Deep Level Transient Spectroscopy

Semetrol’s Deep Level Transient Spectroscopy (DLTS) system is designed to obtain detailed results very efficiently. Place the probes on your sample, test various measurement conditions, and then start measuring. Measurements may be made very quickly for a survey of traps, or with more extensive transient averaging for very high sensitivity. The entire transient is recorded at each temperature step, so there is no need for repeated temperature scans as with conventional boxcar averaging or correlators.

Installation includes training, an extensive manual with examples, and on-line instructions that will allow a new user to produce results quickly with thermal spectroscopy. The data acquisition interface allows the user to set up measurements and see the results of changes in the measurement conditions very quickly. Similarly, the data analysis software performs the tasks very efficiently, automatically generating rate windows over the widest range possible, and automatically finding peaks.

DLTS uses a semiconductor structure in which there is a region that is depleted of mobile charges, and measurement conditions may be changed to allow trapping of carriers. The most straightforward method is to use a diode and decrease the width of the depletion region to fill traps that may be present, followed by an increase in the depletion width and measure the rate of emission of the trapped charges. The filling pulse results in charges trapped on defects. The measurement bias places the traps in a nonequilibrium condition. The charges are released at a rate depending on the trap energy and capture cross section, and the temperature, leading to a change in capacitance. The change in capacitance is the measured transient. The set of transients is processed over the range of measured temperature for trap characteristics.

DLTS analysis of several spectra of defects in a GaN semiconductor sample

Several spectra from deep level defects in GaN with the emission rates, or rate windows, listed in the upper right.

With the full DLTS data set available, post processing can be done in a number of different ways, with user-selected degrees of noise filtering. Analysis includes rate window peak detection, Gaussian deconvolution and transient fitting. Peak detection finds the highest peaks and plots the results on an Arrhenius plot for the energy and capture cross section. Gaussian deconvolution can be used for more detailed fits of the rate window spectra for multiple components. Transient fitting analyzes the data in the time-domain, extracting multiple emission rates.

Semetrol pioneered advances in state of the art DLTS analysis, extending to temperature and time domain (3D) fitting. The user selects several transients for fitting over a temperature range of interest. The fit determines energy, capture cross section, and concentration directly from the selected transients. The analysis program gives the user options to test initial conditions for several trap components, or use automatically generated initial conditions with user selected maximum number of trap components.

DLTS analysis using an Arrhenius plot determining the energy and capture cross section of a sample.

Energy and capture cross section are determined from an Arrhenius plot of T²/e versus 1/kT. The Arrhenius points are generated automatically.

Verify the results from any of the analysis methods with simulation of spectra from fresh analysis results, or simulate a full data set. Full data set can be used to check results against measured data, or to generate spectra from results in published journal articles.

Get detailed results from variations of DLTS. Field dependence from Double Pulse DLTS provides information on acceptor or donor nature of a specific trap. Deep level profiling is done easily with another version of Double Pulse DLTS to provide the spatial profile of a defect of interest. Find out if a trap is related to a dislocation, or lattice relaxation, by using a program that varies the filling pulse duration automatically over a range that you specify.