Ultra-shallow Depth Profiling Research Articles

Synchrotron-radiation XPS analysis of ultra-thin silane films: Specifying the organic silicon

The analysis of chemical and elemental in-depth variations in ultra-thin organic layers with thicknesses below 5nm is very challenging. Energy- and angle-resolved XPS (ER/AR-XPS) opens up the possibility for non-destructive chemical ultra-shallow depth profiling of the outermost surface layer of ultra-thin organic films due to its exceptional surface sensitivity. For common organic materials a reliable chemical in-depth analysis with a lower limit of the XPS information depth z95 of about 1nm can be performed. As a proof-of-principle example with relevance for industrial applications the ER/AR-XPS analysis of different organic monolayers made of amino- or benzamidosilane molecules on silicon oxide surfaces is presented. It is demonstrated how to use the Si 2p core-level region to non-destructively depth-profile the organic (silane monolayer) – inorganic (SiO2/Si) interface and how to quantify Si species, ranging from elemental silicon over native silicon oxide to the silane itself. The main advantage of the applied ER/AR-XPS method is the improved specification of organic from inorganic silicon components in Si 2p core-level spectra with exceptional low uncertainties compared to conventional laboratory XPS.

Roughness development in the depth profiling with 500 eV O 2+ beam with the combination of oxygen flooding and sample rotation

Roughness development is one of the most often addressed issues in the secondary ion mass spectrometry (SIMS) ultra-shallow depth profiling. The effect of oxygen flooding pressure on the roughness development has been investigated under the bombardment of 500 eV O 2 + beam with simultaneous sample rotation. Oxygen flooding had two competing effects on the surface roughening, i.e., enhancement of initiating roughening and suppression of roughening development, which were suggested to be described by the onset depth z on and transient width w tr of surface roughening. Both z on and w tr decreased as oxygen flooding pressure increased. As the result, surface roughening was most pronounced at the intermediate pressure from 4.4E−5 Pa to 5.8E−5 Pa. The surface roughening is negligible while without flooding or with flooding at the saturated pressure. No flooding is preferable for depth profiling ultra-shallow B implantation because of the better B profile shape and short analysis time.

Ultra Shallow Depth Profiling with SIMS

Secondary Ion Mass Spectrometry (SIMS) is frequently used in the characterization of thin films, coatings, diffusion processes, materials composition and in the analysis of implants. The SIMS technique has been continuously developed for more than 30 years. One of the main drivers was semiconductor technology. Standard implants in Si like B, As and P, implanted with a few keV to MeV energy are routinely measured with high precision. But nowadays with implant energies of 500 eV and below, when ultra shallow structures are examined, the desired information is in the first few nm to some tens of nm. This has a great impact on the analytical requirements and quantification procedures. Some of these aspects will be examined in this contribution.

Ultrashallow depth profiling using SIMS and ion scattering spectroscopy

AbstractThe accuracy of ultrashallow depth profiling was studied by secondary ion mass spectrometry (SIMS) and high‐resolution Rutherford backscattering spectroscopy (HRBS) to obtain reliable depth profiles of ultrathin gate dielectrics and ultrashallow dopant profiles, and to provide important information for the modeling and process control of advanced complimentary metal‐oxide semiconductor (CMOS) design. An ultrathin Si3N4/SiO2 stacked layer (2.5 nm) and ultrashallow arsenic implantation distributions (3 keV, 1 × 1015 cm−2) were used to explore the accuracy of near‐surface depth profiles measured by low‐energy O2+ and Cs+ bombardment (0.25 and 0.5 keV) at oblique incidence. The SIMS depth profiles were compared with those by HRBS. Comparison between HRBS and SIMS nitrogen profiles in the stacked layer suggested that SIMS depth profiling with O2+ at low energy (0.25 keV) and an impact angle of 78° provides accurate profiles. For the As+‐implanted Si, the HRBS depth profiles clearly showed redistribution in the near‐surface region. In contrast, those by the conventional SIMS measurement using Cs+ primary ions at oblique incidence were distorted at depths less than 5 nm. The distortion resulted from a long transient caused by the native oxide. To reduce the transient behavior and to obtain more accurate depth profiles in the near‐surface region, the use of O2+ primary ions was found to be effective, and 0.25 keV O2+ at normal incidence provided a more reliable result than Cs+ in the near‐surface region. Copyright © 2007 John Wiley & Sons, Ltd.

Shave-Off Depth Profiling for Nano-Devices

Shave-off depth profiling completely differs from ultra shallow depth profiling and has an absolute depth scale and a 50 µm wide dynamic range. In this paper we report on shave-off depth profiling of a nano-device with a complex nano-structure consisting of dynamic random access memory and introduces an analytical tool for advancing of nano-devices.

Ultrashallow profiling using secondary ion mass spectrometry: Estimating junction depth error using mathematical deconvolution

As implant energies get lower and lower, significant errors can be present in junction depth measurements in secondary ion mass spectrometry (SIMS) ultrashallow depth profiling. Primary beam ion mixing is one of the main sources of errors leading to overestimation of junction depths in SIMS measurements. In this article, we systematically study the correlations between the implant profile trailing edge, junction depth and primary ion beam energy for low energy boron and arsenic implants. Using a mathematical deconvolution model proposed by Yang and Odom [Mater. Res. Soc. Symp. Proc. 669, J4–16 (2001)], we are able to estimate the error of the junction depth and consistently improve the accuracy of junction depth measurements using SIMS.

A-Si Capping SIMS for shallow dopant profiles

Capping a sample surface using amorphous-Si (a-Si) prior to SIMS analysis is a common method for improving ultra-shallow boron depth profiling, however, details of the limitations of a-Si capping have not been thoroughly investigated. In this study we have investigated the problems and/or limitations of a-Si capping method when applied to SIMS. Capping Si layers were grown using Si evaporation in UHV (∼2×10 −8 Pa) at less than 100 °C during growth. Using this a-Si capping SIMS method we have successfully measured shallow depth profiles for ultra-shallow B and P.

Sub-keV secondary ion mass spectrometry depth profiling: comparison of sample rotation and oxygen flooding

Following the increasingly stringent requirements in the characterization of sub-micron IC devices, an understanding of the various factors affecting ultra shallow depth profiling in secondary ion mass spectrometry (SIMS) has become crucial. Achieving high depth resolution (of the order of 1 nm) is critical in the semiconductor industry today, and various methods have been developed to optimize depth resolution. In this paper, we will discuss ultra shallow SIMS depth profiling using B and Ge delta-doped Si samples using low energy 0.5 keV O 2 + primary beams. The relationship between depth resolution of the delta layers and surface topography measured by atomic force microscopy (AFM) is studied. The effect of oxygen flooding and sample rotation, used to suppress surface roughening is also investigated. Oxygen flooding was found to effectively suppress roughening and gives the best depth resolution for B, but sample rotation gives the best resolution for Ge. Possible mechanisms for this are discussed.

Characterization of dopant profiles produced by ultra-shallow As implantation and spike annealing using medium energy ion scattering

Medium energy ion scattering (MEIS) combining a toroidal electrostatic analyzer with an energy resolution (d E/ E) of 4 × 10 −3 has been used for ultra-shallow depth profiling of As implanted into Si at 1, 2 and 5 keV to a dose of 1.2 × 10 15 ions/cm 2 before and after spike annealing at 1075 °C. Depth profiling results extracted from MEIS spectra were compared with those of simulation and SIMS measurement. The arsenic re-distribution close to the surface after spike annealing was found by MEIS and SIMS measurements.

Ultra-shallow depth profiling with secondary ion mass spectrometry

Estimation methods for ultra-shallow profiling with secondary ion mass spectrometry (SIMS) were investigated. The depth and concentration of ultra-shallow profiles were calibrated using multi-delta-doped samples and bulk-doped samples. Boron profiles, whose implantation energy is 200 eV or less, were measured by backside SIMS analysis in order to minimize the atomic mixing effect. This analysis enabled accurate junction depth measurements for even 200 eV boron implanted samples when the primary oxygen ion energy was 500 eV or less, but the sample preparation time was relatively long. SIMS depth resolution functions were then extracted from surface-side and backside (true) profiles in order to deconvolute degraded surface-side profiles. This deconvolution analysis of SIMS (surface-side) profiles, using the depth resolution functions, provided profiles of similar quality to those obtained by backside SIMS analysis and was a relatively quick process.

Low-Temperature Growth of Au on H-Terminated Si(111): Instability of Hydrogen at the Au/Si Interface Revealed by Non-Destructive Ultra-Shallow H-Depth Profiling

The initial stage of Au/Si(111)-interface formation has been investigated by hydrogen-specific ultra-shallow depth profiling using nuclear reaction analysis (NRA) via the 1H(15N,αγ)12C reaction in grazing incidence, which achieves submonolayer depth resolution. Gold was deposited on the H-terminated Si(111)(1×1) substrate at 110 K to suppress Au-silicide formation known to occur at room temperature. The experimental NRA spectrum clearly rules out the prevalence of H at the Au/Si(111) interface, even though the reactive Au-Si segregation is inhibited at the low deposition temperature. Nevertheless only ∼30% of the initial H termination layer is desorbed upon deposition of 4.1 Å Au. The shading of the remaining H from the incident 15N ions by the Au film is analyzed with aid of NRA spectrum simulations, which allows estimating the average of the Au particle size distribution and for exclusion of layer-by-layer growth of Au independent from morphological information by surface-imaging.

Evaluation of the sputtering rate variation in SIMS ultra‐shallow depth profiling using multiple short‐period delta layers

AbstractWe have developed multiple short‐period delta layers as a reference material for SIMS ultra‐shallow depth profiling. Boron nitride delta layers and silicon spacer layers were sputter‐deposited alternately, with a silicon spacer thickness of 1–5 nm. These delta‐doped layers were used to measure the sputtering rate change in the initial stage of oxygen ion bombardment. A significant variation of sputtering rate was observed in the initial 3 nm or less. The sputtering rate in the initial 3 nm was estimated to be about four times larger than the steady‐state value for 1000 eV oxygen ions. Copyright © 2003 John Wiley & Sons, Ltd.

A systematic study of the surface roughening and sputter rate variations occurring during SIMS ultrashallow depth profile analysis of Si with Cs +

Sputter rate variations and surface roughening occurring during the initial stages of secondary ion mass spectrometry ultrashallow depth profiling of Si with Cs + are examined for a range of primary ion incident energies (0.25–5.0 keV) and incident angles (40–80°). Sputter rate variations were derived over the 4.15–36.4 nm range through the analysis of a Boron delta layered structure. A reduction in the sputter rate with sputtering time was noted, with the variation increasing with decreasing primary ion energy and increasing incident angle. From this data set, a relation expressing these sputter rate variations as a function of the primary ion energy and incident angle was derived which, in turn, allows for depth and intensity scale corrections to be carried out. Surface roughening culminating in ripple formation was also observed using atomic force microscopy. These were observed under conditions resulting in excessive sputter rate variations.

Ultrathin oxide interfaces on 6H–SiC formed by plasma hydrogenation: Ultra shallow depth profile study

Silicon oxide ultrathin films grown on silicon carbide (6H–SiC) by plasma hydrogenation have been studied using ultrashallow depth profiling with time–of–flight secondary ion mass spectrometry. Plasma hydrogenation gives rise to an epitaxial 3×3R30° silicate structure on 6H–SiC(0001) and 6H–SiC(0001̄). By selecting appropriate sputtering conditions, an ultrathin and atomically abrupt interface delineating the boundary between the silicate epilayer (SiO+,Si2O+, and SiO3−,SiO2−) and bulk silicon carbide (SiC+) was observed on both C(0001̄) and Si(0001) faces. Differences in the sputtering profile between the C and Si faces suggest an enrichment of the interface stoichiometry by Si and O on the Si face. Our results support the structural models of the silicate on the C and Si-face 6H–SiC(0001) proposed by Starke [Appl. Phys. Lett. 74, 1084 (1999); J. Vac. Sci. Technol. A 17, 688 (1999)].

Deuterium-oxygen exchange on diamond (100)—a study by ERDA, RBS and TOF-SIMS

The exchange of radio-frequency plasma-excited atomic O with chemisorbed D, and vice versa, on single crystalline diamond (100) 2×1 has been investigated by elastic recoil detection analysis (ERDA), Rutherford backscattering spectrometry (RBS) and time-of-flight secondary ion mass spectrometry (TOF-SIMS). It was found that the O-D, as well as D-O, exchange processes were thermally activated by the diamond substrate. The surface D was only partially exchanged by atomic O at room temperature. At elevated temperatures, the replacement of D by O was greatly enhanced, with 80% substitution by O at 300 °C. It was also observed that the uptake of O proceeded more readily on the pre-deuterated surface compared to the clean surface. Ultra-shallow depth profiling revealed that atomic beam treatment of single crystalline samples at 800 °C resulted in only superficial uptake of D and O, with no surface incorporation within the shallow analysis depth. The replacement of pre-adsorbed O by RF-excited atomic D was also studied by TOF-SIMS. Pre-adsorbed O was relatively stable and resisted removal at 300 °C. D dosed onto the oxygenated surface was found to co-adsorb with O, possibly as surface bound OD species. Mechanisms for the O-D and D-O exchange processes were discussed in connection with the atomic structure of the C(100) surface.

Ultra shallow secondary ion mass spectrometry depth profiling: Limitation of sample rotation in improving depth resolution

Depth resolution deterioration attributed to crater bottom roughness induced by 500 eV O 2 + ion bombardment at 56° incidence angle is investigated. During this secondary ion mass spectrometry (SIMS) depth profiling process, sample rotation and oxygen flooding were applied to assess their effectiveness in suppressing surface roughness. Oxygen flooding was found to effectively suppress roughening and gives the best depth resolution at the energy and incident angle used. Although sample rotation was found to suppress roughening, the depth resolution was still degraded, which is counter-intuitive. Profiles fitted to the Hofmann model suggest that ion beam mixing may limit the depth resolution in low energy ion profiling.

Effects of oxygen flooding on crater bottom composition and roughness in ultrashallow secondary ion mass spectrometry depth profiling

The effect of oxygen flooding during ultrashallow depth profiling using secondary ion mass spectrometry (SIMS) was studied on a silicon sample implanted with 2 keV boron. SIMS depth profiles were obtained on a Cameca IMS6f using low energy (1 keV) O2+ primary beams at 56° incident angle. Different oxygen flooding conditions were used to investigate the dependence of crater bottom composition and roughening on oxygen partial pressure. The development of surface oxidation state and the thickness of the silicon oxide layer formed at the crater bottom during sputtering were determined using small area high resolution x-ray photoelectron spectroscopy. It is shown that the oxidation states during sputtering are dominated by Si0 (elemental Si) and Si4+(SiO2) with small contributions from other silicon suboxides, i.e., Si1+(Si2O), Si2+(SiO), and Si3+(Si2O3). The calculated equilibrium oxide thickness was found to increase with oxygen partial pressure. The sputtering induced surface roughening on the crater bottom was characterized using atomic force microscopy as a function of depth and oxygen partial pressure. Surface roughening appears to be suppressed under higher oxygen partial pressures, and is related to the nature of the oxide formed.

Transient effects in ultra shallow depth profiling of silicon by secondary ion mass spectrometry

The significant and often unpredictable variations, or transient effects, observed in the secondary ion intensities of O± and Si± during the initial stages of depth profiling with Cs+ have been studied. These were found to be primarily due to two competing effects: (a) the steady accumulation of Cs in the substrate as a function of sputtering time and (b) the varying oxygen content from the native oxide as a function of depth. These effects prevail over depths approximated by ∼2Rnorm, where Rnorm is the primary ion range normal to the surface. The Cs+ induced effects are consistent with a work function controlled resonance charge transfer process. A method for controlling these effects, namely the prior evaporation of Cs and use of an O2 leak during analysis is described. Doped (As and Sb) and undoped Si wafers with ∼0.9 nm thick native oxides were analyzed using 0.75 and 1 keV Cs+ beams incident at 60°. The more intense polyatomic AsSi− and SbSi− emissions did not exhibit these effects, although other relatively minor intensity fluctuations were still noted over the first ∼0.5 nm.

Study of pre-equilibrium sputter rates for ultrashallow depth profiling with secondary ion mass spectrometry

Sputter rates for sub-keV primary beam energies are investigated with atomic force microscopy measurements of crater step heights. Using a 800 eV O2+ sputtering beam at an incidence angle of 50° with oxygen flooding, the surface swells as the implanted oxygen forms an altered layer. After a fluence of 1E17 O/cm2, the sputter rate reaches equilibrium. Characteristic parameters such as the depth scale offset, zd0, and the apparent initial silicon surface zSi0, are extracted from the data (0.6 and −1.3 nm, respectively), and the magnitude of the profile shift is discussed in terms of the final profile accuracy.

Depth profiling of ultrashallow implanted P using NRA

Depth profiles of ultrashallow implanted P in Si at a dose of 3×1014 cm−2 in the energy range of 5–80 keV were investigated by nuclear reaction analysis (NRA) using a 31P(α,p)34S reaction. Both projected ranges (Rp) and range straggling (ΔRp) are in good agreement with those calculated by TRIM. The accuracy of the dose measurement including reproducibility of this method is better than ±5%. This method is very useful to improve the quantitative accuracy in ultrashallow depth profiling of P in combination with SIMS.