Silicon carbonitride by remote microwave plasma CVD from organosilicon precursor: Growth mechanism and structure of resulting Si:C:N films

Silicon carbonitride by remote microwave plasma CVD from organosilicon precursor: Growth mechanism and structure of resulting Si:C:N films

Applied Surface Science 253 (2007) 7211–7218 www.elsevier.com/locate/apsusc Silicon carbonitride by remote microwave plasma CVD from organosilicon pr...

1MB Sizes 3 Downloads 47 Views

Applied Surface Science 253 (2007) 7211–7218 www.elsevier.com/locate/apsusc

Silicon carbonitride by remote microwave plasma CVD from organosilicon precursor: Growth mechanism and structure of resulting Si:C:N films I. Blaszczyk-Lezak a, A.M. Wrobel a,*, M.P.M. Kivitorma b, I.J. Vayrynen b, A. Tracz a a

Centre of Molecular and Macromolecular Studies, Polish Academy of Sciences, Sienkiewicza 112, PL-90-363 Lodz, Poland b Department of Physics, University of Turku, FIN-20014 Turku, Finland Received 30 January 2007; received in revised form 27 February 2007; accepted 27 February 2007 Available online 3 March 2007

Abstract The remote microwave hydrogen plasma chemical vapor deposition (RP-CVD) from bis(dimethylamino)methylsilane precursor was used for the synthesis of silicon carbonitride (Si:C:N) films. The effect of thermal activation on the RP-CVD process was examined by determining the mass- and the thickness-based film growth rate and film growth yield, at different substrate temperature (TS). It was found that the mechanism of the process depends on TS and for low substrate temperature regime, 30 8C  TS  100 8C, RP-CVD is limited by desorption of film-forming precursors, whereas for high substrate temperature regime, 100 8C < TS  400 8C, RP-CVD is a non-thermally activated and mass-transport limited process. The Si:C:N films were characterized by X-ray photoelectron and Fourier transform infrared spectroscopies, as well as by atomic force microscopy. The increase of TS enhances crosslinking in the film via the formation of nitridic Si–N and carbidic Si–C bonds. On the basis of the structural data a hypothetical crsosslinking reactions contributing to silicon carbonitride network formation have been proposed. # 2007 Elsevier B.V. All rights reserved. Keywords: Remote hydrogen plasma CVD; Bis(dimethylamino)methylsilane precursor; Silicon carbonitride film; Film chemical structure; Film surface morphology

1. Introduction Silicon carbonitride (Si:C:N) films are interesting and relatively unexplored group of silicon-based materials, which are expected to combine the properties of silicon nitride and silicon carbide. The combination of these in one Si:C:N material is very promising because of the observed many unique properties, e.g.: extremely strong resistance to oxidation at high temperatures up to 1600 [1] and a hardness comparable with that of superhard material like cubic boron nitride [2–4]. In view of the literature data [5–25], Si:C:N films can effectively be produced from a number of organosilicon precursors using various chemical vapor deposition (CVD) techniques, such as: thermal CVD from ethylcyclosilazanes [5], Ar ion beam-induced CVD from hexamethyldisilazane [6], direct plasma CVD (DPCVD) from hexamethyldisilazane [7–10], bis(trimethylsilyl)carbodiimide [10,11], and bis(dimethylamino)dimethylsilane [12], as well as remote plasma CVD (RP-CVD) from

* Corresponding author. Tel.: +48 426818952; fax: +48 426847126. E-mail address: [email protected] (A.M. Wrobel). 0169-4332/$ – see front matter # 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.apsusc.2007.02.193

hexamethyldisilazane [13–16], 1-dimethylsilyl-2,2-dimethylhydrazine [17,18], dimethylbis(2,2-dimethylhydrazino)silane [17,18], 1,1,3,3-tetramethyldisilazane [19–21], (dimethylamino)dimethylsilane [22–24], and tris(dimethylamino)silane [25]. Among these methods RP-CVD appeared to be extremely useful technique offering well-controlled deposition conditions. In particular, some damaging effects arising from direct interaction of the plasma and the growing film, namely, charged-particle bombardment or high-energy ultraviolet irradiation [26], which are inherently present in DP-CVD [27], can be avoided. This technique substantially differs from the conventional DP-CVD in two major aspects. The first is that the plasma generation and film deposition takes place in spatially separated regions. Second is that the plasma is induced in a region free of a source compound, unlike DPCVD, using a non-film-forming gas, that is, hydrogen. The electrically neutral hydrogen atoms, effectively generated in the hydrogen plasma, are fed, through the remote section (trap for the electrons, ions, excited atoms, and photons), to the CVD reactor. Thereby, RP-CVD is exclusively induced via chemical reaction between the uncharged ground-state hydrogen atoms and the source compound molecules,

7212

I. Blaszczyk-Lezak et al. / Applied Surface Science 253 (2007) 7211–7218

resulting in the formation only radical active species. Due to a significantly lower concentration of active species contributing to RP-CVD, compared to that of DP-CVD, the growth step in the gas phase is markedly reduced. This prevents the formation of powder particles, which often contaminate the DP-CVD films [28]. In view of these beneficial features RPCVD is very promising for the production of defect-free Si:C:N thin-film coatings [20–23]. In this paper we report the results of the study on the formation of Si:C:N films by RP-CVD using hydrogen as an upstream gas for plasma generation and bis(dimethylamino)methylsilane (BDMAMS) as a novel single-source precursor. BDMAMS, owing to the presence of the hydrosilyl group, SiH, in its molecule (Fig. 1), is assumed to be very reactive in the investigated RPCVD. Moreover, it is a carrier of the C–N and Si–N units with tertiary nitrogen atom, which may readily be incorporated into the film structure. The presence of tertiary nitrogen atom in the BDMAMS molecule is important for the formation of Si:C:N films with good mechanical properties. This is confirmed by the results of our recent study [22,24], which proved that Si:C:N films produced by RP-CVD from a (dimethylamino)dimethylsilane precursor exhibited excellent mechanical properties and they are useful for the modification of surface mechanics of engineering materials for advanced technology. For example, such films may significantly increase hardness and reduce friction of metal surfaces, thus improving their wear resistance. Moreover, the Si:C:N films seem to be a suitable components for the formation of multi-layer coatings together with silicon carbide or silicon nitride films (e.g.: SiCN/SiC/SiCN/SiC. . . or SiCN/SiNx/SiCN/ SiNx. . .) of superior mechanical strength. In the present work we have investigated the mechanism of Si:C:N film growth process in terms of the effect of thermal activation. The chemical structure and surface morphology of Si:C:N films, deposited at various substrate temperatures, have been examined. On the basis of the results of structural study, the chemistry involved in the silicon carbonitride network formation is proposed. The results structural study presented in this report are used for determining the structure–property relationships for the investigated films, which are the subject of the second part of this work. 2. Experimental 2.1. Film deposition procedure Si:C:N films were produced using the RP-CVD apparatus already presented and described in detail elsewhere [20]. The

Fig. 1. Molecular structure of a bis(dimethylamino)methylsilane (BDMAMS) precursor.

apparatus consisted of three major parts: the plasma generation section coupled with a 2.45 GHz microwave power supply unit and fed with hydrogen as an upstream gas; the remote section, equipped with a Wood’s horn photon trap; and the CVD reactor, fed with the source compound. Deposition experiments were performed at total pressure p = 90–95 Pa = 0.68–0.71 Torr; flow rate of upstream gas fed to plasma generation section, F(H2) = 100 sccm; and microwave power input P = 120 W. A BDMAMS precursor was evaporated at 25 8C and fed through a needle valve into the CVD reactor with a flow rate, F(BDMAMS) = 2.9  0.7 mg min 1 = 0.54  0.07 sccm. The distance between the plasma edge and the precursor inlet was 40 cm. No film deposition was observed in the plasma section, indicating that there was no back diffusion of the precursor. Films were deposited on a Fisher microscope cover glass plates (45 mm  50 mm  0.2 mm; for gravimetric measurements of the deposit mass) and on p-type c-Si wafers (3 cm  3 cm  0.2 mm) at the deposition (or substrate) temperature TS = 30–400 8C. As film deposition did not occur without using plasma the contribution of thermally induced CVD process can be excluded. A BDMAMS liquid precursor (b.p. 112 8C), supplied by ABCR, was purified prior to the RP-CVD experiments by distillation in argon atmosphere, following which its purity was tested by gas chromatography. The hydrogen upstream gas was of 99.99% purity. 2.2. Film examination techniques Thickness of the films deposited on c-Si wafers was measured ellipsometrically using a Nippon Infrared Industrial Co. EL-101D ellipsometer equipped with a 632.8 nm He–Ne laser. For each film sample the average thickness value was calculated from at least five ellipsometric measurements. The measured values of film thickness were in the range of 0.4– 1.4 mm. Fourier transform infrared (FT-IR) absorption spectra of the films deposited on c-Si wafers were recorded in a transmission mode on a FT-IR-Infinity ATI Matson spectrophotometer. The resolution of the spectral envelopes into individual absorption bands was performed using Gaussian-type functions for curve fitting. X-ray photoelectron spectroscopic (XPS) analyses of the films were performed with a Perkin-Elmer ESCA 5400 system using Al Ka X-ray photoexcitation source with an electron takeoff angle of 458 from the surface normal. Prior to the XPS analysis, the surface of the film sample was subjected to mild cleaning by sputter-etching for 30 min with a 1 kV Ar+ beam and the beam current less than 0.5 mA. In such mild sputteretching conditions the effect of this process on the elemental composition of the film was minimized. The flood gun of low energy electrons was applied to neutralize the surface charge of film sample. Surface morphology of the films deposited on c-Si substrate was examined by scanning electron microscopy (SEM) using a JEOL MP 41080 electron microscope and atomic force microscopy (AFM) using a Nanoscope IIIa

I. Blaszczyk-Lezak et al. / Applied Surface Science 253 (2007) 7211–7218

7213

Digital Instruments/Veeco, Santa Barbara, CA, equipped with JV scanner operated in tapping mode. The AFM examinations were performed using V-shaped Si3N4 cantilevers with a tip radius curvature of 50 nm. Height and deflection images were recorded simultaneously. All scans were carried out in air with a scan frequency below 5 Hz. Image analysis was performed using a standard software provided with the instrument. The experimental data points in the presented relationships were fitted with the guidelines plotted using a PC SigmaPlot program. 3. Results and discussion 3.1. Rate and yield of RP-CVD The RP-CVD process has been characterized by determining the mass- and the thickness-based film growth rate (rm and rd, respectively) and film growth yield (km and kd, respectively). The latter parameter is defined as km = rm/F and kd = rd/F, respectively, where F denotes the mass-based precursor flow (or feeding) rate. In physical terms, parameter k expresses mass or thickness of the deposit per unit mass of the precursor fed into the reactor. This parameter is very sensitive to the molecular structure of the precursor and may be used for characterizing reactivity of the precursors either in RP-CVD [26,29,30] or DP-CVD [31]. Fig. 2 shows the mass-based (a) and the thickness-based (b) growth rate r and growth yield k of Si:C:N films as a function of substrate temperature. The observed decrease of r and k with increasing TS in low substrate temperature range, 30 8C  TS  100 8C, suggest that RP-CVD is limited by desorption of film-forming precursors, the rate of which exceeds the rate of surface reactions converting them to deposit. No effect of TS on r and k noted for high substrate temperature range, 100 8C < TS  400 8C, proves that the rate and yield of RP-CVD are independent of the temperature and the process is mass-transport limited. It is worth to mention that the substrate temperature dependencies of r and k determined for remote hydrogen plasma CVD from (dimethylamino)dimethylsilane [22,23] revealed the same character as those in Fig 2a and b, respectively. In contrast, the remote hydrogen plasma CVD from various alkylsilane precursors, such as: dimethylsilane [30,32], trimethylsilane [30,32], hexamethyldisilane [30,32– 34], and tertrakis(trimethylsilyl)silane [26,29,33], was found to be a non-thermally activated process for the whole range of investigated substrate temperature (TS = 30–400 8C).

Fig. 2. Mass-based (a) and thickness-based (b) deposition rates, rm and rd (*), respectively, and deposition yields, km and kd (~), respectively, of Si:C:N films as a function of substrate temperature.

As can be noted from these plots, the ratio C/Si decreases markedly with increasing TS, whereas the ratio N/Si slightly rises, and both ratios tend to the value near 0.5 at TS = 400 8C. It is worth to mention, that the values of the N/Si and C/Si ratios

3.2. Compositional parameters The Si:C:N films were characterized by the compositional parameters expressing the atomic concentration ratios N/Si and C/Si, calculated from the XPS data. The values of these ratios remain in close relation with the crosslinking via the formation of the Si–N and Si–C networks, respectively (see Section 3.4). Fig. 3 shows the ratios N/Si and C/Si plotted in a function of TS.

Fig. 3. XPS atomic concentration ratios N/Si (*) and C/Si (&) of Si:C:N films as a function of substrate temperature.

7214

I. Blaszczyk-Lezak et al. / Applied Surface Science 253 (2007) 7211–7218

determined for Si:C:N films (Fig. 3) are much lower than those of the BDMAMS precursor (N/Si = 2 and C/Si = 5). The changes observed for the ratios C/Si and N/Si account for the elimination of organic groups and subsequent formation of the Si–C and Si–N crosslinks. 3.3. Structure The structure of Si:C:N films was examined using FT-IR spectroscopy. Fig. 4 shows the FT-IR spectra of a liquid BDMAMS precursor and Si:C:N films deposited at various TS. For clarity, the film spectra for TS = 50, 250, and 350 8C have been omitted in this figure. The particular absorption bands were identified according to the literature data [35,36]. The film spectra reveal the presence of the following absorption bands:  The broad intense band in the wavenumber range of 1300– 600 cm 1 (Fig. 4) consists of overlapping component bands from –Me deformation mode in the SiMe groups (1273– 1261 cm 1), the C–N stretching and/or N–H bending modes in Si–NH–C and/or Si–NH–Si (1198–1156 cm 1), the Si–O stretching mode and/or the –CH2– wagging mode in Si–CH2– Si (1100–1072 cm 1), the Si–N stretching mode (959– 925 cm 1), and the Si–C stretching mode (830–786 cm 1).  The band at 2350 cm 1 (spectrum for TS = 100 8C) from the CBBN stretching mode in Si–CN.  The band in the range of 2960–2790 cm 1 (spectra for TS = 30 and 100 8C) arising from the stretching modes of C– H.  The band at 3400 cm 1 (spectra for TS = 30, 100, and 200 8C) originating from the –NH– stretching mode in Si–NH–Si and/ or Si–NH–C.

Fig. 4. FT-IR transmission spectra of a liquid BDMAMS precursor and Si:C:N films deposited on c-Si wafer at various deposition temperatures TS = 30, 100, 200, 300, and 400 8C.

An interesting feature of the film spectra in Fig. 4 is the absence of the absorption band corresponding to stretching mode of the Si–H units, which is very intense in the spectrum of a BDMAMS precursor at 2102 cm 1. This accounts for the high reactivity of hydrosilyl unit of the precursor with atomic hydrogen, as mentioned earlier. As can be noted from Fig. 4 the increase in substrate temperature involves substantial changes in the film spectra. A significant drop in the intensity of absorption bands from C–H (2960–2790 cm 1) and SiMe (1273–1261 cm 1) is observed. This is due to thermally induced rupture of the C–H and Si–C bonds in the methylsilyl groups. The resolution of IR absorption envelope in the range of 1300–600 cm 1 (Fig. 4) into the component absorption bands has provided the quantitative information about the evolution of film structure. Fig. 5 exemplifies the resolved FT-IR spectrum of 0.4 mm thick film produced at substrate temperature TS = 400 8C, which includes the components bands from the SiMe (1269 cm 1), C–N and/or N–H (1185 cm 1), Si–O and/or Si–CH2–Si (1099 cm 1), Si–N (975 cm 1), and Si–C (833 cm 1) units. The band at 1099 cm 1 cannot be resolved into the Si–O and Si–CH2–Si components due to their extremely strong overlap. Oxygen contamination, revealed by the Si–O band contribution may originate from two potential sources. The first may be etching of the glass walls of the CVD system by highly reactive atomic plasma species, resulting in incorporation of oxygen-containing etch products into the growing film [29,37]. The second source may arise from the reactions of long-lived dangling bonds in the deposit with atmospheric oxygen or moisture, which may occur after film is exposed to the ambient environment. The oxygen content in the examined films, evaluated by XPS, was in the range of 6–16 at.%. It is worth to mention that the presence of the distinct IR band from the Si–O bonds has been detected for Si:C:N films formed by DP-CVD from bis(trimethylsilyl)carbodiimide on the substrate located on the grounded electrode, which was heated at

Fig. 5. Resolved FT-IR spectrum of Si:C:N film deposited at substrate temperature TS = 400 8C.

I. Blaszczyk-Lezak et al. / Applied Surface Science 253 (2007) 7211–7218

7215

proposed previously for RP-CVD from (dimethylamino)dimethylsilane [22]. The marked rise in the intensities of the Si–N and Si–C IR bands observed with increasing TS in Fig. 6 accounts for two competitive crosslinking processes involved in the formation of Si–N and Si–C crosslinks, respectively. The condensation reaction of the dimethylaminosilane groups incorporated from the precursor molecules and placed in the vicinal segments in the film may proceeds with increasing TS (Eq. (1)) [38]: (1)

Fig. 6. Relative integrated intensities of the IR absorption bands from the Si–N (*), Si–C (&), Si–O/Si–CH2–Si (+), C–N/N–H (~), and SiMe (^) units in Si:C:N films, as a function of substrate temperature.

TS = 300 8C [11]. The oxygen content in these films deposited at different RF power input ranged from 14 to 19 at.%, as determined by glow discharge optical emission spectroscopy [11]. The relative integrated intensities of the component Si–N, Si–C, Si–O and/or Si–CH2–Si, C–N, and SiMe bands (determined by the ratio of particular component band area to its envelope area) are shown in Fig. 6 as a function of TS. The rise of TS increases markedly the Si–C and Si–N bands intensities and decreases those of the Si–O/Si–CH2–Si and SiMe bands to low values at TS = 400 8C. The intensity of the C–N/N–H band rises to a maximum value at TS = 250 8C and then decreases with increasing substrate temperature. It should be noted that reasonable correlation exists between the IR and XPS data presented in Figs. 6 and 3, respectively. The increase in the content of the Si–N bonds with rising TS, revealed by the Si–N IR curve (Fig. 6), well agrees with rising ratio N/Si (Fig. 3), whereas the decay of the content of the SiMe groups, demonstrated by the SiMe curve (Fig. 6), is consistent with decreasing ratio C/Si (Fig. 3). The IR data in Fig. 6 account for thermally enhanced crosslinking, and resulting formation of silicon carbonitride network structure in the films deposited at higher temperatures. 3.4. Silicon carbonitride network formation On the basis of the results of structural study and the literature data referring to the chemistry of alkylaminosilanes [38] and the thermochemistry of a low-pressure pyrolysis of poly(methylsilazanes) [39–43] and poly(methylcarbosilanes) [41–44] we postulate a hypothetical mechanisms of the elementary processes proceeding in the bulk of Si:C:N film, which contribute to silicon carbonitride network formation. The initiation and growth steps involved in the present film formation process are assumed to be essentially similar as those

This reaction is slightly endothermic with a heat of 23 kJ mol 1 in the case of dimethylaminotrimethylsilane [38]. The N-methyl substituted disilazane crosslinks formed in Eq. (1) may undergo transamination reactions resulting in the formation of Si3N units (Eqs. (2) and (3)) [38–40, 43]:

(2)

(3)

These reactions due to their exothermic character [36], proceed spontaneously and contribute to the formation of the Si–N network in the film. The methylsilyl groups may undergo thermally enhanced crosslinking reactions according to the following equations [41–44]: (4)

(5)

Eqs. (4) and (5) account for the elimination of the methylsilyl groups as revealed by the SiMe IR curve in Fig. 6. The reactive hydrogen atoms bonded to secondary and tertiary carbon atoms in the Si2CH2 and Si3CH units, respectively, may contribute to the crosslinking via formation of the C–N bonds as exemplified by the following equations:

(6)

7216

I. Blaszczyk-Lezak et al. / Applied Surface Science 253 (2007) 7211–7218

Fig. 7. SEM images of the surface of Si:C:N films deposited on c-Si wafer at various substrate temperatures: 30 8C (a) and 300 8C (b).

(7)

Fig. 8. 3D and 2D AFM images and cross-sectional surface profiles of Si:C:N film deposited on c-Si substrate at TS = 300 8C. Root mean square value of surface roughness: Rrms = 0.8 nm.

IR intensity curve (Fig. 6). The drop in the content of the Si– CH2–Si units, as inferred by the Si–CH2–Si/Si–O IR intensity curve in Fig. 6, may be due to the reactions expressed by Eqs. (5) and (6). 3.5. Surface morphology

The presented crosslinking reactions eliminate the organic groups and lead to the formation of three dimensional network structure in the Si:C:N film. They are in agreement with thermally induced compositional and structural changes shown in Figs. 3 and 6, respectively. Table 1 illustrates the effect of the discussed crosslinking reactions on the atomic ratios C/Si and N/Si calculated for reacting structural units in the Si:C:N film (excluding volatile by-products). In view of the data listed in Table 1, Eqs. (1), (2), (4), and (5)–(7) cause the marked drop of the atomic ratio C/Si as observed from the XPS plot in Fig. 3. The low values of the XPS ratio N/Si found for Si:C:N films (0.3–0.4, Fig. 3), with respect to that of the BDMAMS precursor (N/Si = 2), are consistent with the low N/Si values resulting from Eqs. (1) and (2) (Table 1). Referring to the structural data, Eqs. (4) and (5) well describe the decay of methylsilyl groups revealed by the SiMe

The results of the SEM study performed for Si:C:N films produced at different TS value (30, 100, 200, 300, and 400 8C) Table 1 Effect of crosslinking reactions on the atomic ratios C/Si and N/Si evaluated for reacting structural units in Si:C:N film Reaction equation

1 2 3 4 5 6 7

C/Si

N/Si

Before reaction

After reaction

Before reaction

After reaction

2 1 0.5 1 0.7 0.5 0.4

0.5 0 0.5 0.5 0.3 0.3 0.2

1 0.7 0.5 – – 0.3 0.2

0.5 0.3 0.5 – – 0.3 0.2

I. Blaszczyk-Lezak et al. / Applied Surface Science 253 (2007) 7211–7218

7217

of the Si–C and Si–N networks in the film. In view of the SEM and AFM studies, the Si:C:N films are morphologically homogeneous materials exhibiting very small surface roughness, which increase in a narrow range of values: 0.3 nm  Rrms  0.8 nm, with rising TS from 30 to 400 8C. This indicates that film surface morphology is only slightly affected by the substrate temperature. Acknowledgements This work has been supported by Polish Ministry of Scientific Research and Information Technology in a frame of the research project No. 3T08C00728. The authors thank Dr. R.P. Socha (Institute of Catalysis and Surface Chemistry PAS) for kind assistance in the XPS analyses and Ms. E. Szkudlarek (MSc) for the SEM examination. Fig. 9. Root mean square surface roughness of Si:C:N film as a function of substrate temperature.

indicate that, on the micrometer scale, the substrate temperature does not influence the surface morphology of the deposit. The film surface, being very smooth and defect-free, exhibited an excellent morphological homogeneity, irrespective of the deposition temperature. This is demonstrated by SEM images in Fig. 7, which exemplify the surface of Si:C:N films deposited at two widely different deposition temperatures: TS = 30 8C (a) and 300 8C (b). Fig. 8 exemplifies 3D and 2D AFM images and crosssectional surface profile of Si:C:N film deposited at TS = 300 8C. As can be noted from the AFM quantitative data in Fig. 9 the root mean square (Rrms) of surface roughness vary in a very narrow range of small values: 0.3  Rrms  0.8 nm with increasing TS from 30 to 400 8C. These results prove that the surface morphology of investigated Si:C:N films is only slightly affected by substrate temperature. For comparison with the present AFM results, Si:N:C films deposited by electron cyclotron resonance plasma CVD from SiH4–CH4–N2 mixture [45] and by rf sputter deposition using N2–Ar mixture as sputtering gas and SiC target [46] exhibited much rougher surfaces with the roughness varying in the ranges: Rrms = 3.5–6.3 nm [45] and Rrms = 15–65 nm [46]. 4. Conclusions The temperature dependencies of the mass- and thicknessbased deposition rate and yield show that in low substrate temperature range, 30 8C  TS  100 8C, film growth is limited by desorption of film-forming precursors, whereas in high substrate temperature range, 100 8C < TS  400 8C, film growth is independent of the temperature and RP-CVD is masstransport limited process. The substrate temperature was found to influence strongly the chemical composition and structure of the Si:C:N films. The XPS compositional and IR structural data show that the increase in TS leads to the elimination of the organic groups and subsequent crosslinking via the formation

References [1] R. Riedel, H. Kleebe, H. Schoenfelder, F. Aldinger, Nature 374 (1995) 526. [2] A. Badzian, T. Badzian, W.D. Drawl, R. Roy, Diamond Relat. Mater. 7 (1998) 1519. [3] A. Badzian, T. Badzian, R. Roy, W.D. Drawl, Thin Solid Films 354 (1999) 148. [4] A. Badzian, J. Am. Ceram. Soc. 85 (2002) 16. [5] Y.B. Bae, H. Du, B. Gallois, K.E. Gonsalves, B.J. Wilkens, Chem. Mater. 4 (1992) 478. [6] T. Matsutani, T. Asanuma, C. Liu, M. Kiuchi, T. Takeuchi, Surf. Coat. Technol. 169–170 (2003) 624. [7] R. Heyner, G. Marx, Thin Solid Films 258 (1995) 14. [8] G. Marx, K.U. Koerner, P. Hager, Steel Res. 72 (2001) 518. [9] D.H. Kuo, D.G. Yang, Thin Solid Films 374 (2000) 92. [10] D. Probst, H. Hoche, Y. Zhou, R. Hauser, T. Stelzner, H. Scheerer, E. Broszeit, C. Berger, R. Riedel, H. Stafast, E. Kroke, Surf. Coat. Technol. 200 (2005) 355. [11] Y. Zhou, D. Probst, A. Thissen, E. Kroke, R. Riedel, R. Hauser, H. Hoche, E. Broszeit, P. Kroll, H. Stafast, J. Eur. Ceram. Soc. 26 (2006) 1325. [12] R. Di Mundo, R. d’Agostino, F. Fracassi, F. Palumbo, Plasma Process. Polym. 2 (2005) 612. [13] N.I. Fainer, Y.M. Rumyantsev, M.L. Kosinova, G.S. Yurjev, E.A. Maximovski, F.A. Kuznetsov, Appl. Surf. Sci. 114 (1997) 614. [14] N.I. Fainer, M.L. Kosinova, Y.M. Rumyantsev, F.A. Kuznetsov, J. Phys. IV 9 (1999) 769. [15] M.L. Kosinova, N.I. Fainer, Y.M. Rumyantsev, M. Terauchi, K. Shibata, F. Satoh, F.A. Kuznetsov, J. Phys. IV 11 (2001) 987. [16] N.I. Fainer, Y.M. Rumyantsev, A.N. Golubenko, M.L. Kosinova, F.A. Kuznetsov, J. Cryst. Growth 248 (2003) 175. [17] T.P. Smirnova, A.M. Badalyan, V.O. Borisov, L.V. Yakovkina, V.V. Kaichev, A.N. Shmakov, A.V. Nartova, V.I. Rakhlin, A.N. Fomina, J. Struct. Chem. 44 (2003) 169. [18] T.P. Smirnova, A.M. Badalian, L.V. Yakovkina, V.V. Kaichev, V.I. Bukhtiyarov, A.N. Shmakov, I.P. Asanov, V.I. Rachlin, A.N. Fomina, Thin Solid Films 429 (2003) 144. [19] A.M. Wrobel, A. Walkiewicz-Pietrzykowska, M. Stasiak, J.E. KlembergSapieha, D.M. Bielin´ski, T. Aoki, Y. Hatanaka, J. Wide Bandgap Mater. 8 (2000) 3. [20] A.M. Wrobel, I. Blaszczyk, A. Walkiewicz-Pietrzykowska, A. Tracz, J.E. Klemberg-Sapieha, T. Aoki, Y. Hatanaka, J. Mater. Chem. 13 (2003) 731. [21] A.M. Wrobel, I. Blaszczyk-Lezak, A. Walkiewicz-Pietrzykowska, D.M. Bielin´ski, T. Aoki, Y. Hatanaka, J. Electrochem. Soc. 151 (2004) C723. [22] I. Blaszczyk-Lezak, A.M. Wrobel, M.P.M. Kivitorma, I.J. Vayrynen, Chem. Vap. Depos. 11 (2005) 44.

7218

I. Blaszczyk-Lezak et al. / Applied Surface Science 253 (2007) 7211–7218

[23] I. Blaszczyk-Lezak, A.M. Wrobel, M.P.M. Kivitorma, I.J. Vayrynen, T. Aoki, Diamond Relat. Mater. 15 (2006) 1484. [24] I. Blaszczyk-Lezak, A.M. Wrobel, D.M. Bielinski, Diamond Relat. Mater. 15 (2006) 1650. [25] T. Aoki, T. Ogishima, A.M. Wrobel, Y. Nakanishi, Y. Hatanaka, Vacuum 51 (1998) 747. [26] A.M. Wrobel, S. Wickramanayaka, Y. Hatanaka, J. Appl. Phys. 76 (1994) 558. [27] A.M. Wrobel, G. Czeremuszkin, Thin Solid Films 216 (1992) 203. [28] A.M. Wro´bel, M.R. Wertheimer, Plasma polymerized organosilicons and organometallics, in: R. d’Agostino (Ed.), Plasma Deposition, Treatment, and Etching of Polymers, Academic Press, Boston, MA, 1990(Chapter 31). [29] A.M. Wrobel, S. Wickramanayaka, Y. Nakanishi, Y. Fukuda, Y. Hatanaka, Chem. Mater. 7 (1995) 1403. [30] A.M. Wrobel, A. Walkiewicz-Pietrzykowska, Chem. Vap. Depos. 4 (1998) 133. [31] A.M. Wrobel, W. Stanczyk, Chem. Mater. 6 (1994) 1766. [32] A.M. Wrobel, A. Walkiewicz-Pietrzykowska, M. Stasiak, T. Aoki, Y. Hatanaka, J. Szumilewicz, J. Electrochem. Soc. 145 (1998) 1060. [33] A.M. Wrobel, S. Wickramanayaka, Y. Nakanishi, Y. Hatanaka, S. Pawlowski, W. Olejniczak, Diamond Relat. Mater. 6 (1997) 1081. [34] A.M. Wrobel, A. Walkiewicz-Pietrzykowska, J.E. Klemberg-Sapieha, Y. Hatanaka, T. Aoki, Y. Nakanishi, J. Appl. Polym. Sci. 86 (2002) 1445.

[35] D.R. Anderson, Infrared, Raman, and ultraviolet spectroscopy, in: A.L. Smith (Ed.), Analysis of Silicones, Wiley/Interscience, New York, 1974 (Chapter 10). [36] G. Sokrates, Infrared Characteristics Group Frequencies, Wiley/Interscience, Chichester, 1994 (Chapter 18). [37] A.M. Wrobel, A. Walkiewicz-Pietrzykowska, D.M. Bielinski, J.E. Klemberg-Sapieha, Y. Nakanishi, T. Aoki, Y. Hatanaka, Chem. Mater. 15 (2003) 1757. [38] R. Walsh, in: S. Patai, Z. Rappoport (Eds.), The Chemistry of Organic Silicon Compounds, Wiley, New York, 1989 , pp. 382–383 (Chapter 5). [39] N.S. Choong Kwet Yive, R.J.P. Corriu, D. Leclercq, P.H. Mutin, A. Vioux, Chem. Mater. 4 (1992) 141. [40] N.S. Choong Kwet Yive, R.J.P. Corriu, D. Leclercq, P.H. Mutin, A. Vioux, Chem. Mater. 4 (1992) 1263. [41] S. Yajima, Y. Hasegawa, J. Hayashi, M. Imura, J. Mater. Sci. 13 (1978) 2569. [42] Y. Hasegawa, K. Okamura, J. Mater. Sci. 18 (1983) 3633. [43] M. Birot, J.P. Pillot, J. Dunogues, Chem. Rev. 95 (1995) 1443. [44] R.M. Laine, F. Babonneau, Chem. Mater. 5 (1993) 260. [45] D.H. Zhang, Y. Gao, J. Wei, Z.Q. Mo, Thin Solid Films 377–378 (2000) 607. [46] K.B. Sundram, Z. Alizadeh, R.M. Todi, V.H. Desai, Mater. Sci. Eng. A 368 (2004) 103.