How does deep-level transient spectroscopy (DLTS) detect deep-level defects in InP?

Blog / Admin: SemiCrystal Limited / Date: September 9, 2026

## Introduction to DLTS Characterization of InP


Deep-Level Transient Spectroscopy (DLTS) is a highly sensitive, high-resolution electrical characterization technique uniquely capable of identifying and quantifying deep-level defects within the bandgap of semiconductor materials. In Indium Phosphide (InP), deep levels—originating from native point defects, dislocations, or unintentional impurities like iron, oxygen, and copper—act as potent carrier traps or generation-recombination centers. These centers directly impact the performance of high-frequency electronic components and optoelectronic devices, including Heterojunction Bipolar Transistors (HBTs), avalanche photodiodes, and lasers. While general optical methods may fail to measure non-radiative centers, DLTS directly probes the thermal emission processes of trapped carriers within a device's space-charge region. By measuring capacitance or current transients as a function of temperature, DLTS reveals vital physical parameters such as activation energy, capture cross-section, and trap concentration.

## The Dynamics of Space-Charge Regions and Carrier Trapping


The fundamental mechanism of DLTS relies on manipulating the occupancy of deep-level traps within the depletion region of a p-n junction, Schottky diode, or heterojunction fabricated on an InP substrate. Under a steady-state reverse bias, the space-charge region is depleted of free carriers. Within this zone, the electric field is strong, and any deep-level traps present are forced into an empty state because the rate of carrier emission far exceeds the rate of capture.
To initiate a measurement, the system applies a short forward-voltage filling pulse, reducing the reverse bias and shrinking the depletion width. This collapse allows free majority carriers (electrons in n-type InP, holes in p-type InP) to flood the region, where they are rapidly captured by empty deep levels according to the trap’s capture cross-section. Once the filling pulse ends and the device returns to its original reverse bias, the captured carriers are trapped above or below the Fermi level. The traps must then return to thermal equilibrium by releasing their bound carriers into the conduction or valence band via thermal emission, a process that forms the basis of the transient detection cycle.

## Thermal Emission Physics and Capacitance Transients


Following the restoration of the reverse bias, the trapped carriers are emitted from the deep levels at a rate governed by the Boltzmann distribution. The thermal emission rate for electrons escaping to the conduction band is mathematically expressed as a function of the trap's signature attributes: the activation energy, the capture cross-section, and the temperature. Because the emission rate is exponentially dependent on temperature, the trap remains stable at very low temperatures but releases its carriers rapidly as the sample warms up.
As these carriers escape from the traps and are swept out of the space-charge region by the electric field, the localized space-charge density changes. This alteration directly impacts the high-frequency capacitance of the diode. The resulting return of the capacitance to its steady-state value manifests as an exponential capacitance transient. The time constant of this transient is inversely proportional to the thermal emission rate. By measuring this capacitance transient over a broad temperature range, the DLTS system isolates the distinct thermal signatures of individual defect species embedded in the InP matrix.

## The Rate Window Concept and Spectral Peak Generation


Extracting useful defect data from a raw, continuously changing capacitance transient requires a signal processing method known as the rate window concept. Historically implemented via Boxcar integrators, this technique samples the capacitance transient at two discrete times, designated as t1 and t2, after the end of the filling pulse. The DLTS signal is defined as the difference between the capacitance measured at t1 and the capacitance measured at t2.
When the sample temperature is very low, the emission rate is near zero, the transient is extremely slow, and the difference between the two sampling points is negligible. At exceptionally high temperatures, the emission is nearly instantaneous, meaning the transient finishes before t1, yielding a signal of zero. However, at a specific intermediate temperature, the emission rate matches the selected instrumental rate window defined by t1 and t2. At this exact point, the capacitance difference reaches a maximum, generating a sharp, symmetrical DLTS peak against the temperature axis. Modern digital systems often replace the Boxcar method with Fourier transform or Laplace transform algorithms, which collect the entire transient to enhance the signal-to-noise ratio and resolve closely spaced defect states.

## Constructing Arrhenius Plots to Extract Fingerprint Parameters


To extract the exact physical parameters of a defect, a series of temperature scans are performed using different rate windows. Changing the values of t1 and t2 shifts the peak position to a different temperature. By plotting the logarithm of the emission rate, normalized by the square of the temperature, against the inverse absolute temperature (1000/T), an Arrhenius plot is constructed.
The slope of the linear fit on the Arrhenius plot yields the precise thermal activation energy of the trap, representing the position of the defect state relative to the conduction band minimum for electron traps or the valence band maximum for hole traps. The intercept of the line determines the apparent capture cross-section, which reflects the physical size of the potential well and its ability to capture moving carriers. Furthermore, the absolute amplitude of the DLTS peak is directly proportional to the defect concentration. If the peak height is small relative to the background doping concentration of the InP wafer, simple linear scaling provides an accurate quantification of the trap density down to concentrations below one part in ten thousand of the background doping.

## Identifying Key Deep-Level Traps in Indium Phosphide


Applying DLTS to InP has historically revealed several critical native and extrinsic defect centers that dominate device performance. In n-type InP, a prominent native electron trap known as the "E1" center frequently appears, featuring an activation energy roughly 0.60 eV to 0.65 eV below the conduction band. This deep state is often associated with phosphorus vacancies or complexes involving indium antisites, acting as a lifetime-limiting recombination center.
Another highly critical deep level is introduced by iron (Fe) doping. Iron is intentionally added during crystal growth to produce semi-insulating InP substrates by compensating for shallow background donors. The Fe acts as a deep acceptor, introducing a level approximately 0.63 eV above the valence band. DLTS is routinely used to monitor the active concentration of this iron level, ensuring it is uniformly distributed to prevent leakage currents in high-speed waveguides and HBT sub-collectors without causing excessive carrier trapping in neighboring active layers. Hole traps in p-type InP, such as those induced by zinc diffusion or electron beam irradiation, are similarly classified, providing an atomic-scale map of how processing history affects the material's electrical behavior.




SemiCrystal Limited (SemiCrystal.com)
https://www.semicrystal.com/blog/how-does-deep-level-transient-spectroscopy-dlts-detect-deep-level-defects-in-inp.html
Related Article