![]() |
ИСТИНА |
Войти в систему Регистрация |
ИСТИНА ИНХС РАН |
||
Sputtering of materials by energetic ions is one of the key processes used in micro and nano technology. Usually used it is carried out by ions of sufficiently high energies, hundreds of eV and more. For this, both ion beams and directed fluxes of ions from the low-pressure discharge plasma are used. However, high-energy ions are capable to create sufficiently deep defect layers of nm and even tens of nm. Modern nanotechnologies, for example, like plasma enhanced ALE and ALD (PEALE, PEALD), require control of material processing with precision, in fact, at the level of one atomic layer. In both cases, ion bombardment of the surface turns out to be important both for the functionalization of the surface state and for the stimulation of needed surface reactions. It is generally believed that ion sputtering has significant effect for high ion energies. However, this is true only in the case of the so-called "kinetic" sputtering mechanism, when the surface atom leaves it due to the part of the kinetic ion energy transferred to a surface. As is known, this mechanism prevails for high-energy ions more than hundreds of eV but almost stops reaching the threshold which is high enough (tens of eV). But when using ions of low energies (< 30-50 eV), nevertheless sputtering (though with a low rate) is still observed even below the "kinetic" sputtering threshold. Unfortunately, at present, there are almost no data on ion sputtering in the low-energy region. Moreover the few data that there are accessible may differ from each other by more than an order of magnitude. For example, there is no reliable data on the sputtering of SiO2 at low energy, a dielectric that is widely used in microtechnology. Although for the region of high energies, above ~200 eV, and ions, such as Ar+, the data of different authors agree satisfactory with each other. In this work, we studied the SiO2 sputtering by ions of the noble gases, Ne+, Ar+, Kr + and Xe +, at low ion energies. Sputtering of SiO2 was studied by in-situ laser ellipsometry in rf (13.56 MHz) ICP low-pressure plasma (< 10 mTorr). Particular attention has been paid to precise measurement of the energy spectrum and ion flux reaching the SiO2 surface, which is essential for accurate determination of the absolute sputtering “Yield” (at/ion). It should be noted that in the overwhelming majority of studies, the accuracy of Yield determination is extremely low, mainly just due to inaccurate measurement of the ion flux. The ion energy spectrum was controlled by applying an rf bias (12 MHz) to the electrode, where a Si sample with a SiO2 film (~500 nm) was placed, and was measured by an rf compensated retarded field energy analyzer (RFEA). Due to the spatial asymmetry (the area of the rf electrode is noticeably smaller than the area of the grounded plasma reactor walls), the ion spectrum is sufficiently peaked with a width of ~ 7-15 eV depending on energy. The ion flux was measured both by the method of the dynamic self-bias on the electrode after fast switching off of the rf bias, and was also determined from the plasma density measured simultaneously by a Langmuir probe and mw hairpin probe. Measurement of sputtering rate versus ion energy, i.e. Yield(Eion), actually showed different behavior at low and high ion energies that indirectly indicates on two sputtering mechanisms. The first can be associated with the classical "kinetic" mechanism with an energy threshold of about 40 eV. Above this threshold, there is observed a fast growth of Yield with increasing ion energy. The second is a low-threshold mechanism (with Eth ~ 10-15 eV depending on ion) with a slow growth of Yield with ion energy. This mechanism cannot be associated only with the direct transfer of the kinetic energy of ions to surface atoms. It is partly "potential" with including mechanism of surface ion-electron neutralization with the change of the local surface potential and the possible transfer of excitation energy to the localized surface atom, which can lead to the release of the latter. With a low probability, but nevertheless it is possible. Such a mechanism can be important for many surface processes with low-energy ions. It in turn can be of importance for plasma technologies at the atomic level. This work was supported by the RFBR grant 18-29-27003.