Silicon Nitride Passivation Film Research Articles

Rapid progression and subsequent saturation of polarization-type potential-induced degradation of n-type front-emitter crystalline-silicon photovoltaic modules

In this study, we investigated progression of potential-induced degradation (PID) in photovoltaic modules fabricated from n-type-based crystalline-silicon cells with front p+ emitters. In PID tests in which a bias of −1000 V was applied to the modules, they started to degrade within 5 s and their degradation saturated within 60 s. This behavior suggested that the PID was caused by positive charge accumulation in the front passivation films. Performing PID tests with a bias of −1500 V revealed that the degradation rate strongly depended on the applied bias whereas the saturation value was independent of the applied bias. Regeneration tests on degraded modules previously subjected to PID tests for durations of 5 and 10 min were performed by applying a positive bias of +1000 V. All the degraded modules completely recovered their performance losses within 60 s regardless of the degradation test duration. On the basis of these results, we proposed that these positive charges originate from positively charged K centers formed by extracting electrons from neutral and negatively charged K centers. This model readily explains the observed degradation and regeneration behavior. To test our model, we determined the fixed positive charge densities (Qf) of a silicon nitride passivation film before and after PID, for which it was found that Qf showed similar saturation behavior. Additionally, the saturated Qf value was of the same order as K center density. These results support our model involving a charging process of K centers.

Direct observation of changes in the effective minority-carrier lifetime of SiNx-passivated n-type crystalline-silicon substrates caused by potential-induced degradation and recovery tests

We directly observed reductions in the effective minority-carrier lifetime (τeff) of n-type crystalline silicon (c-Si) substrates with silicon-nitride passivation films caused by potential-induced degradation (PID). We prepared PID-test samples by encapsulating the passivated substrates with standard photovoltaic-module encapsulation materials. After PID tests applying −1000V to the c-Si samples from the glass surface, the τeff was decreased, which probably pertains to Na introduced into the c-Si. After PID tests applying +1000V, the sample, on the other hand, showed a considerably rapid τeff reduction, probably associated with the surface polarization effect. We also performed recovery tests of predegraded samples, by applying a bias opposite to that used in a degradation test. The τeff of a sample predegraded by applying +1000V was rapidly completely recovered by applying −1000V, while those of predegraded by applying −1000V show only slight and insufficient τeff recovery.

Impact of the silicon-nitride passivation film thickness on the characteristics of InAlAs/InGaAs InP-based HEMTs

In this research, the DC and RF characteristics for our own InAlAs/InGaAs InP-based high-electron-mobility transistors (HEMTs) with 0-nm, 50-nm, and 200-nm silicon-nitride (Si3N4) passivation films have been investigated comprehensively and systematically. With increasing Si3N4 passivation film thickness, the device exhibits an apparent improvement in the output conductance and a distinct drift of the pinch-off voltage toward a positive value. Especially, the device passivated by using a 50-nm Si3N4 film demonstrates the highest extrinsic transconductance and channel current. The appeared changes of the DC properties can be interpreted as due to the passivationinduced positively-charged surface state increasing the sheet carrier density with the carrier mobility and parasitic resistance of device being deteriorated by long-time high-stress-intensity ion bombardment. Additionally, the degradations of the current-gain cutoff frequency (f T) and maximum oscillation frequency (f max) can be attributed to the drastically-increased parasitic capacitance and resistance induced by the passivation process. This work will be of great importance in fabricating InP HEMTs with high performances.

High‐power GaN‐HEMT with low current collapse for millimeter‐wave amplifier

AbstractThere are high expectations for GaN high electron mobility transistors (GaN‐HEMT) to serve as a key device in the production of a high power and high efficiency millimeter‐wave amplifier. Until now, an extremely low three‐terminal pinched‐off‐current of less than 10‐6 A/mm was achieved by applying band engineering. It improved the on‐state breakdown voltage to 50 V and the off‐state breakdown voltage to over 100 V (Makiyama et al., Phys. Status Solidi C 6(S2), S1012 (2009) [1]). However, the output power of a GaN‐HEMT decreases due to current collapse. In particular, the influence of current collapse increases in a GaN‐HEMT with a short gate‐length. In this work, technology was developed to reduce this amount of current collapse. By utilizing thin silicon nitride passivation film that contained many Si‐H bonds, the current collapse was remarkably improved. A 3‐stage common source amplifier MMIC was fabricated that used the current collapse reducing technology. The Pin vs. Pout characteristics were measured at the W‐band. Excellent power characteristics, such as Pout of 1.3 W at 75 GHz at 35 V operation, and a frequency bandwidth of 13 GHz were demonstrated. (© 2011 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)

The Effect of Surface Cleaning on Current Collapse in AlGaN/GaN HEMTs

The phenomenon of current collapse is a limiting factor in the performance of AlGaN/GaN high electron mobility transistors (HEMTs), and can be ameliorated by the deposition of a silicon nitride passivation film on the surface. The effect of three types of surface cleaning prior to the application of a silicon nitride passivation layer are studied. The best results were obtained when the wafers were cleaned using an air plasma descum followed by an HCl dip prior to the deposition of the silicon nitride passivation. Auger electron spectroscopy depth profiling indicated that the degree of collapse was correlated with the amount of residual carbon contamination at the silicon nitride/AlGaN interface.

Effects of Silicon Nitride Passivation Layer on Mean Dark Current and Quantum Efficiency of CMOS Active Pixel Sensors

AbstractIn the present study, we have analyzed the influence of the silicon nitride (SiN) passivation layer properties on the mean dark current and the quantum efficiency of CMOS APS (Active Pixel Sensor) through electrical characterization of lot wafers processed with three different SiN passivation films. The SiN layers were characterized by Spectroscopic Ellipsometry (SE) and Fourier Transform Infra Red (FTIR) spectroscopy to get the optical indices and the hydrogen content of the films, respectively. Hydrogen desorption was also studied by Thermal Desorption Spectrometry (TDS) experiments.The different chemical bond concentrations enable to explain the device performance. It is shown that high [Si-H] and low [Si-N] bonds concentrations lead to high hydrogen desorption from the SiN films. Thus, the lowest dark current values have also been obtained with such silicon nitride passivation layers. Consequently, results are in agreement with the hydrogen passivation of defects being responsible of thermally generated carriers. Concerning the quantum efficiency (QE), it is highly influenced by the optical indices of the SiN passivation layer.Actually, the thickness and complex refractive index of the SiN layer affect the light reflection, while the light absorption, in the visible range, is controlled by the [Si-Si] bonds concentration. There are also competing requirements to minimize dark current and to maximize quantum efficiency. In fact, high [Si-H] and low [Si-N] bonds concentrations (for good passivation) are generally observed in silicon-rich SiN films, which also contain high amount of [Si-Si] bonds (inducing high absorption, thus low QE). Increasing simultaneously the [Si-H] and [N-H] bonds concentrations in the passivation layer can be a way to have high hydrogen desorption during passivation anneal just preserving a transparent layer.

Effects of surface treatment on the velocity–field characteristics of AlGaN/GaN heterostructures

AlGaN/GaN test structures were fabricated with an etched constriction. A nitrogen plasma treatment was used to remove the disordered layer, including natural oxides on the AlGaN surface, before the growth of the silicon nitride passivation film on several of the test structures. A pulsed voltage input, with a 200 ns pulse width, and a four-point measurement were used in a 50 Ω environment to determine the room temperature velocity–field characteristic of the structures. The samples performed similarly over low fields, giving a low-field mobility of 545 cm2 V−1 s−1. The surface treated sample performed slightly better at higher fields than the untreated sample. The highest velocity measured was 1.25 × 107 cm s−1 at a field of 26 kV cm−1.

X-cut LiNbO<sub>3</sub> optoelectronic device passivated by silicon nitride film

Silicon nitride (Si3N4) passivation was attempted on x-cut LiNbO3 10-Gb/s optical integrated modulators. Although silicon nitride films are commonly used to chemically protect semiconductor device surfaces,1 to our knowledge, applicability of the film to rf optoelectronic devices such as high-speed LiNbO3 modulators has not been reported. The purpose of our investigation is to confirm that the silicon nitride passivation films do not degrade the electro-optic (EO) characteristics and bias-drift performance of LiNbO3 modulators.

Adhesion Quantification of Post-CMP Copper to Amorphous SiN Passivation by NanoIndentation

ABSTRACTIn this study, film interfacial fracture is induced by nanoindentation to quantify the practical work of adhesion of a post-CMP copper film to an amorphous silicon nitride passivation film. Poor adhesion of electrodeposited copper to SiN passivation is observed following CMP due to copper oxide growth prior to plasma enhanced silicon nitride deposition. Four point bend testing has shown that failure by brittle fracture of test structures occurs at the Cu/CuO interface. Hydrogen, ammonia, and nitrogen plasma treatments of the post-CMP copper surface can be used to remove the oxide, shown by auger electron spectroscopy, and to increase the surface roughness of copper, shown by atomic force microscopy. Both effects can be used to improve the Cu/SiN adhesion. Nanoindentation with a conical indenter (1.59 μm tip radius) was used to induce SiN film delaminations from Cu, the sizes of which were measured and correlated with the practical work of adhesion.[1,2] In order to more reliably and repeatably produce these delaminations a TiW (10wt% Ti) superlayer was sputter deposited on to the test structures.[2,3] Mechanical properties, including elastic modulus and hardness of SiN, electrodeposited copper, and TiW measured by nanoindentation are also reported here.

The Influence of Plasma Silicon Nitride Passivation Film Quality on Aluminum Void Formation

Voiding in wide aluminum interconnects at high‐temperature heat‐treatment has been studied in conjunction with plasma silicon nitride passivation film formation conditions. Void formation largely depends on the plasma silicon nitride deposition temperature, lower void density for higher temperature deposited films, and higher density voiding for lower temperature deposited films, which cannot be explained by a simple stress model. Experiments are described to study these dependencies and the results are explained by proposing a model. The combination of the different temperature dependence of the transient pseudo‐elastic stress change and the successive plastic stress relaxation of plasma silicon nitride film induces different stress states in aluminum lines when samples are heated up. The dependence of aluminum void density on the plasma silicon nitride formation condition can be explained with this stress difference.

Correlation between Mechanical Stress and Hydrogen-Related Effects on Radiation-Induced Damage in MOS Structures

Correlation between mechanical stress and hydrogen effects on radiation damage in polycide-gate MOS capacitors was investigated as a function of gate-oxide thickness. The compressive stress magnitude was altered by varying the silicide (TiSi2 or WSi2) thickness in the polycide-gate electrode, and hydrogen introduction into gate-SiO2 film was carried out by diffusion from plasma-deposited silicon-nitride passivation film (SiN-Cap). In a MOS capacitor without passivation film (No-Cap sample), it was found that compressive stress on gate-SiO2 reduces both positive charge build-up (#x00394;Qot) and interface-trap generation (#x00394;Dit). Radiation induced shift, #x00394;Qot exhibits a smaller stress effect as compared witni #x00394;Dit. As gate-SiO2 thickness decreases, the stress effect on #x00394;Qot increases, while this effect on #x00394;Dit remains nearly constant. This compressive stress effect was inhibited by hydrogen diffusion, especially in #x00394;Dit. In addition, both stress effects on #x00394;Qot and #x00394;Dit in SiN-Cap samples show a gate-oxide Thickness dependence similar to that on #x00394;Qot in No-Cap samples. The stress effect observed in No-Cap samples can be explained on the basis of the bond reformation process at the SiO2/Si interface and near the electrode/SiO2 interface. In a SiN-Cap sample, hydrogen is highly distributed at the SiO2/Si interface region and hydrogen-related interface-trap generation may occur at this interface region. As a result, only the stress effect near the gate-electrode was observed in SiN-Cap samples.