- Email: [email protected]
Preparation of metal oxide nanoparticles by gas aggregation cluster source
Accepted Manuscript Preparation of metal oxide nanoparticles by gas aggregation cluster source Artem Shelemin, Ondřej Kylián, Jan Hanuš, Andrei Choukourov, Iurii Melnichuk, Anton Serov, Danka Slavínská, Hynek Biederman PII:
S0042-207X(15)30012-9
DOI:
10.1016/j.vacuum.2015.07.008
Reference:
VAC 6744
To appear in:
Vacuum
Received Date: 29 December 2014 Revised Date:
11 July 2015
Accepted Date: 13 July 2015
Please cite this article as: Shelemin A, Kylián O, Hanuš J, Choukourov A, Melnichuk I, Serov A, Slavínská D, Biederman H, Preparation of metal oxide nanoparticles by gas aggregation cluster source, Vaccum (2015), doi: 10.1016/j.vacuum.2015.07.008. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Preparation of metal oxide nanoparticles by gas aggregation cluster source
Artem Shelemin, Ondřej Kylián1, Jan Hanuš, Andrei Choukourov, Iurii Melnichuk, Anton Serov, Danka
RI PT
Slavínská and Hynek Biederman
Department of Macromolecular Physics, Faculty of Mathematics and Physics, Charles University, V
SC
Holešovickách 2, 180 00 Praha 8, Czech Republic
M AN U
Keywords: nanoparticles; metal oxides; gas aggregation sources; TiOx; AlxOy
Abstract
Al/AlxOy and Ti/TiOx nanoparticles have been fabricated using gas aggregation cluster source (GAS) equipped
TE D
with 3-inch planar magnetrons using. The main attention has been devoted to the evaluation of the role of oxygen admixture on the deposition rate and properties of formed nanoparticles. It has been found that oxidized nanoparticles are deposited when oxygen is present in the working gas mixture. Highest deposition rates for
EP
production of Al/AlxOy or Ti/TiOx nanoparticles were observed when 3% or 1.5% oxygen was present in the working gas mixture, respectively. Mean diameters of about 16 nm for the AlxOy nanoparticles and 30 nm for
AC C
TiOx have been found under these deposition conditions.
1
Corresponding author. Phone: (+420) 22191 2258, Fax: (+420) 22191 2350, e-mail: [email protected]
1
ACCEPTED MANUSCRIPT 1 Introduction The fabrication of metal oxide nanoparticles (NPs) receives nowadays increasing attention. The main reason of this is a huge application potential of such materials that covers for instance water treatment, cosmetics, medicine, photovoltaic, semiconductor packing materials, fuel cells or catalysts [1-4]. Probably the most studied
RI PT
metal oxide NPs are titanium oxide that are employed in cosmetics, food, paints, personal care products or photovoltaic applications [5-7] and aluminium oxide NPs that are used in high-performance materials to increase their ductility, scratch resistance and toughness [8]. Another perspective field, where alumina NPs are expected to play an important role, is their use for production of antimicrobial materials [9].
SC
Although nanoparticles may be produced by techniques based on purely chemical synthesis (e.g. [10]), there is a growing interest in methods utilizing vacuum technologies, such as for example vacuum evaporation or
M AN U
sputtering at elevated pressures (typically tens to hundreds of Pa). Under these conditions the NPs are formed by sputtered vapor in the gas phase as the result of three step process that consist of nucleation, coagulation and attachment of atoms and ions onto created NPs. A recent example of this approach is nanoparticle synthesis by highly ionized plasmas produced by high power pulsed hollow cathode [11]. It was reported that the size distribution of produced NPs can be tuned in certain range by adjusting the pulse parameters, such as frequency,
TE D
pulse width, or peak current [12]. An alternative approach is based on the use of gas aggregation sources (GAS) of nanoparticles. In this case the magnetron is inserted into a cooled aggregation chamber that is separated from the rest of the deposition chamber by a small output orifice [13]. This arrangement assures sufficiently high
EP
pressure in the aggregation chamber needed for the effective production of NPs that are subsequently dragged by working gas through an output orifice into the vacuum deposition chamber and reach the substrate in the form of
AC C
a nanoparticle beam. The key advantage of GAS is complete separation of the NPs production from the rest of the deposition chamber, which facilitates combination of GAS with other vacuum based techniques of material deposition. This in turn makes it possible to produce either nanostructured coatings by overcoating NPs by a thin film of the same or other material (e.g. [14-16]), use NPs as seeds for formation of columnar structures prepared by glancing angle deposition [17] or various kinds of nanocomposite materials (e.g. [18,19]). Different GAS systems were already identified as reliable and highly effective sources for deposition of metallic NPs such as Ag, Cu, Pt, Ti or Al (e.g. [18-26]) or plasma polymerized NPs (e.g. [27,28]). In addition, it was shown that gas aggregation sources may also be used for production of metal-oxide NPs (e.g. [23,29-33]). Two strategies were followed. In the first one, the metallic nanoparticles are after the deposition oxidized on ambient
2
ACCEPTED MANUSCRIPT air (e.g. [29]). However, the necessity of the oxidation step makes this approach not suitable for combination of GAS with other vacuum based deposition methods needed for instance for production of nanocomposites. Because of this, there were attempts to produce metal-oxide NPs directly in the GAS. In this case small amount of oxygen was added to the inert working gas [30-33]. According to the reported results, the presence of oxygen influences not only the chemical structure of produced NPs, but it has also strong impact on the deposition rate
RI PT
and the stability of NPs formation. The main objective of this study is to evaluate the influence of amount of added oxygen on the process of TiOx and AlxOy NPs formation. In order to meet this aim, the measurements of deposition rate of NPs were accompanied by optical emission spectroscopy of plasma inside the aggregation
SC
chamber as well as by evaluation of chemical composition and sizes of NPs produced at steady-state condition.
M AN U
2 Materials and Methods
The scheme of apparatus used for production of AlxOy nanoparticles is depicted in Figure 1. It consisted of a gas aggregation source, which was attached onto a deposition chamber pumped by rotary and diffusion pumps to a base pressure 10-4 Pa. The used GAS was based on stainless steel cylindrical aggregation chamber (diameter
TE D
110 mm, length 450 mm) ending in an orifice 3 mm in diameter. The source of material for nanoclusters production was DC, water-cooled, planar, 3-inch magnetron equipped with 3 mm thick Al target (purity 99.99%) that was operated in constant current mode (200 mA). The distance between the target and the output orifice was 150 mm and was kept constant during all experiments. The aggregation chamber was equipped with diagnostic
EP
ports that enabled us to monitor plasma composition during the deposition of nanoclusters by means of optical emission spectroscopy (AvaSpec-3648-2-USB2). In addition, the magnetron and exit orifice can be moved
AC C
against diagnostic ports during the deposition (i.e. in-situ) while keeping the constant distance of 15 cm. This is a special ”diagnostic” variation of our simple GAS that enables e.g. OES measurements in an arbitrary part of the aggregation chamber.
3
RI PT
ACCEPTED MANUSCRIPT
SC
Figure 1. Scheme of deposition set-up with moveable Gas Aggregation Source used for Al and AlxOy NPs.
M AN U
Similar GAS was used for production of Ti and TiOx nanoparticles. In contrast to the source employed for fabrication of AlxOy NPs the aggregation chamber of this system was not moveable and the diagnostic ports were thus at a fixed 3 cm distance from the 3-inch magnetron. The diameter of the output orifice was 1.5 mm and all
TE D
experiments reported in this study were performed at constant applied 300 mA DC current.
Argon (99.99%) with an admixture of oxygen (up to 20%) was used in both cases as working gas. The flow of Ar (6 sccm in case of Al/AlxOy NPs and 1.5 sccm in case of Ti/TiOx NPs) was regulated by flow controller
EP
(MKS), while the amount of oxygen was tuned using a calibrated needle valve (Pfeiffer). The total pressure in the gas aggregation chamber was kept constant and equal to 40 Pa in experiments with Al and AlxOy NPs and 28
AC C
Pa in case of Ti an TiOx NPs. The pressure in the main deposition chamber was 0.1 Pa.
The deposition rate of produced NPs was measured by a quartz crystal microbalance. Chemical composition of produced nanoparticles was determined by means of X-ray Photoelectron Spectroscopy (XPS, Phobios 100, Specs) performed in-situ, i.e. without exposing produced NPs films to the ambient atmosphere. Spectra were acquired at constant take-off angle of 90° using Al Kα X-rays source (1486.6 eV, 200W, Specs). Wide spectra were acquired for binding energies in the range of 0 - 1100eV at a pass energy of 40 eV (dwell time 100 ms, step 0.5 eV). High resolution spectra were obtained at pass energy of 10 eV and were charge referenced to
4
ACCEPTED MANUSCRIPT adventitious carbon at 285.0 eV. Spectral analysis was performed using CASA XPS software. Finally, scanning electron microscopy (SEM, Tescan Mira 3) was used for the study of morphology of deposited coating.
Adhesion of NPs to Si/SiO2 substrates was studied by an AFM (Ntegra Prima, NT-MDT) both in a semi-contact and in a contact mode using diamond-like-carbon coated cantilevers (ContDLC, NanoandMore) with a
RI PT
guaranteed tip radius of 15 nm. The spring constant of 0.26 N/m was measured by the Sader method and the cantilever deflection detection system was calibrated before the measurements as described in [34]. First, the surface of the sub-monolayer of the NPs was scanned in the semicontact mode to identify an individual particle
SC
to be studied. Then the AFM was switched to the contact mode and the profile of the particle was continuously scanned with the increasing contact force until it was moved away. Finally, the AFM was switched back to the
3 Results and discussion 3.1 Al and AlxOy nanoparticles
M AN U
semi-contact mode and the entire area was scanned again to confirm that the particle was removed.
TE D
The first goal of this study was the identification of optimum experimental conditions which would provide stable and effective deposition of alumina nanoparticles. It was found that at constant magnetron current and pressure in the aggregation chamber, the deposition rate of NPs is strongly dependent on the used working gas
EP
mixture.
AC C
First, it was observed that the presence of oxygen significantly improves the stability of the process of formation of nanoparticles (see Figure 2): whereas in case of using Ar/O2 working gas mixture the amount of deposited NPs increased almost linearly with deposition time over the period of more than 30 minutes, the deposition rate, in agreement with a previous study [19], gradually decreased and finally stopped after prolonged operation time when only pure Ar was used.
5
SC
RI PT
ACCEPTED MANUSCRIPT
M AN U
Figure 2. Long-term stability of the deposition process in dependence on the presence of oxygen in the working gas mixture. Magnetron current 200 mA, pressure in the aggregation chamber 40 Pa. The Ar flow was 6 sccm.
Further experiments were performed after both the plasma parameters (magnetron voltage, intensity of emission spectral lines) and the deposition rate were stabilized. The temporal instabilities of deposition process in the
TE D
initial period, observed also in previous studies in case of Ti NPs [31], lasted typically 20-30 min depending on the deposition conditions and the history of the magnetron. After this phase the deposition rate, magnetron voltage and intensities of spectral lines become temporally stable: the variation of magnetron voltage was in the
EP
range ±3 V, and the variations of the deposition rate and intensities of spectral lines were less than 5%. As can be seen in Figure 3a, the deposition rate after the plasma stabilization in terms of deposited mass expressed in the
AC C
frequency shift of quartz crystal per minute initially rapidly increases with increasing fraction of oxygen in the working gas mixture and reaches its maximal value for about 3-5% of O2. Further increase of O2 fraction subsequently leads to gradual decrease of the deposition rate.
The dependence of the NPs deposition rate on the oxygen fraction in the working gas mixture may be explained as follows. On starting to add the small amount of oxygen to the working gas mixture an increase of magnetron voltage takes place due to oxygen chemisorption [35] (see Figure 3b). Since the O atom, upon an impact of an energetic ion, will contribute less to electron emission than an Al atom, the presence of adsorbed oxygen atoms will lower overall electron emission yield [36]. This in turn results in the increase of magnetron voltage. The
6
ACCEPTED MANUSCRIPT increase of magnetron voltage subsequently rises the energy of Ar ions bombarding the target, which leads to enhancement of the number of emitted Al atoms and thus also probability of NPs formation. This is consistent with the results of optical emission spectroscopy that showed rapid rise of the intensity of neutral Al atomic spectral lines with the addition of a small amount of O2 to working gas mixture (see Figure 3c) which reflects the increase of the density of Al atoms that are present in the plasma . In addition, it has to be stressed that the
RI PT
presence of oxygen in the gas phase facilitates formation of stable embryonic clusters [19] and thus promotes further growth of NPs.
100
a)
0 320
Magnetron voltage
300 280
b)
Intensity [a.u.]
260 600
Intensity [a.u.]
900
Al atom line
TE D
400 200 0
600 300
EP
Voltage [V]
SC
Deposition rate 200
M AN U
∆f [Hz/min]
300
O atom line
0 0.00
0.05
0.10
0.15
c)
d)
0.20
AC C
[O2]/[Ar]+[O2]
Figure 3. Dependence of a) deposition rate, b) magnetron voltage, c) intensity of Al spectral line at wavelength 396 nm and d) intensity of O atom spectral line at 777 nm on oxygen fraction in the working gas mixture. Magnetron current 200 mA, pressure in the aggregation chamber 40 Pa and Ar flow was 6 sccm.
Increasing the oxygen fraction in the working gas mixture subsequently causes the oxidation of Al target. Since the electron emission yield of Al oxide is higher than that of Al (e.g. [37]), the formation of aluminium oxide
7
ACCEPTED MANUSCRIPT layer may explain the observed decrease of the magnetron voltage as well as lowering of the number of Al atoms emitted from the magnetron target. The latter in turn causes a decrease in the rate of NPs formation.
The chemical structure of produced nanoparticle films prepared at 0.5% and 3% of O2 in the working gas
RI PT
mixture was determined by XPS without exposing the NPs to ambient air. The analysis of high resolution
spectra (see Figure 4) revealed that Al 2p peak is dominated by aluminium oxide peak positioned at binding energy 74.5 eV [38]. Moreover, it can be seen that the peak at binding energy 72 eV that belongs to metallic Al decays when oxygen fraction higher than 3% is used. However, also in case of a low amount of admixed
SC
oxygen to working gas mixture the contribution of metallic Al to Al 2p (15%) is still considerably lower as compared to the value obtained in the case of deposition of Al NPs in pure Ar atmosphere, which was 76% [19].
M AN U
In other words, addition of O2 into argon causes a decrease of metallic fraction of prepared NPs. Taking into account the mean size of produced NPs (see following text) and the information depth of XPs that is 7-8 nm, we
AC C
EP
TE D
can conclude that AlxOy NPs are produced when the amount of oxygen is higher than 3%.
8
ACCEPTED MANUSCRIPT Oxygen admixture: 0.5%
Al2O3 - 85 %
78
76
74
72
70
a)
Binding energy (eV) Al2O3 - 100%
78
76
74
72
Binding energy (eV)
M AN U
Counts
SC
Oxygen admixture: 3.0%
RI PT
Counts
Al - 15 %
70
b)
TE D
Figure 4. High resolution XPS spectra of Al 2p peak measured on nanoparticle films deposited in Ar/O2 mixture containing a) 0.5% of oxygen and b) 3% of oxygen. Magnetron current 200 mA, pressure in the aggregation
EP
chamber 40 Pa, Ar flow 6 sccm.
In order to see the development of nanoparticles inside the aggregation chamber, Si substrates were placed
AC C
inside in certain distances from the magnetron and then the collected nanoparticles were studied by means of SEM (Figure 5). One can immediately notice that the size of produced NPs increases as the NPs travel from the magnetron towards the output orifice of the aggregation chamber. It has been found that the mean diameter of NPs rises from 6.1 ± 1.1 up to 16 ± 3 nm when the distance from the magnetron increases from 4 to 10 cm. This suggests that the particles that are created in the vicinity of the magnetron continue in their growth also in the magnetron afterglow region. Furthermore, it was found that AlxOy nanoparticle films deposited in the main deposition chamber, i.e. after passage through the orifice of GAS, are composed of individual nanoparticles whose mean diameter is close to the one of NPs detected in the gas aggregation chamber at the highest distance from the magnetron (see Figure 6).
9
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
Figure 5. SEM images and corresponding size histograms of NPs deposited inside the aggregation chamber at
EP
different distances from the target. a) 4cm b) 8 cm and c) 10 cm. Magnetron current 200 mA, pressure in the
AC C
aggregation chamber 40 Pa, 3% of oxygen. Ar flow 6 sccm.
In addition to SEM analysis of NPs, analytical GAS was used also for the characterization of plasma inside the aggregation chamber by means of optical emission spectroscopy. Emission spectra were recorded at the pressure 40 Pa, magnetron current 200 mA and for Ar/O2 mixture containing 3% of oxygen. However, discharge excitation processes that are obviously most intense close to the magnetron target dye down quit fast when one moves out of the target. This can be seen in Figure 7 as a spatial dependence of atomic Al, Ar and O lines on the distance from the magnetron to the output orifice of the aggregation chamber. Nevertheless, presence of atomic oxygen spectral lines in the emission spectra recorded at the large distance from the magnetron indicates non-
10
ACCEPTED MANUSCRIPT negligible amount of highly reactive O atoms in the magnetron afterglow region that suggests that oxidation of
M AN U
SC
RI PT
Al NPs may occur also in this part of the aggregation chamber.
Figure 6. SEM image of AlxOy NPs film deposited in the main deposition chamber Magnetron current 200 mA,
AC C
EP
TE D
pressure in the aggregation chamber 40 Pa, 3% of oxygen. Ar flow 6 sccm.
11
ACCEPTED MANUSCRIPT 30x
Distance from target 2 cm
10x
Al line
Ar lines
400
O line
Ar lines 200
a)
Intensity [a.u.]
0
Distance from target 5 cm
600
30x
10x 400
Ar lines
O line
200
400
420
830
440
840
SC
b)
0 850
Wavelength [nm]
O atom line Al atom line Ar atom line
10
M AN U
Intensity [a.u.]
1000
100
RI PT
Intensity [a.u.]
600
c)
1
0
2
4
Distance from magnetron [cm]
TE D
Figure 7. Sections of optical emission spectra recorded in the aggregation chamber at different distances from the magnetron Ar/O2 mixture containing 3% of oxygen, magnetron current 200 mA, pressure in the aggregation chamber 40 Pa and Ar flow 6 sccm. Distances from the magnetron were a) 2 cm and b) 5 cm. Variations of
AC C
c).
EP
intensities of spectral lines of atomic oxygen, Al and Ar with the distance from the magnetron are presented in
3.2 Ti and TiOx nanoparticles
The second part of this study was to test the possibility of titania nanoparticles production. In contrast to the situation when Al was used as starting material, it was possible to maintain a stable deposition of NPs using pure Ar as a working gas mixture, which is consistent with previous findings reported with the 2-inch magnetron [24]. However, similarly to the case of AlxOy nanoparticles and in agreement with previous study [31] there was a transition period, in which the deposition rate of nanoparticles was not stable and evolved with time. Therefore, in all experiments reported in this study, the plasma was operated for 30 min prior each measurement of deposition rate, magnetron voltage and optical emission spectra. It was found that the deposition rate at steady-
12
ACCEPTED MANUSCRIPT state conditions slightly increases with an addition of a small amount of oxygen to working gas mixture (around 1.5%, see figure 8a). Further addition of oxygen then leads to a rapid decrease of deposition rate of NPs and for O2 fraction in the working gas mixture equal to or higher than 3% the deposition stopped completely.
RI PT
As can be seen in Figure 8, the maximal value of deposition rate coincides with the oxygen concentration at which spectral lines of atomic oxygen become detectable, as well as with the oxygen fraction for which
magnetron voltage starts to rise steeply. These findings are in agreement with the assumption of Polonskyi et al. [39], who ascribed the observed dependence of deposition rate of Ti/TiOx on oxygen admixture to interplay
SC
between poisoning and cleaning of the target. In case of a low amount of O2, energetic ions are capable of
continuously cleaning the target surface. Such target cleaning does not allow the creation of a fully oxidized
M AN U
layer on it and thus also the sputtering rate and voltage on the magnetron remain stable. The oxygen present in the plasma may consequently interact with Ti and form stable TiOx seeds that facilitate the growth of NPs. In addition, TiOx seeds may be produced also by sputtering of titanium oxide fragments from partially oxidized Ti target. In this oxygen-deficient regime all O atoms are most likely captured by produced TiOx nanoparticles and thus no O atom emission lines are detected in the emission spectra at a distance of 3 cm from the magnetron. By
TE D
contrast, at high fractions of oxygen the poisoning of the target is faster than its cleaning. As a consequence, oxide layer is formed on the target surface. Because titanium oxide has lower ion induced secondary electron emission coefficient than metallic Ti [40], the presence of the oxidized layer causes a substantial increase of
EP
magnetron voltage. This increase is furthermore accompanied by an abrupt decrease of the sputtering rate, which in turn lowers the amount of material available for NPs creation and a decrease of deposition rate is observed.
AC C
Moreover, at this stage not all oxygen is consumed by growing NPs and thus O atom lines start to be visible in the emission spectra.
13
ACCEPTED MANUSCRIPT
Deposition rate
28 14 0
a)
250
Magnetron voltage
Intensity [a.u.]
200
800
800
b)
600 400 200
O atom line
0
400 200
M AN U
Ti atom line
600
0 0.00
c)
SC
300
RI PT
350
Intensity [a.u.]
Voltage [V]
∆f [Hz/min]
42
d)
0.02
0.04
[O2] / ([Ar]+[O2])
TE D
Figure 8. Dependence of a) deposition rate, b) magnetron voltage, c) intensity of Ti atom spectral line at 521 nm intensity and d) of O spectral line at wavelength 777 nm on oxygen fraction in the working gas mixture. Magnetron current 300 mA, pressure in the aggregation chamber 28 Pa. Ar flow was 1.5 sccm.
EP
Investigation of the chemical composition of nanoparticles was done for two samples deposited with and without oxygen in the working gas mixture. Both samples were transferred to the XPS measuring chamber without
AC C
breaking the vacuum in order to prevent additional oxidation and contamination from open air. The XPS analysis shows that the nanoparticles have a completely different chemistry. The results are shown in Figure 9. In case of NPs deposited in pure Ar atmosphere a well-separated signal from metallic Ti 1/2 and Ti 3/2 (92 %) and a small contribution from Ti bound to oxygen were observed. The Ti 2p peaks of the sample prepared in oxygen containing gas mixture was fitted by four components TiO2, TiOx, TiO and metallic Ti. As can be seen in this case the titanium-oxygen bonds dominate (97%), which indicates the oxide character of the formed nanoparticles.
14
ACCEPTED MANUSCRIPT
Oxygen amount: 0%
CPS
Ti
TiO2
TiO2
a) Oxygen amount: 1.4%
TiO
TiOxTiO
Ti
M AN U
TiOx
CPS
Ti TiO2
SC
TiO2
RI PT
Ti
b)
470
465
460
455
Binding energy [eV]
450
TE D
Figure 9. High resolution XPS spectra of Ti 2p peak. a) pure Ar or b) Ar/O2 mixture with 1.4% of O2. Magnetron current 300 mA, pressure in the aggregation chamber 28 Pa. Ar flow was 1.5 sccm.
EP
An effect of oxygen on the size of produced nanoparticles was also observed. As depicted in Figure 10, the nanoparticles are smaller when oxygen is present in the working gas mixture i.e. their mean size is ~27 nm in
AC C
case of oxygen addition and ~35 nm without oxygen.
15
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
Figure 10. SEM images and corresponding size histograms of NPs deposited using a) pure Ar or b) Ar/O2
1.5 sccm.
TE D
mixture with 1.4% of O2. Magnetron current 300 mA, pressure in the aggregation chamber 28 Pa. Ar flow was
EP
The NPs produced by GAS deposit on the substrate in a soft-landing regime and therefore they are held on the surface by weak van der Waals forces only. This was confirmed by scanning the sub-monolayer of Ti/TiOx NPs
AC C
by AFM in the contact mode. The smallest force of 1 nN reliably achievable with the cantilever used was capable of shifting the 25 nm particle from its position and even of removing it from the Si/SiO2 surface. This may limit practical usability of NP deposits. 4 Conclusions
It was demonstrated that the type of a simple gas aggregation source (GAS) may be utilized for production of alumina and titania nanoparticles with the mean diameter around 16 and 30 nm, respectively. In addition, the optimum working gas mixture from the point of view of deposition rate was identified as that containing 3% and 1.5 % of oxygen, respectively. A correlation was found between the rate of NPs production and the intensities of
16
ACCEPTED MANUSCRIPT spectral lines of atomic Ti and Al measured inside the aggregation chamber. Although the optical emission spectroscopy is not the direct method to quantify densities of atomic species presented in the plasma, observed dependences suggests that the rate at which NPs are produced is highly sensitive to the amount of atoms available for the NPs formation. The dependences of both emission spectra and NPs deposition rate on the amount of oxygen were explained by the interaction of oxygen with sputtered target that governs the magnetron
RI PT
voltage and thus also the sputtering yield. In addition, small fraction of oxygen in the working gas mixture
promotes formation of embryonic proto-particles that acts as seeds for further growth of NPs. Furthermore, it was found that in case of alumina the presence of oxygen not only promotes formation of AlxOy NPs, but it also
SC
stabilizes the NPs production.
Moreover, in case of AlxOy NPs, it was possible to determine evolution of the size of produced NPs in
M AN U
dependence on the position inside the gas aggregation chamber of the GAS. It was shown that nanoparticles in the aggregation chamber increase their size with the distance from the magnetron. This is experimental confirmation of previously reported theoretical models of NPs formation.
Finally, the metal oxide character of both types of fabricated NPs was verified by XPS analysis of deposited
TE D
nanoparticle films.
The possibility to fabricate metal oxide NPs by means of GAS is of high importance since this method is wellsuited for combination with other low-pressure deposition techniques and thus offers the possibility to produce
EP
nanocomposite materials. Moreover, deposition of plasma polymer overlayers may anchor otherwise loosely
AC C
bound NPs on the surface.
Acknowledgements
This work was supported by the Grant GACR 13-09853S from the Grant Agency of the Czech Republic.
References
[1] Gupta AK, Gupta M. Synthesis and surface engineering of iron oxide nanoparticles for biomedical applications. Biomaterials 2005;26:3995–4021.
17
ACCEPTED MANUSCRIPT [2] Gordon T, Perlstein B, Houbara O, Felner I, Banin E, Margel S. Synthesis and characterization of zinc/iron oxide composite nanoparticles and their antibacterial properties. Colloids and Surfaces A: Physicochemical and Engineering Aspects 2011;374:1–8. [3] Bandyopadhyay P, Dey A, Basu R, Das S, Nandy P. Synthesis and characterization of copper doped zinc oxide nanoparticles and its application in energy conversion. Current Applied Physics 2014;14:1149– 1155.
RI PT
[4] Ren G, Hu D, Cheng EWC, Vargas-Reus MA, Reip P, Allaker RP. Characterisation of copper oxide
nanoparticles for antimicrobial applications. International journal of antimicrobial agents 2009;33:587– 90.
[5] Richards BS. Comparison of TiO2 and other dielectric coatings for buried-contact solar cells: a review.
SC
Progress in Photovoltaics: Research and Applications 2004;12:253–281.
[6] Weir A, Westerhoff P, Fabricius L, Hristovski K, von Goetz N. Titanium dioxide nanoparticles in food and personal care products. Environmental science & technology 2012;46:2242–50.
M AN U
[7] Peters RJB, van Bemmel G, Herrera-Rivera Z, Helsper HPFG, Marvin HJP, Weigel S, et al. Characterization of titanium dioxide nanoparticles in food products: analytical methods to define nanoparticles. Journal of agricultural and food chemistry 2014;62:6285–93.
[8] Ash BJ, Siegel RW, Schadler LS. Mechanical Behavior of Alumina/Poly(methyl methacrylate) Nanocomposites. Macromolecules 2004;37:1358–1369.
[9] Mukherjee A, I MS, Prathna TC, Chandrasekaran N. Antimicrobial activity of aluminium oxide nanoparticles
TE D
for potential clinical applications. Science against microbial pathogens: communicating current research and technological advances, Ed.: Antonio Méndez-Vilas, Formatex Research Center, 2011;245–251. [10] DeFriend KA, Wiesner MR, Barron AR. Alumina and aluminate ultrafiltration membranes derived from alumina nanoparticles. Journal of Membrane Science 2003;224:11–28.
EP
[11] Pilch I, Söderström D, Hasan MI, Helmersson U, Brenning N. Fast growth of nanoparticles in a hollow cathode plasma through orbit motion limited ion collection. Applied Physics Letters 2013;103:193108.
AC C
[12] Pilch I, Söderström D, Brenning N, Helmersson U. Size-controlled growth of nanoparticles in a highly ionized pulsed plasma. Applied Physics Letters 2013;102: 033108.
[13] Haberland H, Karrais M, Mall M, Thurner Y. Thin films from energetic cluster impact: A feasibility study. Journal of Vacuum Scince & Technology A 1992;10: 3266-3271.
[14] Kylián O, Polonskyi O, Kratochvíl J, Artemenko A, Choukourov A, Drábik M, et al. Control of Wettability of Plasma Polymers by Application of Ti Nano-Clusters. Plasma Processes and Polymers 2012;9:180–7. [15] Kylián O, Petr M, Serov A, Solař P, Polonskyi O, Hanuš J, et al. Hydrophobic and super-hydrophobic coatings based on nanoparticles overcoated by fluorocarbon plasma polymer. Vacuum 2014;100:57–60. [16] Kuzminova A, Shelemin A, Kylián O, Petr M, Kratochvíl J, Solař P, Biederman H. From super-hydrophilic to super-hydrophobic surfaces using plasma polymerization combined with gas aggregation source of nanoparticles. Vacuum 2014;110: 58-61.
18
ACCEPTED MANUSCRIPT [17] Solař P, Kylián O, Petr M, Hanuš J, Choukourov A, Slavínská D, Biederman H. Glancing Angle Deposition of Silver Promoted by Pre-Deposited Nanoparticles. Plasma Processes and Polymers 2015; DOI: 10.1002/ppap.201400246. [18] Polonskyi O, Solař P, Kylián O, Drábik M, Artemenko A, Kousal J, et al. Nanocomposite metal/plasma polymer films prepared by means of gas aggregation cluster source. Thin Solid Films 2012;520:4155– 4162.
RI PT
[19] Polonskyi O, Kylián O, Drábik M, Kousal J, Solař P, Artemenko A, et al. Deposition of Al nanoparticles and their nanocomposites using a gas aggregation cluster source. Journal of Materials Science 2014;49:3352–3360.
[20] Ganeva M, Peter T, Bornholdt S, Kersten H, Strunskus T, Zaporojtchenko V, et al. Mass Spectrometric
Contributions to Plasma Physics 2012;52: 881-889.
SC
Investigations of Nano-Size Cluster Ions Produced by High Pressure Magnetron Sputtering.
[21] Majumdar A, Ganeva M, Köpp D, Datta D, Mishra P, Bhattacharayy S, et al. Surface morphology and
M AN U
composition of films grown by size-selected Cu nanoclusters. Vacuum 2008;83:719–723. [22] Kylián O, Valeš V, Polonskyi O, Pešička J, Čechvala J, Solař P, et al. Deposition of Pt nanoclusters by means of gas aggregation cluster source. Materials Letters 2012;79:229–231. [23] Marek A, Valter J, Kadlec S, Vyskočil J. Gas aggregation nanocluster source — Reactive sputter deposition of copper and titanium nanoclusters. Surface and Coatings Technology 2011;205:S573–S576. [24] Drábik M, Choukourov A, Artemenko A, Kousal J, Polonskyi O, Solař P, et al. Morphology of Titanium
2011;8:640–650.
TE D
Nanocluster Films Prepared by Gas Aggregation Cluster Source. Plasma Processes and Polymers
[25] Gracia-Pinilla M, Martínez E, Silva Vidaurri G, Pérez-Tijerina E. Deposition of Size-Selected Cu Nanoparticles by Inert Gas Condensation. Nanoscale Research Letters 2009;5:180-188.
EP
[26] Straňák V, Block S, Drache S, Hubička Z, Helm CA, Jastrabík L, et al. Size-controlled formation of Cu nanoclusters in pulsed magnetron sputtering system. Surface and Coatings Technology 2011; 205:
AC C
2755–2762.
[27] Polonskyi O, Kylián O, Solař P, Artemenko A, Kousal J, Slavínská D, Choukourov A, Biederman H. Nylon-sputtered nanoparticles: fabrication and basic properties. Journal of Phyics D: Applied Physics 2012; 45: 495301.
[28] Drábik, M, Serov, A, Kylián, O, Choukourov A, Artemenko A, Kousal J, et al. Deposition of Fluorocarbon Nanoclusters by Gas Aggregation Cluster Source. Plasma Processes and Polymers 2012; 9: 390–397. [29] Shyjumon I, Gopinadhan M, Helm CA, Smirnov BM, Hippler R. Deposition of titanium/titanium oxide clusters produced by magnetron sputtering. Thin Solid Films 2006;500:41 – 51.
19
ACCEPTED MANUSCRIPT [30] Peter T, Polonskyi O, Gojdka B, Ahadi AM, Strunskus T, Zaporojtchenko V, Biederman H, Faupel F. Influence of reactive gas admixture on transition metal cluster nucleation in a gas aggregation cluster source. Journal of Applied Physics. 2012; 112: 114321. [31] Ahadi AM, Zaporojtchenko V, Peter T, Polonskyi O, Strunskus T, Faupel F. Role of oxygen admixture in stabilizing TiOx nanoparticle deposition from a gas aggregation source. Journal of Nanoparticles
RI PT
Research 2013;15:2125. [32] Peter T, Polonskyi O, Gojdka B, Ahadi AM, Strunskus V., Zaporojtchenko V et al. Influence of reactive gas admixture on transition metal cluster nucleation in a gas aggregation cluster source Journal of Applied Physics 2012;112:114321.
SC
[33] Ahadi AM, Polonskyi O, Schürmann U, Strunskus T, Faupel F. Stable production of TiOx nanoparticles with narrow size distribution by reactive pulsed dc magnetron sputtering. Journal of Physics D: Applied
M AN U
Physics 2015;48: 035501.
[34] Choukourov A, Gordeev I, Arzhakov D, Artemenko A, Kousal J, Kylián O, et al. Does Cross-Link Density of PEO-Like Plasma Polymers Influence their Resistance to Adsorption of Fibrinogen? Plasma Processes and Polymers 2012;9:48–58.
[35] Depla D, Buyle G, Haemers J, De Gryse R. Discharge voltage measurements during magnetron sputtering. Surface and Coatings Technology 2006;200:4329–4338.
TE D
[36] Benka O, Steinbatz M. Oxidation of aluminum studied by secondary electron emission. 2003;525:207–214. [37] Phelps A, Petrovic Z. Cold-cathode discharges and breakdown in argon: surface and gas phase production of secondary electrons. Plasma Sources Science and Technology 1999;8:R21–R44. [38] Strohmeier BR. Comment on “‘aluminum deposition on polyimides: The effect of in situ ion
1990;8:3347.
EP
bombardment’.”Journal of Vacuum Science & Technology A: Vacuum, Surfaces, and Films
AC C
[39] Polonskyi O, Peter T, Mohammad Ahadi A, Hinz A, Strunskus T, Zaporojtchenko V, et al. Huge increase in gas phase nanoparticle generation by pulsed direct current sputtering in a reactive gas admixture. Applied Physics Letters 2013;103:033118.
[40] Depla D, Heirwegh S, Mahieu S, Haemers J, De Gryse R. Understanding the discharge voltage behavior during reactive sputtering of oxides. Journal of Applied Physics 2007;101:013301.
20
ACCEPTED MANUSCRIPT Al/AlxOy and Ti/TiOx nanoparticles (NPs) were produced by gas aggregation source
•
Dependence of the deposition rate on O2 fraction in the gas mixture was found
•
Optimal Ar/O2 working gas mixtures for production of nanoparticles were identified
•
O atoms were detected by OES even in the afterglow at a long distance from the target
•
Higher amount of O2 in the working gas mixture led to production of metal oxide NPs
AC C
EP
TE D
M AN U
SC
RI PT
•
S0042-207X(15)30012-9
DOI:
10.1016/j.vacuum.2015.07.008
Reference:
VAC 6744
To appear in:
Vacuum
Received Date: 29 December 2014 Revised Date:
11 July 2015
Accepted Date: 13 July 2015
Please cite this article as: Shelemin A, Kylián O, Hanuš J, Choukourov A, Melnichuk I, Serov A, Slavínská D, Biederman H, Preparation of metal oxide nanoparticles by gas aggregation cluster source, Vaccum (2015), doi: 10.1016/j.vacuum.2015.07.008. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Preparation of metal oxide nanoparticles by gas aggregation cluster source
Artem Shelemin, Ondřej Kylián1, Jan Hanuš, Andrei Choukourov, Iurii Melnichuk, Anton Serov, Danka
RI PT
Slavínská and Hynek Biederman
Department of Macromolecular Physics, Faculty of Mathematics and Physics, Charles University, V
SC
Holešovickách 2, 180 00 Praha 8, Czech Republic
M AN U
Keywords: nanoparticles; metal oxides; gas aggregation sources; TiOx; AlxOy
Abstract
Al/AlxOy and Ti/TiOx nanoparticles have been fabricated using gas aggregation cluster source (GAS) equipped
TE D
with 3-inch planar magnetrons using. The main attention has been devoted to the evaluation of the role of oxygen admixture on the deposition rate and properties of formed nanoparticles. It has been found that oxidized nanoparticles are deposited when oxygen is present in the working gas mixture. Highest deposition rates for
EP
production of Al/AlxOy or Ti/TiOx nanoparticles were observed when 3% or 1.5% oxygen was present in the working gas mixture, respectively. Mean diameters of about 16 nm for the AlxOy nanoparticles and 30 nm for
AC C
TiOx have been found under these deposition conditions.
1
Corresponding author. Phone: (+420) 22191 2258, Fax: (+420) 22191 2350, e-mail: [email protected]
1
ACCEPTED MANUSCRIPT 1 Introduction The fabrication of metal oxide nanoparticles (NPs) receives nowadays increasing attention. The main reason of this is a huge application potential of such materials that covers for instance water treatment, cosmetics, medicine, photovoltaic, semiconductor packing materials, fuel cells or catalysts [1-4]. Probably the most studied
RI PT
metal oxide NPs are titanium oxide that are employed in cosmetics, food, paints, personal care products or photovoltaic applications [5-7] and aluminium oxide NPs that are used in high-performance materials to increase their ductility, scratch resistance and toughness [8]. Another perspective field, where alumina NPs are expected to play an important role, is their use for production of antimicrobial materials [9].
SC
Although nanoparticles may be produced by techniques based on purely chemical synthesis (e.g. [10]), there is a growing interest in methods utilizing vacuum technologies, such as for example vacuum evaporation or
M AN U
sputtering at elevated pressures (typically tens to hundreds of Pa). Under these conditions the NPs are formed by sputtered vapor in the gas phase as the result of three step process that consist of nucleation, coagulation and attachment of atoms and ions onto created NPs. A recent example of this approach is nanoparticle synthesis by highly ionized plasmas produced by high power pulsed hollow cathode [11]. It was reported that the size distribution of produced NPs can be tuned in certain range by adjusting the pulse parameters, such as frequency,
TE D
pulse width, or peak current [12]. An alternative approach is based on the use of gas aggregation sources (GAS) of nanoparticles. In this case the magnetron is inserted into a cooled aggregation chamber that is separated from the rest of the deposition chamber by a small output orifice [13]. This arrangement assures sufficiently high
EP
pressure in the aggregation chamber needed for the effective production of NPs that are subsequently dragged by working gas through an output orifice into the vacuum deposition chamber and reach the substrate in the form of
AC C
a nanoparticle beam. The key advantage of GAS is complete separation of the NPs production from the rest of the deposition chamber, which facilitates combination of GAS with other vacuum based techniques of material deposition. This in turn makes it possible to produce either nanostructured coatings by overcoating NPs by a thin film of the same or other material (e.g. [14-16]), use NPs as seeds for formation of columnar structures prepared by glancing angle deposition [17] or various kinds of nanocomposite materials (e.g. [18,19]). Different GAS systems were already identified as reliable and highly effective sources for deposition of metallic NPs such as Ag, Cu, Pt, Ti or Al (e.g. [18-26]) or plasma polymerized NPs (e.g. [27,28]). In addition, it was shown that gas aggregation sources may also be used for production of metal-oxide NPs (e.g. [23,29-33]). Two strategies were followed. In the first one, the metallic nanoparticles are after the deposition oxidized on ambient
2
ACCEPTED MANUSCRIPT air (e.g. [29]). However, the necessity of the oxidation step makes this approach not suitable for combination of GAS with other vacuum based deposition methods needed for instance for production of nanocomposites. Because of this, there were attempts to produce metal-oxide NPs directly in the GAS. In this case small amount of oxygen was added to the inert working gas [30-33]. According to the reported results, the presence of oxygen influences not only the chemical structure of produced NPs, but it has also strong impact on the deposition rate
RI PT
and the stability of NPs formation. The main objective of this study is to evaluate the influence of amount of added oxygen on the process of TiOx and AlxOy NPs formation. In order to meet this aim, the measurements of deposition rate of NPs were accompanied by optical emission spectroscopy of plasma inside the aggregation
SC
chamber as well as by evaluation of chemical composition and sizes of NPs produced at steady-state condition.
M AN U
2 Materials and Methods
The scheme of apparatus used for production of AlxOy nanoparticles is depicted in Figure 1. It consisted of a gas aggregation source, which was attached onto a deposition chamber pumped by rotary and diffusion pumps to a base pressure 10-4 Pa. The used GAS was based on stainless steel cylindrical aggregation chamber (diameter
TE D
110 mm, length 450 mm) ending in an orifice 3 mm in diameter. The source of material for nanoclusters production was DC, water-cooled, planar, 3-inch magnetron equipped with 3 mm thick Al target (purity 99.99%) that was operated in constant current mode (200 mA). The distance between the target and the output orifice was 150 mm and was kept constant during all experiments. The aggregation chamber was equipped with diagnostic
EP
ports that enabled us to monitor plasma composition during the deposition of nanoclusters by means of optical emission spectroscopy (AvaSpec-3648-2-USB2). In addition, the magnetron and exit orifice can be moved
AC C
against diagnostic ports during the deposition (i.e. in-situ) while keeping the constant distance of 15 cm. This is a special ”diagnostic” variation of our simple GAS that enables e.g. OES measurements in an arbitrary part of the aggregation chamber.
3
RI PT
ACCEPTED MANUSCRIPT
SC
Figure 1. Scheme of deposition set-up with moveable Gas Aggregation Source used for Al and AlxOy NPs.
M AN U
Similar GAS was used for production of Ti and TiOx nanoparticles. In contrast to the source employed for fabrication of AlxOy NPs the aggregation chamber of this system was not moveable and the diagnostic ports were thus at a fixed 3 cm distance from the 3-inch magnetron. The diameter of the output orifice was 1.5 mm and all
TE D
experiments reported in this study were performed at constant applied 300 mA DC current.
Argon (99.99%) with an admixture of oxygen (up to 20%) was used in both cases as working gas. The flow of Ar (6 sccm in case of Al/AlxOy NPs and 1.5 sccm in case of Ti/TiOx NPs) was regulated by flow controller
EP
(MKS), while the amount of oxygen was tuned using a calibrated needle valve (Pfeiffer). The total pressure in the gas aggregation chamber was kept constant and equal to 40 Pa in experiments with Al and AlxOy NPs and 28
AC C
Pa in case of Ti an TiOx NPs. The pressure in the main deposition chamber was 0.1 Pa.
The deposition rate of produced NPs was measured by a quartz crystal microbalance. Chemical composition of produced nanoparticles was determined by means of X-ray Photoelectron Spectroscopy (XPS, Phobios 100, Specs) performed in-situ, i.e. without exposing produced NPs films to the ambient atmosphere. Spectra were acquired at constant take-off angle of 90° using Al Kα X-rays source (1486.6 eV, 200W, Specs). Wide spectra were acquired for binding energies in the range of 0 - 1100eV at a pass energy of 40 eV (dwell time 100 ms, step 0.5 eV). High resolution spectra were obtained at pass energy of 10 eV and were charge referenced to
4
ACCEPTED MANUSCRIPT adventitious carbon at 285.0 eV. Spectral analysis was performed using CASA XPS software. Finally, scanning electron microscopy (SEM, Tescan Mira 3) was used for the study of morphology of deposited coating.
Adhesion of NPs to Si/SiO2 substrates was studied by an AFM (Ntegra Prima, NT-MDT) both in a semi-contact and in a contact mode using diamond-like-carbon coated cantilevers (ContDLC, NanoandMore) with a
RI PT
guaranteed tip radius of 15 nm. The spring constant of 0.26 N/m was measured by the Sader method and the cantilever deflection detection system was calibrated before the measurements as described in [34]. First, the surface of the sub-monolayer of the NPs was scanned in the semicontact mode to identify an individual particle
SC
to be studied. Then the AFM was switched to the contact mode and the profile of the particle was continuously scanned with the increasing contact force until it was moved away. Finally, the AFM was switched back to the
3 Results and discussion 3.1 Al and AlxOy nanoparticles
M AN U
semi-contact mode and the entire area was scanned again to confirm that the particle was removed.
TE D
The first goal of this study was the identification of optimum experimental conditions which would provide stable and effective deposition of alumina nanoparticles. It was found that at constant magnetron current and pressure in the aggregation chamber, the deposition rate of NPs is strongly dependent on the used working gas
EP
mixture.
AC C
First, it was observed that the presence of oxygen significantly improves the stability of the process of formation of nanoparticles (see Figure 2): whereas in case of using Ar/O2 working gas mixture the amount of deposited NPs increased almost linearly with deposition time over the period of more than 30 minutes, the deposition rate, in agreement with a previous study [19], gradually decreased and finally stopped after prolonged operation time when only pure Ar was used.
5
SC
RI PT
ACCEPTED MANUSCRIPT
M AN U
Figure 2. Long-term stability of the deposition process in dependence on the presence of oxygen in the working gas mixture. Magnetron current 200 mA, pressure in the aggregation chamber 40 Pa. The Ar flow was 6 sccm.
Further experiments were performed after both the plasma parameters (magnetron voltage, intensity of emission spectral lines) and the deposition rate were stabilized. The temporal instabilities of deposition process in the
TE D
initial period, observed also in previous studies in case of Ti NPs [31], lasted typically 20-30 min depending on the deposition conditions and the history of the magnetron. After this phase the deposition rate, magnetron voltage and intensities of spectral lines become temporally stable: the variation of magnetron voltage was in the
EP
range ±3 V, and the variations of the deposition rate and intensities of spectral lines were less than 5%. As can be seen in Figure 3a, the deposition rate after the plasma stabilization in terms of deposited mass expressed in the
AC C
frequency shift of quartz crystal per minute initially rapidly increases with increasing fraction of oxygen in the working gas mixture and reaches its maximal value for about 3-5% of O2. Further increase of O2 fraction subsequently leads to gradual decrease of the deposition rate.
The dependence of the NPs deposition rate on the oxygen fraction in the working gas mixture may be explained as follows. On starting to add the small amount of oxygen to the working gas mixture an increase of magnetron voltage takes place due to oxygen chemisorption [35] (see Figure 3b). Since the O atom, upon an impact of an energetic ion, will contribute less to electron emission than an Al atom, the presence of adsorbed oxygen atoms will lower overall electron emission yield [36]. This in turn results in the increase of magnetron voltage. The
6
ACCEPTED MANUSCRIPT increase of magnetron voltage subsequently rises the energy of Ar ions bombarding the target, which leads to enhancement of the number of emitted Al atoms and thus also probability of NPs formation. This is consistent with the results of optical emission spectroscopy that showed rapid rise of the intensity of neutral Al atomic spectral lines with the addition of a small amount of O2 to working gas mixture (see Figure 3c) which reflects the increase of the density of Al atoms that are present in the plasma . In addition, it has to be stressed that the
RI PT
presence of oxygen in the gas phase facilitates formation of stable embryonic clusters [19] and thus promotes further growth of NPs.
100
a)
0 320
Magnetron voltage
300 280
b)
Intensity [a.u.]
260 600
Intensity [a.u.]
900
Al atom line
TE D
400 200 0
600 300
EP
Voltage [V]
SC
Deposition rate 200
M AN U
∆f [Hz/min]
300
O atom line
0 0.00
0.05
0.10
0.15
c)
d)
0.20
AC C
[O2]/[Ar]+[O2]
Figure 3. Dependence of a) deposition rate, b) magnetron voltage, c) intensity of Al spectral line at wavelength 396 nm and d) intensity of O atom spectral line at 777 nm on oxygen fraction in the working gas mixture. Magnetron current 200 mA, pressure in the aggregation chamber 40 Pa and Ar flow was 6 sccm.
Increasing the oxygen fraction in the working gas mixture subsequently causes the oxidation of Al target. Since the electron emission yield of Al oxide is higher than that of Al (e.g. [37]), the formation of aluminium oxide
7
ACCEPTED MANUSCRIPT layer may explain the observed decrease of the magnetron voltage as well as lowering of the number of Al atoms emitted from the magnetron target. The latter in turn causes a decrease in the rate of NPs formation.
The chemical structure of produced nanoparticle films prepared at 0.5% and 3% of O2 in the working gas
RI PT
mixture was determined by XPS without exposing the NPs to ambient air. The analysis of high resolution
spectra (see Figure 4) revealed that Al 2p peak is dominated by aluminium oxide peak positioned at binding energy 74.5 eV [38]. Moreover, it can be seen that the peak at binding energy 72 eV that belongs to metallic Al decays when oxygen fraction higher than 3% is used. However, also in case of a low amount of admixed
SC
oxygen to working gas mixture the contribution of metallic Al to Al 2p (15%) is still considerably lower as compared to the value obtained in the case of deposition of Al NPs in pure Ar atmosphere, which was 76% [19].
M AN U
In other words, addition of O2 into argon causes a decrease of metallic fraction of prepared NPs. Taking into account the mean size of produced NPs (see following text) and the information depth of XPs that is 7-8 nm, we
AC C
EP
TE D
can conclude that AlxOy NPs are produced when the amount of oxygen is higher than 3%.
8
ACCEPTED MANUSCRIPT Oxygen admixture: 0.5%
Al2O3 - 85 %
78
76
74
72
70
a)
Binding energy (eV) Al2O3 - 100%
78
76
74
72
Binding energy (eV)
M AN U
Counts
SC
Oxygen admixture: 3.0%
RI PT
Counts
Al - 15 %
70
b)
TE D
Figure 4. High resolution XPS spectra of Al 2p peak measured on nanoparticle films deposited in Ar/O2 mixture containing a) 0.5% of oxygen and b) 3% of oxygen. Magnetron current 200 mA, pressure in the aggregation
EP
chamber 40 Pa, Ar flow 6 sccm.
In order to see the development of nanoparticles inside the aggregation chamber, Si substrates were placed
AC C
inside in certain distances from the magnetron and then the collected nanoparticles were studied by means of SEM (Figure 5). One can immediately notice that the size of produced NPs increases as the NPs travel from the magnetron towards the output orifice of the aggregation chamber. It has been found that the mean diameter of NPs rises from 6.1 ± 1.1 up to 16 ± 3 nm when the distance from the magnetron increases from 4 to 10 cm. This suggests that the particles that are created in the vicinity of the magnetron continue in their growth also in the magnetron afterglow region. Furthermore, it was found that AlxOy nanoparticle films deposited in the main deposition chamber, i.e. after passage through the orifice of GAS, are composed of individual nanoparticles whose mean diameter is close to the one of NPs detected in the gas aggregation chamber at the highest distance from the magnetron (see Figure 6).
9
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
Figure 5. SEM images and corresponding size histograms of NPs deposited inside the aggregation chamber at
EP
different distances from the target. a) 4cm b) 8 cm and c) 10 cm. Magnetron current 200 mA, pressure in the
AC C
aggregation chamber 40 Pa, 3% of oxygen. Ar flow 6 sccm.
In addition to SEM analysis of NPs, analytical GAS was used also for the characterization of plasma inside the aggregation chamber by means of optical emission spectroscopy. Emission spectra were recorded at the pressure 40 Pa, magnetron current 200 mA and for Ar/O2 mixture containing 3% of oxygen. However, discharge excitation processes that are obviously most intense close to the magnetron target dye down quit fast when one moves out of the target. This can be seen in Figure 7 as a spatial dependence of atomic Al, Ar and O lines on the distance from the magnetron to the output orifice of the aggregation chamber. Nevertheless, presence of atomic oxygen spectral lines in the emission spectra recorded at the large distance from the magnetron indicates non-
10
ACCEPTED MANUSCRIPT negligible amount of highly reactive O atoms in the magnetron afterglow region that suggests that oxidation of
M AN U
SC
RI PT
Al NPs may occur also in this part of the aggregation chamber.
Figure 6. SEM image of AlxOy NPs film deposited in the main deposition chamber Magnetron current 200 mA,
AC C
EP
TE D
pressure in the aggregation chamber 40 Pa, 3% of oxygen. Ar flow 6 sccm.
11
ACCEPTED MANUSCRIPT 30x
Distance from target 2 cm
10x
Al line
Ar lines
400
O line
Ar lines 200
a)
Intensity [a.u.]
0
Distance from target 5 cm
600
30x
10x 400
Ar lines
O line
200
400
420
830
440
840
SC
b)
0 850
Wavelength [nm]
O atom line Al atom line Ar atom line
10
M AN U
Intensity [a.u.]
1000
100
RI PT
Intensity [a.u.]
600
c)
1
0
2
4
Distance from magnetron [cm]
TE D
Figure 7. Sections of optical emission spectra recorded in the aggregation chamber at different distances from the magnetron Ar/O2 mixture containing 3% of oxygen, magnetron current 200 mA, pressure in the aggregation chamber 40 Pa and Ar flow 6 sccm. Distances from the magnetron were a) 2 cm and b) 5 cm. Variations of
AC C
c).
EP
intensities of spectral lines of atomic oxygen, Al and Ar with the distance from the magnetron are presented in
3.2 Ti and TiOx nanoparticles
The second part of this study was to test the possibility of titania nanoparticles production. In contrast to the situation when Al was used as starting material, it was possible to maintain a stable deposition of NPs using pure Ar as a working gas mixture, which is consistent with previous findings reported with the 2-inch magnetron [24]. However, similarly to the case of AlxOy nanoparticles and in agreement with previous study [31] there was a transition period, in which the deposition rate of nanoparticles was not stable and evolved with time. Therefore, in all experiments reported in this study, the plasma was operated for 30 min prior each measurement of deposition rate, magnetron voltage and optical emission spectra. It was found that the deposition rate at steady-
12
ACCEPTED MANUSCRIPT state conditions slightly increases with an addition of a small amount of oxygen to working gas mixture (around 1.5%, see figure 8a). Further addition of oxygen then leads to a rapid decrease of deposition rate of NPs and for O2 fraction in the working gas mixture equal to or higher than 3% the deposition stopped completely.
RI PT
As can be seen in Figure 8, the maximal value of deposition rate coincides with the oxygen concentration at which spectral lines of atomic oxygen become detectable, as well as with the oxygen fraction for which
magnetron voltage starts to rise steeply. These findings are in agreement with the assumption of Polonskyi et al. [39], who ascribed the observed dependence of deposition rate of Ti/TiOx on oxygen admixture to interplay
SC
between poisoning and cleaning of the target. In case of a low amount of O2, energetic ions are capable of
continuously cleaning the target surface. Such target cleaning does not allow the creation of a fully oxidized
M AN U
layer on it and thus also the sputtering rate and voltage on the magnetron remain stable. The oxygen present in the plasma may consequently interact with Ti and form stable TiOx seeds that facilitate the growth of NPs. In addition, TiOx seeds may be produced also by sputtering of titanium oxide fragments from partially oxidized Ti target. In this oxygen-deficient regime all O atoms are most likely captured by produced TiOx nanoparticles and thus no O atom emission lines are detected in the emission spectra at a distance of 3 cm from the magnetron. By
TE D
contrast, at high fractions of oxygen the poisoning of the target is faster than its cleaning. As a consequence, oxide layer is formed on the target surface. Because titanium oxide has lower ion induced secondary electron emission coefficient than metallic Ti [40], the presence of the oxidized layer causes a substantial increase of
EP
magnetron voltage. This increase is furthermore accompanied by an abrupt decrease of the sputtering rate, which in turn lowers the amount of material available for NPs creation and a decrease of deposition rate is observed.
AC C
Moreover, at this stage not all oxygen is consumed by growing NPs and thus O atom lines start to be visible in the emission spectra.
13
ACCEPTED MANUSCRIPT
Deposition rate
28 14 0
a)
250
Magnetron voltage
Intensity [a.u.]
200
800
800
b)
600 400 200
O atom line
0
400 200
M AN U
Ti atom line
600
0 0.00
c)
SC
300
RI PT
350
Intensity [a.u.]
Voltage [V]
∆f [Hz/min]
42
d)
0.02
0.04
[O2] / ([Ar]+[O2])
TE D
Figure 8. Dependence of a) deposition rate, b) magnetron voltage, c) intensity of Ti atom spectral line at 521 nm intensity and d) of O spectral line at wavelength 777 nm on oxygen fraction in the working gas mixture. Magnetron current 300 mA, pressure in the aggregation chamber 28 Pa. Ar flow was 1.5 sccm.
EP
Investigation of the chemical composition of nanoparticles was done for two samples deposited with and without oxygen in the working gas mixture. Both samples were transferred to the XPS measuring chamber without
AC C
breaking the vacuum in order to prevent additional oxidation and contamination from open air. The XPS analysis shows that the nanoparticles have a completely different chemistry. The results are shown in Figure 9. In case of NPs deposited in pure Ar atmosphere a well-separated signal from metallic Ti 1/2 and Ti 3/2 (92 %) and a small contribution from Ti bound to oxygen were observed. The Ti 2p peaks of the sample prepared in oxygen containing gas mixture was fitted by four components TiO2, TiOx, TiO and metallic Ti. As can be seen in this case the titanium-oxygen bonds dominate (97%), which indicates the oxide character of the formed nanoparticles.
14
ACCEPTED MANUSCRIPT
Oxygen amount: 0%
CPS
Ti
TiO2
TiO2
a) Oxygen amount: 1.4%
TiO
TiOxTiO
Ti
M AN U
TiOx
CPS
Ti TiO2
SC
TiO2
RI PT
Ti
b)
470
465
460
455
Binding energy [eV]
450
TE D
Figure 9. High resolution XPS spectra of Ti 2p peak. a) pure Ar or b) Ar/O2 mixture with 1.4% of O2. Magnetron current 300 mA, pressure in the aggregation chamber 28 Pa. Ar flow was 1.5 sccm.
EP
An effect of oxygen on the size of produced nanoparticles was also observed. As depicted in Figure 10, the nanoparticles are smaller when oxygen is present in the working gas mixture i.e. their mean size is ~27 nm in
AC C
case of oxygen addition and ~35 nm without oxygen.
15
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
Figure 10. SEM images and corresponding size histograms of NPs deposited using a) pure Ar or b) Ar/O2
1.5 sccm.
TE D
mixture with 1.4% of O2. Magnetron current 300 mA, pressure in the aggregation chamber 28 Pa. Ar flow was
EP
The NPs produced by GAS deposit on the substrate in a soft-landing regime and therefore they are held on the surface by weak van der Waals forces only. This was confirmed by scanning the sub-monolayer of Ti/TiOx NPs
AC C
by AFM in the contact mode. The smallest force of 1 nN reliably achievable with the cantilever used was capable of shifting the 25 nm particle from its position and even of removing it from the Si/SiO2 surface. This may limit practical usability of NP deposits. 4 Conclusions
It was demonstrated that the type of a simple gas aggregation source (GAS) may be utilized for production of alumina and titania nanoparticles with the mean diameter around 16 and 30 nm, respectively. In addition, the optimum working gas mixture from the point of view of deposition rate was identified as that containing 3% and 1.5 % of oxygen, respectively. A correlation was found between the rate of NPs production and the intensities of
16
ACCEPTED MANUSCRIPT spectral lines of atomic Ti and Al measured inside the aggregation chamber. Although the optical emission spectroscopy is not the direct method to quantify densities of atomic species presented in the plasma, observed dependences suggests that the rate at which NPs are produced is highly sensitive to the amount of atoms available for the NPs formation. The dependences of both emission spectra and NPs deposition rate on the amount of oxygen were explained by the interaction of oxygen with sputtered target that governs the magnetron
RI PT
voltage and thus also the sputtering yield. In addition, small fraction of oxygen in the working gas mixture
promotes formation of embryonic proto-particles that acts as seeds for further growth of NPs. Furthermore, it was found that in case of alumina the presence of oxygen not only promotes formation of AlxOy NPs, but it also
SC
stabilizes the NPs production.
Moreover, in case of AlxOy NPs, it was possible to determine evolution of the size of produced NPs in
M AN U
dependence on the position inside the gas aggregation chamber of the GAS. It was shown that nanoparticles in the aggregation chamber increase their size with the distance from the magnetron. This is experimental confirmation of previously reported theoretical models of NPs formation.
Finally, the metal oxide character of both types of fabricated NPs was verified by XPS analysis of deposited
TE D
nanoparticle films.
The possibility to fabricate metal oxide NPs by means of GAS is of high importance since this method is wellsuited for combination with other low-pressure deposition techniques and thus offers the possibility to produce
EP
nanocomposite materials. Moreover, deposition of plasma polymer overlayers may anchor otherwise loosely
AC C
bound NPs on the surface.
Acknowledgements
This work was supported by the Grant GACR 13-09853S from the Grant Agency of the Czech Republic.
References
[1] Gupta AK, Gupta M. Synthesis and surface engineering of iron oxide nanoparticles for biomedical applications. Biomaterials 2005;26:3995–4021.
17
ACCEPTED MANUSCRIPT [2] Gordon T, Perlstein B, Houbara O, Felner I, Banin E, Margel S. Synthesis and characterization of zinc/iron oxide composite nanoparticles and their antibacterial properties. Colloids and Surfaces A: Physicochemical and Engineering Aspects 2011;374:1–8. [3] Bandyopadhyay P, Dey A, Basu R, Das S, Nandy P. Synthesis and characterization of copper doped zinc oxide nanoparticles and its application in energy conversion. Current Applied Physics 2014;14:1149– 1155.
RI PT
[4] Ren G, Hu D, Cheng EWC, Vargas-Reus MA, Reip P, Allaker RP. Characterisation of copper oxide
nanoparticles for antimicrobial applications. International journal of antimicrobial agents 2009;33:587– 90.
[5] Richards BS. Comparison of TiO2 and other dielectric coatings for buried-contact solar cells: a review.
SC
Progress in Photovoltaics: Research and Applications 2004;12:253–281.
[6] Weir A, Westerhoff P, Fabricius L, Hristovski K, von Goetz N. Titanium dioxide nanoparticles in food and personal care products. Environmental science & technology 2012;46:2242–50.
M AN U
[7] Peters RJB, van Bemmel G, Herrera-Rivera Z, Helsper HPFG, Marvin HJP, Weigel S, et al. Characterization of titanium dioxide nanoparticles in food products: analytical methods to define nanoparticles. Journal of agricultural and food chemistry 2014;62:6285–93.
[8] Ash BJ, Siegel RW, Schadler LS. Mechanical Behavior of Alumina/Poly(methyl methacrylate) Nanocomposites. Macromolecules 2004;37:1358–1369.
[9] Mukherjee A, I MS, Prathna TC, Chandrasekaran N. Antimicrobial activity of aluminium oxide nanoparticles
TE D
for potential clinical applications. Science against microbial pathogens: communicating current research and technological advances, Ed.: Antonio Méndez-Vilas, Formatex Research Center, 2011;245–251. [10] DeFriend KA, Wiesner MR, Barron AR. Alumina and aluminate ultrafiltration membranes derived from alumina nanoparticles. Journal of Membrane Science 2003;224:11–28.
EP
[11] Pilch I, Söderström D, Hasan MI, Helmersson U, Brenning N. Fast growth of nanoparticles in a hollow cathode plasma through orbit motion limited ion collection. Applied Physics Letters 2013;103:193108.
AC C
[12] Pilch I, Söderström D, Brenning N, Helmersson U. Size-controlled growth of nanoparticles in a highly ionized pulsed plasma. Applied Physics Letters 2013;102: 033108.
[13] Haberland H, Karrais M, Mall M, Thurner Y. Thin films from energetic cluster impact: A feasibility study. Journal of Vacuum Scince & Technology A 1992;10: 3266-3271.
[14] Kylián O, Polonskyi O, Kratochvíl J, Artemenko A, Choukourov A, Drábik M, et al. Control of Wettability of Plasma Polymers by Application of Ti Nano-Clusters. Plasma Processes and Polymers 2012;9:180–7. [15] Kylián O, Petr M, Serov A, Solař P, Polonskyi O, Hanuš J, et al. Hydrophobic and super-hydrophobic coatings based on nanoparticles overcoated by fluorocarbon plasma polymer. Vacuum 2014;100:57–60. [16] Kuzminova A, Shelemin A, Kylián O, Petr M, Kratochvíl J, Solař P, Biederman H. From super-hydrophilic to super-hydrophobic surfaces using plasma polymerization combined with gas aggregation source of nanoparticles. Vacuum 2014;110: 58-61.
18
ACCEPTED MANUSCRIPT [17] Solař P, Kylián O, Petr M, Hanuš J, Choukourov A, Slavínská D, Biederman H. Glancing Angle Deposition of Silver Promoted by Pre-Deposited Nanoparticles. Plasma Processes and Polymers 2015; DOI: 10.1002/ppap.201400246. [18] Polonskyi O, Solař P, Kylián O, Drábik M, Artemenko A, Kousal J, et al. Nanocomposite metal/plasma polymer films prepared by means of gas aggregation cluster source. Thin Solid Films 2012;520:4155– 4162.
RI PT
[19] Polonskyi O, Kylián O, Drábik M, Kousal J, Solař P, Artemenko A, et al. Deposition of Al nanoparticles and their nanocomposites using a gas aggregation cluster source. Journal of Materials Science 2014;49:3352–3360.
[20] Ganeva M, Peter T, Bornholdt S, Kersten H, Strunskus T, Zaporojtchenko V, et al. Mass Spectrometric
Contributions to Plasma Physics 2012;52: 881-889.
SC
Investigations of Nano-Size Cluster Ions Produced by High Pressure Magnetron Sputtering.
[21] Majumdar A, Ganeva M, Köpp D, Datta D, Mishra P, Bhattacharayy S, et al. Surface morphology and
M AN U
composition of films grown by size-selected Cu nanoclusters. Vacuum 2008;83:719–723. [22] Kylián O, Valeš V, Polonskyi O, Pešička J, Čechvala J, Solař P, et al. Deposition of Pt nanoclusters by means of gas aggregation cluster source. Materials Letters 2012;79:229–231. [23] Marek A, Valter J, Kadlec S, Vyskočil J. Gas aggregation nanocluster source — Reactive sputter deposition of copper and titanium nanoclusters. Surface and Coatings Technology 2011;205:S573–S576. [24] Drábik M, Choukourov A, Artemenko A, Kousal J, Polonskyi O, Solař P, et al. Morphology of Titanium
2011;8:640–650.
TE D
Nanocluster Films Prepared by Gas Aggregation Cluster Source. Plasma Processes and Polymers
[25] Gracia-Pinilla M, Martínez E, Silva Vidaurri G, Pérez-Tijerina E. Deposition of Size-Selected Cu Nanoparticles by Inert Gas Condensation. Nanoscale Research Letters 2009;5:180-188.
EP
[26] Straňák V, Block S, Drache S, Hubička Z, Helm CA, Jastrabík L, et al. Size-controlled formation of Cu nanoclusters in pulsed magnetron sputtering system. Surface and Coatings Technology 2011; 205:
AC C
2755–2762.
[27] Polonskyi O, Kylián O, Solař P, Artemenko A, Kousal J, Slavínská D, Choukourov A, Biederman H. Nylon-sputtered nanoparticles: fabrication and basic properties. Journal of Phyics D: Applied Physics 2012; 45: 495301.
[28] Drábik, M, Serov, A, Kylián, O, Choukourov A, Artemenko A, Kousal J, et al. Deposition of Fluorocarbon Nanoclusters by Gas Aggregation Cluster Source. Plasma Processes and Polymers 2012; 9: 390–397. [29] Shyjumon I, Gopinadhan M, Helm CA, Smirnov BM, Hippler R. Deposition of titanium/titanium oxide clusters produced by magnetron sputtering. Thin Solid Films 2006;500:41 – 51.
19
ACCEPTED MANUSCRIPT [30] Peter T, Polonskyi O, Gojdka B, Ahadi AM, Strunskus T, Zaporojtchenko V, Biederman H, Faupel F. Influence of reactive gas admixture on transition metal cluster nucleation in a gas aggregation cluster source. Journal of Applied Physics. 2012; 112: 114321. [31] Ahadi AM, Zaporojtchenko V, Peter T, Polonskyi O, Strunskus T, Faupel F. Role of oxygen admixture in stabilizing TiOx nanoparticle deposition from a gas aggregation source. Journal of Nanoparticles
RI PT
Research 2013;15:2125. [32] Peter T, Polonskyi O, Gojdka B, Ahadi AM, Strunskus V., Zaporojtchenko V et al. Influence of reactive gas admixture on transition metal cluster nucleation in a gas aggregation cluster source Journal of Applied Physics 2012;112:114321.
SC
[33] Ahadi AM, Polonskyi O, Schürmann U, Strunskus T, Faupel F. Stable production of TiOx nanoparticles with narrow size distribution by reactive pulsed dc magnetron sputtering. Journal of Physics D: Applied
M AN U
Physics 2015;48: 035501.
[34] Choukourov A, Gordeev I, Arzhakov D, Artemenko A, Kousal J, Kylián O, et al. Does Cross-Link Density of PEO-Like Plasma Polymers Influence their Resistance to Adsorption of Fibrinogen? Plasma Processes and Polymers 2012;9:48–58.
[35] Depla D, Buyle G, Haemers J, De Gryse R. Discharge voltage measurements during magnetron sputtering. Surface and Coatings Technology 2006;200:4329–4338.
TE D
[36] Benka O, Steinbatz M. Oxidation of aluminum studied by secondary electron emission. 2003;525:207–214. [37] Phelps A, Petrovic Z. Cold-cathode discharges and breakdown in argon: surface and gas phase production of secondary electrons. Plasma Sources Science and Technology 1999;8:R21–R44. [38] Strohmeier BR. Comment on “‘aluminum deposition on polyimides: The effect of in situ ion
1990;8:3347.
EP
bombardment’.”Journal of Vacuum Science & Technology A: Vacuum, Surfaces, and Films
AC C
[39] Polonskyi O, Peter T, Mohammad Ahadi A, Hinz A, Strunskus T, Zaporojtchenko V, et al. Huge increase in gas phase nanoparticle generation by pulsed direct current sputtering in a reactive gas admixture. Applied Physics Letters 2013;103:033118.
[40] Depla D, Heirwegh S, Mahieu S, Haemers J, De Gryse R. Understanding the discharge voltage behavior during reactive sputtering of oxides. Journal of Applied Physics 2007;101:013301.
20
ACCEPTED MANUSCRIPT Al/AlxOy and Ti/TiOx nanoparticles (NPs) were produced by gas aggregation source
•
Dependence of the deposition rate on O2 fraction in the gas mixture was found
•
Optimal Ar/O2 working gas mixtures for production of nanoparticles were identified
•
O atoms were detected by OES even in the afterglow at a long distance from the target
•
Higher amount of O2 in the working gas mixture led to production of metal oxide NPs
AC C
EP
TE D
M AN U
SC
RI PT
•