linnros, J. Colloidal suspensions of silicon nanocrystals: from single nanocrystals to photonic structures. Opt. Mater. 2005, 27, 1046–1049. [CrossRef] Švrˇcek, V.; Rehspringer, J.L.; Slaoui, A.; Pivac, B.; Muller, J.C. Clustering/declustering of silicon nanocrystals in spin-on glass solutions. Semicond. Sci. Technol. 2005, 20, 314–319. [CrossRef] Chen, X.Y.; Lu, Y.F.; Wu, Y.H.; Cho, B.J.; Liu, M.H.; Dai, D.Y.; Song, W.D. Mechanisms of photoluminescence from silicon nanocrystals formed by pulsed-laser deposition in argon and oxygen ambient. J. Appl. Phys. 2003, 93, 6311–6319. [CrossRef] Ali, A.M.; Kobayashi, H.; Inokuma, T.; Al-Hajry, A. Morphological, luminescence and structural properties of nanocrystalline silicon thin films. Mater. Res. Bull. 2013, 48, 1027–1033. [CrossRef]
Appl. Sci. 2017, 7, 72
12.
13.
14. 15. 16. 17.
18.
19. 20. 21.
22.
23.
24.
25. 26.
27. 28. 29.
30.
31.
12 of 13
Godefroo, S.; Hayne, M.; Jivanescu, M.; Stesmans, A.; Zacharias, M.; Lebedev, O.I.; Van Tendeloo, G.; Moshchalkov, V.V. Classification and control of the origin of photoluminescence from Si nanocrystals. Nat. Nanotechnol. 2008, 3, 174–178. [CrossRef] [PubMed] Dohnalová, K.; Poddubny, A.N.; Prokofiev, A.A.; De Boer, W.D.; Umesh, C.P.; Paulusse, J.M.; Zuilhof, H.; Gregorkiewicz, T. Surface brightens up Si quantum dots: Direct bandgap-like size-tunable emission. Light Sci. Appl. 2013, 2. [CrossRef] Wolkin, M.V.; Jorne, J.; Fauchet, P.M.; Allan, G.; Delerue, C. Electronic States and Luminescence in Porous Silicon Quantum Dots: The Role of Oxygen. Phys. Rev. Lett. 1999, 82, 197–200. [CrossRef] ˚ Dohnalová, K.; Gregorkiewicz, T.; Kusová, K. Silicon quantum dots: Surface matters. J. Phys. Condens. Matter 2014, 26. [CrossRef] [PubMed] Skuja, L. Optically active oxygen-deficiency-related centers in amorphous silicon dioxide. J. Non-Cryst. Solids 1998, 239, 16–48. [CrossRef] Spallino, L.; Vaccaro, L.; Sciortino, L.; Agnello, S.; Buscarino, G.; Cannas, M.; Gelardi, F.M. Visible-ultraviolet vibronic emission of silica nanoparticles. Phys. Chem. Chem. Phys. 2014, 16, 22028–22034. [CrossRef] [PubMed] Vaccaro, L.A.; Cannas, M.A.; Boscaino, R.O. Luminescence features of nonbridging oxygen hole centres in silica probed by site-selective excitation with tunable laser. Solid State Commun. 2008, 146, 148–151. [CrossRef] Carrada, M.; Wellner, A.; Paillard, V.; Bonafos, C.; Coffin, H.; Claverie, A. Photoluminescence of Si nanocrystal memory devices obtained by ion beam synthesis. Appl. Phys. Lett. 2005, 87. [CrossRef] Limpens, R.; Lesage, A.; Fujii, M.; Gregorkiewicz, T. Size confinement of Si nanocrystals in multinanolayer structures. Sci. Rep. 2015, 5. [CrossRef] [PubMed] Aceves-Mijares, M.; Espinosa-Torres, N.D.; Flores-Gracia, F.; González-Fernández, A.A.; López-Estopier, R.; Román-López, S.; Pedraza, G.; Domínguez, C.; Morales, A.; Falcony, C. Composition and emission characterization and computational simulation of silicon rich oxide films obtained by LPCVD. Surf. Interface Anal. 2014, 46, 216–223. [CrossRef] Hessel, C.M.; Summers, M.A.; Meldrum, A.; Malac, M.; Veinot, J.G. Direct Patterning, Conformal Coating, and Erbium Doping of Luminescent nc-Si/SiO2 Thin Films from Solution Processable Hydrogen Silsesquioxane. Adv. Mater. 2007, 19, 3513–3516. [CrossRef] Lin, G.R.; Lin, C.J.; Lin, C.K.; Chou, L.J.; Chueh, Y.L. Oxygen defect and Si nanocrystal dependent white-light and near-infrared electroluminescence of Si-implanted and plasma-enhanced chemical-vapor deposition-grown Si-rich SiO2 . J. Appl. Phys. 2005, 97. [CrossRef] Švrˇcek, V.; Sasaki, T.; Shimizu, Y.; Koshizaki, N. Colloidal blue and red luminescent silicon nanocrystals and their elaboration in pure and doped spin on glasses. Phys. E Low Dimens. Syst. Nanostruct. 2007, 40, 293–296. [CrossRef] Švrˇcek, V.; Sasaki, T.; Shimizu, Y.; Koshizaki, N. Silicon nanocrystals formed by pulsed laser-induced fragmentation of electrochemically etched Si micrograins. Chem. Phys. Lett. 2006, 429, 483–487. [CrossRef] Svrcek, V.; Sasaki, T.; Shimizu, Y.; Koshizaki, N. Blue luminescent silicon nanocrystals prepared by nanosecond laser ablation and stabilized in electronically compatible spin on glasses. J. Appl. Phys. 2008, 103. [CrossRef] Švrˇcek, V.; Slaoui, A.; Rehspringer, J.L.; Muller, J.C. Photoluminescence studies from silicon nanocrystals embedded in spin on glass thin films. J. Lumin. 2003, 101, 269–274. [CrossRef] Nayfeh, M.H.; Rao, S.; Nayfeh, O.M.; Smith, A.; Therrien, J. UV Photodetectors with Thin-Film Si Nanoparticle Active Medium. IEEE Trans. Nanotechnol. 2005, 4, 660–668. [CrossRef] Cortazar, O.; Vasquez-A, M.A.; Aceves-Mijares, M. Effect of the Annealing Atmosphere and Temperature on the Photoluminescence of Si Nanocrystal Films Covered with Spin-On Glass. J. Nanoelectron. Optoelectron. 2015, 10, 93–98. [CrossRef] Comedi, D.; Zalloum, O.H.; Irving, E.A.; Wojcik, J.; Roschuk, T.; Flynn, M.J.; Mascher, P. X-ray-diffraction study of crystalline Si nanocluster formation in annealed silicon-rich silicon oxides. J. Appl. Phys. 2006, 99. [CrossRef] Vásquez-A., M.A.; Águila Rodríguez, G.; García-Salgado, G.; Romero-Paredes, G.; Peña-Sierra, R. FTIR and photoluminescence studies of porous silicon layers oxidized in controlled water vapor conditions. Rev. Mex. Fís. 2007, 53, 431–435.
Appl. Sci. 2017, 7, 72
32. 33. 34.
35.
36. 37.
38.
39. 40. 41. 42. 43.
44. 45. 46. 47.
48.
49. 50.
13 of 13
Lau, W.S. Infrared Characterization for Microelectronics; World Scientific: Singapore, 1999. Chou, J.S.; Lee, S.C. Effect of porosity on infrared-absorption spectra of silicon dioxide. J. Appl. Phys. 1995, 77, 1805–1807. [CrossRef] Hayashi, S.; Tanimoto, S.; Yamamoto, K. Analysis of surface oxides of gas-evaporated Si small particles with infrared spectroscopy, high-resolution electron microscopy, and X-ray photoemission spectroscopy. J. Appl. Phys. 1990, 68, 5300–5308. [CrossRef] Daldosso, N.; Das, G.; Larcheri, S.; Mariotto, G.; Dalba, G.; Pavesi, L.; Irrera, A.; Priolo, F.; Iacona, F.; Rocca, F. Silicon nanocrystal formation in annealed silicon-rich silicon oxide films prepared by plasma enhanced chemical vapor deposition. J. Appl. Phys. 2007, 101. [CrossRef] Borghesi, A.; Piaggi, A.; Sassella, A.; Stella, A.; Pivac, B. Infrared study of oxygen precipitate composition in silicon. Phys. Rev. B 1992, 46, 4123–4127. [CrossRef] Chayani, M.; Caquineau, H.; Despax, B.; Bandet, J.; Berjoan, R. Variations in the physico-chemical properties of near-stoichiometric silica deposited from SiH4 –N2 O and SiH4 –N2 O–He radiofrequency discharges. Thin Solid Films 2005, 471, 53–62. [CrossRef] Alayo, M.I.; Pereyra, I.; Scopel, W.L.; Fantini, M.C. On the nitrogen and oxygen incorporation in plasma-enhanced chemical vapor deposition (PECVD) SiOx Ny films. Thin Solid Films 2002, 402, 154–161. [CrossRef] Kumar, V. Nanosilicon; Elsevier: Amsterdam, The Netherlands, 2011. Neamen, D.A. Semiconductor Physics and Devices; McGraw-Hill Higher Education: New York, NY, USA, 2003. Dinh, L.N.; Chase, L.L.; Balooch, M.; Siekhaus, W.J.; Wooten, F. Optical properties of passivated Si nanocrystals and SiOx nanostructures. Phys. Rev. B 1996, 54, 5029–5037. [CrossRef] Nishikawa, H.; Watanabe, E.; Ito, D.; Sakurai, Y.; Nagasawa, K.; Ohki, Y. Visible photoluminescence from Si clusters in γ-irradiated amorphous SiO2 . J. Appl. Phys. 1996, 80, 3513–3517. [CrossRef] Vaccaro, L.; Popescu, R.; Messina, F.; Camarda, P.; Schneider, R.; Gerthsen, D.; Gelardi, F.M.; Cannas, M. Self-limiting and complete oxidation of silicon nanostructures produced by laser ablation in water. J. Appl. Phys. 2016, 120. [CrossRef] Nishikawa, H.; Watanabe, E.; Ito, D.; Takiyama, M.; Ieki, A.; Ohki, Y. Photoluminescence study of defects in ion-implanted thermal SiO2 films. J. Appl. Phys. 1995, 78, 842–846. [CrossRef] Zhao, X.; Schoenfeld, O.; Kusano, J.I.; Aoyagi, Y.; Sugano, T. Observation of Direct Transitions in Silicon Nanocrystallites. Jpn. J. Appl. Phys. 1994, 33, L899–L901. [CrossRef] Chou, S.T.; Tsai, J.H.; Sheu, B.C. The photoluminescence in Si+ -implanted SiO2 films with rapid thermal anneal. J. Appl. Phys. 1998, 83, 5394–5398. [CrossRef] Nishikawa, H.; Nakamura, R.; Tohmon, R.; Ohki, Y.; Sakurai, Y.; Nagasawa, K.; Hama, Y. Generation mechanism of photoinduced paramagnetic centers from preexisting precursors in high-purity silicas. Phys. Rev. B 1990, 41, 7828–7834. [CrossRef] Tugay, ˘ E.; Turan, R. Investigation of Photoluminescence Mechanisms from SiO2 /Si:SiO2 /SiO2 Structures in Weak Quantum Confined Regime by Deconvolution of Photoluminescence Spectra. J. Nanosci. Nanotechnol. 2016, 16, 4052–4064. [CrossRef] [PubMed] Fauchet, P.M. Photoluminescence and electroluminescence from porous silicon. J. Lumin. 1996, 70, 294–309. [CrossRef] ˚ Dohnalová, K.; Kusová, K.; Pelant, I. Time-resolved photoluminescence spectroscopy of the initial oxidation stage of small silicon nanocrystals. Appl. Phys. Lett. 2009, 94. [CrossRef] © 2017 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC-BY) license (http://creativecommons.org/licenses/by/4.0/).