Sputtering yields and angular distributions have been measured as functions of sample preparation techniques and incident ion-beam orientation with respect to the crystal axes for 100 keV Cu-ion beams on Cu crystals and polycrystalline samples. The angular distributions have structure requiring an nth order cosine with two Gaussians superimposed to fit the data; strong peaking is observed near the backscatter direction. The yield is dependent on the beam to crystal and beam to polycrystalline-rod axis orientation, on the grain size of the polycrystals and on sample-preparation techniques. Yield measurements vary by as much as a factor of 4. Lattice-damage differences, measured with alpha particle channeling, are much smaller and seem to be saturated by fluences of the order of 1 × 1016/cm2.

The sputtering yield and the angular distribution of sputtered atoms have been measured as a function of copper-ion-beam orientation with respect to copper single-crystal targets. Alpha-particle channeling has been employed to measure the lattice damage induced during the sputtering experiments. The experimental results have been compared with results from the cascade simulation program MARLOWE. Both planar and axial channeling of incident ions have been shown to significantly reduce the sputtering yield relative to a psuedo-random irradiation direction. Channeling of the incident particles also increases the depth to which lattice damage is produced. Surface preparation is found to strongly influence the orientation effects, with mechanical polishing reducing the channeling of incident ions, relative to electropolishing.

The sputtering yield, joint polar and azimuthal angular distributions and the energy distribution of atoms sputtered from copper single crystals by 100 keV copper ions has been calculated as a function of crystal orientation using the cascade simulation code MARLOWE. The calculated yields and angular distributions are compared with experimental results for ions incident parallel to a <110> close-packed axial channelling direction, parallel to a <100> plane, and along a pseudo-random direction. Significant channeling effects are exhibited in both calculations and experiment. Strong peaking in the backscattered direction is calculated and observed for the axial irradiations alone. Primary range distributions and damage profiles are calculated for the three crystal orientations and compared with lattice damage measurements. The distributions of sputtered ions in both energy and angle are presented.

Copper targets were implanted with 125 keV gold and tantalum ions at selected incident angles of the beam relative to the target normal to examine the effect of the incident angle on the sputtering yields, the dose retained in the modified layer, and the atomic surface fractions. Sputtered particles were collected on aluminum foils to measure the angular distribution of the sputtered particles and to determine the partial sputtering yields of the beam and target species. A 1.5 MeV xenon beam implanted markers into the copper targets to determine the average sputtering yield for the surface during the implantation. Theoretical predictions from a surface binding energy correction model were compared with the experimental results. Although the predictions from the model were not exact, the trends of the data were in agreement with the model. The measured total sputtering yield increases with increasing incident beam angle, and, correspondingly, the fluence retained decreases with incident beam angle.

A physical model is presented for calculating infrared reflection interference spectra from ion implanted and annealed crystalline materials. The utility of the method is illustrated by presenting best fit spectra for a <111> silicon sample implanted with 2.7 MeV phosphorous to a fluence of 1.74 × 1016 ions/cm2 and isothermally annealed at 500°C. Non-linear least-squares fitting of reflection data yields structural and electrical information about the implanted region with reasonable precision. The physical quantities determined are (i) the depth of the amorphous layer produced by implantation both before and during isothermal annealing, the thickness of the recrystallized material, and the widths of any transition regions, (ii) the dielectric properties of the amorphous and recrystallized material, and (iii) the characteristics of the free carrier plasma which yield the carrier density profile, the mobility near the carrier density maximum, and the carrier activation efficiency. Up to nine fitting parameters are necessary to describe these physical quantities. A critical discussion of the sensitivity of data fit to variation in the parameters is given to establish the uniqueness of fitted parameters. The infrared method is non-destructive, is applicable to other dopants and semiconductors, and provides information complementary to both ion channeling and resistivity profiling techniques.

The present experiments were performed to investigate the effect of a factor of 100 difference in damage dose rate on /gamma/sub p/rime/-precipitates in Ni-6.84Al. Prior to irradiation, the alloy was solution-treated and aged to produce /gamma/sub p/rime/-precipitates with a mean cube edge of 33.7 nm. Specimens were examined by transmission electron microscopy following irradiation with 2.8 MeV/sup 58/Ni/sup +/ ions to peak damage doses of 0.81, 2.5 and 8.1 displacements per atom (dpa) at an irradiation temperature of 725 degree C for damage dose rates of 4.4 X 10/sup -4/ dpa/s. Analysis of the observed behavior by exisitng theories in the literature indicated that all of the effects noted could not be explained by any single, currently available theory. Satisfactory agreement was found with a treatment based on a combination of several previously proposed growth, dissolution, and reprecipitation mechanisms.

Single crystal silicon has been implanted with nitrogen and phosphorus ions at MeV energies to fluences between 1016 and 1.6 × 1018 ions/cm2. Infrared transmission and reflection spectra in the range of 1.25 to 40 μm were measured for as-implanted samples and after various annealing treatments. Interference fringes were observed in the IR spectra which are produced by the interference of light which has been multiply reflected between the front surface and the buried layers. By detailed theoretical analyses of the interference fringe structure, we obtained refractive index profiles, which, under suitable interpretation, provide accurate measurements and several quantities of interest. These quantities are the range and straggling of the implanted ions, the depth of disordered layers, and the width of the order-disorder transition. Mechanisms for the refractive index changes which have been identified include amorphization of the implanted silicon, bulk compositional change in the buried layer, localized vibrational mode dispersion, and free electron dispersion. Experimental results and theoretical predictions are presented, demonstrating each of these mechanisms.

Single-crystal silicon has been implanted with nitrogen ions at MeV energies, to fluences between 0.25×1018 and 1.65×1018 ions/cm2 at a substrate temperature of 700 °C. Infrared transmission and reflection spectra in the range of 1.25–40 μm were measured and interference fringes were observed which are produced by the interference of light which has been multiply reflected between the front surface and the buried layers. By detailed theoretical analyses of the interference fringes we obtain refractive-index profiles, which, under suitable interpretation, provide accurate measurements of the range and straggling of the implanted ions. Rutherford backscattering measurements on the same samples confirm this interpretation. Between the energies of 0.67 and 3.17 MeV, the measured values of the projected range agree with theory after adjusting the electronic stopping power, but the straggling measurements are lower by ∼30%. It is demonstrated that the asymmetry of the range distribution can be measured with this technique as well.

Void and dislocation structures in an Fe-25Ni-15Cr alloy were studied following irradiation with 2.8 MeV58Ni+ ions at temperatures between 600 and 750°C (1112 and 1382°F) to maximum damage levels up to 80 displacements per atom (dpa). Void formation was observed at all the temperatures investigated, with the maximum swelling between 650 and 700 °C (1202 and 1292°F). The swelling versus dose relationships exhibited an incubation dose followed by swelling at a rate that increased with increasing damage level. These data were consistent with previous swelling results for austenitic alloys irradiated with charged particles, which indicate that the swelling should become linear with irradiation dose at higher damage levels. Tangled dislocation networks were observed to form at low doses and to be fairly stable up to the highest damage levels examined. With the assumption of the observed stable dislocation networks, the dose dependence of swelling could be explained by a general form of the chemical rate theory for swelling due to void growth.

Nickel samples have been irradiated at two different dose rates with the same dose of 2.8-MeV 58Ni+ to a peak damage level of 13 dpa. The peak damage dose rates were ∼7 × 10−2 dpa sec−1 (HDR) and ∼7 × 10−4 dpa sec−1 (LDR). An upward shift of ∼75°C in the low-temperature limit for swelling has been observed for this increase of a factor of 100 in dose rate. The HDR results showed a peak swelling of 1.2% at 625°C, while the LDR results gave a peak swelling of 2.4% at 550°C. The LDR samples contained larger voids at lower densities than the HDR samples at all temperatures at which voids were observed. The results were compared with models for void nucleation and growth from which it was concluded that the HDR and LDR swelling curves both exhibited a low temperature regime which was void- growth-limited and a high temperature regime which was void-nucleation-limited. When temperature independent void and dislocation parameters were employed, the Brails ford and Bullough (BB) void growth model predicted that both LDR and HDR swelling would extend to much higher temperatures than was observed. When the experimentally measured parameters were employed at each point, the temperature dependence of the calculated swelling rate and that of the observed swelling agreed well for the LDR data but the resulting curves differed by ∼50°C for the HDR results. Thus, the BB theory has little quantitative predictive capacity because at present the temperature and dose rate variations of the void and dislocation parameters must be determined experimentally and even the use of these data does not necessarily give quantitative agreement. The various BB temperature shift models similarly lack quantitative predictive value since similar void and dislocation parameters enter these expressions. The temperature dependences of the void number density, the mean void diameter, and the dislocation density were only qualitatively similar when compared for the two dose rates, and a simple temperature shift did not produce reasonable coincidence between the values of these quantities obtained at the two different dose rates. This direct measurement of dose rate effects in nickel-ion-irradiated nickel indicates that, for the present, experimental measurement of dose rate effects appear to be needed in a given material before reasonably accurate predictions of swelling behavior as a function of damage dose rate can be made; further evaluation and improvement of theories of swelling is called for.

Swelling in reactor materials can be simulated, at least qualitatively, by irradiation with heavy ions in a small fraction of the time required with neutron irradiation. However, the 103−105 higher production rate of initial damage which makes simulation feasible may also obscure subtle microstructural changes. Thus, the dose rate effects need to be understood. The present investigation obtained a quantitative comparison between the damage parameters in 2.8−MeV nickel−ion−irradiated nickel, where the nickel ion flux was the only variable, being changed by a factor of 100. Preliminary results of studies of some of the samples of this investigation were described at a recent conference.1 The paper presents further analyses of these and other samples and comparison with theories of void nucleation and void growth.