Modulated Photocurrent Experiments Research Articles

Hydrogenated amorphous silicon characterization from steady state photoconductive measurements

We propose to use steady state measurements and a teaching-learning-based optimization (TLBO) algorithm to get the complete set of material transport parameters of disordered semiconductors, taking undoped hydrogenated amorphous silicon as an example. First, the steady-state conductivity under illumination and the ambipolar diffusion length (Lamb) are measured for several temperatures and generation rates. The steady-state photocarrier grating technique is used for the evaluation of Lamb. Then, the TLBO algorithm is used for the obtainment of the material parameters that best satisfy the charge neutrality and the continuity equations. The use of this algorithm allowed us to get an excellent estimation of the valence band tail slope, as compared to the one obtained from measurements of the absorption coefficient by Fourier transform photocurrent spectroscopy and transmittance/reflectance. The dangling bonds and the conduction band tail parameters were also found to be in very good agreement with those measured from high frequency modulated photocurrent experiments. Numerical simulations show that the capture coefficients of the band tails and defects states can also be estimated, although with less precision than the DOS parameters.

Modulated photocurrent experiments-comparison of different data treatments

Modulated Photocurrent experiments provide a powerful tool to characterize localized defect states within the band gap of semiconductors. However, in recent years, different data analysis methods have been established. The method proposed by Herberholz et al. identify defect positions and attempt-to-escape frequencies considering the phase shift between excitation flux and measured photo current only. In contrast, Brüggemann et al. propose to use phase shift and amplitude of the modulated photo current to resolve an energy dependent density of trap states. This work compares both analysing procedures. We demonstrate that for simple and dominant defects, such as acceptor or donor like monovalent traps, both methods give equivalent results. However, the Herberholz method fails for more complex defect arrangements as exemplified on hydrogenated amorphous silicon or polycrystalline CuInSe2.

Nature of gap states in GeSbTe phase change memory materials

We have calculated the states associated with valence alternation pair defects in a Te-rich amorphous GeTe random network. Two-fold Te sites and four-fold Ge sites are found to introduce states that are resonant with the valence and conduction bands, respectively. The system has a band gap, but these defect states are in the same energy range as those found by modulated photocurrent experiments. The absence of electron spin resonance signals is still not explained.

Metastable defect in CuInSe2 probed by modulated photo current experiments above 390 K

Modulated photocurrent experiments have been widely used to study defects in semiconductors. Previous studies have found a number of defects in CuInSe2, which is used as an absorber in solar cells. We apply a method of analysis, which has previously not been used for Cu(In,Ga)Se2 semiconductors and which allows the determination of defect concentrations in addition to defect energies. We found that at least one of the previously discovered defects shows a metastable behaviour, increasing in concentration, and can be related to the efficiency loss in corresponding solar cells.

Defects in amorphous phase-change materials

Abstract

The influence of a temperature dependent bandgap on the energy scale of modulated photocurrent experiments

Amorphous semiconductors and chalcogenide glasses exhibit a high density of localized states in their bandgap as a consequence of structural defects or due to a lack of long range order. These defect states have a strong influence on the electronic transport properties. Thus, many theories attribute the “resistance drift” or the “threshold switching” effects, both observed in amorphous phase change alloys, to defects within the bandgap. The energetic distribution of states within the bandgap can be probed via modulated photocurrent (MPC) experiments that enable a spectroscopy of the relative density of these defect states by varying the modulation frequency at various temperatures T. It is also a common feature that the bandgap decreases with temperature. Nevertheless, the consequences of a shrinking bandgap with increasing temperature have been neglected in the classical analysis of MPC experiments. In this paper, we propose to add correction terms to the classical MPC energy scaling to take the temperature dependence of the bandgap of the studied material into account to improve the accuracy of the determination of the defect distribution. We illustrate the efficacy of our proposed corrections by applying it to the study of disordered materials such as hydrogenated amorphous silicon a-Si:H, a-GeTe and a-Ge2Sb2Te5.

Investigation of defect states in the amorphous phase of phase change alloys GeTe and Ge2Sb2Te5

AbstractIn amorphous materials localized states within the band gap govern the electronic transport properties. This work presents the first measurement of the density of states within the band gap of amorphous phase change alloys using modulated photo current experiments. Modulated photo current measurements performed on a‐Ge50Te50 and a‐Ge2Sb2Te5 show that the defect state density of both compositions is quite similar consisting of a pronounced valence band tail and at least two defect states (© 2010 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)

Deep defects and their electron-capture cross sections in polymorphous silicon-germanium thin films

Hydrogenated silicon-germanium alloys (SiGe:H) are deposited by plasma-enhanced chemical vapor deposition in a plasma regime close to that of the formation of powder. It is thus possible to obtain nanostructured materials that we call polymorphous materials, pm-Si1−xGex:H. Studies of space-charge-limited currents and space-charge relaxation allow to get information on midgap states originating from Ge. It is observed that the electron-capture cross section of states at the Fermi level increases linearly with their concentration. This is supported by modulated photocurrent experiments. Finally, it is shown that the variations of both the Ge dangling bond concentration and their electron-capture cross section with Ge content account for the photoresponse in these materials.

Light-induced changes in the gap states above midgap of hydrogenated amorphous silicon

The energy dependence of the capture coefficients and the density of states (DOS) above midgap of hydrogenated amorphous silicon (a-Si:H) in annealed and light exposed states are examined by modulated photocurrent experiments. In the annealed and light exposed states, the electrons are found to interact with two different kinds of gap states through trapping and thermal release. The densities of both gap state distributions at trap depths shallower than 0.67eV below the conduction band decrease rapidly during the initial stage of light degradation. The DOS of the annealed and light exposed states are found to cross at about 0.67eV. In addition, a parallel increase in the capture coefficients of the gap states at trap depths lower than 0.67eV is observed indicating a light-induced disorder. The above behavior is explained with the conversion of strained Si–Si and Si–H–Si bonds, which become highly strained during illumination, into dangling bonds near midgap.

Effects of proton irradiation on the photoelectronic properties of microcrystalline silicon

We report on proton irradiation of microcrystalline silicon with 2 MeV protons and find a drop in the mobility–lifetime product related to an increase in the density of gap states as revealed by modulated photocurrent experiments. The dark conductivity also drops due to a shift in the Fermi energy by defect creation and re-balance of charge. The impact on the ambipolar diffusion length, representing the minority-carrier properties, is much smaller. The electronic properties of P and B doped samples are hardly affected. We discuss the implications for microcrystalline silicon pin solar cells.

Density of states in the gap of amorphous semiconductors determined from modulated photocurrent measurements in the recombination regime

An experimental technique to study the energy profile of localized states in the gap of amorphous semiconductors is proposed. The method is based on the relationship between the recombination lifetime and the density of states (DOS) at the quasi-Fermi level for trapped carriers. We use the modulated photocurrent experiment in the recombination-limited regime as a convenient method to measure the recombination lifetime. Measurements performed as a function of temperature allow the DOS above the Fermi energy to be determined. The accuracy and limitations of the method are studied by means of computer simulations. The experimental technique is applied to obtain the density of defect states of a hydrogenated amorphous silicon sample.

Microcrystalline silicon deposited by the hot-wire CVD technique

Hot-wire chemical vapour deposition (HW-CVD) is a well-known technique to deposit amorphous silicon with high deposition rates from the decomposition of silane and hydrogen gases. By changing the hydrogen and silane flow rates, it is possible to observe a transition from amorphous silicon (a-Si:H) to microcrystalline silicon (μc-Si:H). In this study, structural and electrical properties of layers deposited as a function of silane concentration in the gas are presented. Ellipsometry and X-ray diffractometry have been used to assess the structure of the films. Steady-state photoconductivity, steady-state photocarrier grating, and modulated photocurrent experiments have been carried out to characterise both majority and minority carrier transport properties. Finally, the addition of phosphine and diborane in the reactor allows the deposition of n- and p-type layers with conductivities up to 10 and 1 s cm −1, respectively, making possible the realisation of HW-CVD solar cells.

From amorphous to polycrystalline thin films: dependence on annealing time of structural and electronic properties

Some new results about the amorphous to polycrystalline transition of silicon thin films obtained by low pressure chemical vapor deposition (LPCVD) at 550°C are presented. From in situ (monitored during the crystallization annealing) conductance, electron spin resonance, photoluminescence and modulated photocurrent experiments, the density of states is shown to increase in the so-called nucleation regime and reach a maximum just before crystal grain growth starts. These results indicate that crystallization needs the creation of a defected material with large dangling bond densities and wide band-tails. It appears that the electronic quality of the final polycrystalline material is linked to the maximum density of states reached just at the end of the nucleation phase. This level may be associated to a viscous structure.

Deep Defect Relaxation in Hydrogenated Amorphous Silicon: New Experimental Evidence and Implications

ABSTRACTWe review modulated photocurrent experiments which indicate that thermal emission rate for Do defects in intrinsic samples varies in response to changes in the Fermi-level or quasi-Fermi position. This apparent shift in energy threshold is confirmed using time resolved sub-band-gap spectroscopy. We also demonstrate that such a variation of emission rate with changes in the Fermi-level position, if present within the depletion region near a barrier junction, is consistent with the details of the temperature dependence of the junction capacitance in intrinsic samples.

Density of states and capture cross-sections in annealed and light-soaked hydrogenated amorphous silicon layers

The density of states of a lightly p-doped hydrogenated amorphous silicon sample and the related transport behaviours were extensively studied and compared with those of intrinsic samples both in the annealed and light-soaked states. It is observed that p-doping increases the density of positively charged dangling bonds in the annealed samples, in agreement with the predictions of the defect-pool model. After light-soaking, a decrease of positively charged dangling bonds and an increase of neutral ones in the p-doped sample are observed, which is still in qualitative agreement with the predictions of the model. However, quantitatively, the results of modulated photocurrent experiments cannot be reproduced on the basis of this model. It is proposed that light-soaking either induces a variation of the defect capture cross-sections or creates a distribution of states in the gap for which the defect-pool model does not account.

Trapping and recombination via dangling bonds in amorphous and glassy semiconductors under steady-state conditions: Application to the modulated photocurrent experiment.

Interesting features of the modulated photocurrent experiment have been emphasized in a recent paper where we have shown that, if the localized states in the gap belong to the same species of monovalent centers, then it is possible to deduce both an order of magnitude of the capture cross section and the absolute value of the density of these states by means of this experiment. In the present paper, we extend the calculation of the dc and ac photocurrent to the case where distributions of correlated states associated with the well-known dangling bonds (DB) are present in the material. This calculation includes the contributions of both holes and electrons and takes into account the interactions of both types of carriers with distributions of monovalent as well as correlated states in the gap of a semiconductor. We concentrate in particular on the signature of the DB states in the data analysis, the correlation energy being either positive or negative. We first derive the variations with respect to energy of the occupation functions of any distribution of DB states for both signs of the correlation energy under steady-state conditions. We show that he concept of quasi-Fermi-levels for trapped carriers introduced by Simmons and Taylor for monovalent states has to be reconsidered for the correlated states and we underline the differences between the cases of positive and negative correlation energies.We give a clear and comprehensive scheme of the different recombination paths for the correlated states and derive the correct expressions of the electron and hole lifetimes related to these kinds of states. The results of the steady-state-regime study are then used in a second part to identify how the DB states modify the modulated photocurrent. Simplified expressions of the modulus of the modulated photocurrent and of its phase shift with reference to the ac excitation light are given in two cases: a recombination-limited regime and a trapping- and release-limited regime. The behaviors of the modulated photocurrent related to the presence of DB states are also underlined in both regimes by means of a numerical simulation. It is shown that a distribution of DB centers exhibiting a positive correlation energy roughly behaves as two distributions of monovalent states whereas a distribution of DB centers with a negative correlation energy roughly behaves as a single distribution of monovalent states. An important property of the DB states with a positive correlation energy is that they can give a significant response in the trapping- and release-limited regime even if their ground-energy level is below the Fermi level. It is also shown that if there are both monovalent- and correlated-state distributions in the gap, it is possible to derive an order of magnitude of the lowest capture cross sections from the transition between the recombination-limited and trapping- and release-limited regimes.

Study of the density of states of hydrogenated amorphous silicon from time-of-flight and modulated photocurrent experiments

The densities of states (DOS) deduced from the modulated photocurrent (MPC) and the time of flight (TOF) techniques performed on two films of a-Si:H deposited in the same run are compared. For the MPC experiment the sample was built in the coplanar geometry whereas a Schottky diode was used for the TOF experiment. This comparison suggests that the MPC technique applied to a coplanar sample is sensitive to the electron trapping states. The influences on the MPC results of the flux and of the wavelength of the excitation light are studied. Probing the DOS closer to the illuminated interface by decreasing the wavelength, we observe a slight increase of the DOS, the increase on the a-Si:H/air side being more important than on the glass/a-Si:H side.