Preparation of Plasma-polymerized SiOx-like Thin Films From a Mixture

download Preparation of Plasma-polymerized SiOx-like Thin Films From a Mixture

of 6

Transcript of Preparation of Plasma-polymerized SiOx-like Thin Films From a Mixture

  • 8/19/2019 Preparation of Plasma-polymerized SiOx-like Thin Films From a Mixture

    1/6

    Preparation of plasma-polymerized SiO x-like thin films from a mixture

    of hexamethyldisiloxane and oxygen to improve the corrosion behaviour 

    E. Vassalloa,*, A. Cremonaa , L. Laguardiaa , E. Mesto b

    a  Istituto di Fisica del Plasma-CNR, Milano, Italy b Dipartimento di Geomineralogia, Università degli Studi di Bari, Italy

    Received 27 July 2004; accepted in revised form 1 November 2004

    Available online 2 December 2004

    Abstract

    Chemical and electrochemical properties of plasma polymerized SiO x-like thin films have been studied to search for an alternative to the

    anticorrosive coatings produced by chromate conversion treatment. The films were deposited on steel substrates by means of 

    hexamethydisiloxane/oxygen-fed plasmas ignited in an RF capacitive coupled parallel plates reactor. Their surface chemical characterization

    was carried out by means of X-ray photoelectron spectroscopy (XPS) and infrared spectroscopy (FT-IR). The protective abilities of SiO  x-like

    films as a function of RF power and gas feed composition in aerated 1 M Na 2SO4   solution were examined by electrochemical methods.

    Alone and in combination, these techniques help us to understand the relationship between the deposition parameters and the properties of the

    deposited films.

    D  2004 Elsevier B.V. All rights reserved.

     Keywords:  Corrosion; PECVD; Plasma polymerization

    1. Introduction

    Corrosion is one of the most serious problems for 

    metallic materials utilised for many application. From many

    years, zinc particles have been used as a metal coating on

    steel to reduce corrosion. However, when materials covered

    with zinc are exposed to the atmosphere or aqueous

    solutions a corrosion product, Zn(OH)2, is rapidly gener-

    ated. Therefore, chromate chemical conversion (CCC) films

    have been widely used for many engineering materials to

     provide improved corrosion resistance. Chromium chemicalanalysis has shown that the corrosion resistance of the CCC

    films is due to the inhibiting effect resulting from the

    dissolution of Cr(VI) ions   [1]. Unfortunately, Cr(VI) ions

    that remain in the establishment of the chromium plating are

    remarkably toxic and carcinogenic for humans. According

    to Appendix II of the European Directive 2000/53/CE, as

    from 1st July 2007 coatings containing hexavalent chro-

    mium shall no longer be used for corrosion prevention. In

    the EU Directive on Waste Electrical and Electronic

    Equipment, the use of hexavalent chromium is also

    forbidden as from 1st January 2006. The dates quoted in

    the legislation require immediate action by manufacturers of 

    coating materials, by coating plants, by fastener manufac-

    turers and by vehicle manufacturers; consequently, an

    alternative surface treatment is necessary [2].

    Vapour phase deposition of thin films from a plasma

    environment is a flexible strategy to modify the surface of materials   [3].   Several surface chemical treatments are

     possible and a wide range of substrate materials can be

    coated. The application of these coating includes corrosion

     protection, biomaterials, vapours barrier and area where

    surface modification can enhance performance. In the

     plasma deposition process, a gaseous or volatile compound

    is introduced into a reaction chamber, fragmented and/or 

    ionized in a glow discharge plasma, and reassembled on the

    surface of the sample. The variables that can be readily

    controlled during this ionization–deposition process are

    0257-8972/$ - see front matter  D  2004 Elsevier B.V. All rights reserved.

    doi:10.1016/j.surfcoat.2004.11.001

    * Corresponding author. Tel.: +39 2 66173245; fax: +39 2 66173203.

     E-mail address:  [email protected] (E. Vassallo).

    Surface & Coatings Technology 200 (2006) 3035– 3040

    www.elsevier.com/locate/surfcoat 

  • 8/19/2019 Preparation of Plasma-polymerized SiOx-like Thin Films From a Mixture

    2/6

     power density, precursor gas flow rate, pressure, reactor 

    geometry and plasma frequency. The ionized species formed

    in the process acquire energies typically in the range of 0–2

    eV, while electrons and metastables can achieve energies up

    to 20 eV [4]. The processes that simultaneously occur in the

    reaction chamber are ionization, neutralisation, recombina-

    tion, etching and polymerisation. For these reasons, plasma-enhanced chemical vapour deposition (PECVD) appears to

     be an important technology for the surface modification of 

    metallic materials in order to increase their corrosion

    resistance. The deposited films are usually branched, highly

    crosslinked, insoluble, pinhole-free and adhere well to the

    substrates. Organosilicon precursors can be utilised to

    deposit SiO x   thin films, which are a good candidate for 

     protective coating on metallic and other substrates   [5–9].

    This monomer has been chosen for the flexibility of the Si– 

    O–Si bond in the backbone [7]  and also for the resistance to

    water permeation characteristics of polisiloxane films [10].

    In order to provide an adequate corrosion protection, thecoating must be uniform and well adhered on the substrate;

    it has been demonstrated that Fe–O–Si oxane bonds, created

    at the interface between the plasma-polymer and the steel

    surface, enhance the adhesion of the film  [11].   In any case

     preview study demonstrated that hydrogen pre-treatment 

    removes surface contamination layer and improves the

    coating-substrate performances. In fact, hydrogen removes

    carbon from the surface and creates on the surface of the

    substrate catalytic active points, so that the surface reactions

     between the adsorbed precursor fragments and the substrate

    are facilitated   [12],  also at low temperatures. The goal of 

    this paper is to evaluate the effect of several deposition

     parameters on the electrochemical behaviour of the depos-

    ited films.

    2. Experimental

    Plasma polymerization has been carried out in a stainless

    steel reactor with a parallel plate configuration and a 5 cm

    inter-electrode gap. The upper electrode (diameter=15 cm)

    is RF driven, the lower ground steel electrode (diameter 20

    cm) holds the substrate to be processed. The reactor was

     pumped by a turbomolecular and a rotary pump. The flow

    rates of the feed gases and of the HMDSO vapour werecontrolled by means of MKS mass flow meters, and the

     pressure was monitored by an MKS capacitive gauge. SiO xfilms were deposited on iron dishes. The experiments were

     performed varying the HMDSO–O2   flow rate ratio and the

    input power, keeping constant the pressure at 100 mTorr.

    After cleaning the sample using acetone, the substrates were

    in situ pre-treated with hydrogen plasma at 50 W and 500

    mTorr to give an oxygen-free surface and a buffer layer to

    enhance the film adhesion.

    SiO x   films were also deposited onto polished Si

    substrates for ex situ FTIR chemical characterization

     performed by means of a Perkin Elmer Spectrum One IR 

    Fourier spectrometer under N2  purging and in the presence

    of a desiccant reducing water adsorption on the samples. In

    order to gather information about the chemical composition

    of deposited SiO x   films, X-ray photoelectron spectroscopy

    (XPS) analysis was performed by a Leybold LHS10

    spectrometer using an unmonochromatised Al K a   source.

    Base vacuum was 109 mbar or better. The energy scale of the spectrometer was calibrated using Cu and Au by the

     binding energies Cu2p3/2=932.6 eV and Au4f 7/2=84.0 eV.

    Survey spectra were acquired in fixed retarding ratio mode

    (retarding ratio=3) and high resolution spectra were

    recorded in fixed analyser transmission mode, with a pass

    energy of 30 eV. Calibration of binding energy scales was

     performed by taking the alkyl component of the C1s

     photoelectron peak (BE=284.8 eV). Hardware and software

    for spectra acquisition and processing are described in

    Desimoni and Malitesta (1986) and Malitesta et al. (1989).

    In particular background subtraction is performed by the

    Shirley’s method (Shirley, 1972). As far as the fitting procedure is concerned, a non-linear least squares fitting

     program was employed to fit as many peaks as the guessed

    chemical species to experimental spectrum. In this process,

     peak parameters (position, height, width, etc.) are optimised

    starting from initial estimates. Quantitative analysis was

     performed by using high-resolution spectra. Atomic ratios

    were obtained by dividing each peak area by empirically

    derived sensitivity factor. Polarization measurements of the

    electrodes, both bare and covered with SiO x   film, were

    carried out potentiodynamically in aerated 1 M Na2SO4solution at pH 5.3 and at the laboratory temperature of 

    25F1   8C. After the immersion of the electrode in the

    solution for the time necessary to achieve the stability of the

    open-circuit potential (generally 3 h), the potential of the

    electrode was swept at a rate of 0.166 mV/s from the initial

     potential of 250 mV vs. E corr  to the final potential of 1000

    mV vs.   E corr . A three-electrode electrochemical configu-

    ration was used with a clamp on electrolyte cell attached by

    a rubber O-ring. The exposure area of the samples was 1

    cm2. A Silver–Silver chlorine (Ag–AgCl) electrode as

    reference and a titanium counter electrode were employed.

    3. Results and discussion

    3.1. SiO x   film characterization

    3.1.1. Infrared spectroscopy

    In   Fig. 1,   the infrared spectra of the deposited films

    (about 1200 nm thick) are shown as a function of the O2/ 

    monomer ratio in the plasma feed gas. The characteristic

    features of the Si–O–Si group   [13–20]   are evident:

    asymmetric stretching and bending mode at around 1070

    and 800 cm1, respectively. The adsorption bands at 2970

    and 1260 cm1, relative to the stretching of CH x   and the

     bending modes of methyl groups in Si(CH3) x, mark the

     presence of organic components in the film.

     E. Vassallo et al. / Surface & Coatings Technology 200 (2006) 3035–30403036

  • 8/19/2019 Preparation of Plasma-polymerized SiOx-like Thin Films From a Mixture

    3/6

    Another important feature is t he absorption of the

    characteristic bands of Si–OH bonds [13–23], OH stretching

    in H-bonded silanol at around 3500 cm

    1

    and the bendingmode at 930 cm1. It can be observed that, in low dilution

    conditions, the film infrared spectra show the characteristic

    features of carbon-containing moieties (in particular, Si–C

     bonds). This is in agreement with the kinetics described by

    Lamendola et al.  [24], which states that under low dilution

    conditions the deposition precursors are mainly SiC xH y  and

    CH x radicals, and the film stoichiometry can be expressed as

    SiC xH y O z . As the O2/monomer ratio increases, the

    enhanced oxidation efficiency produces more SiO radicals

    and CO2volatile molecules than Si and CH, leading to a

    marked inorganic character of the film (Si–O–Si functional

    groups prevail over Si(CH3) x), thus the film chemistry

    approaches the SiO2 one. Unfortunately, this can result in acontemporaneous increase of the silanol concentration [25– 

    27], reducing the chemical stability of the coating and

    therefore its corrosion resistance. This inconvenience can be

    solved rising the input power, in order to reduce the SiOH

    concentration as well as the residual organic groups. The

    effect of the power has been investigated at a constant O2/ 

    HMDSO ratio of 6,6. In   Fig. 2,   the infrared adsorption

    spectra of the plasma-deposited films are shown at different 

     power conditions.

    It is interesting to observe that the presence of carbon

    within the film is detected only for a power value of 50 W.The SiOH-related bands decrease when power increases.

    This trend can be explained with a higher input energy,

    leading to a higher ion bombardment of the growing film,

    which can enhance the condensation of vicinal silanol

    groups to form Si–O–Si bridges.

    3.1.2. X-ray photoelectron spectroscopy

    The chemical composition of the deposited SiO x   films

    has been investigated by XPS. In particular, it was studied

    the intensity of the peaks Si2p, Si2s, C1s, and O1s. This

    analysis, as well as infrared, permits to follow the evolution

    of the concentration of diverse elements as a function of 

     preparation conditions. Fig. 3 shows the XPS spectrum of afilm obtained by an RF plasma of HMDSO mixed with

    oxygen (oxygen percentage 58.8%).

    X-ray photoelectron spectroscopy has shown that oxy-

    gen content rises from 43% to 65%, while the silicon

    atomic concentration remains almost constant at 29%, when

    the O2/HMDSO ratio varies in the range from 1.4 to 13.3

    (see Fig. 4).

    Fig. 2. Effect of RF power on the FTIR spectra obtained at 100 mTorr and

    O2/HMDSO ratio 6.6.

    Fig. 3. XPS spectra of films deposited from HMDSO and 58.8% O 2, at 200

    W and 5101 mbar.

    Fig. 4. Surface composition as a function of the O2/HMDSO ratio for 

    samples obtained at 100 W and 5102 mbar.

    Fig. 1. Effect of feed gas composition on FTIR spectra obtained at 100 W

    and 100 mTorr (the O2/HMDSO ratios are marked on the spectra).

     E. Vassallo et al. / Surface & Coatings Technology 200 (2006) 3035–3040   3037

  • 8/19/2019 Preparation of Plasma-polymerized SiOx-like Thin Films From a Mixture

    4/6

    Moreover, the carbon content drops to a nearly negligible

    value, indicating the decay of the organic moieties as the

    oxygen concentration in the plasma increases. The carbon/ 

    silicon (C/Si) and oxygen/silicon (O/Si) atomic ratio trends

    are plotted in   Fig. 5a against the discharge power.

    Considering that the C/Si and O/Si ratio are res pectively 3

    and 0.5 for the monomer and 2 and 1 for PDMS  [28], the C/ Si trend indicates that when power increases an abstraction

    of the methyl groups, substantial loss of carbon on deposited

    film occurs, while as Si–O bonds did not seem to undergo

    scission to an appreciable extent. On the other hand, the O/ 

    Si trend suggests that slight oxidation process takes place

    when the power rises.

    The experimental determined O1s binding energy of 

    532.2 eV is consistent with the O1s signal in siloxane (Si– 

    O–Si) by comparison with earlier studies (Gengenbach and

    Griesser, 1999 and references therein). The carbon peaks

    have been decomposed in two component, a high contribu-

    tion, attributed to carbon bound to hydrogen (C–H) and tocarbon bound to silicium (C–Si), both considered at a

     binding energy of 284.8 eV, and a minor component at 

    higher binding energy, consistent with carbon singly bound

    to oxygen such as ether, hydroxyl or epoxy carbon

    functionary at a binding energy of 286.3 eV. The inves-

    tigation of the Si2p signal results in a peak shift towards

    higher binding energy when power increases. The high

    value of its FWHM indicates several environments of 

    silicon atoms. We assumed that this signal is a compound

    line with unresolved contributions from chemically different 

    species and assuming that all the silicon atoms in the deposit 

    have a valence of four. The major factor in determining theSi2p binding energy is the number of oxygen atom. On the

    other hand, the secondary effect upon the position of Si2p

    signal due to various combination of carbon and hydrogen

    atoms fulfilling the remaining bonding requirements of the

    silicon has been assumed negligible in comparison to the

    shift caused by an oxygen. The spectra data has been fitted

     by four component peak within the Si2p envelope: SiOC3(BE=101.5 eV), SiO2C2 (BE=102.1 eV), SiO3C (BE=102.8

    eV) and SiO4   (BE=103.4 eV). Each component has been

    fitted as symmetrical peak. The Gaussian to Lorentzian ratio

    is kept constant at 85%. The fitting   of Si2 p spectra from

    HDMSO/O2  deposits is illustrated in  Fig. 5 b.The effect of power in glow discharges on the

    distribution of the four silicon environments is reported in

    Fig. 5c. When power increases, the proportions of SiOC3and SiO2C2  decrease in favour of the more highly oxidized

    SiO3C and SiO4 species. This is indicative of the role of the

    Fig. 5. (a) C/Si and O/Si trend as function of the power from O2/HDMSO=1.4 plasma deposits. (b) Curve fit of the Si2p peak of a deposit formed from O2/ 

    HDMSO=1.4 plasma at 100 mTorr 50 W (a) and 100 W (b). (c) Functional composition of the deposited films determined by fitting of Si2p peak plotted

    against the glow discharge power.

     E. Vassallo et al. / Surface & Coatings Technology 200 (2006) 3035–30403038

  • 8/19/2019 Preparation of Plasma-polymerized SiOx-like Thin Films From a Mixture

    5/6

    discharge power in the process. An increase of power results

    in the formation of CO x  gaseous species and the formation

    of silicon–oxygen moieties in the deposit.

    3.2. Corrosion properties of iron with SiO x-deposited films

    The potentiodynamic polarization behaviour of the

    coated and uncoated iron af ter 1   h exposure in a 1 M

     Na2SO4  solution is shown in Fig. 6.  The corrosion current 

    measured of the uncoated iron is 2.0102 AA.

    In the potential range examined, the polarization curve

    generally has three distinct regions: the active dissolution

    region (Tafel region), the active–passive transition region

    and the limiting current region. Especially, a linear  E /log   i

    relation with a Tafel slope of approximately 60 mV/decade

    was obtained. This slope value indicates that the dissolution

     process is determined by the rate of diffusion of soluble

    species for the electrode surface into the bulk solution.

    Fig. 7 shows the variation of the corrosion properties of an iron disk covered with a 1150 nm thick SiO x-like films

    deposited by plasma with different O2/HMDSO ratios, after 

    1 h of immersion in 1 M Na2SO4 solution. The process was

     performed at 100 W of input power after hydrogen plasma

     pre-treatment.

    In comparison with the bare metal, a significant decreaseof the current of corrosion is observed, especially for film

    deposited by an oxygen-rich plasma. In these conditions,

    inorganic SiO x-like coatings are obtained, as infrared

    spectroscopy reveals. The best approach to obtain inorganic

    deposits from organosilicon precursors is to increase   the

    oxygen content of the feed gas and the power ( Fig. 8).

    4. Conclusions

    XPS, IR and electrochemical methods have been used to

    characterize the plasma deposition of SiO x   films for iron

    coatings with corrosive resistance capabilities. Increasing of 

     power or oxygen flow has an appreciable effect on the film

    composition and on its properties. In particular, the

    increasing of power in the glow discharge results in the

    formation of CO x   gaseous species and of higher oxidised

    silicon species in the film. On the other hand, a major 

    oxygen content in plasma phase produces more inorganic,

    compact and cross-linked films, with good anticorrosive

     properties. Nevertheless, our results have demonstrated that 

    this is true only when the O2/HMDSO ratio is lower than

    13.3, while for values greater than 13.3, when power 

    increases the current corrosion increases. Probably in these

    conditions the coating-metal adhesion is worse, the coatingis stressed and the film defectiveness increases. As a matter 

    of fact, samples coated with SiO x  films obtained in plasma

    fed with 5 O2/HMDSO ratio at 200 W show a significant 

    improvement of corrosion resistance.

    In conclusion, our corrosion measurements, in 1 M

     Na2SO4  with pH 5.3, showed that PECVD appears to be a

     promising technique in enhancing the corrosion properties of 

    materials. Oxygen and power play an important role in

    determining the electrochemical behaviour of deposited film.

    However, other experiments at pH 10 and pH 2 revealed that 

    SiO x   thin films, obtained by means of the PECVD

    technology, exhibit also considerable corrosion protection

    Fig. 6. Potentiodynamic curve of uncoated iron.

    Fig. 7. Variation of the corrosion behaviour of the iron disks covered with

    SiO x-like films.

    Fig. 8. Effect of the RF power on the corrosion properties of iron coated

    with SiO x   films.

     E. Vassallo et al. / Surface & Coatings Technology 200 (2006) 3035–3040   3039

  • 8/19/2019 Preparation of Plasma-polymerized SiOx-like Thin Films From a Mixture

    6/6

    in mild basic and acid media [29]. To make an in-depth study

    of the effect of SiO x   films on corrosion behaviour,

    impedance measurements for the coated and uncoated iron

    are in progress. Moreover, electrochemical measurements

    will be supported by scanning electron microscopy (SEM),

    FTIR and XPS before and after corrosion test.

    Acknowledgements

    The authors of this work would like to thank Prof. C.

    Malitesta for XPS measurements and for technical assistance.

    References

    [1] Z.L. Long, Y.C. Zhou, L. Xiao, Appl. Surf. Sci. 218 (2003) 123.

    [2] What Alternatives to Hexavalent Chromium? December 4, 2003, Jolly

    Hotel, Bologna.

    [3] P. Favia, R. d’Agostino, Surf. Coat. Technol. 98 (1–3) (1998January).

    [4] Annemie Bogaerts, Erik Neyts, Renaat Gijbels, Joost van der Mullen,

    Spectrochim. Acta, Part B: Atom. Spectrosc. 57 (2002) 609.

    [5] D.L. Cho, H. Yasuda, J. Appl. Polym. Sci., Appl. Polym. Symp.

    (1988) 233.

    [6] H.P. Schreiber, M.R. Wertheimer, A.M. Wrobel, Thin Solid Films 72

    (1980) 487.

    [7] K.D. Conners, W.J. van Ooij, S.J. Clarson, A. Sabata, J. Appl. Polym.

    Sci., Appl. Polym. Symp. 54 (1994) 167.

    [8] E. Sacher, J.E. Klemberg-Sapieha, H.P. Schreiber, M.R. Wertheimer,

    J. Appl. Polym. Sci., Appl. Polym. Symp. 38 (1984) 163.

    [9] W.J. van Ooij, A. Sabata, I.H. Loh, Pro. Eng. Symp. (1994) 253.

    [10] W.J. van Ooij, N. Tang, Polym. Mater. Sci. Eng. (1996) 155.

    [11] C. Vautrin-Ul, C. Boisse-Laporte, N. Benissad, A. Chausse, P.

    Leprince, R. Messina, Prog. Org. Coat. 38 (2000) 9.

    [12] E. Angelini, S. Grassini, F. Rosalbino, F. Fracassi, R. d’Agostino,

    Prog. Org. Coat. 46 (2003) 107.

    [13] J.T. Felts, in: Society of Vacuum Coaters, Proc. 36th Ann. Tech.

    Conf., vol. 324, 1993.

    [14] G. Czeremuszkin, M.R. Wertheimer, M. Latrèche, A.S. Da Silva

    Sorbrinho, Plasmas Polym. 6 (2001) 107.

    [15] A.S. Da Silva Sorbrinho, G. Czeremuszkin, M. Latrèche, M.R.

    Wertheimer, in: W.W. Lee, R. D’Agostino, R. Wertheimer (Eds.),Material Symposium Society Proceeding, Plasma Deposition and

    Treatment of Polymers, vol. 544, Material Research Society Publisher,

    Werrendale, PA, 1999.

    [16] A. Roulin, B. Jaccoud, Patent no. WO 94/29103 (1994).

    [17] J.W. Coburn, M. Chen, J. Appl. Phys. 51 (3134) (1980).

    [18] R. D’Agostino, F. Cramarossa, S. de Benedictis, G. Ferraro, J. Appl.

    Phys. 52 (1259) (1981).

    [19] N. Selamoglu, J.A. Mucha, D.A. Ibbotson, D.L. Flamm, J. Vac. Sci.

    Technol., B 7 (1345) (1989).

    [20] C. Bourreau, Y. Catherine, P. Garcia, Plasma Chem. Plasma Process.

    10 (247) (1990).

    [21] A. Goullet, C. Vallè, A. Granier, G. Turban, Plasma Sources Sci.

    Technol., A 18 (2452) (2000).

    [22] K. Aumaille, C. Valleè, A. Granier, A. Goullet, F. Gaboriau, G.

    Turban, Thin Solid Films 359 (188) (2000).[23] C. Valleè, A. Granir, A. Goullet, K. Aumaille, C. Carinau, N. Coulon,

    G. Turban, Appl. Surf. Sci. 57 (138–139) (1999).

    [24] R. Lamendola, R. d’Agostino, F. Fracassi, Plasmas Polym. 2 (147)

    (1997).

    [25] L. Agres, Y. Segui, R. Delsol, P. Raynaud, J. Appl. Polym. Sci. 61

    (2015) (1998).

    [26] W.A. Pliskin, J. Vac. Sci. Technol. 14 (1064) (1977).

    [27] C. Valleè, A. Goullet, A. Granire, A. van der Lee, C.J. Durand, J.

     Non-Cryst. Solids 272 (163) (2000).

    [28] G. Beamson, D. Briggs, High Resolution XPS of Organic Polymers— 

    The ESCA 300 Database, vol. 72, Wiley, 1992.

    [29] G. Schmitt, J.-W. Schltze, F. Fah, H . Lqth, Electrochim. Acta 44

    (1999) 3865.

     E. Vassallo et al. / Surface & Coatings Technology 200 (2006) 3035–30403040