Coverage, Composition and Band Gap Analysis of ALD-Grown Ultra Thin Films

Material development for semiconductor and thin film materials requires a full understanding of the material characteristics and how they impact each other.  The Thermo Scientific™ Nexsa™ XPS System integrates multiple techniques making it ideal for correlative analysis and thin film material development. See how in this application note.

The Nexsa™ makes correlative analysis possible because it uses multiple methods, namely, UV photoelectron spectroscopy (UPS), ion scattering spectroscopy (ISS), reflected energy loss spectroscopy (REELS) and Raman spectroscopy. Taking advantage of this capability, the Nexsa™ was used to investigate various properties of a set of samples of HfO2. Each sample was made up of a thin layer of HfO2 laid down using a graded number of atomic layer deposition (ALD) cycles. Initially, the mass of Hf laid down on the substrate was measured quantitatively using XPS, as well as the thickness of the layers of HfO2 and SiO2. ISS and REELS were then used to carry out further analysis to obtain surface coverage and band gap readings respectively.

Thermo Scientific Nexsa XPS System

Thermo Scientific Nexsa XPS System

Analysis of the Thickness and Composition of the Film

The investigation began with XPS measurements since this is the primary capability of the Nexsa™. This high-end XPS device is used in this experiment to obtain data about the composition and the thickness of the overlayer. The technique to quantify how much HfO2 was deposited in one ALD process relies on the XPS spectrum obtained from the sample of HfO2 on SiO2/Si following each cycle.

XPS survey spectra showing increasing HfO2 composition as a function of ALD cycles

Figure 1. XPS survey spectra showing increasing HfO2 composition as a function of ALD cycles

Following XPS acquisition, the data is analyzed by the Avantage Data System of the Thermo Scientific™ device to obtain the information about the thickness of the film, using an equation modified from the

Beer-Lambert law. This needs only data on material densities and bandgap readings to be fed in by the user, and will generate the attenuation length separately for each material layer. This information will be used finally by Avantage to output the thickness of the individual layers, making use of the comparative strengths of each XPS signal. In the present example, Hf films were found to be between 0 and 10 nm based on the relative XPS signal intensity of Hf and Si signals.

This kind of measurement helps ensure that the ALD film is of specified thickness (taking it for granted that each cycle deposits film in a uniform manner) and so maintains a constant and accurate control on quality standards during the process.

Coverage Analysis

The problem in real-time production is the lack of complete uniformity of deposition leading to uneven coverage of the film. However, this can be compensated for by the Nexsa™, using sample coverage as the base. The measurements of film coverage for each sample are taken using ISS. This technique utilizes only the most superficial layer of the sample, one molecule thick, to obtain the spectrum, unlike XPS, because it uses a beam of helium ions at much lower energy. These are therefore scattered by the atoms at the top of the sample, and fail to penetrate this layer, accounting for the lack of effect from any but the topmost layer.

The ions lose kinetic energy through this scattering, the degree of loss depending upon the ionic mass, the mass of the sample atom, and the angle of scattering. Of these three, the only variable is the mass of the sample atom, and therefore the energy of the scattered helium ions is measured and analyzed by the Nexsa™ to clarify the nature of the sample’s top monolayer. As a result, if the ISS spectra show the presence of signals originating from Si atoms, it may be concluded that the surface coverage by the HfO2 film was patchy.

As shown in Figure 2, the Si peak declines in amplitude with an increase in the number of ALD cycles. Once 50 cycles are completed, the Si peak is no longer visible within the spectra. In other words, 20-50 ALD cycles are necessary to produce full surface coverage.

ISS spectra of the HfO2 sample over 100 ALD cycles (right) and a plot indicating the level of surface coverage by ALD cycle (left).

Figure 2. ISS spectra of the HfO2 sample over 100 ALD cycles (right) and a plot indicating the level of surface coverage by ALD cycle (left).

Band Gap Measurements

The band gap is a vital characteristic of high-k dielectric materials, and especially of gate components. This term is used to define the difference in the energy of the valence band and conduction band of any material. It is important because its magnitude determines whether the material can be used in many applications of great interest, such as the LED, photovoltaic or solar cell. Its precise determination is therefore vital. In most materials, the XPS technique is useful for this purpose, but for high-k dielectric materials, such as HfO2 in this experiment, the XPS peaks show significant overlap within the region of the band gap energy, giving rise to potential inaccuracy. This problem is anticipated and corrected in the Nexsa™ with the REELS system which can be used to obtain REELS spectra from the same spot on the surface of the sample from which the earlier spectral measurements were generated. Avantage is then able to calculate the band gap of the sample material from this data by an automated process, by measuring at what level of loss of energy the electrons begin to scatter in inelastic fashion. Another advantage is the use of a single dual-source flood gun for charge compensation and REELS measurement, precluding the necessity of a separate electron source for this technique.

REELS spectrum of the HfO2 sample after 100 ALD cycles with the band gap measurement illustrated.

Figure 3. REELS spectrum of the HfO2 sample after 100 ALD cycles with the band gap measurement illustrated.

Summary

The Thermo Scientific™ Nexsa™ XPS System not only provides data on the sample composition and thickness using XPS technology, but utilizes ISS and REELS technology to produce information about the completeness of the coverage and band gap measurements respectively. This means the Nexsa™ is capable of generating a full picture regarding the key properties of a material which could be used to produce ultra-thin films for semiconductors and related applications.

This information has been sourced, reviewed and adapted from materials provided by Thermo Fisher Scientific – Materials & Structural Analysis.

For more information on this source, please visit Thermo Fisher Scientific – Materials & Structural Analysis.

Ask A Question

Do you have a question you'd like to ask regarding this article?

Leave your feedback
Submit