Atomic layer deposition of TiN films: Growth and electrical behavior down to sub-nanometer scale
- Author
- Hao Van Bui
- Year
- 2013
- Abstract & Cover
- Source of Information
- Fred Roozeboom
- University
- University of Twente (Enschede, Netherlands)
- External Link
- Read Thesis
During the last several decades, titanium nitride (TiN) has gained much interest because of its low resistivity, chemical inertness and compatibility with complementary metaloxide- semiconductor (CMOS) technology. Thin films of TiN are commonly used as diffusion barrier and gate material for CMOS devices. Mostly sputtering and chemical vapor deposition (CVD) techniques were employed to deposit the TiN films. However, to achieve thin conformal films with accurate thickness control and excellent step coverage, atomic layer deposition (ALD) has become an ideal choice for making such high quality films in the thickness range of a few tens of nanometers. In this work, the growth mechanism and electrical properties of ALD TiN films are investigated. The films are grown on SiO2 substrate using the TiCl4/NH3 chemistry in the pressure range of 2.6−3.2 ×10-2 mbar and temperature range of 350−425 oC. The growth of the films is monitored by in situ spectroscopic ellipsometry (SE). The films are characterized by other complementary analysis techniques, including atomic force microscopy (AFM), high-resolution scanning/transmission electron microscopy (HRSEM/ TEM), X-ray fluorescence (XRF) and X-ray photoluminescence spectroscopy (XPS). We fabricated test structures to characterize electrical properties of the thin films. The main results of this research are presented in this thesis. In Chapter 2, the application of SE in studying the optical functions and measuring film thickness of TiN thin films is presented. We apply the Drude−Lorentz model to parameterize the dielectric functions of TiN. The film thickness obtained by SE is compared with the results obtained from HR-TEM/SEM and XRF measurements. The results show good agreement in a wide thickness range (up to 40 nm), indicating the reliable applicability of the SE technique and the Drude−Lorentz model. SE is employed to in situ study the growth (Chapter 3), the resistivity (Chapter 4) and the real-time oxidation of ALD TiN films (Chapter 5). In Chapter 3, the growth mechanism of ALD TiN films on SiO2 substrate is investigated. We have observed that the growth obeys Stranski−Krastanov model. The growth starts with a 2D mode (continuous layers) followed by a 2D-3D transition (onset of islanding). This transition is temperature independent and occurs at a film thickness of about 0.7 nm. This equals 3 monolayers of TiN. The growth of the 3D islands (on the continuous layers) eventually leads to their coalescence which occurs at 2.5 nm and 3.5 nm for the growth at 350 oC and 425 oC, respectively. Before the coalescence, new nuclei are constantly formed during the growth. Hereafter, the film grows with a constant (ALD) growth rate of 0.02 nm/cycle at both temperatures. In Chapter 4, the resistivity, temperature coefficient of resistance (TCR) and field effect in thin TiN films down to sub-nanometer scale are presented. The resistivity of TiN is determined by both spectroscopic ellipsometry (SE) and electrical test structures. The latter includes circular transfer length method (CTLM) and linear test structures. The results show that the resistivity increases significantly with decreasing film thickness. We compared the resistivity values obtained by the optical and electrical techniques. For films thicker than 4 nm, the values show good agreement. However, for films thinner than 4 nm, much higher values are found from the electrical measurements. This is attributed to the scattering effects at interfaces and grain boundaries that cannot be fully taken into account by the optical measurements. The TCR of TiN is determined from the electrical measurements in temperature range of 25−150 oC for different film thicknesses (0.65– 8.3 nm). With decreasing film thickness, the TCR values change sign from positive to negative. This change occurs at a thickness of about 2.5 nm. This effect is attributed to the metal-semimetal transition in these films. In the last part of this chapter, the electric field effect in ultra-thin TiN films in the metallic and the semimetallic states is presented. Upon electrical field, a change of the conduction current up to 22% is found for the films in the semimetallic state. In Chapter 5, the influence of native oxidation on electrical behavior of thin ALD TiN films is reported. For films thinner than 5 nm, this oxidation changes the I-V behavior from linear to non-linear. It shows an aging effect that causes an irreversible degradation of TiN. The native oxidation does not influence the linearity of the I-V characteristics for films thicker than 5 nm. Nevertheless, it reduces the electrical thickness of TiN. The kinetics and mechanism of thermal oxidation in the temperature range of 325−425 oC of films with a thickness of 5 and 15 nm are intensively studied. The oxidation in oxygen ambient was carried out directly after the TiN deposition in the same reactor (without vacuum break) and in situ spectroscopic ellipsometry monitoring was used. The oxidation of the films can be described as a four-stage process: (1) surface oxidation, (2) diffusion of oxygen along grain boundaries and lateral oxidation and sharpening the grains, (3) rapid oxidation of the sharp grains, and (4) diffusion controlled oxidation. Finally, in Chapter 6, the generation of atomic hydrogen (H) is presented. Atomic hydrogen is made by the dissociation of molecular hydrogen (H2) upon collision with a tungsten (W) filament kept at a high temperature (T ≈ 1600−1900 oC). We have demonstrated the ability to create atomic hydrogen and to introduce it in short pulses by experiments on etching of tellurium (Te) films in the reactor chamber. The generated atomic hydrogen (H) is further utilized to explore its impact on surface reactions in the TiCl4/NH3 precursor system. Atomic hydrogen is introduced in pulses additionally to TiCl4 and NH3 with different pulse sequences. The results show that for the TiCl4/NH3/H sequence, there is no influence on the process compared to the ALD without H-pulses. The growth rate remains at 0.02 nm/cycle and the oxygen content - at 3−5 at%. For the TiCl4/H/NH3 pulse sequence, the growth rate decreases to 0.01 nm/cycle and the oxygen content increases to 30−35 at%. This indicates the reduction of Ti-Cl surface groups by H and the creation of very reactive Ti surface. This surface can be easily oxidized by residual gases from the reactor and process gases. TiCl4/H pulses only result in no growth after the formation of approximately one monolayer.
