Determination of thicknesses of oxide films grown on titanium under argon irradiation by spectroscopic ellipsometry

Abstract

In this article we present a study of the oxidation of pure titanium bulk samples under argon ion irradiation at 500 °C under rarefied air. In particular we follow the dependence of the oxide thickness as a function of the energy of argon ions. The novelty of this study consists in the range of ion energy explored, from 1 to 9 MeV. Until very recently it was commonly accepted that metal surfaces were transparent to ion beams in this low energy range (few MeV), and no surface modifications were expected. In a previous paper by the authors of this work, the formation of shallow craters in the surface of titanium was reported as a result of argon ion bombardment with energies of 2, 4 and 9 MeV under the same environmental conditions. We show here that around 3 MeV the oxide growth is unexpectedly enhanced. We think that an interplay of electronic excitations and nuclear ballistic collisions could possibly explain this enhanced oxide growth. We have used spectroscopic visible ultraviolet ellipsometry and XPS to determine the thickness of the oxide layers and characterize their optical properties. From the optical properties of the oxides we observed that for ion energies below 3–4 MeV the oxides show a dielectric-like behavior, whereas for ion energies above 3–4 MeV the oxides show a metal-like behavior. These findings indicate also that ion bombardment in this energy range may change substantially the oxygen-to-titanium ratio in the oxide films grown under irradiation leading to the formation of titanium sub-oxides.

Introduction

In this paper, we report the result of our studies on oxide film growth on the surface of bulk titanium under MeV argon irradiation. We paid special attention to characterize the thickness and the optical properties of the oxide films as a function of the energy of the Ar ions, which was varied between 1 and 9 MeV. In a previous work we realized, after examining irradiated zirconium [1] or titanium [2], [3], [4], [5], [6] samples with Ar ions in similar environmental conditions, that their surfaces were damaged leading to either an unexpected oxide growth or cratering, as function of ion energy. In particular, we observed the occurrence of craters whose size and areal density distribution depend on the ion energy. Because the irradiation experiments were performed in a rarefied atmosphere (5 × 10⁻³ Pa) containing traces of oxygen and water vapour, we found that a thin oxide was formed on the damaged surfaces. We explained this oxidation by a modification of the surface reactivity of metals under the effect of ion irradiation. To the best of our knowledge, such an effect of ion irradiation on thermal oxidation has not been observed before. In our opinion, this is partially explained by the fact that it is extremely rare to find reliable ion sources equipped with an adequate environmental chamber operating in the low MeV energy range. For instance, in France there is only one facility at the Nuclear Physics Institute of Lyon (IPNL), equipped with an environmental chamber. Another possible explanation may be the fact that there is a general belief that there is not a particular effect of corrosion under irradiation in the concerned energy range, therefore no specific studies have been devoted to it.

There is an additional difficulty when it comes to interpret the experimental results of surface corrosion under ion bombardment. When we checked the literature to look for a theoretical explanation of our observations, we found that the interaction of energetic ions with metal (or semiconductor) surfaces in the energy range of interest is not well understood. Indeed, there is a whole body of literature on metals and semiconductors oxidation following ion irradiation and implantation, but for ion energies limited to the range of ten to a few hundreds of keV essentially. In this low energy range, the implantation profile remains near the surface of the irradiated/implanted material, within one hundred or a few hundred nanometres depending on the characteristics of the projectile/target couple. The surface modifications produced by such ion beams were largely investigated in the past and continue to be studied today. This is not the case in the specifically studied energy range, for essentially two reasons. First, in the MeV range, according to SRIM-2011 computer code simulations [7], the ion penetration depth, or more precisely, the projected range (R_p), may attain a few micrometers, the sputtering effects are largely reduced by comparison with the keV range and the target surface is supposed left largely undamaged. Second, in the low MeV range, superficial radiation damages (whatever they may be, here we are concerned with an apparent oxidizing effect of irradiation) should a priori result of combined projectile slowing down effects due to nuclear ballistic collisions and electronic excitations, contrary to what happens either in the low energy range (keV range) where the damage production is controlled by ballistic collisions [8], [9], [10], [11] or in the high energy range (a few MeV per nucleon) where the damage production is controlled by electronic excitations [12], [13], in other words contrary to what happens both in the low and high energy range for which only one slowing down mechanism is involved. The lack of well-established models may be partially explained by the lack of experimental data. Moreover, the fact is that all of those models describing radiation effects on metal surfaces do not consider the effect of the environment (presence of oxygen or other oxidizing species), which is of extreme importance in the case of our study.

Therefore, we think that a thorough study of the oxides formed in our samples can bring experimental evidence to the modification of surface reactivity under MeV ion irradiation. This information should be helpful to propose and test models to explain the interaction of energetic ions with metals in an oxidizing atmosphere.

As mentioned previously, along our studies we reported on the cratering and roughening of titanium surfaces as a result of MeV argon ion irradiation under slightly oxidizing environment [2], [3]. At the same time we also noticed an unexpected modification of the titanium oxide film thickness depending on the argon ion energy [3]. In the present paper we report on a thorough analysis of the oxidation of titanium under the same irradiation conditions (3 × 10¹⁰ ions cm⁻² s⁻¹ up to 5 × 10¹⁴ ions cm⁻² – 3 h) and the same environmental conditions (5 × 10⁻³ Pa of circulating dry air at 500 °C), by spectroscopic ellipsometry (SE).

Spectroscopic ellipsometry is a non-destructive, non-invasive and contactless optical technique, which is commonly used to characterize the thickness and the dielectric function of thin films [14], [15]. Standard ellipsometric measurements are commonly performed in external configuration, which means that a light beam propagating in air (or vacuum) is reflected by, or transmitted through a sample, and then it propagates again in air (or vacuum) before arriving at the detector. The benefit of ellipsometry when compared to reflectometry is that it can measure simultaneously the modulus and phase of the polarization components of the light whereas reflectivity is only sensitive to the modulus. The sensitivity of phase measurements, exploited to determine the thin film thickness, has its roots in an interferometric effect. The light reflected by the first interface of a layer present in the sample, interferes with the light reflected by the second interface of the layer. The same principle remains valid when a stack of layers is present. The analysis of the spectroscopic dependence dielectric can provide information about the electronic structure of the materials analyzed as will be explained in more detail in a forthcoming section.

In the present study we found two particular results. Firstly, we noticed that Ar ion irradiation enhances surface oxidation no matter the ion energy involved. Secondly we also found that the oxidation enhancement has a particular dependence with ion energy. We found a peak of oxidation for ion energies of about 3 MeV under present environmental conditions. In view of the existing models (which do not take into account the effect of the atmosphere) we think that the oxidation peak could be attributed by an optimal combination of nuclear collision and electronic excitation mechanisms.

In the following, we describe briefly the oxidation under irradiation experiment, continue with a short overview of the SE technique to introduce the physical meaning and the data treatment currently done with SE data. Then we explicit for the particular case of the samples considered in this study the results of the SE measurements, together with the subsequent data treatment and physical interpretation. We discuss the argon energy dependence of the titanium oxide thicknesses and conclude briefly.