AVS2001 Session AS-TuP: Aspects of Applied Surface Analysis II Poster Session

Tuesday, October 30, 2001 5:30 PM in Room 134/135

Tuesday Afternoon

Time Period TuP Sessions | Topic AS Sessions | Time Periods | Topics | AVS2001 Schedule

AS-TuP-1 XPS and XAES Characterisation of the Interaction of Copper-containing Alloys with Ultra-pure Water in Oxidising and Reducing Conditions
J.E. Castle, P.A. Zhdan (University of Surrey, U.K.)
X-ray photoelectron spectroscopy (XPS) and X-ray Auger electron spectroscopy (XAES) were used to characterise changes in the surface chemistry of different copper alloys (admiralty brass, Al brass and 90Cu-10Ni alloy) as a result of their interaction with ultra-pure water at 95C for periods of up to 200 hours. All alloys were treated at pH 8-11.5 using oxygen-containing water (oxidising environment) and oxygen free water (reducing environment). This allowed us to determine the role of the dissolved oxygen in changes of the chemistry of the resulting surface products. The surface composition of representative samples before and after treatment was determined by XPS and XAES using the ESCALAB 2 and the SigmaProbe electron spectrometer (VG Scientific, UK) with monochromatised Al radiation. Experimental results for admiralty brass demonstrated that under reducing conditions the protective surface layer is cuprous oxide while under oxidising conditions the surface oxide contains increasing amounts of cupric oxide (CuO). When a reducing condition is changed to an oxidising condition at an operating temperature of 95C with pH 9 - the surface layer undergoes a conversion from cuprous to cupric oxide. After exposure of admiralty brass at pH 10.5 the surface is mainly composed of zinc oxide and no cupric oxide was present on the surface after its exposure to an oxidising environment at pH 11.5. The role of the alloy composition to chemistry and stability of resulting surface products is discussed and the marked effect of the addition of carbon dioxide on the composition of surface layers formed in oxidising conditions is demonstrated.
AS-TuP-2 Non-destructive Depth Profiling Analysis of β-FeSi2 Formation Process by SR-XPS
T. Saito, H. Yamamoto, H. Asaoka, K. Hojou (Japan Atomic Energy Research Institute); M. Imamura, N. Matsubayashi, H. Shimada (National Institute of Advanced Industrial Science and Technology, Japan)
In recent year, semiconducting iron silicide,β-FeSi2 is extensively attracted because of its semiconducting character. In the present study, the formation processes of β-FeSi2 during solid-phase epitaxy (SPE) from Fe thin films on Si (111) surface have been investigated by means of X-ray photoelectron spectroscopy using synchrotron radiation (SR-XPS). All the experiments were carried out at a soft X-ray beam line (BL-13C) of Photon Factory. Non destructive depth-profiling analysis with changing excitation energy (220 - 900 eV) revealed that the surface composition of the iron deposited substrate was gradually changed with rising annealing temperature indicating aggregation or diffusion of the surface iron. The comparisons of experimental results with simulation results using inelastic mean free path (IMFP) of electron in Si and Fe revealed that the surface iron was diffused into Si bulk gradually with rising annealing temperature. The changes in the core-level Fe 2p photoemission spectra indicated the formation of iron silicide above annealing temperature of 673K. The valence-band photoemission spectra also indicated that metallic surface iron changed into semiconducting β-FeSi2 phase by the annealing. These diffusion and silicide formation behavior were influenced by the thickness of the initial iron layer before annealing suggesting the importance of the iron diffusion in β-FeSi2 formation process.
AS-TuP-3 X-ray Photoelectron Spectroscopy and Auger Electron Spectroscopy Studies of the Effect of a Pre-oxidation Clean on Boron And Residual Fluorine Distributions in Ultra-shallow Junction BF2+ Implants
E.G. Garza, S.N. Raman (Advanced Micro Devices)
The distribution of as-implanted boron and fluorine in ultra-shallow junction 49BF2+ implants was investigated using X-ray photoelectron spectroscopy (XPS) and Auger electron spectroscopy (AES). The effects of a pre-oxidation clean following low energy, high-dose 49BF2+ implants (LEHD) were also investigated. Auger and XPS depth profiles and angle-resolved XPS (ARXPS) studies reveal significant dopant retention in the native oxide and subsequent high degrees of dopant loss due to removal of the native oxide during a standard SPM/APM/HPM clean process. Fluorine Auger parameter analysis was also performed to identify changes in fluorine binding states.
AS-TuP-4 X-ray Photoelectron Spectroscopy Study of As-implanted and Annealed Ultra Shallow Junction 75As+ Implants
S.N. Raman, E.G. Garza (Advanced Micro Devices)
Implant depth and silicon binding states in as-implanted and annealed ultra-shallow junction 75As+ implants were characterized using X-ray photoelectron spectroscopy (XPS). Dopant profiles were merged with secondary ion mass spectrometry (SIMS) depth profiles to complete the dopant distribution in the initial transient region. Doped regions of the implanted silicon substrate were identified using shifts in the Si2p photoelectron energy. The effects of rapid thermal annealing on Si2p energy are also presented.
AS-TuP-5 The Particularities of In-line Control Tools Employing Electron Spectroscopic Methods
L. Vasilyev, C. Bryson, D. Klyachko, B. Linder (Surface Interface Inc.)
The minitiaturization of electronic devices will inevitably lead to the development of the new methods of in-line control. One of the good candidates for such a method of in-line control is electron spectroscopy that is characterized by a better surface sensitivity compared to the currently used optical methods. In addition, electron spectroscopy provides quantitative information regarding the chemical composition of thin films at the surface. The specific requirements to the in-line control tools affect the design of electron spectrometers now mostly used in the research laboratory environment. 1. The in-line tool is application specific. It is optimized to perform a limited number of tasks with a best possible reproducibility and a required precision. 2. The tool must be simple in use and can be operated by a technician with a rather low education level. This requires high degree of automation that includes control of the system status, in-line data analysis and data reduction, aiming to detect and eliminate wrong data, quantify the results and perform self-diagnostics and calibration. 3. The software is an integral part of the tool that is specific for a given type of the hardware. By limiting its flexibility it is possible to reduce the time and cost of its development and to customize for a particular task. 4. High throughput and low price are the key issues for the in-line tool to be accepted in industry. Non-traditional and application specific design of the vacuum chamber and the measurement units could substantially reduce their price and increase the throughput. 5. The whole power of the electron spectroscopy (not only AES and XPS) must be utilized for these applications. For example, Beta backscattering can be utilized for film thickness control in hard drive industry. Here we describe our experience of the development of in line control electron spectrometer utilized for thin carbon based overcoat control.
AS-TuP-6 High Spatial Resolution Auger Microanalysis of FIB Prepared Samples
A.L. Linsebigler, M. Larsen (General Electric Corporate Research and Development Center)
In the past several years, the realm of Auger microanalysis has been expanded to include analysis of nanometer range thin films and particles. Auger analysis is ideal for fast and simple elemental determination of thin film and particle composition on a submicron level. In most cases, the area of interest is buried below the surface of the sample out of the range of the Auger electron sampling depth. Traditional metallographic sample preparation techniques can be used to prepare a cross section of a sample to expose buried layers and particles. The surfaces of samples can also be polished to produce high spatial resolution (< 0.3um) Auger elemental maps. Traditional metallographic polishing techniques require that the sample be mounted in nonconductive materials and require the use of liquid polishing media, which can absorb into microcracks in the mounting material and sample. This in turn, leads to considerable outgassing under the ultrahigh vacuum conditions needed for Auger analysis. In addition, the traditional techniques can miss the submicron area of interest, take a considerable amount of time, and lead to smearing and rounding of interfaces. FIB sample techniques are UHV compatible, are well controlled to prepare the submicron area of interest, and certain FIB sample preparation methods are much faster than traditional polishing methods. Auger microanalysis of a cross-sectional surface of buried submicron thin films prepared with the box, staircase, and TEM lift out FIB techniques will be presented and the results will be compared to the same surfaces prepared by traditional metallographic polishing techniques. A comparison between the Auger interfacial resolution, surface roughness, and surface cleanliness obtained with each technique will be presented.
AS-TuP-7 Applications of AFM/SCM in Process Control and Failure Analysis of Semiconductor Devices
K.-J. Chao, J.R. Kingsley, H. Ho, H. Shen, I.D. Ward (Charles Evans & Associates)
As the minimum feature size of a semiconductor device continues to shrink, analyses of the engineered structures and materials of semiconductor devices have been very critical in device manufacturing and the development of new generation devices. Furthermore, the performance of a modern device is strongly influenced by its dopant distributions in depth and laterally. Scanning capacitance microscopy (SCM) is one of the techniques that have been developed to reveal the dopant distribution two-dimensionally. In this work, AFM/SCM was applied to both process control and failure analysis of semiconductor devices. Two types of semiconductor devices, Si based and GaAs based, were studied. In the applications of process control, two examples are presented. The first example is to investigate the uniformity of the As layer of a silicon substrate. This sample has a layer of uniformly distributed As region. X-sectioned SCM/AFM study found local non-uniformity in the As layer. The second example is to determine the lateral diffusion length of dopants after the thermal annealing process. A GaAs substrate is masked and patterned by a layer of Si3N4. Zn dopants were thermally dif-fused through a line-opening in the Si3N4 layer into the GaAs substrate. By overlaying the AFM image on top of the SCM image, the lateral diffusion length of the Zn dopants was found to be about 2.2µm away from the edges of the exposed GaAs region. Another important application of AFM/SCM is in the device failure analysis. A failed p-channel transistor was first studied by AFM/SCM. Two levels of dopant concentration were found in the corresponding source and drain regions by SCM. Another way to identify the cause of the failure is to directly compare the failed device with a properly functioning one. We applied AFM/SCM to study the same N-well struc-tures in two devices, one good and the other failed, and found that the depth of the N-well for the failed device is about 0.4µm shorter.
AS-TuP-8 Surface Analysis of Shock-compacted Nd-system Superconductor
H. Kezuka (Tokyo University of Technology, Japan); M. Kikuchi, K. Fukuoka, E. Ohshima (Tohoku University, Japan); S. Yoshizawa (Meisei University, Japan); T. Suzuki (Tokai University, Japan)
Shock compaction experiments are performed in vacuum for Nd-system superconductor particles (NdBa2Cu3O7-x) which has a peak effect in the high magnetic field region of near 2T. The as-shocked specimen shows an c-axis oriented crystal structures with the lattice constant of the orthorhombic cell:a=0.5787nm, b=0.5874 and c=1.17462nm calculated by XRD(X-ray powder diffraction method)-analysis just after the shock compaction under 5.7Gpa. From SEM(scanning electron microscopy)- and AFM (atomic force microscopy)-observations, the surface of the specimen for as-shocked Nd-system has larger grains of 4-10 micrometer in length with growth steps caused mainly frictions among the particles in shock compaction.
AS-TuP-9 Surface Oxidation of NiTi Shape Memory Alloy
R.G. Vitchev (Katholieke Universiteit Leuven, Belgium); G.S. Firstov (On leave of absence from National Academy of Sciences of Ukraine); H. Kumar, B. Blanpain, J. Van Humbeeck (Katholieke Universiteit Leuven, Belgium)
NiTi shape memory alloys are often used as materials for medical implants. It is known that the biocompatibility of the implants fabricated from NiTi depends on their surface composition since nickel often causes allergic and toxic reactions. Protective titanium oxide layers on the NiTi surface are often created since they can prevent the contact of Ni with the body tissues. The purpose of this study is to determine the oxidation mechanism of the NiTi alloy. The samples were oxidised in air in the temperature range 473-1073 K. Scanning electron microscopy was used to study the morphology and composition of the oxidised surfaces. The kinetics of oxidation was investigated by means of thermo-gravimetry. X-ray photoelectron spectroscopy and Auger electron spectroscopy combined with sputter depth profiling were used to investigate the surface composition and chemical state of the constituent atoms and their depth distribution. Grazing incidence X-ray diffractometry and Raman spectroscopy were implemented to determine the phases formed in the oxide layers. Thermodynamic calculations were performed to explain the oxidation behaviour of the NiTi alloy.
AS-TuP-10 Structure and Corrosion Properties of PVD CrN Coatings
C. Liu (Loughborough University, UK); Q. Bi (Hull University, UK); H. Ziegele (BMW Group, Germany); A. Leyland, A. Matthews (Hull University, UK)
Corrosion of PVD CrN coated steels in an aqueous solution is usually by galvanic attack due to the difference in electro-potentials between the steel substrate and the coating material which causes corrosion activity in through-coating defects (e.g. pinholes). The structure related variables which can determine corrosion performance of CrN coatings include: (i) Surface discontinuities and uniformity of coverage; (ii) Open and through-coating porosity; (iii) Film density and chemical stability; (iv) Growth stresses and associated strain; (v) Interfacial adhesion and the properties of intermediate layers; (vi) Coating thickness; (vii) Substrate properties (e.g. activity and morphology); (viii) Coating composition. In this study, PVD CrN coatings were prepared with different composition, thickness, and surface roughness by changing the nitrogen flow rate, applying multilayering techniques and changing the substrate finish prior to coating, respectively. It has been found that the microstructure of CrN coatings can vary with the N content in the film, thus their corrosion performance can be affected significantly by nitrogen flow rate during coating preparation. The steel substrate surface finish (i.e. roughness) can affect the uniformity and coverage of PVD coatings, as the grooves and inclusions on the original substrate can raise the susceptibility of the coated systems to crevice corrosion. Increased film thickness can greatly reduce the through-coating porosity; this makes the coatings less permeable in terms of solution penetration, such that the corrosion resistance of the system can be enhanced significantly.
AS-TuP-11 Analysis of Boron-, Germanium- and Fluorine Diffusion through a SiO2 Gate Oxide Into Silicon using Secondary Ion Mass Spectrometry
F. Persson, H. Svensson, U. Södervall, M. Willander (Chalmers University of Technology, Sweden)
This project work is a study of diffusion using in-situ B-doped, B- and BF-implanted samples as sources for boron. The main aim of this report is to see how fast boron and germanium moves through a SiO2 gate oxide into a Si-matrix. To get an answer to this and other questions, we have with a SIMS analyzed annealed specimens at different times and temperatures. Using Fick's law and data from the SIMS, the diffusion constants have been determined. An interesting fact emerging is the dependence of the oxide thickness as a barrier against penetration. SiO2<2,5 nm let through atoms at 950°C and 240 s while SiO2<3 nm is pervious first at 1050°C and 60 s. Another observation is the anomalously high values of boron after the annealing, especially where germanium is involved. Besides, boron penetrates faster from a SiGe gate than from pure Si. The fast influx of fluorine from the BF-implanted samples is also worth mentioning. According to our results, in-situ B-doped gate is preferable to B-implanted gate as gate material. In-situ doping provides a highly uniform profile from surface to the SiO2 interface. By implantation it is difficult to achieve a uniform doping profile, which can lead to gate-depletion or by annealing to an increased boron penetration through the oxide.
Time Period TuP Sessions | Topic AS Sessions | Time Periods | Topics | AVS2001 Schedule