Issue
Eur. Phys. J. Appl. Phys.
Volume 89, Number 2, February 2020
Disordered Semiconductors: Physics and Applications
Article Number 20401
Number of page(s) 6
Section Nanomaterials and Nanotechnologies
DOI https://doi.org/10.1051/epjap/2020190263
Published online 15 April 2020

© EDP Sciences, 2020

1 Introduction

Silicon nanocrystals (Si-NCs) exhibiting efficient room-temperature photoluminescence have been intensively studied in the last three decades. Their potential applications range from optoelectronics, sensing, and energetics to medicine and biology [1,2]. Laboratory prototypes of light emitting diodes [3], photovoltaic applications [4], lithium-ion battery anodes [5], and sensors [6] based on Si-NCs have already been demonstrated. Si-NCs exhibit low inherent toxicity, good biocompatibility, and suitable biodegrability, which makes them an ideal candidate for various biological applications such as medical imaging [7]. Furthermore, the photoluminescence of Si-NCs depends on their size [8,9] and it exhibits a shift towards desired shorter wavelength with a decline in a Si-NC diameter. Moreover, a transformation of the silicon indirect bandgap towards the direct one in Si-NCs with dimensions below roughly 3 nm due to quantum confinement and tensile strain has been shown recently [10], making them a good candidate for instance for the integration of optoelectronics on a silicon wafer. However, those Si-NCs were prepared by electrochemical wet etching of porous Si and followed by a time-consuming chemical treatment [10] yielding only a relatively small amount of the product.

Recently, an effective glow discharge synthesis of Si-NCs has been developed by Kortshagen [11,12]. This method enables the synthesis of sufficient amounts of small Si-NCs with the size going down to 2 nm [12] for their systematic exploitation and with good reproducibility. While, this technique using micro-plasma systems is currently well established, using high diameter flow-through reactors (diameter > 20 mm) that allows maximal yield production as well as the mastering of the after preparation surface termination of produced nanocrystals is still a highly challenging task. Many previous studies of Si-NCs plasmatic synthesis have suggested that synthesis conditions can significantly affect the size and luminescence of Si-NCs [11,12], yet these parameters have not been systematically studied in the literature. Unfortunately, the plasma synthesis of Si-NC is accompanied by silyl surface termination (especially with SiH3 groups) [13] making them insoluble in water and degrading without proper surface treatment (sush as hydrosilylation). Therefore, the subsequent Si-NC surface modification is necessary.

In this work we study an effect of deposition parameters on a synthesis of silicon nanoparticles (Si-NPs) in a flow-through glass tube reactor. The synthesised Si-NPs are modified by the second plasma in the vacuum system during the synthesis or on air after the synthesis. While the former aims to the alkyl group surface termination, the latter is supposed to lead to the surface oxidation and thus the reduction of the core Si-NPs diameter.

2 Experimental

Si-NPs have been synthesised in low-pressure non-thermal plasma in a continuous flow-through reactor. In this study, we have adapted the Kortshagen method for the plasma synthesis of the Si-NPs [11,12] (schematically shown in Fig. 1a). In particular, the system is based on an RF glow discharge in a glass tube. As a source material for nucleation of silicon clusters and their subsequent growth to nanoparticles is used silane that is mixed with working gas–argon. The growing particles, which are supposed to be in an amorphous phase at the beginning of the plasma region [13], are carried by the gas stream through the plasma, they get a negative charge from the plasma which stabilises them from agglomeration and coagulation so they can grow only through surface reactions with plasma until the precursor gas is depleted or until they leave plasma. Due to interactions with plasma, the growing particles are heated and crystallise [13]. The length and volume of the plasma region in the glass tube, the gas flow and pressure are the main parameters that define the residence time, which describes how long the particles stay in the plasma.

The deposition parameters have been as follows: argon and silane flow 105 sccm and 3 sccm, respectively; pressure in the glass tube in ranges from 300 to 800 Pa; the glass tube inner diameter of 23 mm and the length of plasma region ranges from 6 to 30 cm in dependence on the applied RF power of 40–100 W. The 13.56 MHz source with an automatic matching controller from Coaxial Power Systems has been used to obtain stable plasma conditions during the few-minute long synthesis. The Si-NPs are sprayed through the aperture on the glass substrate in the vacuum chamber with approximately ten times lower pressure than in the glass tube. The production rate of synthesised Si-NPs was roughly in units of milligrams per minute.

Samples for photoluminescence characterization were prepared as follows: Si-NPs of approximately 0.3–1.5 mg were mixed with 1 ml of spectrophotometric grade toluene (Rotisolv > 99.8%, UV/IR − Grade, Roth) and homogenized in standard laboratory ultrasonic bath. The resulting suspensions were diluted to obtain a specific value of absorbance (same for all samples) at the wavelength of 325 nm, thus, the measured luminescence intensity of the samples can be used for comparison of photoluminescence efficiency between the samples.

The photoluminescence spectra were excited using a HeCd laser (continuous, wavelength 325 nm (3.6 mW), mode TEM00, linear polarization). The diluted samples were placed into a quartz cuvette (Hellma, Suprasil® quartz, 10 × 4 mm, and chamber volume 1500 µL). The photoluminescence was collected using a quartz optical cable in a 90 deg angle with respect to the excitation. The detection system consisted of a CCD camera (Andor Newton DU970P-UVB, 200–1050 nm) coupled to imaging spectrometer Triax 320 (Horiba Jobin Yvon). All spectra were corrected for the spectral response of the optical system and measured at room temperature.

Raman spectra were measured on a dry powder of Si-NPs in a Renishaw setup using a 442 nm HeCd continuous excitation laser with a power of as low as 0.33 mW to avoid a laser-induced crystallisation of the silicon powder.

Conventional TEM was carried out on a Jeol 2200FS equipped with an EDS detector from Oxford Instruments. The TEM was operated at accelerating voltage of 200 kV.

thumbnail Fig. 1

(a) Schematic of the plasma system for the Si-NPs synthesis. (b) Photoluminescence spectra of Si-NPs prepared by glow discharge with SiH4 and Ar flow of 3 sccm and 105 sccm, respectively. The plasma source power varied from 40 W to 120 W, the pressure in a glass tube was 700 Pa for the samples prepared at power below 90 W and 400 Pa for the samples prepared at power above 100 W. (c) Normalised Raman spectra of Si-NPs from (b) synthesised at power of 60 W (open square), 90 W (circle) and 110 W (triangle), respectively. (d) TEM figures and diffraction patterns of Si-NPs from (b) synthesised at power of 60 W, 90 W and 110 W, respectively.

3 Results

Firstly, we have studied influence of the deposition parameters on the photoluminescence and structural properties of the resulting Si-NPs. For the photoluminescence measurement the Si-NPs were dispersed in spectroscopic grade toluene and ultrasonicated directly after being removed from the vacuum chamber. The other methods used Si-NPs in dry form. Figure 1b shows the influence of the plasma power on the photoluminescence of the synthesised Si-NPs. The plasma length increased from 6 cm at the power of 40 W to 20 cm at 120 W. Note that these Si-NPs are without any additional functionalisation. The maximum of the luminescence shifts to the shorter wavelength with a decrease in plasma power. Low plasma power produces a glow discharge of only a short length in the glass tube, which seems to be sufficient for the synthesis of only the small silicon nanoparticles (with blue-shifted photoluminescence spectra). On the contrary, high plasma power enlarges the area of the glow discharge and leads to the growth of the larger nanoparticles and increased heating and crystallisation by the interactions with plasma.

Figure 1c shows Raman scattering spectra of a few samples from the power series above. Samples synthesised at a power of 60 W, 90 W and 110 W are shown as they represent those with photoluminescence in the visible range, at the edge between red and infrared and in the infrared region. Si-NPs with luminescence in the visible range show in the Raman scattering only an amorphous band at 480 cm−1. Si-NPs with luminescence in the infrared region show a crystalline band at 518 cm−1 which is very close to the Raman shift of 520 cm−1 attributed to the crystalline silicon. Si-NPs with luminescence in between visible and infrared region show both amorphous and crystalline Raman band. This suggests that the power below 110 W was probably not sufficiently high to crystallise all the synthesised silicon nanoparticles and the power below 90 W was not sufficient to crystallise the nanoparticles at all. However, it has been recently shown that the surface and subsurface atoms of Si-NCs act differently from the bulk and small Si-NCs with the size below 6 nm do not necessarily exhibit bulk quality [14], thus the structural classification of such small Si nanoparticles can be misleading.

Figure 1d shows TEM pictures and diffraction patterns of samples synthesised at a power of 60 W, 90 W and 110 W corresponding to those measured by Raman scattering in the previous paragraph. These results partially correspond with the Raman measurement. Small Si-NPs synthesised at low powers with luminescence in the visible range do not exhibit traditional diffraction patterns in a routine TEM measurement (i.e. without any surface treatment), whereas the large Si-NCs synthesised at high powers with luminescence in the infrared range are clearly crystalline. The average diameter of these Si-NCs is 17 nm. The Si-NPs synthesised at conditions in between have some crystalline fraction based on the diffraction pattern and their average size is 10 nm.

The deposition parameters, which have influence the plasma length in the glass tube, have a similar effect on the synthesised Si-NPs. For instance, the increase in silane concentration reduces the length of the plasma (and hence the residence time), so the Si-NPs synthesised in the conditions of higher silane concentration have luminescence at shorter wavelengths, suggesting that they are smaller than those synthesised at lower silane concentrations. The photoluminescence of the Si-NPs synthesised under silane concentration ranging from 1% to 5% (1–5 sccm of SiH4 in 105 sccm of Ar) is shown in Figure 2a. A similar effect has been achieved by the changes in pressure. The decrease in pressure in the glass tube increases the length of the plasma and shifts the luminescence of synthesised Si-NPs to the longer wavelengths.

The addition of a small amount of hydrogen does not change the length of the plasma zone, but it still changes the synthesised Si-NPs. We have added the hydrogen with various flows in range 0–15 sccm into the Ar:SiH4 mixture with flows remained unchanged at values of 105 sccm and 3 sccm, respectively. The photoluminescence of Si-NPs synthesised in the presence of hydrogen is shown in Figure 2b. The increase in hydrogen concentration leads in a small blue-shift and a slight increase in the photoluminescence. This effect might be connected with the atomic hydrogen being responsible for the etching processes in the so-called growth zone of silicon [15,16] on the surface of the Si-NPs. It can be expected that the hydrogen treatment of Si-NPs results in production of smaller Si-NPs possesing more homogenous surface with less concentration of photoluminescence quenching centers that is in good agreement with observed photoluminescence dependence. Suprisingly, the presence of hydrogen in gas mixture doest not cause significant blue-shift as was presented recently [14].

The synthesised Si-NPs can be modified chemically after they are removed from the vacuum chamber or directly in the vacuum chamber right after the synthesis. First, we have studied the oxidation of Si-NPs surface both by storing them in the dry air and treating them in the air with an atmospheric plasma. Two samples of Si-NPs with photoluminescence at around 600 nm were tested. Sample 1 was kept at ambient atmosphere with controlled humidity (56%) for 5 days, while sample 2 atmospheric pressure non-thermal plasma generated by positive corona discharge (with an electric current in discharge ∼ 0.1 mA) [17] for 4 hours. Both samples were dispersed and characterized in toluene afterward and both modifications lead in similar blue-shift of the photoluminescence shown in Figure 3a, from which we estimate the slow oxidation of Si-NP surface and the reduction of Si-NC core size down to approximately 2 nm. The application of air plasma made this oxidation and the photoluminescence blue-shift to proceed much faster.

The monitoring of Si-NPs stability shows that while their dispersed in toluene slows their degradation they are still not perfectly protected from all devaluing processes. Slow ageing accompanied by a slight blue-shift in photoluminescence was observed for samples stored in toluene in the dark. This can be connected with the slow surface oxidation of the Si-NPs dispersed in toluene. The photoluminescence of small Si-NPs shifted from the initial value of around 600 nm to 500 nm within two weeks and after that Si-NPs in toluene disappeared as they probably completely change to silica particles. On the other hand, large Si-NCs with photoluminescence at around 1000 nm within a few weeks shifted only imperceptibly; but their luminescence intensity has been dramatically increasing (Fig. 3b), suggesting that the oxidation passivated some defects on the Si-NCs surface.

The modification of Si-NCs directly in the vacuum chamber, where they are synthesised and collected, is the most promising treatment avoiding the direct contact of bare Si-NCs with the air. We have used the second plasma ignited in the vacuum chamber, where the Si-NCs are collected (out of the influence of the primary discharge). The separate flow of methylsilane (CH3SiH3) has been added to the flow of unprocessed gases coming from the glass tube through the aperture. By this, the pressure in the vacuum chamber increased up to 100 Pa (with the partial pressure of the CH3SiH3 of around 30 Pa).

The photoluminescence spectra of Si-NCs fabricated with the second plasma of CH3SiH3 are shown in Figure 4. As a starting point, we have used the Si-NCs with photoluminescence at around 1000 nm. The addition of 0.7 sccm of methylsilane to the vacuum chamber and the ignition of the weak second plasma in the vacuum chamber with a power of 2.5 W lead to the increase in the luminescence intensity suggesting the passivation of the Si-NCs surface by the methylsilane.

The high power CH3SiH3 plasma in the second vacuum chamber lead to an increase and significant blue-shift in photoluminescence. The original and the modified Si-NCs are crystalline and they exhibit the same Raman band centred at 520 cm−1, but unlike the low-power methylsilane plasma, the Si-NCs modified by high-power methylsilane plasma exhibit lower Raman cross-section (the Rama signal is weak and noisy) which is typical for smaller nanoparticles. This suggests that the high-power second CH3SiH3 plasma does not only passivate the Si-NCs surface, but possibly lead to decrease in the Si-NC size. Further study is needed to confirm this.

thumbnail Fig. 2

Influence of silane flow and the addition of hydrogen on photoluminescence spectra of Si-NPs prepared by plasma power of 60 W and argon flow of 105 sccm and the pressure of 700 Pa. (a) The flow of silane was varied between 1 and 5 sccm while no hydrogen was used. (b) The flow of hydrogen was varied between 0 and 15 sccm while the silane flow was set to 3 sccm.

thumbnail Fig. 3

(a) Influence of ambient conditions and atmospheric presure non-thermal plasma on photoluminescence spectra of Si-NPs prepared by 60W, 105 sccm of Ar and 3 sccm of SiH4, 700 Pa. (b) Influence of ambient conditions on photoluminescence spectra of Si-NCs prepared by 110 W.

thumbnail Fig. 4

Photoluminescence (a) and Raman measurement (b) of Si-NCs synthesised by high power of 110 W (with SiH4 and Ar flows and pressure same as shown in Fig. 1) and modified by methylsilane (MMS) either low power or high power plasma in the second vacuum chamber. The MMS flow and plasma power indicated in the figure (a) are valid for both graphs.

4 Summary

The photoluminescence of plasma synthesised Si-NPs strongly depends on the length of plasma in the glass tube, which can be affected with the RF power, gas concentrations or pressure. The lower power, higher silane concentration or higher pressure reduces the length of plasma in the glass tube and leads to the synthesis of Si-NPs with photoluminescence at shorter wavelengths. By varying these deposition parameters we have synthesised Si-Ns with photoluminescence ranging from blue to IR spectral region.

The synthesised Si-NCs have been modified both chemically after they were removed from the vacuum chamber and plasmatically directly in the vacuum chamber. The oxidation in the air or by air plasma led to a blue-shift and increase in photoluminescence of those Si-NPs. This suggests that as prepared Si-NCs suffer from defective surface which is being passivated by the oxidation. The air treatment also seems to reduce the Si-NP size and is in combination with the high material yield of plasma techniques a good candidate on mass production of small luminescent Si-NPs. Also, the second plasma of methylsilane leads to photoluminescence increase and blue-shift of synthesised Si-NCs, therefore the methylsilane seems perspective for the in-situ passivation or functionalisation of Si-NCs.

Acknowledgments

This work was supported by DAAD grant n. 19-01 and Czech Science Foundation (GAČR project No. 18-05552S).

Author contribution statement

Martin Müller: Si-NP synthesis, Raman measurements and data analysis, manuscript writing. Pavel Galář: photoluminescence measurements and data analysis. Jiří Stuchlík: synthesis of Si-NPs with methylsilane plasma, Jan Kočka: data analysis. Jonáš Kupka: Si-NP synthesis. Kateřina Kůsová: project supervision.

References

  1. K. Herynková, E. Podkorytov, M. Šlechta, O. Cibulka, J. Leitner, I. Pelant, Nanoscale Res. Lett. 9 , 478 (2014) [PubMed] [Google Scholar]
  2. K. Kůsová, K. Dohnalová, Organically capped silicon nanocrystals, in Silicon Nanomaterials Sourcebook: Low-Dimensional Structures, Quantum Dots, and Nanowires (2017), Vol. 1, pp. 367–398 [Google Scholar]
  3. S. Zhao, X. Liu, X. Pi, D. Yang, J. Semicond. 39 , 6 (2018) [Google Scholar]
  4. M. Dutta, L. Thirugnanam, P.V. Trinh, N. Fukata, ACS Nano 9 , 7 (2015) [CrossRef] [PubMed] [Google Scholar]
  5. H. Wu, G. Yu, L. Pan, N. Liu, M.T. McDowell, Z. Bao, Y. Cui, Nat. Commun. 4, 1943 (2013) [PubMed] [Google Scholar]
  6. C.M. Gonzalez, J.G. Veinot, J. Mater. Chem. C 4 , 22 (2016) [Google Scholar]
  7. F. Erogbogbo, K.T. Yong, I. Roy, R. Hu, W.C. Law, W. Zhao, H. Ding, F. Wu, R. Kumar, M.T. Swihart, P.N. Prasad, ACS Nano 5 , 413 (2011) [Google Scholar]
  8. M.L. Mastronardi, F. Maier-Flaig, D. Faulkner, E.J. Henderson, C. Kübel, U. Lemmer, G.A. Ozin, Nano Lett. 12 , 1 (2012) [CrossRef] [PubMed] [Google Scholar]
  9. L.M. Wheeler, N.C. Anderson, P.K. Palomaki, J.L. Blackburn, J.C. Johnson, N.R. Neale, Chem. Mater. 27 , 19 (2015) [Google Scholar]
  10. K. Kusova, P. Hapala, J. Valenta, P. Jelinek, O. Cibulka, L. Ondic, I. Pelant, Adv. Mater. Interfaces 1 , 2 (2014) [Google Scholar]
  11. U.R. Kortshagen, R.M. Sankaran, R.N. Pereira, S.L. Girshick, J.J. Wu, E.S. Aydil, Chem. Rev. 116 , 1811061 (2016) [Google Scholar]
  12. L. Mangolini, E. Thimsen, U. Kortshagen, Nano Lett. 5 , 4 (2005) [Google Scholar]
  13. T. Lopez, L. Mangolini, J. Vacuum Sci. Technol. B 32 , 061802 (2014) [CrossRef] [Google Scholar]
  14. A.N. Thiessen, M. Ha, R.W. Hooper, H. Yu, A.O. Oliynyk, J.G. Veinot, V.K. Michaelis, Chem. Mater. 31 , 3 (2019) [Google Scholar]
  15. C.C. Tsai, G.B. Anderson, R. Thompson, J. Non-Cryst. Solids 137, 673 (1991) [CrossRef] [Google Scholar]
  16. H.V. Nguyen, I. An, R.W. Collins, Y. Lu, M. Wakagi, C.R. Wronski, Appl. Phys. Lett. 65 , 26 (1994) [Google Scholar]
  17. P. Horák, J. Khun, J. Phys: Conf. Ser. 223, 012006 (2010) [CrossRef] [Google Scholar]

Cite this article as: Martin Müller, Pavel Galář, Jiří Stuchlík, Jan Kočka, Jonáš Kupka, Kateřina Kůsová, Synthesis and surface modification of light emitting silicon nanoparticles using non-thermal plasma techniques, Eur. Phys. J. Appl. Phys. 89, 20401 (2020)

All Figures

thumbnail Fig. 1

(a) Schematic of the plasma system for the Si-NPs synthesis. (b) Photoluminescence spectra of Si-NPs prepared by glow discharge with SiH4 and Ar flow of 3 sccm and 105 sccm, respectively. The plasma source power varied from 40 W to 120 W, the pressure in a glass tube was 700 Pa for the samples prepared at power below 90 W and 400 Pa for the samples prepared at power above 100 W. (c) Normalised Raman spectra of Si-NPs from (b) synthesised at power of 60 W (open square), 90 W (circle) and 110 W (triangle), respectively. (d) TEM figures and diffraction patterns of Si-NPs from (b) synthesised at power of 60 W, 90 W and 110 W, respectively.

In the text
thumbnail Fig. 2

Influence of silane flow and the addition of hydrogen on photoluminescence spectra of Si-NPs prepared by plasma power of 60 W and argon flow of 105 sccm and the pressure of 700 Pa. (a) The flow of silane was varied between 1 and 5 sccm while no hydrogen was used. (b) The flow of hydrogen was varied between 0 and 15 sccm while the silane flow was set to 3 sccm.

In the text
thumbnail Fig. 3

(a) Influence of ambient conditions and atmospheric presure non-thermal plasma on photoluminescence spectra of Si-NPs prepared by 60W, 105 sccm of Ar and 3 sccm of SiH4, 700 Pa. (b) Influence of ambient conditions on photoluminescence spectra of Si-NCs prepared by 110 W.

In the text
thumbnail Fig. 4

Photoluminescence (a) and Raman measurement (b) of Si-NCs synthesised by high power of 110 W (with SiH4 and Ar flows and pressure same as shown in Fig. 1) and modified by methylsilane (MMS) either low power or high power plasma in the second vacuum chamber. The MMS flow and plasma power indicated in the figure (a) are valid for both graphs.

In the text

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.