Abstract
To meet various application requirements, it is important to enable an improvement of a-C structure and properties, such as hardness, adhesion, and wear resistance. In this study, we used the Raman spectroscopy to investigate the a-C thin films structure dependence on the different deposition parameters. The effect of nitrogen, argon, and hydrogen gas flow rate was analyzed to determine the influence on the film properties. The change in the gas type, combination, and flow had a significant influence on the D and G bands of the a-C Raman spectra. The addition of N2 into the chamber promoted the sp2 creation, while with adding hydrogen the layer contained more sp3 bonds. The depositions of a-C thin films were carried out in pulsed arc discharge vacuum installation. Micro-Raman measurements of the deposited materials were performed using an ISA Dilor-Jobin Yvon-Spex Labram confocal system with 632.8 nm radiation from a He-Ne laser using a back-scattering geometry.
1. Introduction
In recent years, a lot of work has been focused on the synthesis of novel carbon thin films materials, especially crystalline diamond, amorphous carbon (a-C), and carbon nitride (a-C:N) which offers excellent tribological, optical, electrical, and other properties with the additional advantage that carbon is a biocompatible material. The huge range of properties achievable in carbon coatings is mainly due to the ability of carbon to form different types of interatomic bonds, to take up different sites, and to adopt different structures. However, for each application, there are different requirements on the film properties, for example, the adhesion level achievable and coating cost. Therefore, several methods to influence the a-C films structure, composition, and thus the properties, such as variation of ion energy, plasma treatment, pressure, and the use of different gases were presented [1–6].
Raman spectroscopy can be applied as a simple and accurate identification method of different carbon phases within the film. Due to its sensitivity to variation of translation symmetry, Raman spectroscopy allows distinguishing several types of carbon such as diamond, graphite, diamond-like carbon, and carbon nitride. The Raman spectra of nanocrystalline and amorphous carbon are dominated by the D (D for disorder) and G (G for graphite) peaks with varied intensity, position, and band width [7–11]. The G band at approximately 1540–1600 cm−1 corresponds to the symmetric E2g mode in graphite-like materials, while the D band at 1350 cm−1 arises from the limitations in the graphite domain size, induced by grain boundaries or imperfections, such as substitutional N atoms, sp3 carbon, or impurities. The D peak is not present in perfect single crystal graphite and becomes active only in the presence of disorder. For visible excitation, the G and D peaks are due to sp2 sites only. The sp2 sites have such a high cross-section that they dominate the spectra, the sp3 sites are invisible, and the spectrum responds only to the configuration or order of the sp2 sites. The Raman spectra of a-C depend on (1) clustering of the sp2 phase; (2) bond length and bond angle disorder; (3) presence of sp2 rings or chains; (4) the sp2/sp3 ratio. The spectra directly depend on the quality or configuration of the sp2 phase and only indirectly on the quantity of the sp2 phase. Most times the sp2 configuration varies consistently with the sp2 fraction. However, in some cases, the sp2 quality can be changed independently from the sp2/sp3 ratio [12–17]. A weak peak at 2200 cm−1 is attributed to triple C–N bonds and appears in the case of high nitrogen doping. The second-order silicon peak which appeared at 960 cm−1 can be used to measure the transparency of the film. The ratio increases with the increase of N content in the film, revealing a reduction in the optical band gap when more N atoms are incorporated in films [18]. In this study, we used the Raman spectroscopy to investigate a-C thin films structure and its dependence on the gas type and flow.
2. Experimental
The depositions of amorphous carbon thin films were carried out in a vacuum installation type UVNIPA-1-001 described previously [19]. The sputtering frequency of the arc source pulses was 5 Hz. Background pressure was 10−4 Pa and working pressure was maintained around 1 Pa according to gas flow. Argon and nitrogen gas flow was varied from 0 to 100 sccm and the flow of hydrogen from 20 to 60 sccm. The deposition temperature was kept below 100°C. Mirror polished silicon substrates were used for deposition of the coatings. Before deposition of the layer, substrates were cleaned for 10 min with Ar ions within one vacuum cycle. The effects of nitrogen, argon, and hydrogen gas flow on the film properties were investigated by Raman spectroscopy. Micro-Raman measurements of the deposited materials were performed using an ISA Dilor-Jobin Yvon-Spex Labram confocal system with 632.8 nm radiation from a He-Ne laser using a back-scattering geometry. Microscope objective ×80 was used to focus the laser beam onto a spot of approximately 1–5 μm in diameter and to collect the scattered light, which then passed through the spectrometer onto a CCD detector. The acquired Raman spectra were fitted with a Gaussian line to illustrate the G peak position and ratio. The XPS investigations were performed in an Omicron UHV system equipped with an EA 125 hemispherical electron analyzer using monochromated AlK radiation (excitation energy eV). The samples have been annealed by radiative heating from the backside at 500°C for 10 min in UHV conditions. The calibration of the energy scale was made by comparison to reference measurements on a polycrystalline silver sample and HOPG.
3. Results and Discussion
Raman spectra of the films deposited with various argon or nitrogen flow during the deposition process are shown in Figure 1. We can see that, with changing both Ar and N2 flow the intensity maximum of all three depicted bands (Si, D, G band) changes significantly. The maximum intensity in Si band falls with increasing the process gas flow from 0 to 100 sccm. This indicates lowered transparency of the films due to the growth of a thicker layer and also due to more graphitic character of the films deposited with higher Ar or N2 flow [18]. Figure 2 shows the ratio and G peak position dependence of the films from Figure 1. For the layers deposited in nitrogen atmosphere both / and G peak position increases as the amount of N2 in the chamber during deposition rises. The ratio is 0.25 for pure a-C film and 1.35 for film deposited with 100 sccm of N2 flow. The G peak position rises from 1518 cm−1 for a-C to 1574 cm−1 for a-C:N with 100 sccm of nitrogen flow. The simultaneous growth of both ratio and G peak position with increasing of nitrogen flow is a reliable indicator of ratio decrease and corresponds to another research [20]. From the spectra, we can assume that, nitrogen creates rather sp2 than sp3 bonds and the films become more graphitic when the N2 flow rises, which is similar to other studies [11, 14, 18].
(a)
(b)
The ratio and G peak position of the films deposited with various Ar flow shows slight grow from 0 to 30 sccm, drop around 60 sccm and sharp increase for films deposited in atmosphere with 100 sccm of Ar flow. The films around 60 sccm Ar flow have the lowest ratio and it seems such a moderate Ar flow is optimal for deposition of the films with higher ratio. Further increasing of Ar flow caused more graphitic bonds within the films.
Raman spectra of the a-C:N films deposited with combined Ar/N2 flow are shown in Figure 3. The Ar flow was set constant to the values with best results from previous depositions (40 and 60 sccm) while the N2 flow was changed from 40 to 100 sccm to estimate the influence of nitrogen in the combined gas flow. From ratio and G peak position dependence shown in the Figure 4 it is obvious that, for the films deposited with 40 and 60 sccm argon flows, the maximum of G peak position is around 80 and 60 sccm of nitrogen flow, respectively. The values of ratio and G peak position for all gas combined depositions are higher than for layers deposited with only one gas type. Therefore, we can assume that, the further Ar/N2 admixture into the chamber during deposition process promotes the creation of sp2 bonding in the thin film. Besides the main D and G broad peaks, with increasing the N2 flow we can observe also a small local maximum at approximately 2200 cm−1 attributed to triple C–N bonds. This band is characteristic with very small intensity in the Raman spectra made using visible laser excitation [21, 22].
(a)
(b)
Figure 5 shows the Raman spectra and data abstracted from fitting of the films deposited with constant Ar/N2 flow while adding H2 into the gas mixture. As we can see, the ratio decreased from 1.6 for nonhydrogenated a-C:N to 1.1 due to the presence of hydrogen in the gas mixture. The G peak position is changing only slightly towards higher values. The creation of sp2 phase is due to the hydrogen addition lower. This may be due to hydrogen affecting as an etchant of graphitic bonds during growth of the film causing the higher sp3 content.
(a)
(b)
For comparison, XPS analysis of the a-C:N samples deposited with 30, 50, and 70 sccm of N2 flow was made (Figure 6). The resulting ratio decreased with nitrogen flow increasing. The results correspond to the Raman measurements made on these samples and are in agreement with other studies [13, 15, 18].
(a)
(b)
(c)
4. Conclusions
We deposited continuous, homogeneous, and adhesive a-C, a-C:N and a-C:N:H thin films on Si substrates using pulsed vacuum arc discharge technique. Following the Raman spectra we found that, with adding different gases with various concentrations the film structure and properties change significantly. The addition of N2 and/or Ar caused the simultaneous growth of both ratio and G peak position which is a reliable indicator of ratio decrease and increase of graphitic contents in the films. The films with only a moderate Ar flow had lower ratio and this was found to be optimal for deposition of the films with high ratio. The addition of hydrogen into the gas mixture caused the decrease in ratio and thus we can expect lower amount of graphitic sp2 bonding. We found that, the maximum intensity of Si band falls with increasing of the process gas flow. This indicates lowered transparency of the films due to the growth of a thicker layer and also due to the more graphitic character of films deposited with higher Ar and/or N2 flow.
Acknowledgments
This work was done in Center of Excellence CENAMOST (Slovak Research and Development Agency Contract no. VVCE-0049-07) and was financially supported also by Grants APVV-0548-07, LPP-0149-09, DAAD 507 552 00, and VEGA 1/1103/11, 1/1102/11, 1/0264/11.