Near-surface Defects Research Articles

Relation between swelling and embrittlement during post-irradiation annealing and instability of helium-vacancy complexes in nickel

Abstract An original cross-section technique in combination with TEM has been used to study near-surface defects produced in nickel by 40 keV He+-ion irradiation up to a dose of 5 × 1021 ions m−2 at T ≤ 323 K. The evolution of the damage structure in these specimens after fixing the irradiated surface with nickel electrodeposit and subsequent one-hour annealing at 773, 873 and 973 K was also studied. Quantitative data on the dislocation structure, swelling and volume content of helium trapped in bubbles along the whole ion range have been obtained and compared to the theoretical predictions. The well-known conceptions concerning helium-vacancy complexes were used to explain: (1) the observed dramatic swelling in nickel after irradiation at annealing temperatures above 0.5 Tm; and (2) the established correlation between the onset temperature of high-temperature radiation-induced embrittlement (in the course of post-irradiation tests) and the dissociation temperature of the most simple and stable helium-vacancy complexes in a given material, namely, HeV. (3) Experimentally verified has been the SIA-capture model of bubble nucleation and growth by low-temperature He+-ion irradiation advanced by Baskes et al. (1981).

A variable-energy positron beam for low to medium energy research

A positron beam facility is described which provides a monoenergetic beam (± 1 eV) which is variable in energy from a few eV up to a maximum of ∼ 80 keV. The positron moderation process, beam transport, and design of the target chamber are discussed in detail. Some of the research being done with the facility is summarized, including scattering and energy-loss measurements, near-surface defect profiling studies, and low energy positron channeling studies. An ultrahigh-vacuum 2-axis goniometer is described, which is used in the channeling work.

Surface states and the exponential valence-band tail in a-Si:H.

We use total-yield photoelectron spectroscopy to measure the density of occupied states in undoped and lightly boron-doped a-Si:H. We observe a large density of states in the gap of undoped a-Si:H, which we ascribe to near-surface defect states. Incorporation of boron removes these states and allows an unobstructed view of the valence-band tail. This is shown to be exponential over more than 3 orders of magnitude in the density of states with an inverse logarithmic slope of 45 meV, in good agreement with the results of dispersive transport measurements.

Physical nature of the InP near-surface defect acceptor and donor states.

The initial stage [0.001--30 monolayers (ML)] of room-temperature Schottky-barrier formation of In on n- and p-type InP(110) interfaces has been studied by photoemission using a newly designed ultralow-coverage metal evaporator which can reproducibly evaporate metals with coverages as low as 0.0001 ML. It is found that the In/n-type InP(110) interface band bending does not start until the In coverage reaches about 0.3 ML, while the In/p-type InP(110) band bending is almost saturated at 0.3 ML. The heating effect on the band bending of clean cleaved n- and p-type InP(110) surfaces is also studied by using photoemission. It is found that heating has an irreversible band-bending effect on the p-type InP(110) but not on the n-type InP(110) interface. Based on these two striking differences in the band-bending behavior of n- and p-type InP, it is proposed that the physical nature of InP near-surface defect acceptor and donor levels may be different. (Our results are discussed and compared with calculated defect levels of Daw and Smith [Solid State Commun. 37, 205 (1981)] and Allen and Dow [Phys. Rev. B 25, 1423 (1982)]).

Low-energy hydrogen ion bombardment damage in silicon: An i n s i t u optical investigation

I n situ and spectroscopic ellipsometry measurements have been applied to probe damage mechanisms induced by low-energy hydrogen ion bombardment of single-crystal silicon. Ion beam voltages from 20 to 1500 V have been employed, covering the range found to passivate grain boundaries in polycrystalline Si and near-surface defects in dry etched single-crystal Si. The spectroscopic data for room-temperature samples after bombardment provide information needed to model the in situ data, obtained using a fixed photon energy of 3.4 eV at a nominal sample temperature of 200 °C. Models for the in situ ellipsometry data for ion energies <500 V allow determination of the evolution of damaged surface layers as a function of exposure time. Two layers dominate the time evolution of the optical data: (1) a polycrystalline Si (p-Si)-like layer, modeled as an effective medium mixture of c-Si, a-Si, and void, which develops on the scale of 1 min (dose: 6×1017 ions/cm2), and (2) a more slowly evolving transparent layer, interpreted as an oxide. The latter layer, also observed in earlier studies, may be attributed to oxidation caused by vacuum residual oxygen ions in the hydrogen ion beam. The oxidation consumes the top-most p-Si-like region of heaviest damage.

Defect site competition for metal atoms in intrinsically gettered silicon: Numerical model

Intrinsic gettering of silicon immobilizes metal atoms at oxide precipitates and other crystal defects in the bulk of a wafer. The competition for metal atoms between these bulk gettering sites and near-surface defect sites is investigated with simple models. Metal-decorated near-surface defect sites lead to dark or leakage currents that limit semiconductor-device performance. The calculations indicate the circumstances under which bulk gettering sites are predicted to be successful in competing for metal atoms. Numerical results are shown for variations in the number of near-surface sites, the metal atom capacity of a bulk site, and the number of initial metal atoms. Predicted profiles of free and trapped metal atoms are presented.

Scanning electron acoustic microscopy and its applications

Scanning electron acoustic microscopy (SEAM) is a relatively new technique for imaging and characterization of thermal and elastic property variations on the scale of a few micrometres. A megahertz-chopped, focused electron beam in a scanning electron microscope (SEM) generates sound waves in the sample, and the signal from a transducer attached to the specimen is used to form a scanned image in parallel with the normal SEM image. Although the acoustic wavelengths are typically several millimetres, lateral and depth resolution may be only a few micrometres. This is because image contrast is mainly derived from the micrometre-sized acoustic generation volume just under the beam. In this generation volume, both elastic variations and thermal scattering of the critically damped ‘thermal waves’, with wavelength of a few micrometres (due to the periodic beam heating), may lead to contrast. There is also evidence for non-thermoelastic contrast generation mechanisms, and these show finer resolution, SEAM shows both advantages and drawbacks compared to conventional scanning acoustic microscopy (SAM). Although understanding of the detailed contrast mechanisms in seam is at present only at a qualitative level, it is clear that they are generally different from those in SAM. The technique is now beginning to attract commercial attention, and applications include imaging of cracks and voids, grains and grain boundaries and second phases in polycrystalline materials, regions of plastic deformation, and magnetic domain structures. Images can be obtained from ics and semiconductor materials which show doped and implanted regions, near-surface manufacturing defects, and even individual dislocations. The acoustic waves excite vibrational patterns in the specimen, and these can be used as low-resolution probes of, for example, bonding integrity.

Photoemission Study of the Physical Nature of the InP Near-Surface Defect States

ABSTRACTThe physical nature of the InP near-surface defect acceptor and donor states are studied by using photoemission spectroscopy. It is found that the In/n-InP(110) interface band bending does not start until the In coverage reaches about 0.3 monolayer (ML), while the In/p-InP(110) band bending is almost saturated at 0.3 ML. The annealing effect on the band bending of clean cleaved n-and p-type InP(llO) surfaces is also studied. It is found that annealing of the clean surface creates an irreversible band bending effect on the p-InP(110), but the n-InP(110) almost does not show any band bending after low temperature annealing. Based on these two striking differences in the band bending behavior of n- and p-type InP, it is proposed that the physical nature of InP near-surface defect acceptor and donor levels may be different and that phosphorus vacancies are the cause of p-InP surface Fermi level pinning.

Near-surface defect profiling with slow positrons: Argon-sputtered Al(110).

Near-surface defect profiling with slow positrons: Argon-sputtered Al(110).

Two-Dimensional Modelling of Compressive Failure in Delaminated Laminates

An analytical model is developed to assess the compressive strength criticality of near-surface interlaminar defects in laminated composites.

Review of some recent advances in quantitative ultrasonic NDT

Nondestructive testing is moving from being an art to a quantitative science. The paper first reviews the ultrasonic NDT literature; this is followed by consideration of the contribution which can be made to the understanding of analytically intractable wave-defect interaction problems by the use of numerical modelling based on explicit finite-difference schemes. Examples are presented of a developing family of ultrasonic NDT techniques which extract additional information from the wave field by combining time-of-flight and mode conversion information with ultrasonic spectroscopy. Immersion schemes, based on leaky Rayleigh waves for surface and near-surface defects are considered. These techniques are shown to extend the scope of ultrasonic NDT with improved reliability of detection and the potential to provide quantitative defect characterisations.

Helium implantation in metals investigated by monoenergetic positrons

Abstract Helium-implanted metals and alloys were investigated with a positron beam system capable of producing positrons with energies ranging between 150 eV and 28 keV. Such a system is superior to the classical positron annihilation methods using high energy positrons from radioactive sources, since, for the first time, near-surface defects and defect agglomerates at high concentrations can be studied. Results are reported on crystalline nickel and amorphous nickel alloys irradiated at room temperature with helium up to doses of 1.8 × 1018 cm−2 at different energies. The Doppler broadening technique of the annihilation radiation is applied, and the lineshape parameter Ic is used to characterize the various defect configurations. In some favourable cases information about defect concentrations could be obtained.

Dynamic photoelasticity as an aid in developing new ultrasonic-test methods

A prerequisite for the development of quantitative ultrasonic-inspection techniques for surface flaws is a thorough understanding of the ways in which elastic waves interact with defects. Analytical and numerical approaches are presently inadequate. Experimental methods are needed for a better understanding of wave interactions with real geometries. This paper describes how dynamic photoelasticity was used to study the interaction between Rayleigh waves and slots. To fully interpret the interactions between an incident Rayleigh wave and a surface slot, the problem was subdivided as follows: first, the reflections and mode conversions of a Rayleigh wave at a corner were studied. This simulated the Rayleigh-wave interaction with a slot opening. Then, the interaction when a Rayleigh wave ran off the tip of a slot was observed, and, finally, the total interaction with slots perpendicular to the surface was studied. The results for these three cases are presented. It is suggested that the most important property of a Rayleigh wave that can be used to size surface and near-surface defects is the subsurface particle motions. These motions persist up to a depth of the order of a wavelength. The shape (that is, the frequency spectrum of the transmitted wave) should, therefore, be affected by the depth of the slot. Spectroscopic analysis is applied to the photoelastic data to develop a simple method for sizing slots. Results from ultrasonic tests on slots in steel confirm the validity of the suggested method. By applying contemporary concepts of signal processing to photoelastic data, a powerful new area of experimental investigation is introduced. It promises to overcome the current inability of scatter theories to predict the interactions between real-life defects and acoustic waves as used in ultrasonic testing. Applications of this approach will improve the quantitative ability of ultrasonic-inspection methods.

Methods for determining the depth of near-surface defects

In previous work by the authors,(1,6) it was demonstrated that the presence of near-surface defects could be detected reliably, even though the defect echo was contained within the near-surface echo. The algorithm consists of examining the variation in the composite (near-surface plus defect) response after it has been deconvolved from a near-surface response known to be defect-free. This paper presents two algorithms that have been developed subsequent to the work presented in ref. (6) for estimating thedepth of a near-surface defect, given that its presence has already been detected. One algorithm uses complex frequency domain techniques, and the other uses time domain analysis. Both procedures operate on the surface-plus-defect signal, using reference signals containing surface-only and defect-only responses. The defect signal is extracted from the composite signal. Defect depth is then computed from the time difference between the centers of the front-surface and extracted defect responses. A mean absolute depth error of 0.015 in. was obtained by applying the algorithms to experimental data containing depths from 0.020 to 0.130 in. below the near-surface.

Ion Milling of Magnetic Oxide Platelets for the Removal of Surface and Near-Surface Imperfections and Defects

Magnetic oxide platelets, useful for memory and logic circuits, are presently prepared by diamond abrasive polishing techniques. Processed platelets, however, usually contain thin damage layers and microscopic scratches resulting in increased domain wall coercivity. Platelets less than 1 mil in thickness and of desirable flatness are difficult to prepare using these techniques. Ion milling has been successfully used to rapidly remove abrasive and polishing induced defects and at the same time produce flat surfaces of much less than 1 mil in thickness. The measured removal rate was 0.5 μ/h/side; however 500-μ/h/side rates are attainable. Final platelets are scratch free, flat, and highly reflecting, having lower domain wall coercivity than diamond polished plates. No change in mobility was observed following ion milling. The process is adaptable to large areas and a wide variety of related applications.

Infrared Detection of Fatigue Cracks and Other Near-Surface Defects

This paper describes a noncontacting, nondestructive inspection system that can rapidly locate, in real time, the position and extent of fatigue cracks in aircraft and missile structures. The automated ir system described was developed for the USAF for use in their Sonic Test Facility at Wright-Patterson AFB. This facility is capable of subjecting large aircraft or missile structures to the intense sound fields experienced in actual flight or launch. In addition to presenting the basic concepts of the method, this paper (1) describes briefly the design of the system that was developed for the USAF, (2) discusses test results, and (3) presents conclusions and indicates future applications of the technique.

Occurrence and Mechanism of Rebinder Effects in CaF2

A study has been made of the influence of adsorbed species on the room-temperature mobility of near-surface dislocations introduced into freshly cleaved {111} CaF2 surfaces by a diamond indenter. It was found that, in general, adsorbed surface-active ions or molecules decrease dislocation mobility, i.e., increase microhardness. Studies of the relaxation behavior of near-surface dislocations in CaF2 exposed to solutions of DMF in DMSO revealed that these environments also can influence dislocation mobility in the absence of stresses imposed by a loaded indenter. Moreover, for any particular DMSO-DMF environment, a correlation exists between the extent of dislocation motion induced by a loaded indenter and the amount of relaxation which occurs after the indenter is removed. From these and other observations it is concluded that indenter-lubrication effects do not play a fundamental role in the occurrence of Rebinder phenomena in CaF2. The results of this work are consistent, however, with the view that Rebinder effects in ionic solids are associated with chemisorption-induced changes in the state of ionization of near-surface point defects and dislocations. Such changes introduce variations in their mutual interactions which are then manifested as environmentally induced changes in the mobility of near-surface dislocations, and hence in microhardness.

A practical introduction to magnetic particle testing: Basic rules for satisfactory detection

Magnetic particle testing is a means of testing iron and steel components for surface and near-surface defects such as cracks, laps and inclusions. It can provide information which makes it possible to avoid premature failures and to correct manufacturing faults such as too harsh grinding. It requires equipment which is comparatively cheap and is capable of testing items as small as a needle or as large as the Queen Mary to a high degree of sensitivity. It will almost certainly produce economies well above the running cost.

16. Gases and solids

The areal distribution of near-surface defects created by the implantation of O+ ions into 6H–SiC through chemically-deposited masks has been studied by reflection-geometry positron re-emission microscopy (PRM). PRM images were obtained for ion doses of 1×1013 and 1×1014 cm−2. The ratios of re-emitted positron intensities from unmasked and masked regions of the two samples were measured to be 78(3)% and 45(3)%, respectively. These results are consistent with the trapping of thermalised positrons prior to reaching the exit surface; the sharpness of images obtained from the low fluence sample is also consistent with work-function, rather than epithermal, emission from the sample. Removal of the instrumental resolution function from the image from the higher fluence sample shows evidence of lateral subsurface vacancy migration characterised by a broadening function of 44(5) μm F.W.H.M.