We have measured the electrical and optical properties of blue light-emitting diodes ͑LEDs͒ based on III-V nitrides. The current-voltage characteristic is described by means of the relation IϭI 0 exp(␣V). In this equation ␣ is temperature independent, suggesting a process of conduction by tunneling, as was recently reported also for blue-green LEDs based on III-V nitrides ͓Appl. Phys. Lett. 68, 2867 ͑1996͔͒. We explain the differences between blue and blue-green devices taking into account the tunneling process across semiconductor interfaces, in which a great number of defects is present. The light output intensity of the LED as a function of junction-voltage data reveals a dependence on the junction-voltage of the type LϭL 0 exp͑qV/1.4 KT͒, indicating that the radiative recombination path is via deep levels located at the forbidden gap. Furthermore, we find that the light output-current characteristic follows a power law like LϰI p . From the analysis of data it appears that, contrary to expectations, the nonradiative centers are saturated at very low current values that are comparable to the values at which this saturation takes place in LEDs based on III-V arsenides with a low content of defects. © 1997 American Institute of Physics. ͓S0021-8979͑97͒05304-8͔Blue light-emitting diodes ͑LEDs͒ based on III-V nitrides have recently shown high brightness characteristics 1 in spite of the presence of an extremely high density of dislocations ͑in the 2-10ϫ10 10 cm Ϫ2 range 2 ͒ that is several orders of magnitude higher than those measured in LEDs based on III-V arsenides and phosphides. Blue and blue-green devices are commercially available from Nichia Chemical Industries. 1,3 The main difference between them is the composition of the active layer ͑In 0.06 Ga 0.94 N for blue devices and In 0.23 Ga 0.77 N for the blue-green emitters͒.Here, we present a study of the influence of defects on the transport properties and on the optical characteristics of the Nichia blue LEDs and we compare the results with those reported in blue-green LEDs from the same manufacturer. 4 The blue LED analyzed in this study is device part No. NLPB500. The physical structure is composed of different layers, all of them grown by metalorganic chemical vapor deposition ͑MOCVD͒. Briefly, the structure is as follows: a n-GaN bottom contact layer ͑ϳ4 m͒, followed by a double heterostructure ͑DH͒ composed of an n-Al 0.15 Ga 0.85 N carrier confinement layer ͑ϳ0.15 m͒, an In 0.06 Ga 0.94 N active layer, Si and Zn codoped ͑0.05 m͒, and a p-Al 0.15 Ga 0.85 N carrier confinement layer ͑0.15 m͒. Last is a p-GaN top contact layer ͑0.5 m͒.We have measured the current-voltage (I -V) characteristic and the light output intensity as a function of junctionvoltage (L -V) and LED output current (L -I). The I -V and L -V data are influenced by the presence of a series resistance at high voltages. We have suppressed this influence following the procedure described in Ref. 5. Over the entire current range, where the influence of series resistance is present ͑4ϫ10 Ϫ...

The influence of rapid thermal annealing treatments on the interface characteristics of Al/SiNx:H/InP devices was analyzed. The insulator was obtained by an electron cyclotron resonance plasma method at a 200 °C-deposition temperature. The films were deposited in a single deposition run but in two steps: first, we deposited the bottom layer with a film composition of x=1.55 and then the top layer with x=1.43. Total film thickness was 500 Å in one set of samples and 200 Å in the other one. Annealings were conducted in Ar atmosphere for 30 s in a temperature range between 400 and 800 °C. To characterize the electrical behavior of these devices, capacitance–voltage (C–V) and deep level transient spectroscopy (DLTS) measurements have been performed on each sample. This last characterization shows the presence of features in the spectra at Ec−0.2 eV, Ec−0.25 eV, Ec−0.38 eV. The last one is due to phosphorus vacancies, VP. Devices with 200-Å-thick insulator present the minimum interface trap densities. According to the DLTS analysis, this minimum (3×1011 cm−2 eV−1) is achieved on the 400 °C-annealed samples. A tentative explanation of these results is given in terms of a possible InP surface passivation due to the fact that nitrogen atoms coming from the insulator can fill phosphorus vacancies, giving rise to a low defective insulator/semiconductor interface. This process is enhanced by rapid thermal annealing treatments at moderate temperatures (400–500 °C).

A minimum interface trap density of 10 12 eV Ϫ1 cm Ϫ2 was obtained on SiN x :H/InP metalinsulator-semiconductor structures without InP surface passivation. The SiN x :H gate insulator was obtained by the electron cyclotron resonance plasma method. This insulator was deposited in a single vacuum run and was composed of two layers with different nitrogen-to-silicon ratios. The first layer deposited onto the InP was grown with a nitrogen-to-silicon ratio of N/Siϭ1.55, whereas the second one was grown with a N/Si ratio of N/Siϭ1.43. After the insulator deposition, rapid thermal annealing of the devices was performed at a constant annealing time of 30 s. The interface trap density minimum value was obtained at an optimum annealing temperature of 500°C. Higher annealing temperatures promote thermal degradation of the interface and a sharp increase in the trap density. © 1999 American Institute of Physics. ͓S0003-6951͑99͒02007-0͔There is considerable interest in InP-based field-effect transistor technology ͓both metal-semiconductor field-effect transistors ͑MESFET͒ and metal-insulator-semiconductor field-effect transistor ͑MISFET͔͒ due to its high electron mobility and saturated electron drift velocity. However, there are two main limiting factors of device performance: the low Schottky barrier height on InP ͑MESFET͒ and the high surface state density at the insulator-InP interface ͑MISFET͒.

Ex situ deposited SiN x :H/In 0.53 Ga 0.47 As metal-insulator-semiconductor devices, with a minimum of interface state density of 3.5 × 10 11 eV −1 cm −2 have been obtained by electron cyclotron resonance plasma method at a low substrate temperature (200 • C), after a rapid thermal annealing treatment. The effects of annealing temperature on interfacial and bulk electrical properties have been analysed using the C-V high-low frequency method and I-V measurements. The results show that, up to 600 • C, the annealing procedure gradually improves the interface properties of the devices. The frequency dispersion, the hysteresis and the interface trap density diminish, while the resistivity and the electrical breakdown field of the insulator film increase up to values of 8 × 10 15 cm and 4 MV cm −1 , respectively. We explain this behaviour in terms of the thermal relaxation and the reconstruction of the SiN x :H lattice and its interface with the In 0.53 Ga 0.47 As. At higher annealing temperatures, a sharp degradation of the structure occurs.

We present a laboratory experiment to show the current - voltage (I - V), light output - voltage (L - V) and light output - current (L - I) characteristics of light-emitting diodes (LEDs) and semiconductor laser (SLs). The experiment allows us to compare the processes of light emission in LEDs and lasers below the threshold (spontaneous emission) and above it (stimulated emission). We establish unambiguously the analogies of the characteristics of LED emission in comparison with laser emission, when this occurs in the regime of spontaneous emission. We also determine the differences between LED and laser operation (when this occurs in the regime of stimulated emission) in a different way than that usually found in textbooks. Resumen. En este trabajo se presenta un sistema experimental destinado a mostrar las relaciones entre las caracteristicas I - V, L - V y L - I de diodos emisores de luz (LED) y láseres de semiconductor. La experiencia permite comparar los procesos de emisión de luz de los LEDs y láseres por debajo del umbral (emisión espontánea) y por encima de él (emisión estimulada). Mostramos de forma clara las analogías entre el LED y el láser en la zona de emisión espontanea. Cuando el láser trabaja en la zona de emisión estimulada, las diferencias se caracterizan de forma distinta a la habitual de los libros de texto.