Gaseous phase monitoring during reactive sputtering of yttrium in glow discharges

Abstract

543.42The emission and fluorescence spectra at wavelengths of 580-650 nm during sputtering of yttrium, zirconium, and their alloys are studied for the purpose of developing spectroscopic methods for monitoring the composition of the gaseous phase during cathode sputtering of these metals in glow discharges in oxygen-argon atmospheres. The 584.2, 585.9, 587.4; 613.2, 614.8, 616.5, 617.9; 647.1, 648.7, and 650.3 nm lines are identified as molecular lines of yttrium oxide. The corresponding electronic transition energies are calculated. A correlation is found between the behavior of the emission and fluorescence spectra of the gaseous phase within this wavelength range during sputtering of yttrium in glow discharges and changes in the concentration of the oxygen-argon mixture. The dependence of the fluorescence and emission line intensities of the yttrium oxide molecule on the partial pressure of oxygen is studied. Introduction.One of the important problems in microelectronics technology is the preparation of high purity films with various compositions and for different purposes by sputtering a target-sample (magnetron or cathode sputtering in glow discharges) with subsequent deposition of sputtered gas-phase molecules on a substrate. Sputtering can take place in a vacuum or in a reactive (e.g., oxygen) atmosphere. The properties of the resulting films depend on their stoichiometric composition and structural perfection, which, in turn, is determined by the composition of the gaseous phase and the sputtering conditions. Thus, films of intermediate vanadium oxide (VO 2 ) with thermochromism and a high temperature coefficient of electrical resistivity undergo a change in resistance by a factor of three when the concentration of oxygen in argon is changed by 1% [1]. The dielectric properties and surface morphology of laser deposited epitaxial Ba(Zr 0.3 Ti 0.7 )O 3 films depend substantially on the partial pressure of oxygen [2]. For the same reason, the phase composition and crystallite dimensions of zirconium oxide films change during film deposition [3]. Thus, finding optimum discharge conditions for creating a constant gaseous phase composition during sputtering-deposition processes means that the task of monitoring of this composition is extremely important.Various methods, direct and indirect, can be used for monitoring film deposition processes. The existing methods for direct in situ monitoring of the composition of sputtered films including x-ray structure analysis [4], x-ray photoelectron spectroscopy, Raman scattering [5], ellipsometry, pyrometry, etc.[6], require quite complicated and costly equipment. Therefore, in most cases indirect monitoring techniques are used, including the following: measuring the intensities of metal and oxygen emission lines [7], monitoring the cathode potential [8,9], measuring the IR optical absorption [6], measuring the partial pressure of oxygen in the vessel [1], and stabilizing the temperature [3] in correlation with the film quality. These monitoring techniques, however, do ...

Laser-induced fluorescence monitoring of the gas phase in a glow discharge during reactive sputtering of vanadium

“…The advantages of LFS compared with related spectroscopic methods include not only high sensitivity and selectivity but also an insensitivity of the results to the excitation conditions of atoms and molecules in the analytical zone. The capabilities of spectroscopic methods were studied earlier [13]. They were shown to be promising for monitoring the gas-phase composition during cathode diode sputtering of yttrium.…”

“…The experimental set up included a magnetron-type glow discharge atomizer, a pumping system and fine gas bleed, and laser fluorescence and emission spectrometers. The atomizer described before [13,14] equipped additionally with a strong transverse magnetic-field source was used for sample sputtering. The plasma during magnetron sputtering was concentrated near the cathode×target surface owing to the magnetic system placed on the opposite side of the cathode.…”

“…A two-stage gas bleed system was used to feed the reactive gas into the atomizer. The presence of two stages and the stabilization of the gas pressure in the intermediate stage enabled the gas to be bled reproducibly into the system at pressures of 10 -2 -1 Pa. Fluorescence spectra of the glow discharge were measured using an improved atomic-fluorescence spectrometer with laser fluorescence excitation as described before [13,14]. Fluorescence was excited by radiation of an LZhI-504 tuned dye laser with pumping by a CL-7020 pulsed excimer laser (2 kW power, 20 ns pulse length, 0.05 nm emission spectral line width, 420-700 nm wavelength tuning without frequency doubling).…”

“…It was turned off during the fluorescence measurement (~1 ms) in order to decrease the discharge emission background. Emission spectra were measured using an atomic-emission spectrometer based on the solid-state CCD detector described in the literature [13]. The optical spectrometer was built according to the PaschenRunge scheme and included a convex diffraction grating (1800 lines/mm, focal distance 150 mm) and four CCD-lines (2048 light-sensitive pixels 14 μm wide) positioned around a Rowland circle.…”