The damage to GaAs crystals caused by helium ion implants has been monitored by changes in the Raman scattering phonon modes, double-crystal x-ray diffraction rocking curves, photoreflectance (PR), and electron beam electroreflectance (EBER) band edge transitions. As the implanted helium ion dose was increased, the various techniques revealed threshold damage behavior at very different levels. Although PR and EBER were the most sensitive to the defects created at the lowest ion doses, all techniques indicated substantial disorder for implants greater than 1014 ions/cm2.

Photoconductivity measurements have been made on three SI GaAs sample types: Ar+ ion etched at leV, 3keV, and unetched. Measurements were made versus time, wavelength, and temperature. Photoquenching was done at 1.19 and 2.0eV energy.

Some distinct changes in photoconductive properties are caused by the ion etch: 1) Increase in dark resistance; 2) presence of persistent photoconductivity; 3) increased photosensitivity; 4) increase In quenching rate under 2eV illumination; 5) destruction of the 0.47eV thermally activated photoconductivity.

All samples exhibit Shockley-Read recombination controlled photoconductivity below a temperature of 125K, with the same apparent trap location, at 0.26eV above EF.

Si atoms were recoil-implanted into GaAs by bombarding neutral (As+) or dopant (Si+) ions through a thin Si cap. The bombarded samples were subsequently rapid thermally or furnace annealed at 815–1000°C in Ar or arsine ambient. The presence of the recoiled Si in GaAs and resulting ni-doping was confirmed by secondary ion mass spectrometry and Hall measurements. It was found that sheet resistances of < 150 Ω/‪ can be achieved by this method. Capless furnace annealing in arsine ambient generally yielded better electrical results (especially for shallow implants, i.e., < 100 nm deep) compared to those obtained by RTA in an inert ambient with a Si cap. In the latter case, electrical activation deteriorated above 900°C due to high As loss and the deterioration was pronounced for shallow implants. Significant Si redistribution occurred during arsine annealing whenever the Si concentration (from recoil or direct implant) in GaAs exceeded 1×1019cm3 and the annealing temperature was > 850°C. Our present electrical data show that the recoil implant method is a viable alternative to direct shallow implant for n+ doping of GaAs.

Molecular ion S2+ and SiFn+ implantations into GaAs have been investigated to form very thin active layers. After implantation, the transient annealing (TA) and furnace annealing (FA) were used. The measurements of activation efficiency, mobility, carrier concentration profiles and PL spectra were carried out. The experiments show that after TA, the activation efficiency, mobility and carrier distribution are almost the same between samples implanted with S+ at an energy of 50KeV to a dose of 3×1013cm−2 and S+2 at 100KeV to 1.5×1013cm−2. It shows that the damage of S2-implanted samples can be removed by TA, and a very thin active layer can be formed by the implantation of S2+ at 50KeV. For SiFn-implanted samples, the activation efficiency and mobility. decrease with increase of the implanted ion mass. As+ co-implantation into SiF-implanted samples has been used to improve both activation efficiency and mobility. After comparison with the properties of the SiFt implantation, S2+implantation is more acceptable to form thin active layers.

As+ implantation has been used to compensate the As loss during annealing for Si- or S-implanted GaAs. The As+ implantation was at an energy range of 25–100KeV to a dose range of 1012-1015cm−2. The Si+ or S+ implantation was at 150 or 50KeV respectively to the same dose, 27times;1013cm−2. By using the measurements of the Hall effect and sheet resistivity, the activation efficiency and mobility of the dually implanted sample with Si+ and As+ (25KeV, 1×1013cm− 2 ) after TA at 1050°C for 6s increase by 21 and 18% respectively over those of the singly Si-implanted sample after TA at 1100°C for 6s. Those of the dually implanted sample with Si+ and As+ (100KeV, 1×1013cm− 2 ) increase by 63 and 30% respectively over those of the singly Si-implanted sample after FA at 800°C for 30min. Also, the density of SiAs and VGa determined by 20K PL spectra decreases of dually Si- and As-implanted samples. The activation efficiency and mobility of the d8ally implanted sample with S+ and As+ (40KeV, 1×1013cm−2 ) after TA at 1100°C for 12s increase by 17 and 16% respectively over those of the singly Simplanted sample. The annealing behaviour after As+ implantation is discussed.

Annealing of Beryllium implantations into epitaxial InP/InGaAs hetero-structures was investigated. Different ion energies were used to position the profile maximum in the InP cap, at the hetero-interface, or in the underlying InGaAs layer. By measuring the Be atom and carrier profiles, it is shown that annealing conditions necessary for optimum Be activation in the InP cap layer are compatible with good profile control for Be in InGaAs. There is a slight transfer of Be across the interface during annealing, which is important, if the pn-juntion is intended to be close to the hetero-interface.

Transmission electron diffraction patterns of cross-sectional TEM samples of OMVPE ZnSe on GaAs indicate the existence of the hexagonal wurtzite phase in the epitaxial layers. The orientation relationship is (0002)//(111); (1120)//(220). Etching studies indicate the phase is internal not ion milling induced. The average wurtzite particle size is 80Å-120Å. Because of interplanar spacing matches it is easily overlooked. Electrical property measurements show a high resistivity (1010ω/square) which drops by four orders of magnitude upon rapid thermal annealing between 700°C and 900 °C for 3 sec. Implantation of Li and N have little effect on the electrical transport properties. The Li is shown to have a high diffusivity, a solid solubility of ≈1016/cm3 at 800°C and getters to the ZnSeA/aAs interface.

Si29 implants have been performed in InGaAs and InAlAs single layers as well as in InGaAs/InAlAs heterostructures. Nearly 100% activation has been obtained in InGaAs after conventional furnace annealing or rapid thermal annealing. On the other hand, a low activation efficiency (30%) has been observed in InAlAs. A preliminary photoluminescence measurements study shows that an appreciable fraction of Si exists in the form of complex centers. Hall effect and specific contact resistivity maps carried out in the heterostructures indicate a very good homogeneity of the electrical parameters over a 4 cm2 sample and give a satisfactory value of the specific contact resistivity (10−7 ω.cm2)

The doping efficiencies obtained in phosphorus-implanted cadmium telluride and copper indium disulfide annealed by pulse electron beam were higher than those annealed by conventional thermal method. To get insights into this phenomenon, electron paramagnetic resonance measurements were performed for both crystals at various stages during the doping process. The results indicated that the pulse electron beam annealing could effectively eliminate the phosphorus interstitials in the implanted crystals but the thermal annealing could not. This shows the significant effect of melting crystals by pulse electron beam annealing to obtain high doping efficiencies.

The structures and magnetic characteristics of multilayered samples with alternatively deposited metal layers (Nd and Fe) were studied for 80 keV boron ion implantation to various doses. Boron ion implantation induced amorphization in Nd/Fe multilayers gradually with increasing irradiation dose. The hysteresis loops of magnetization of the samples showed that a sunken curve appear-ed in the second and fourth quadrants of the loops for the depo-sited Nd/Fe films and disappeared when implantation dose was more than 4*1016 B+/cm2. During ion irradiation, the coercive forces of the films were decreased. The saturation magnetization and the residual magnetization decreared first and then increased with increasing ion dose, the squareness ratio S. R increased first and then decreased. Discussion was carried out in this paper.

A key issue in modern ion implantation processing is the requirement for dramatic improvements in the purity of the incident ion beam and reductions in the deposition of foreign materials onto the wafer surface. These deposited materials include particles as well as sputtered and vapor deposited metals and dopants. Physical mechanisms which effect the elemental purity of atoms arriving at the surface of ion implanted wafers and progress towards achieving implantation purity levels of below 100 ppm of the implanted dose for sputtered metal and dopant films are discussed.

We have implanted a number of insulating polymer (PET) thick films with fluences in the range 1 × 1013 to 5 × 1017 ions/cm2 in order to induce electronic activity. Finite electrical conductivities were obtained for fluences as low as 3 × 10 15ions/cm2. The resistivity for these implanted materials increases with decreasing temperature demonstrating highly disordered or semiconducting behavior. Hall effect measurements confirm this result and indicate negatively charged carriers with a carrier density of 1.44 × 10 17cm -3 for an implant dose of 1 × 1017 ions/cm2. The conductivities also seem to be dependent on the implanted ion species. We believe this material will be useful as an accurate temperature sensor near room temperature.

The distribution of Mg+ (a p-type dopant for GaAs) and As+ (an p-type dopant for Si) implanted into both photoresist (PR) and polyimide (PI) have been determined experimentally. Range data of Mg ions at 200 keV and 300 keV and As ions at 150 keV have been measured by Secondary Ion Mass Spectroscopy (SIMS). SIMS values for the projected range Rp and the standard deviation ARp were compared to range profile data calculated using the Projected Range Algorithm (PRAL) of Biersack [1] as well as the standard LSS theory [2]. While the values for Rp calculated from the PRAL model generally agreed within 10% of the SIMS values, the calculations underestimated Rp for PR but were in good agreement for PI. The LSS calculations underestimated Rp in both materials.

We have used high-dose metal ion implantation to ‘fine tune’ the composition of Y-Ba- Cu-O thin films. The films were prepared by either of two rf sputtering systems. One system uses three modified Varian S-guns capable of sputtering various metal powder targets; the other uses reactive rf magnetron sputtering from a single mixed-oxide stoichiometric solid target. Film thickness was typically in the range 2000–5000 A. Substrates of magnesium oxide, zirconia-buffered silicon, and strontium titanate have been used. Ion implantation was carried out using a metal vapor vacuum arc (MEVVA) high current metal ion source. Beam energy was 100–200 keV, average beam current about 1 mA, and dose up to about 1017 ions/cm2. Samples were annealed at 800 – 900°C in wet oxygen. Film composition was determined using Rutherford Backscattering Spectrometry (RBS), and the resistivity versus temperature curves were obtained using a four-point probe method. We find that the zero-resistance temperature can be greatly increased after implantation and reannealing, and that the ion beam modification technique described here provides a powerful means for optimizing the thin film superconducting properties.