Growth Mechanisms Involved in the Synthesis of Smooth and

ARTICLE pubs.acs.org/Langmuir

Growth Mechanisms Involved in the Synthesis of Smooth and Microtextured Films by Acetylene Magnetron Discharges Val
erie De Vriendt,† Sasa M. Miladinovic,‡ Julien L. Colaux,† Fabrizio Maseri,§ Charles L. Wilkins,‡ and St
ephane Lucas*,† †

Research Center for the Physics of Matter and Radiation (PMR), University of Namur (FUNDP), 61 rue de Bruxelles, B-5000 Namur, Belgium ‡ Chemistry and Biochemistry Department, University of Arkansas, Fayetteville, Arkansas 72701, United States § ArcelorMittal Liege (AMLR), Research and Development, Bld de Colonster, B-4100 Liege, Belgium

bS Supporting Information ABSTRACT: The growth of hydrogenated amorphous carbons (a-C:H) produced by continuous or pulsed discharges of acetylene (C2H2) in an unbalanced magnetron setup was investigated. At 5  10
3 Torr, only smooth films are obtained, whereas at 5  10
1 Torr using a pulsed discharge some microtextured films are formed if the duty cycle is low. The morphology of these microtextured films consists of nanoparticles, filamentary particles, and particular agglomerates (“microflowers”). This paper presents a study of acetylene gas phase polymerization by mass spectrometry, and a detailed analysis of bulk structure of films by combining three techniques which include IR spectroscopy, Raman spectroscopy, and laser desorption/ionization Fourier transform mass spectrometry (LDI-FTMS). Finally, based on the study of gas phase and film structure, we propose a model for the growth of both smooth and microtextured films.

’ INTRODUCTION Despite the intensive investigations performed during the past decade, the scientific interest in amorphous carbon (a-C:H) thin films is still growing. The main reason for this is the possibility of their application in different fields of modern technology due to a wide variety of promising properties. a-C:H films are chemically inert to practically any solvent, acid or base, and thus, they can be used as corrosion-resistant coatings.1 The main application of the hard version of these films is their use as protective coatings, for example, on magnetic hard discs.2 Due to the optical transparancy of a-C:H films in the IR region (apart from the absorbing C
H bands), they are also used as protective coatings on barcode scanners and sunglasses.3 Furthermore as a-C:H films are carbon materials and have a low friction coefficient, they can be used as biocompatible coatings.4 Finally, polymeric or soft a-C:H films seem to be good candidates for active materials in photoluminescence devices.5 The most popular techniques for a-C:H films deposition include ion beam deposition,6 radio frequency plasma enhanced chemical vapor deposition (PECVD),7 and sputtering.8 The deposition technique used in this paper is a conventional unbalanced magnetron sputtering system, which is the most common industrial method. However, the new aspect of this work is the use of a carbon target and only one gas precursor (acetylene, C2H2) in the absence of argon. Ions and radicals formed in the gas phase lead to the growth of either smooth films or microtextured ones depending upon the deposition conditions. The morphology of microtextured films consists of nanoparticles, r 2011 American Chemical Society

filamentary particles, and particular agglomerates (“microflowers”).9 To the best of our knowledge, it is the first time that such carbon based microstructures have been observed. In the microelectronics industry, the presence of nanoparticles in thin films is a problem, whereas in the catalytic and electromechanical industry nanoparticles play a positive role due to their large surface to volume ratio. Moreover, in the material science field, nanoparticles and agglomerates can be used to obtain a rough surface that increases the hydrophobic properties of films.10 Thus, for some applications, the presence of nanoparticles in thin films is required, whereas for other applications it is a big problem. In those cases, smooth films are preferred. In order to obtain the right film for the right application, it is of fundamental importance to understand the growth mode of obtained films (smooth or microtextured). Accordingly, this requirement inevitably leads to the need for their characterization. The main aim of most studies concerning a-C:H films characterization is the determination of sp3 content.11
15 This work goes further, while trying to define as well as possible the structure of smooth and microtextured films. It is a great challenge, because, on one hand, carbon based microstructures in the shape of flowers are unknown in the literature and, on the other hand, the smooth film structure can be extremely complex due to superposition Received: February 19, 2011 Revised: May 30, 2011 Published: June 21, 2011 8913

dx.doi.org/10.1021/la2003035 | Langmuir 2011, 27, 8913–8922

Langmuir

ARTICLE

Figure 1. Schematic representation of the various plasma excitation modes used in this work: continuous (DC), unipolar pulsed (UP), and bipolar pulsed (BP).

of manifold bonds, ambiguous interlinkings, intrinsic disorder, and the presence of defect states. In order to obtain some extremely valuable information on the bulk film structure, the literature mentions several techniques such as Raman spectroscopy,12 IR spectroscopy,16 solid-state nuclear magnetic resonance spectroscopy (SSNMR),17 and highresolution transmission electron microscopy (HRTEM) in combination with electron energy loss spectroscopy (EELS).18 However, this latter technique is not always easily available. Concerning SSNMR, a major limitation is the number of samples that can be analyzed in a given period. For one sample, analysis time can be as much as a few days.17 This paper demonstrates that the bulk structure of film can be determined by the use of three techniques including IR spectroscopy, Raman spectroscopy, and laser desorption/ionization Fourier transform mass spectrometry (LDI-FTMS). This latter technique offers advantages because there are few constraints on sample preparation and because high resolution mass spectra are obtained very quickly. Compared to our previous works,9,19 the significant advances of this paper concern, on one hand, the interpretation of LDIFTMS results in terms of film formation mechanisms and, on the other hand, the suggestion of a qualitative model for the growth of smooth and microtextured films obtained by pulsed magnetron discharge in pure acetylene. First, we will describe the deposition and gas phase characterization techniques, followed by a decription of the spectroscopy methods used in this work. Second, we propose a model based on the observations described here but also partly published by our group elsewhere.9,19 While three excitation modes are used in the present study (DC, unipolar, and bipolar), the data were systematically regrouped by pairs: data for smooth and microtextured films, data for low and high duty cycles, and data for low and high pressure conditions.

’ EXPERIMENTAL SECTION Film Deposition and Gas Phase Characterization. Plasma polymerization experiments were carried out in a vacuum chamber (0.05 m3) equipped with an unbalanced magnetron sputtering system

(self-bias estimated to
30 V) working in DC or pulsed DC mode. This chamber was pumped down with a turbomolecular pump, allowing a base pressure of 10
6 Torr. A carbon target of 50 mm diameter was used. The distance from target to sample holder was 150 mm, and, due to the unbalanced magnetron configuration, the plasma extended to the sample surface. The substrate holder was held at ground potential during deposition. When DC power was continuously applied, the substrate temperature measured with a thermocouple never exceeded 105
C. However, in pulsed plasma mode, the maximum substrate temperature was reduced to 78
C. Films were grown on silicon (100) substrates, cleaned with methanol, and dried with a pure inert gas. For infrared (IR) analyses, films were grown on copper substrates etched in situ with 2 keV Ar+ ion bombardment. The acetylene gas (Air Liquide, purity: 99.995%) was fed into the chamber near the sample, at a fixed mass flow rate of 45 and 375 sccm. As mentioned above, no argon was used in these experiments. The working pressure during film deposition was fixed at 5  10
3 or 5  10
1 Torr, measured with a Baratron pressure gauge. The plasma was excited using a combination of power supplies: the output from a DC supply (Maris) connected to the magnetron and set to either 200, 150, 100, or 40 W was modulated at different frequencies by a SPIK 1000 power unit (MELEC GmbH). The pulsing unit was configured in three different modes which were continuous (DC), unipolar (UP), and bipolar (BP) where the discharge voltage was clamped to ground potential in the off time, as schematically shown in Figure 1. Depending on the conditions, the discharge voltage varied from 450 to 480 V. The upper part of Table 1 shows the different experimental conditions used for film synthesis. Various tON and tOFF were used in this work (e.g., fixed tON of 5 μs and tOFF ranging from 10 to 90 μs). Based on our observations, we can summarize the different results in two categories which are low duty cycle (30%). A differentially pumped quadrupole mass spectrometer (Hiden PSM) was connected to vacuum chamber. The aperture was placed at the exact location of the sample. The mass spectrometer was used in residual gas analysis (RGA) mode, in which the neutral flow emanating from the aperture reaches the ionizer of the PSM where the energy of the electrons is set to 70 eV. The ions formed are then transmitted to the quadrupole mass filter (QMF). The QMF identifies various ions according to their mass-to-charge ratio (m/z). The maximum mass measured by the instrument was up to 500 amu (for singly charged ions). 8914

dx.doi.org/10.1021/la2003035 |Langmuir 2011, 27, 8913–8922

Langmuir

ARTICLE

Table 1. Experimental Parameters and Main Observations of the Two Types of Film Discussed in This Work working pressure

10
3 Torr

discharge mode

continuous

duty cycle

10
3 Torr

10
3 Torr

10
3 Torr

10
3 Torr

10
1 Torr

10
1 Torr

UP pulsed

UP pulsed

BP pulsed

BP pulsed

BP pulsed

BP pulsed

high (>30%)

low (30%)

low (30%)

low (