The implanted semiconductors need annealing-induced restoration of luminescent and electrical properties for further applications. However, it is still unknown how the effect of the extra-defect formation near the surface influences on the reduction process, as is the magnitude of this effect in GaAs. To confirm the effect and study its influence on luminescence properties of the implanted GaAs based semiconductors, the room-temperature bandedge photoluminescence (PL) spectra were measured before and after the removal of a near-surface layer similar to 100 nm thick for GaAs implanted with N+ ions at energies of 30-250 keV with fluence similar to 1 x 10(16) cm(-2), followed by annealing at 700 degrees C for several minutes. An increase in the intensity of the PL line by a factor of 1.3-1.6 was found in different samples, indicating the presence of a high concentration of radiation defects in the near-surface layer after annealing. It was shown that the density of the surface defects is several times higher than the maximum density of point defects in the bulk, calculated using the TRIM program and measured previously. Annealing was shown not to restore the surface part of the implanted GaAs layer and its PL. This fact was explained by the clustering of point defects and the oxygen atoms recoiled from native oxide followed by forming chemical bonds with native atoms, which leads to the creation of stable defects irremovable by annealing. The conclusion was made that to increase the PL of implanted GaAs, it is necessary to remove the surface layer.

The properties of a metal contact with a p-GaAs layer ~8 nm thick induced by low-energy Ar+ ions on an n-GaAs wafer as a result of the conduction tipe type conversion have been studied. The metal was deposited according to the standard technology on the surface of the semiconductor p-GaAs with a natural oxide layer, partially restored when the sample was transferred to a deposition setup. To prevent metallization of the nanolayer, the contact was not annealed. Therefore, a Schottky barrier emerged at the interface and a residual oxide layer retained. However, current-voltage characteristics showed that the formed contact is predominantly ohmic. It has been found that a high concentration of ion-induced defects radically reduces the width of the Schottky barrier and ensures the tunneling of holes and electrons of the semiconductor valence band through the barrier in the forward and reverse directions, respectively. It is shown that ion bombardment of the p-GaAs semiconductor surface makes it possible to obtain an ohmic contact with any metal without annealing. It is concluded that the ion-stimulated modification of the semiconductor and the exclusion of annealing make it possible to obtain a tunnel ohmic contact with an extremely thin p-GaAs nanolayer coated with a residual layer of natural oxide.

To elucidate the possibility of creating an ohmic contact with an extremely thin p-GaAs nanolayer, we have studied the properties of the contact deposited without annealing (“cold” contact) with a p-layer $\sim $ 8 nm thick formed on the n-GaAs wafer by low-energy Ar $^{+}$ ions due to the conductivity type conversion (n $\to $ p). Exclusion of annealing prevents metallization of the semiconductor nanolayer. Despite the obvious formation of the Schottky barrier and the presence of a residual oxide layer, the current–voltage characteristics of the ion-induced p-n structure indicates the ohmic nature of the contact. It is shown that high concentration of defects in the irradiated p-layer leads to a decrease in the barrier width down to the value ${W}$ = 0.3 nm, which is much smaller than the de Broglie wavelength of p-layer charge carriers ( $\lambda >$ 2–19 nm). Therefore, the ohmic character of the contact is provided by holes and electrons of the p-layer tunneling through the barrier. It is shown that the “cold” tunneling ohmic contact on p-GaAs can be formed with any metal if a defect or doping density is ${N}_{\text {D}} > 10^{20}$ cm $^{-{3}}$ , which is an order of magnitude higher than the value ( ${N}_{\text {D}} > 10^{19}$ cm $^{-{3}}{)}$ considered as providing efficient tunneling of carriers through a contact with heavily doped n-type semiconductors.

The dark current–voltage characteristics of a p–n structure created on the surface of an n-type GaAs wafer by Ar+ ions with the energy Ei = 2500 eV are studied. To avoid the metallization of a thin (~10 nm) ion-modified p-type layer, multilayer metal contacts are deposited onto both sides of the plate without subsequent annealing. The diode effect with a forward-to-reverse current ratio up to three orders of magnitude is observed in the voltage range as high as 0.7 eV contrary to the unirradiated reference sample. The linear dependence of most current–voltage characteristics of the unirradiated reference sample and coincidence of the experimental current–voltage characteristics with the calculated ones for the ion-induced p–n-structure indicate the predominantly ohmic nature of the metallic contacts and that the observed diode effect is determined by the ion-induced p–n structure. An analysis of two regions with different slopes in the current–voltage characteristics reveals two current-transport mechanisms: recombination and diffusion. The considered effect of the ion-induced formation of the p–n structure limits the use of low-energy argon ions for preparing an atomically clean surface of n-type GaAs-based semiconductors for their study by surface-sensitive methods because the p–n structure is formed on the surface instead of a homogeneous material.

The features of oxidation of the surface of GaAs irradiated by low-energy Ar+ ions is considered based on elemental and chemical-composition analyses, calculations of the concentration profiles for radiation-induced defects, and estimations of radiation-enhanced diffusivities and diffusion lengths. The native oxide layer is revealed to be highly enriched with Ga (by a factor of 1.5) due to the radiation-enhanced diffusion of elemental arsenic through vacancy defects even at room temperature. Elemental arsenic emerging at the interface with the oxide layer moves to a deeper radiation-damaged layer and fills vacancies there. At irradiation doses of Q > 3 × 1014 cm–2, which are sufficient for removal of the oxide layer with 3-keV Ar+ ions, elemental arsenic leaves the oxide layer within one hour, and the diffusion length reaches the thickness of the radiation-damaged layer within one day. The total number of vacancies in the radiation-damaged layer is enough to absorb all elemental arsenic formed during oxidation. The considered radiation-enhanced diffusion can be used to remove elemental arsenic, which is known to form nonradiative recombination centers quenching the luminescence of the underlying bulk layer, from the oxide layer.

The electronic structure and chemical composition of the n-GaAs surface after implantation of N2+ ions with energy Ei = 3000 eV and fluence Q - 3 x 1015 cm-2 were studied by synchrotron-based X-ray photoelectron spectroscopy to clarify effects of low-energy nitrogen ion implantation on A3B5 semiconductor surfaces. Conversion of the conductivity type and creation of a p-n structure on the n-GaAs surface were revealed under N2+ ion irradiation. The conductivity type transformation was shown to occur due to pure mechanical action of nitrogen ions, resulting in formation of Ga anti-site acceptors. The - 10 nm-thick p-layer obtained consisted of concentrated GaAs1-xNx (x - 0.1) alloy whose bandgap width is known to be essentially narrower compared to the pristine GaAs semiconductor. Therefore, the structure formed is a nano-heterosctructure incorporating the semiconductor alloy whose properties are considered to be attractive for infrared applications. The obtained GaAsN electron binding energies (EB (N1s) = 397.0 eV, EB (Ga3d) = 19.42 eV and EB (As3d) = 41.25 eV) can be used for identification of the GaAsN alloy phase. An approach to 3D p-n structures formation was suggested using nitrogen ion beam without wet lithography.

A number of mechanisms of the diffusion of arsenic atoms including the radiation- stimulated vacancy, interstitial, and mixed (Frank—Turnbull mechanism) types are considered to explain the earlier discovered ion-stimulated transformation of a layer of native GaAs oxide into a layer of Ga2O3 at room temperature. An estimate of the diffusion coefficients and lengths makes it possible to conclude that the interstitial diffusion mechanism dominates at a fluence Q < 1015 cm–2. It is found that at room temperature, when the interstitial mechanism is implemented, the diffusion coefficient (DAs ~ 1.3 × 10–16 cm2/s) and the diffusion length (L > 9 nm) are sufficient to remove elemental arsenic formed under the action of argon ions from a layer of native oxide with a thickness of 2.0–2.5 nm in 10 minutes. However, the contribution of the vacancy mechanism increases with increasing irradiation dose due to an increase in the concentration of vacancies. In this case, the diffusion mechanism becomes mixed. At a fluence of Q > 1015 cm–2, the vacancy mechanism provides a diffusion coefficient (DAs ~ 0.7 × 10–17 cm2/s) and diffusion length (L > 2.5 nm) also sufficient to remove elemental arsenic from the oxide layer within 10–20 min. It is shown that the diffusion of arsenic can be significant in the processes of chemical modification of GaAs oxides even at room temperature.

Cleaning the n-GaAs surface with low-energy Ar+ ions, required in X-ray photoelectron spectroscopy (XPS), has been recently shown to drastically change the core-level binding energies (BEs) of the irradiated surface layer, which prevents the diagnostics of the n-GaAs based semiconductors by the ordinary XPS. Synchrotron-based XPS measurements and modeling of As3d and Ga3d photoemission spectra for an Ar+ - etched n-GaAs wafer made it possible to find conditions for XPS testing the unmodified deep n-GaAs bulk. XPS resolution should be better than 0.5 eV and Ar+ ion energy should be less than 0.5 keV. Control of the Ga3d and As3d BEs in unmodified n-GaAs is important for chemical analysis since they are close to Ga2O3 and elemental arsenic BEs, respectively, which appear in oxidation, nitridation and other important chemical processes.

Electronic structure and chemical composition of GaAs-based semiconductors are considered to be well characterized by photoelectron spectroscopy and other surface-sensitive methods. However, effect of ion irradiation on electron binding energies and band structure of semiconductors can drastically distort results of diagnostics. This effect was studied by synchrotron-based XPS applied to n-GaAs wafer after 1250 eV Ar+ ion exposure with fluence Q similar to 1 x 10(15) ions/cm(2) which is typical for preparation of atomically clean surfaces. Mechanical action of ions was shown to change the n-GaAs electron binding energies by the value of the bandgap width due to creation of defect states, and conversion of the conductivity type from n to p. The Ga3d and As3d core-level binding energies for p-GaAs and n-GaAs were measured in one experiment on the p-n plane structure formed due to irradiation: E-B(p/n) = 19.3/20.4 eV and E-B(p/n) = 41.3/42.4 eV. The p-layer nanothickness was determined and the band diagram of the p-n GaAs structure was constructed. The revealed effect may be comparable with core level chemical shifts and should be taken into account to avoid mistakes in the XPS chemical composition diagnostic of GaAs-based semiconductors.

The excitation spectrum of GaAs has been studied by reflection electron energy loss spectroscopy. In addition to dominant collective excitations, a series of single-electron transitions of a core Ga 3d electron to previously unknown unoccupied states located above the Fermi level by 1.25, 3.7, and 6.8 eV have been detected in the spectrum. It has been shown that the detected states appear near the ion core of Ga because of an increase in its effective charge at excitation. Since the detected electronic levels are equidistant, they can be described by a subnanometer spherical quantum dot model. It has been shown that one of the decay channels of the detected states involves the emission of ultraviolet radiation.

A photovoltaic effect was observed under exposure to 810-nm laser light on the atomically clean surface of an n-GaAs wafer etched with Ar+ ions: open-circuit voltage in current–voltage (J–V) light characteristics was as high as 47 mV. The effect is due to the formation of a p–n structure under Ar+ ion bombardment in the near-surface bulk layer (~ 7 nm thick) via the conversion of the conductivity type of the near-surface layer from n to p. The effect of the conductivity type conversion under pure mechanical ion action manifested itself in that the Fermi level approached the valence band top, which was detected by high-resolution photoelectron spectroscopy with the use of a synchrotron radiation. The formation of the p–n junction in the structure was confirmed by dark J–V characteristics, which show a diode effect at voltages in the range (0–0.6) eV with a direct-to-reverse current ratio as large as 103. The results obtained show that the near-surface bulk properties of n-GaAs treated by the most widely employed surface cleaning procedure via etching with a beam of Ar+ ions differ dramatically from the pristine deep bulk.

The thickness, elemental and chemical compositions of the native oxide naturally formed on a perfect GaAs(1 0 0) crystal grown by MBE have been studied by Auger electron spectroscopy (AES) and electron energy-loss spectroscopy (EELS) to specify the oxidation mechanism and to confirm or reject the questionable presence of elemental arsenic in the natural oxide. Elemental arsenic (As degrees) arising in the oxidation process due to reduction of As2O3 by the GaAs substrate was revealed at the Auger energy of 1225.8 eV and shown to reach similar to 16 at% of the oxide whose thickness was determined to be similar to 4 nm. As degrees was shown by EELS to form a segregate with the plasmon energy of 18.1 eV. Room temperature oxygen diffusivity through the finally formed oxide layer was estimated to be low enough (D < 10(-21) m(2)/s) to explain retention of the As degrees in the deepest oxide layers where it cannot oxidize again or diffuse away. The studied natural oxide was shown to have a layered nanostructure consisting of a cap As2O3 layer (similar to 1 nm), a core oxide layer containing a Ga2O3/As2O3 mixture (similar to 2 nm) enriched in Ga by a factor of 1.5, and a deep interface layer of elemental arsenic (similar to 1 nm).

The core-level and valence band electronic structure of the n-GaAs (100) has been studied by synchrotron-based high-resolution photoelectron spectroscopy after irradiation by an Ar+ ion beam with energy Ei = 1500 eV and fluence Q = 1 × 1015 ions/cm2. Conversion of the conductivity type of the surface layer and formation of a p–n structure have been observed. The p-surface layer thickness (d ~ 5.0 nm) and band structure were experimentally determined from the Ga3d photoelectron spectrum by separation and analysis of the low intense n-type bulk contribution from deeper layers. A band diagram of the p–n junction formed on the n-GaAs surface by Ar+ ion bombardment was reconstructed. The p–n junction proved to be unexpectedly narrow compared to the extended tail of the implanted ion depth distribution.

Oxidation specific of the defected GaAs has been considered on the basis of elemental and chemical composition study of the oxide layer naturally emerged on the GaAs surface strongly irradiated by Ar+ ions with energy Ei = 3000 eV and fluence Q ~ 3 × 1015 cm–2. The diffusivity of elemental arsenic known to form an interface layer was shown to increase at room temperature by more than 35 orders of magnitude due to radiation defects and to amount to the value D ~ 1 × 10–17 cm2/s. Efficient room temperature diffusion results in total removal of elemental arsenic from oxide into the bulk, thus curing the damaged substrate.

The concentration profiles of defects produced in structures upon the implantation of nitrogen ions into GaAs epitaxial layers with an uncovered surface and that covered with an AlN film and subsequent annealing are studied. The ion energies and the implantation doses are chosen so that the nitrogen-atom concentration profiles coincided in structures of both types. Rutherford proton backscattering spectra are measured in the random and channeling modes, and the concentration profiles of point defects formed are calculated for the samples under study. It is found that the implantation of nitrogen ions introduces nearly the same number of point defects into structures of both types, and the formation of an AlN film by ion-plasma sputtering is accompanied by the formation of an additional number of defects. However, the annealing of structures of both types leads to nearly the same concentrations of residual defects.

A highly defective ~10-nm-thick layer was fabricated in a high vacuum by 2.5 keV Ar+ ion bombardment of the n-GaAs surface. Valence band photoelectron spectra showed a p-type conductivity of the layer arising due to the high concentration of mechanically created point defects (p-centers). J–V characteristics measured ex situ for the structure consisting of the irradiated p-layer on the n-type substrate revealed a diode effect. Analysis of the data attributes the effect to the formation of a specific p–n junction. Thereby, we demonstrated that Ar+ ion bombardment of the n-GaAs surface results in that a nanostructure with the p–n junction properties is formed. The p–n junction under consideration seems to deserve further study and possible application since it can be formed in high-vacuum clean conditions directly by exposure to a low-energy Ar+ ion beam without wet lithography.

n approach to solving the problem of the in situ bandgap determination in the extremely thin and chemically active nitride nanolayers fabricated in high vacuum on the n -GaAs surface has been suggested. The approach is based on measuring the interband transitions involving the quantum well states by the method of electron energy loss spectroscopy. The bandgap of the nitride layer formed on the GaAs surface by low-energy N_2^ + ion implantation was determined to be 0.2 eV less than that of GaAs, which evidenced for creation of the GaAsN dilute alloy on the GaAs surface.

The core-level and valence band electronic structure of the well-defined near-surface layer of n-GaAs (100) has been studied by synchrotron-based high-resolution photoelectron spectroscopy before and after modification of the layer by an Ar+ ion beam in the 1.5 - 2.5 keV energy range. Conversion of the conductivity type from n into p has been revealed in the irradiated layer several nm thick. The effect manifests itself in shifts of the core-levels and valence band edge by the value comparable to the bandgap width. Transformation on the conductivity type has been assumed to be caused by Ga-antisite point defects generated by ion bombardment. The possibility of local formation of a p-n nanojunction within the ion-beam spot has been shown.

The elemental and chemical compositions throughout the thickness of the GaAs native oxide layer slightly irradiated by Ar + ions have been studied by synchrotron-based photoelectron spectroscopy at different photon energies enabling variation of probing depth. The presence of only two phases was observed: of the gallium oxide Ga 2 O 3 and elementary arsenic As o generated due to complete decay of arsenic oxides under the ion irradiation. Depth composition profiles were determined nondestructively. Despite inhomogeneous depth distribution, these profiles demonstrated domination (90 at %) of the dielectric Ga 2 O 3 phase virtually throughout all the oxide thickness (~2 nm).

Secondary-ion mass spectrometry and Rutherford proton backscattering have been used to measure the concentration profiles of nitrogen atoms and examine the defect structure of epitaxial GaAs layers implanted with 250-keV N+ ions at doses of 5 × 1014–5 × 1016 cm–2. It was found that no amorphization of the layers being implanted occurs at doses exceeding the calculated amorphization threshold, a concentration of point defects that is formed is substantially lower than the calculated value, and a characteristic specific feature of the defect concentration profiles is the high defect concentration in the surface layer.