Search

Discharge current transient spectroscopy (DCTS) is a promising technique for detecting the trap level and density in dielectrics because it is based on a simple emission process. In order to confirm the validity of DCTS, we compare the results from the conventional deep-level transient spectroscopy (DLTS) technique for samples of CVD-grown silicon nitride (SiNx) films. Results indicated that a trap level, about 0.6 eV below the energy level of the conduction band edge in the SiNx thin films estimated by DCTS is in good agreement with that obtained from DLTS analysis, and it is found that the trap density increases with the decreasing N/Si ratio in the SiNx film. As a proposed estimation for energy level in defects, it will be originated from hydrogen-incorporated defects in the SiNx matrix. This study demonstrates that the DCTS method will be a useful electrical method for the evaluation of defects.

The energy level of electron traps in silicon nitride (SiN x ) thin films was investigated by discharging current transient spectroscopy (DCTS). Results indicate that the trap level of the SiN x thin films becomes deeper with decreasing composition (N/Si) and shallower after hydrogen annealing. The dependence of the trap level on the SiN x composition and the modulation of the trap level by hydrogen annealing are possibly related to the change in the number of Si–H bonds in the SiN x thin films.

Electrical characteristics and an origin of traps in silicon nitride film were investigated using discharge current transient spectroscopy (DCTS), electron spin resonance (ESR), and photoluminescence (PL). It was found from DCTS that the density of traps of 1.1 eV from the conduction band edge increased with the increase of the gas-mixture ratio of SiH2Cl2 over NH3 for the fabrication of silicon nitride film. The energy levels of traps in silicon nitride film of 0.25, 0.68, 0.83–0.91, and 1.25–1.3 eV from the conduction band edge were found from photoluminescence observation. The ESR signals of the K-center that indicated a silicon dangling-bond increased with the increase of the gas-mixture ratio of SiH2Cl2 over NH3. It is suggested that the origin of traps of 1.1 eV from the conduction band in silicon nitride film is the silicon dangling-bond.

The gettering behaviors of Fe into Si with and without a p+ layer are investigated by deep-level transient spectroscopy. The samples contaminated with Fe in a wide concentration range were annealed at 600 °C to induce gettering. The surface-layer gettering behaviors of Fe for the sample without the p+ layer strongly depend on the Fe contamination level, in which the surface-layer gettering is not effective for the sample with low-level contamination at a concentration of less than 1 ×1013 cm-3 but effective for the sample with middle-level contamination at a concentration of (1–5) ×1013 cm-3. In contrast, the samples with the p+ layer show effective gettering for low- and middle-level contaminations. The gettering mechanisms of Fe in Si without and with the p+ layer are discussed in detail.

High-resolution photoinduced transient spectroscopy (HRPITS) has been employed to investigate iron-related point defects produced by electron irradiation in high-resistivity silicon. The measurements revealed 16 traps with activation energies ranging from 14 to 530 meV. From the comparison of results, obtained by the HRPITS and ESR measurements performed on the same samples, a trap positioned at Ev+500 meV was attributed to the iron-vacancy pair (Fei)V and that positioned at EV+530 meV to the vacancy complex with two iron atoms (Fei)2V. The former is found to be stable at 400°C, while the latter anneals out at this temperature similarly to the trap at EV+410 meV related to interstitial iron. As a result of the transformation of the (Fei)2V complex and reactions of the interstitial iron with native defects, new iron-related traps with activation energies of 85, 220 and 380 meV are formed.

The rapid degradation of Se-doped GaAs0.6P0.4 light-emitting diodes is of prime importance in reliability studies of optical devices using III–V compound semiconductors. Marked changes are observed at the junction of Se-doped GaAs0.6P0.4 light-emitting diodes, because they degrade rapidly, in contrast to Te-doped GaAs0.6P0.4 light-emitting diodes, which degrade gradually. Here, this rapid degradation is investigated using aging tests under various conditions by a capacitance–voltage method and deep-level transient spectroscopy. We discuss the differences in the degradation mechanisms for the two diodes and propose a model for the rapid degradation.

The defect states of an electron-beam irradiated GaN epilayer were characterized by capacitance–voltage (C–V) and deep-level transient spectroscopy (DLTS) measurements. The electron irradiation was performed using 1 and 2 MeV energy beams with 1 ×1015 and 1 ×1016 cm-2 doses, respectively. The depth penetrated in the GaN epilayer by 1 MeV electrons appeared to be about 450 to 600 nm from the C–V measurement. Five defect states with activation energies of 0.34, 0.49, and 0.65 eV including two broad signals were observed by DLTS measurements, and it was suggested that they might originate from crystal damage during high-energy electron irradiation.

The iron-related deep levels in n-type silicon and their thermal stabilities were investigated by deep-level transient spectroscopy (DLTS). Three deep energy levels at E c − 0.35, E c − 0.41, and E c − 0.48 eV were observed and classified into two types from the annealing behavior at room temperature and a low temperature of 200 °C. We found for the first time that only one iron-related deep level at E c − 0.35 eV was highly stable at room temperature and 200 °C, while other iron-related deep levels were unstable. We also found that the concentration of the deep energy level at E c − 0.41 eV gradually decreased at room temperature. These results suggest that the origin of the thermally stable level at E c − 0.35 eV is attributed to the substitutional iron-related level, and those of the thermally unstable levels at E c − 0.41 and E c − 0.48 eV are attributed to interstitial iron-related complexes such as iron-acceptor pairs in p-type silicon.

The concentration changes of nickel-related species after thermal annealing in Schottky electrode-formed (EL-formed) and electrode-free (EL-free) p-type silicon samples diffused with nickel were measured by deep-level transient spectroscopy. The nickel donor center began to decay at approximately 100 °C with the activation energies of 1.06 and 0.26 eV for the EL-formed and EL-free samples, respectively, which were analyzed as the required energies for the center to form complexes with interstitial nickel (Nii) and hydrogen, respectively. These complexes evolved into extended complexes by further bonding of Nii at higher annealing temperatures. All the complexes above disappeared by evolving into precipitates within temperatures lower than 400 °C without recovering the nickel donor center. The transformation reactions of the complexes progressed at lower temperatures and shorter times in the EL-formed samples than in the EL-free samples because of the electric neutralization of the nickel-related species in the space-charge region of the electrode.

The effects of the oxidation atmosphere and crystal faces on the traps in SiC/SiO2 interfaces close to the conduction band were investigated by deep-level transient spectroscopy. It was found that a high density of traps was observed around the energy of 0.16 eV from the edge of the conduction band in interfaces oxidized in a N2O atmosphere on the Si-face (O1 traps) and C-face (C1 traps). From the comparison of their energies, it was concluded that the O1 traps and C1 traps had the same origin. We also found that C1 traps were eliminated only on the C-face by wet oxidation. The simulation using the obtained energy and density of C1 traps in the interface oxidized in a N2O atmosphere on the C-face showed that the trapping of electrons in C1 was predominant just after the onset of the formation of an inversion layer, where the total electron density generated by the gate voltage was less than 7 × 1011 cm−2.

Diffusion profiles of substitutional nickel atoms in dislocated silicon crystal are investigated using deep-level transient spectroscopy (DLTS). The profiles are analyzed on the basis of the theory of the dissociative mechanism of diffusion. It is found that dislocations play an important role as the sinks and sources of vacancy annihilation and generation, respectively, giving rise to the flat distribution of the U-shaped distribution observed in the crystal bulk.

We report on the first observation of metastable defects in n-type silicon implanted with 90 keV hydrogen ions at 88 K at a dose of 2×1010 cm-2. Schottky contacts were fabricated on the implanted surfaces to study room-temperature stable defects by deep-level transient spectroscopy. After reverse-bias cooling, three new peaks (Ec–0.29, 0.41, 0.55 eV) appear in addition to well-known defects. Metastable defects are hydrogen-related and their production rates are higher in the 88 K implantation than in the room-temperature implantation.

Deep-level transient spectroscopy, thermal admittance spectroscopy (TAS), and deep-level optical spectroscopy techniques have been used to investigate the defect levels in the band gap of a non-doped In0.09Ga0.91N film with a thickness of 300 nm. At temperatures ranging from 10 to 450 K, TAS analysis revealed defect levels at ∼7 meV under the conduction band (E c) which were homogenously distributed in the InGaN film. The defect levels are considered to be attributed to the residual carrier density, probably originating from indium fluctuations and/or nitrogen vacancies (VN). Compared with the defect density of GaN, the defects of the InGaN film corresponding to the cation vacancies (VIII) and/or the VIII–O complex at 2.07 eV, and the shallow acceptor carbon (CN) defect at 3.05 eV, respectively, were remarkably enhanced. The possible increase of these defects is discussed in terms of VN and its complex with VIII that have been induced in the InGaN film.

The characteristics of interface states in a GaAs/GaAs1-x Bi x heterointerface have been evaluated by capacitance–frequency measurements, thermal admittance spectroscopy, and isothermal capacitance transient spectroscopy. The interface states density D it is evaluated to be approximately 9 ×1011 cm-2 eV-1 for the first time. The large density is probably caused by the fact that the surface of GaAs and GaAs1-x Bi x are shown to be nonmetallic and metallic, respectively. The interface states density is reduced by half by insertion of a Bi graded layer into the GaAs/p-GaAs1-x Bi x heterointerface, which is on the same order as other III–V heterointerfaces such as GaAs/GaAs0.97N0.03 and In0.5Ga0.5P/Al0.25Ga0.75As.

Photoinduced current transient spectroscopy (PICTS) has been utilized to study electrically active defects in TlInS2 single crystals doped with Er, B, and Tb. Eight traps with activation energies lying in between 0.21 to 0.55 eV have been detected. The origin of these traps is discussed.

A simple, spectroscopic method is proposed for the characterization of metastable defects in semiconductors using modified photoisothermal capacitance transient spectroscopy (photo-ICTS). A photo-ICTS method using a weighting function is formularized for the conversion process between two defect states. The potential energy barrier separating two configurations of the metastable defect in Cu(In,Ga)Se2 thin films has been successfully analyzed along with a tentatively determined defect concentration.

We have observed the formation of the Cu centers in p-type Si crystals diffused with dilute Cu between 400 and 1000 °C by deep-level transient spectroscopy (DLTS) and photoluminescence methods. For the samples diffused below 800 °C, a DLTS Cu center denoted as the CuDLB center was the main component and was almost uniformly distributed in the bulk. For the samples diffused above 800 °C, a precipitate-related Cu species was the main component at the surface, and low-concentration CuDLB centers and another precipitate-related species were the main components in the bulk. We attributed the inhomogeneous depth profiles of the Cu species in the samples diffused above 800 °C to the out-diffusion of Cu atoms and their precipitation near the sample surface. The dissociation product of the CuDLB center denoted as the CuDLA center was occasionally produced by the diffusion of Cu below 600 °C due to the precipitation of interstitial Cu. On the basis of these observations, we discussed the formation kinetics and structures of the CuDLB and CuDLA centers.

Carrier distribution and defect induction in In0.34Ga0.66As0.98N0.02/GaAs single quantum wells grown by molecular beam epitaxy at low growth rates are investigated by frequency-dependent capacitance–voltage (C–V) and deep-level transient spectroscopy (DLTS). The C–V studies show that lowering the growth rate of the InGaAsN layer splits the carrier accumulation in the well into a central and two side peaks with different frequency dispersions. The DLTS studies show that a continuum of states (0–0.083 eV) and a deep trap at 0.21–0.25 eV are responsible for the central and the side peaks, respectively. A comparison with photoluminescence (PL) spectra shows that these defects are induced by composition fluctuation. Lowering the growth rate degrades composition fluctuation by segregating the material into an InGaAsN phase and an N-depleted phase. Post-growth annealing can remove the deep trap and improve the InGaAsN emission, confirming that the deep trap degrades the InGaAsN phase. The feature of the continuum of states suggests that it may be the structural defects associated with lattice expansion or localized states introduced by composition fluctuation.

Interface states density (N ss) and minority carrier generation lifetime (τg) were evaluated in detail for strained Si/SiGe wafers by deep-level transient spectroscopy and metal-oxide-semiconductor transient capacitance methods. N ss shows a nondependence on strained Si thickness (d Si) within a critical thickness and a clear dependence on Ge fraction (Ge%). τg shows a strong dependence on both d Si and Ge%. The reasons for these dependences are discussed in detail.

The electrical and optical properties of the defect traps, with and without annealing, in InAs/GaAs quantum dots (QDs) emitting at 1.3 µm are investigated by capacitance-voltage (C-V), deep-level transient spectroscopy (DLTS) and photoluminescence (PL). When increasing the InAs thickness to 3 ML, an abnormal temperature dependence of the C-V characteristic was observed in the triple-stack InAs/GaAs QD sample. This temperature dependence is attributed to the defect levels at 0.39 and 0.54 eV observed in DLTS. The level at 0.39 eV, found in the top GaAs barrier, is probably related to the relaxation-induced dislocations. The level at 0.54 eV is found close to the QD region. Rapid thermal annealing can reduce the concentrations of both levels. Comparing with PL result, which shows a blueshift of 140 meV and linewidth narrowing in the QD emission by annealing at 800°C, the level at 0.54 eV is speculated to be strain- or relaxation-related defects in the interface between the QDs and the barrier. Removal of this level by high-temperature RTA is important since this level lies close to the QDs and influences the optical quality of the QDs.

A method for detecting defects such as interface states and a deep trap in Si-on-insulator (SOI) has been proposed, which is based on deep-level transient spectroscopy (DLTS) analysis of the metal-oxide semiconductor (MOS) structure. It is shown, for a bonded p-type SOI wafer consisting of 5 µm top-Si (T-Si) and 1 µm buried oxide (BOX) layers, that the interface state densities between T-Si/BOX and BOX/back-Si (B-Si) can be measured. Furthermore, the validity of this DLTS technique for the detection of a discrete deep level has been confirmed by using a Fe-doped SOI wafer. It is shown that the Fe concentration in B-Si is consistent with that of bulk Si, while the Fe concentration in T-Si shows significant enhancement. Through these investigations of SOI wafers, the validity of this DLTS method has been demonstrated.

Deep levels in AlInP layers, grown by metal-organic chemical vapor deposition (MOCVD) with various V/III mole ratios, have been carefully investigated by deep-level transient spectroscopy (DLTS). A deep level originating from phosphorus vacancy was observed with the activation energy of 0.65 eV. Examining this phosphorus-vacancy-related deep level provided a relatively simple means of understanding the phosphorus vacancy in AlInP, thus allowing us to determine an appropriate V/III mole ratio for growing AlInP.

Admittance spectroscopy is used to study a low-temperature (LT)-grown relaxed In0.12Ga0.88As/GaAs p-i-n superlattice. The capacitance-frequency spectra show two step-like trapping effects which are explained by the existence of two traps (E a=0.73 eV, σ= 4.6 ×10-11 cm2 and E a=0.75 eV, σ=3.7×10-15 cm2) in the LT-grown superlattice region. These two traps are the dominating defects observed in deep-level transient spectroscopy with one being a majority trap and the other being a minority trap. The emission parameters and photo-capacitance quenching effect for the 0.75 eV trap agree with those known for EL2 defects, suggesting that the EL2 defect is strongly enhanced in InGaAs/GaAs by LT growth.

Room temperature storage and/or minority carrier injection behaviors of three prominent majority carrier levels H3 (E v+0.34 eV), H4 (E v+0.39 eV) and H5 (E v+0.58 eV) and three minority carrier levels, 0.22, 0.29 and 0.35 eV below the conduction band edge in metal-organic chemical vapor deposition (MOCVD) grown p-type InP crystals produced by alpha radiation have been studied using deep levels transient spectroscopy. In particular, H5, absent immediately after irradiation, is found to grow with storage at room temperature after irradiation with no change in the other two majority carrier levels. Minority carrier injection saturates H5 while H3 and H4 continue to decay after injection. These observations are interpreted to mean that H4 and H5 are unrelated levels with no or very little mutual conversion proposed in some previous irradiation studies. Detailed production rate data for H3, H4 and H5 (post-injection) over a range of alpha-particle doses have also been provided for the first time.

The influence of different substrates on the deep levels in Hg1-x Cd x Te was investigated by spectral analysis of deep-level transient spectroscopy (SADLTS). In this study, CdZnTe, which had almost the same lattice constant and thermal expansion coefficient as Hg1-x Cd x Te, and Si, which had a larger lattice constant and thermal expansion coefficient than Hg1-x Cd x Te, were used as the substrates. The results of SADLTS show that there are six levels between 27 meV and 56 meV as activation energies in Hg1-x Cd x Te grown on the Si substrate, while there are six levels between 10 meV and 44 meV as activation energies in Hg1-x Cd x Te grown on the CdZnTe substrate. It was confirmed that the Hg1-x Cd x Te grown on the Si substrate had deeper levels than Hg1-x Cd x Te grown on the CdZnTe substrate.

The deep levels in GaTe single crystal have been investigated by deep-level transient spectroscopy measurements. The hole-trapping level is detected at 0.8 eV above the valence band. It is found that the deep level is associated with the defect or defect complex and the concentration increases with increasing annealing temperature.

The application of discharge current transient spectroscopy (DCTS) to the evaluation of the steady-state polarization P s(τ) in dielectrics with continuously distributed dipole relaxation time (τ) is discussed. The method proposed in Jpn. J. Appl. Phys. 35 (1996) 2216 can be used to evaluate P s(τ) only in dielectrics with discrete τ. Using the method discussed here, P s(τ) corresponding to the transient discharge current, which obeys a power law with respect to discharge time, can easily be evaluated.

A simple graphical method for evaluating the dipole relaxation times (τ) in a dielectric is proposed. From isothermal measurements of the transient discharge current due to dipole relaxation, τ can be obtained graphically by means of discharge current transient spectroscopy (DCTS). In other words, the time at which the DCTS signal is a maximum corresponds to τ. Since τ and the steady-state polarization (P s) in dielectrics at electric field (F p) and temperature (T) can be obtained using DCTS, the dependencies of τ and P s on F p or T can be investigated in detail.

Deep levels and resistivity in annealed low-temperature molecular beam epitaxial (LT-MBE) GaAs layers were characterized using n- i-n structure. We found that annealing at 620°C resulted in the highest attainable resistivity of LT GaAs layers. However, deep-level transient spectroscopy (DLTS) and photocapacitance quenching measurements showed no signal of the midgap (EL2) level. Only a “shallower” EL3-like level was observed in the layer of high resistivity. The fact that no midgap level was detected in the highest-resistivity LT GaAs layer means that the “compensation" mechanism does not play an important role in controlling the semi-insulating properties of LT GaAs layers. On the other hand, a midgap level appeared upon annealing at higher than 620°C, although the resistivity of LT GaAs layers decreased more than five orders of magnitude. The photoquenching measurements suggest that this level is identical to the EL2 level. These results support the “buried Schottky barrier" model.

Deep levels and residual impurities in sublimation-grown n-type SiC substrates are investigated. Three electron trap centers at 0.26, 0.35 and 0.72 eV below the conduction band (CB) are observed in 4H-SiC for the first time by deep-level transient spectroscopy (DLTS). One electron trap center at 0.68 eV below the CB is also observed in n-type 6H-SiC. The trap concentration and capture cross section are estimated. The concentration of residual impurities V, Cr, Fe and Ti are found to be high. The origin of observed trap centers is discussed with respect to the residual impurities and available information.

We report on the effects of exposure to a hydrogen plasma (hydrogenation) and of thermal annealing (dehydrogenation) on shallow and deep levels in direct-gap AlGaAs:Si grown by molecular beam epitaxy. Photoluminescence (PL) experiments show that hydrogenation results in the passivation of shallow levels, thus confirming the data of capacitance-voltage measurements, while deep-level transient spectroscopy shows that ME5, ME6 and DX centers are passivated by hydrogenation. Dehydrogenation at 420°C results in an almost complete recovery of the free electron concentration, and restores, to a significant extent, only the DX center. The study of samples grown at different temperatures shows that the significant increase of the PL efficiency after hydrogenation and its decrease after dehydrogenation are consistent with the passivation of the ME5 and ME6 levels and with their partial reactivation, respectively.

The annealing behavior of the native defect EL2 in GaAs grown by the liquid-encapsulated Czochralski (LEC) method has been investigated by means of deep-level transient spectroscopy (DLTS). After annealing in vacuum, the EL2 out-diffusion profiles were strongly dependent on the annealing temperature and time. Applying the three-step model for EL2 out-diffusion, the activation energy of the diffusion for an arsenic vacancy (V As) was estimated to be about 3-4 eV.

Deep-level transient spectroscopy is successfully applied to detect interface states in ZnO/PrCoO x /ZnO thin-film junctions. The interface state is measured at 0.70 eV below the conduction-band edge. The junctions are equivalent to a model of a single grain boundary in a ceramic varistor.

Deep-leve transient spectroscopy (DLTS) of bulk traps and interface states in Si MOS diodes are theoretically studied and energy levels, capture cross-sections and spatial and energy density distributions of majority-carrier traps are measured. In P+-implanted unannealed MOS diodes, four bulk traps are measured at Ec-0.18 eV, Ec-0.20 eV, Ec-0.31 eV and Ec-0.45 eV. Their spatial distributions are found to be the same among them within experimental error and thought to be corresponding to the distribution of implanted ions qualitatively. Bulk traps are distinguished from interface states experimentally. The capture cross-section of interface states in non-implanted MOS diodes are measured to be of the order of 10-16 cm2 in the energy range of Ec-0.15 eV to Ec-0.30 eV. The interface state density measured with DLTS is found to be in a reasonable agreement with those detetmined by other methods.

For the analysis of deep-level parameters in semiconductors, a new DLTS (Deep-Level Transient Spectroscopy) method has been proposed, in which a squaring device for the capacitance-signal processing is used with the conventional DLTS system. This method (C2-DLTS) is simple and is especially effective in the precise analysis of large signals due to deep-level impurities and defects.

Deep impurity levels in InP LEC single crystals have been investigated by deep-level transient spectroscopy (DLTS) and photoluminescence (PL) measurements. The effects of heat treatment on DLTS and PL signals have been investigated, the results indicating that a deep impurity level whose emission activation energy is 0.42 eV is closely related to the well-known PL emission band at 1.1 eV. It seems that these DLTS and PL signals originate from a certain complex including P vacancies and defects, since some behaviours of these signals can be well explained by a configuration-coordinate model. Two other deep levels with emission activation energies of 0.31 and 0.60 eV have also been observed. The origin of these deep impurity levels might be due to native defects in the InP LEC crystals.

Scanning isothermal current transient spectroscopy (SICTS) is proposed for the microscopic deep-level distribution measurement in semiconductors and applied to the spatial distribution measurement of gap states in hydrogenated amorphous silicon films.

A method of stabilizing the microwave frequency modulated for the observation of transient molecular responses has been developed. Good long-term stability of the mean values of the modulated microwave frequencies is achieved; the stability is 3×10-9/day for an NEC 10V13 klystron and 10-8/day for OKI 24V11, 25V12, and 47V12 klystrons. This stabilizer has operated successfully in measuring the relaxation rate constants of weak rotational transitions of OCS and HCOOH, whose absorption coefficients are in the region 10-6-10-7 cm-1.

A general formula for the capacitance transient response in an MIS structure has been developed, including the continuous density distribution of interface states and discrete multi-impurity levels. A new method of spectroscopic measurement is proposed for determining the density distributions and capture cross-sections of interface states and bulk impurity levels under isothermal conditions. The method is then applied to an n-InAs/anodic oxide MIS diode.

A new measurement method for deep levels in semiconductors is proposed, by which the measurement of the transient change of capacitance is performed under an isothermal condition (Isothermal Capacitance Transient Spectroscopy). The method allows us to construct a precise measurement and analysis system by a programmable calculator. Computer simulation and experiment by the method in the case of Au-doped Si are demonstrated. It is shown that the method is one of useful tools for spectroscopic analysis of deep levels in semiconductors.

An analysis was made of the depletion-discharge process observed in amorphous semiconductors, where the dark decay of the surface potential V was caused by a sweep out from the bulk of only one sign of the charge carriers thermally generated from deep states in the bulk and a simultaneously progressive formation of the space charge of the opposite sign. It is shown that the distribution of the density of deep emission states can be directly determined from a plot of t|dV/dt| for tt d vs ln (t) using a deconvolution technique.

The isothermal-capacitance-transient-spectroscopy (ICTS) and the deep-level-transient-spectroscopy (DLTS) data of a-Si:H are analyzed using a model based on the electronic relaxation theory of Ngai and coworkers. The electron capture and emission processes are taken into account in a selfconsistent manner, and are described by relaxation functions with the same fractional exponent, i.e., φ(t)αexp [-(t/τ*)1-λ], where λ=0.2 and τ* is energy dependent in both processes. This approach (i) explains the anomalous energy and temperature dependence of the effective relaxation rate obtained by the ICTS method and (ii) yields an explanation for the energy difference of the D- states between ICTS and DLTS measurements.

A variation of Photo-Induced Transient Spectroscopy (PITS) is described. By elimination of charge exchange with the surface of semi-insulating GaAs, negative peaks in PITS spectra are suppressed. This allows the transient response of bulk EL2 centres to be detected by PITS in undoped LEC GaAs for the first time.

A method is presented for determining the interface-state distribution in the half of the band gap adjacent to the minority-carrier band in MOS structures. Deep-level transient spectroscopy (DLTS) for MOS capacitors was performed under pulsed light illumination with a photon energy above the band gap energy (optical DLTS). Minority carriers are captured at the interface states during the light pulse and thermally emitted after the end of the light pulse. Experimental results are given for both hole and electron traps at the nSi–SiO2 interface.

A computer controlled multi-channel PICTS system which allows one to detect many deep trapping levels in high-resistivity bulk material by only a single thermal scanning process, has been developed. In order to obtain the normalized PICTS signal, an automatic gain controlled pre-amplifier has been used. An expanded-scale technique which allows one to distinguish fine PICTS peaks has been proposed. This expanded-scale technique is very useful especially for the case when many PICTS peaks are closely superposed. In high-resistivity bulk CdS single crystal, it has been found that PICTS spectrum-shape and number of detected peaks depend strongly upon the wavelength of exciting laser light, and that the trap-parameters obtained are not strongly influenced by light intensity. The PICTS method may be superior on many points, especially in peak-separation, to the thermally stimulated current method.

An ICTS (Isothermal Capacitance Transient Spectroscopy) theory for a system of continuously-distributed gap states is given and experimental data by the ICTS applied to P-doped a-Si:H Schottky barrier diode are presented. It is shown that the ICTS is a useful tool for the study of gap states of a-Si:H whose material parameters are strongly temperature-dependent.

Trap levels in plastically-bent vapor-phase epitaxial (VPE) and liquid-phase epitaxial (LPE) GaAs n-layers were investigated by deep-level transient spectroscopy (DLTS). Two hole traps, Level A(E v+0.38 eV) and Level B(E v+0.68 eV), and one electron trap, EL2(E c-0.82 eV), were found after deformation, as well as hole traps due to Cu and Fe contaimnation. No trap levels attributable to the dislocations in GaAs were found. The total amount of additional trap concentrations detected by DLTS measurements fell fairly far short of the reduction in the free carrier concentration, owing to deformation. These findings suggest that dislocation-associated traps in GaAs in themselves do not show a peak in DLTS spectra. This is tentatively discussed in connection with recent reports on the degradation of light-emitting devices.