Plasma Passivation Research Articles

Novel polycrystalline silicon thin film transistors prepared by deposition from Si 2H 6 and subsequent annealing by NH 3 plasma

This work investigated the channel layer of polycrystalline silicon (poly-Si) thin film transistors (TFTs) prepared by amorphous silicon (a-Si) films deposited using Si 2H 6 gas. The recrystallization of channel layers, source/drain, gate electrodes and post implant anneal were performed at the same time. Due to the larger grain size, the device has higher field effect mobility than SiH 4 deposited devices. These devices were also subsequently passivated by NH 3 plasma. The NH 3 plasma significantly improves the n-channel devices; however, the improvement of p-channel devices is limited. Especially, the threshold voltage of n-channel devices is significantly shifted toward the negative gate voltage than the shift magnitude of p-channel devices. To investigate the band gap width and Fermi level by determining the leakage activation energy, it is found that the channel film is changed slightly from p-type to n-type. These results may be attributed to the donor effect by NH 3 plasma passivation.

Effects of etch-induced damage on the electrical characteristics of in-plane gated quantum wire transistors

In-plane gated (IPG) quantum wire transistors were fabricated using dry etching in a Cl2/Ar plasma generated with an electron cyclotron resonance source. The electrical characteristics of the IPG transistors were correlated with the geometrical dimensions as well as the dry etching and passivation conditions. In-plane gates with the width of the channel (Wc) and the width of the gate isolation (Wg) ranging from 100 to 850 nm were studied. Good field-effect transistor characteristics with transconductances up to 371 mS/mm were obtained on these devices. At a gate-source voltage (VGS) of 2 V, the saturated drain-source current (IDSAT) increased from 68 to 153 μA as Wc increased from 440 to 800 nm. No current was measured on IPG transistors with Wc ≤−130 nm. The quasi-one-dimensional channel can be completely pinched off with VGS ≤−1 V. It was found that the gate leakage current decreased with a wider Wg and a deeper depth for the gate isolation. The leakage current at VGS=2 V decreased significantly from 250 to <0.1 pA when the etch depth increased from 320 to 440 nm. The gate leakage current and IDS were also found to increase with rf power used for etching due to additional defects generated at higher ion energy. These defects, however, can be passivated with low energy chlorine species, and reduction of the gate leakage current from 40 to 4.4 nA was observed after a 1 min Cl2 plasma passivation.

Spatial distributions of power and ion densities in RF excited remote plasma reactors

Remote plasma systems operating at moderate pressures (tens to hundreds of milli-Torr) are being developed for use in deposition and etching of microelectronics materials. In particular, remote plasma-enhanced chemical vapour deposition (RPECVD) has been investigated for fabrication of c-Si, and films, as well as for plasma cleaning and passivation. RPECVD reactors typically consist of a narrow upstream plasma zone and a wide downstream deposition chamber. A sub-set of the reactants is made to flow through the upstream plasma zone, creating excited states which are mixed with additional reactants injected into the downstream deposition chamber. RPECVD systems are typically excited by a radio frequency electric field produced by a coil surrounding the upstream plasma zone with the intent of generating a plasma that is well confined to the upstream zone. It is common, however, for the plasma to extend downstream towards the substrate due to stray inductive fields and capacitive coupling. In this paper, a computer model for remote plasma reactors is described, with which the spatial distributions of power deposition and ion densities are investigated. The characteristics of remote plasma reactors are presented and the influences of the operating conditions (geometry, gas pressure and RF frequency) on plasma confinement are investigated for He, and He - mixtures.

Hydrogen plasma passivation of InP: Real time ellipsometry monitoring and ex situ photoluminescence measurements

The remote hydrogen plasma passivation of semi-insulating InP substrates has been investigated by in situ spectroscopic ellipsometry (SE), to characterize the changes of InP morphology, and ex situ photoluminescence (PL) to reveal the passivation effect of nonradiative centers operated by H atoms. The concurrence of the InP native oxide removal without surface damage (phosphorus ablation) and the passivation by H-atoms results in a remarkable enhancement of the PL intensity.

Near-surface dopant passivation after wet-chemical preparation

Recent advances in wet-chemical surface treatment of silicon has resulted in unreconstructed, hydrogen terminated Si(111) surfaces with extremely low surface recombination velocity of only 0.25 cm/s [E. Yablonovitch et al., Phys. Rev. Lett. 57, 249 (1986)]. This points towards a surface free of surface states that normally pin the Fermi level EF on Si(111) somewhere near midgap depending on the surface reconstruction. This expectation is not borne out, however, by experiment. Using Si 2p core level spectra, we find consistently a surface Fermi level position near midgap for 16 heavily doped n- and p-type samples. Checks for surface photovoltage and ion concentrations sufficient to induce the observed band bending were negative. All samples attain their bulk Fermi level position at the surface after annealing to about 400 °C, a temperature at which the surface hydrogen coverage is unchanged as monitored by ultraviolet excited photoemission. Above 400 °C, hydrogen starts to desorb from the surface and only with the onset of the 7×7 reconstruction does EF move back towards midgap. We suggest that the initial position of EF near midgap is due to a hydrogen induced passivation of dopants that leaves an intrinsic surface layer. Annealing at ∼400 °C reactivates the dopants by breaking the B–H and P–H complexes as is known from plasma passivation. The necessary passivation depth has been calculated as a function of bulk doping yielding values of the order of one micrometer.

Influence of hydrogen incorporation into Silicon on the room-temperature photoluminescence

AbstractWe investigated the possibility to use the photoluminescence (PL) emission at room temperature for the characterization of grain boundary and surface passivation of μc-Si and single crystalline silicon. The PL-spectra of c-Si and μc-Si taken at 300 K consist of an emission band around 1.1 and 1 eV, respectively. Measuring the PL intensity we study the influence of electrochemical and plasma passivation by hydrogen. We compare the behavior of the films to that of Si crystals. In this case the PL intensity is high for a well H-terminated c-Si surface and decreases during cathodic polarization (hydrogen evolution) which points to the formation of additional non-radiative recombination centers. Whereas the PL intensity of the as prepared μc-Si films is unchanged, that of the annealed films increases during hydrogen evolution indicating the electrochemical passivation of defects on grain boundaries.

The Influence of Hydrogen Plasma Passivation on Electrical and Optical Properties of Aigan Samples Grown on Sapphire

AbstractHydrogen plasma passivation effects are studied for undoped AlGaN layers grown by MOCVD. Hydrogen treatment at 200°C for lh led to a substantial decrease in carrier concentration accompanied by an increase in electron mobility. The magnitude of the near-band edge absorption tails also dramatically decreased after hydrogen passivation. The effect is explained by pairing of negatively charged hydrogen acceptors with positively charged native donors commonly believed to be related to nitrogen vacancies. The bond strength in such hydrogen complexes increases as the composition of AlGaN moves towards AIN, as revealed by results of post hydrogenation annealing.

Polycrystalline silicon thin film transistors fabricated at reduced thermal budgets by utilizing fluorinated gate oxidation

Polycrystalline silicon thin film transistors have been fabricated at reduced gate oxidation thermal budgets by utilizing NF/sub 3/-enhanced dry oxidation. Good performance TFTs with effective electron mobility values as high as 38 cm/sup 2//V.sec, threshold voltage values near zero, ON/OFF current ratios of up to 5/spl times/10/sup 7/ and subthreshold slopes of 0.3 V/dec have been fabricated at an oxidation temperature of 800/spl deg/C. Stable devices at an electrical stressing field of 3 MV/cm were demonstrated. Thermal gate oxide TFTs have also been fabricated at a maximum temperature of 650/spl deg/C. The effect of hydrogen plasma passivation was found to depend on process conditions and was correlated with the amount of fluorine in the area near the Si-SiO/sub 2/ interface. Passivation at low power was always beneficial. Passivation at high power was highly beneficial for a limited amount of interfacial fluorine, but less beneficial or even detrimental when a large fluorine amount in the near interface area was present.

Plasma passivation of etch-induced surface damage on GaAs

Various plasma passivation techniques for removal of etch-induced damage on GaAs have been studied. It was found that a Cl2 plasma generated with an electron cyclotron resonance source can efficiently remove dry etch-induced damage with minimal etching of the GaAs. Complete recovery of the electrical characteristics of both the Schottky diodes and unalloyed transmission lines was found with a 30 s Cl2 plasma passivation at 25 °C. The chlorine reactive species used for passivation were generated with 50 W microwave power at 2 mTorr without any rf power applied at the stage. The Cl2 passivated surface was thermally stable up to 450 °C. Similar recovery was also observed for diodes passivated with a N2 plasma. Compared to Cl2, however, N2 plasma passivation requires a higher temperature (350 °C) and higher microwave power (500 W). Capacitance–voltage measurements show that the presence of H2 in the plasma during passivation results in dopant depletion near the surface, but the dopants can be reactivated after annealing at temperature ≥450 °C for 3 min. Plasma passivation with H2S was found to result in a partial recovery of the electrical characteristics for the etched diodes and transmission lines. Annealing at 300 °C is also required after H2S plasma passivation to desorb the excess S on the GaAs surface. Changes in the defect density as a function of the conditions used for passivation have been correlated to Schottky diode characteristics.

The effects of NH3 plasma passivation on polysilicon thin-film transistors

The NH <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sub> plasma passivation has been performed for the first time on the polycrystalline silicon (poly-Si) thin-film transistors (TFT's). It is found that the TFT's after the NH <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sub> plasma passivation achieve better device performances, including the off-current below 0.1 pA/μm and the on/off current ratio higher than 10/sup 8/, and also better hot-carrier reliability as well as thermal stability than the H <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> -plasma devices. These improvements were attributed to not only the hydrogen passivation of the grain-boundary dangling bonds, but also the nitrogen pile-up at SiO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> /poly-Si interface and the strong Si-N bond formation to terminate the dangling bonds at the grain boundaries of the polysilicon films.

Efficient combination of surface and bulk passivation schemes of high-efficiency multicrystalline silicon solar cells

Conventional and electromagnetically casted multicrystalline silicon solar cells are fabricated following different passivation schemes. Thin layers (∼100 Å) of thermal dry and plasma-enhanced chemical-vapor-deposition (PECVD) SiO2 are implemented for surface oxide passivation of multicrystalline silicon solar cells and compared. It is found that growing thin layers of thermal dry oxide results in efficient surface passivation. However, for thin PECVD SiO2 layers it is necessary to perform low-temperature forming gas anneal, postdeposition, in order to observe the surface passivation effect. In addition, hydrogen plasma passivation has been optimized for achieving deep penetration of atomic hydrogen in the material (≳30 μm) and as a consequence very effective bulk passivation of multicrystalline silicon solar cells. By combining front and back thermal dry SiO2 passivation with hydrogen remote plasma treatment, a cell efficiency of 17% (independently confirmed) on 4 cm2 area and 180 μm thickness is realized without any Al gettering. On the other hand, the cell efficiencies obtained using thin layers of PECVD SiO2 are found to be very comparable to the efficiency of the cells fabricated with thermal dry SiO2 layers provided that PECVD Si3N4/SiO2 are used as a double-layer antireflection coating.

Characterization of [formula omitted] plasma passivation process for poly-Si thin film transistors (TFTs)

The effects of nitrogen additives on the plasma hydrogenation of polycrystalline silicon thin film transistors (poly-Si TFTs) have been investigated with various radio-frequency (RF) power densities, substrate heating temperature, gas flow rates, as well as chamber pressures. The nitrogen-containing hydrogen ( H 2 N 2 ) plasma treatments show better passivation effects on the electrical characteristics of the poly-Si TFTs than the pure H 2 hydrogenation. It is attributed to the passivation effect of the nitrogen radicals themselves and the promotion of the hydrogen plasma generation due to the radical collision. The passivation effects have been enhanced by properly chosen RF power density, gas flow rate and chamber pressure. Furthermore, the H 2 N 2 plasma were also utilized to passivate the poly-Si TFTs with different grain structures.

Effects of proton implantation and hydrogen plasma passivation on electrical properties of InGaAlP and InGaP

It is shown that both acceptors and donors can be passivated by hydrogen in InGaP and InGaAlP and that high-resistivity (about 10 3 Ω cm) layers of p-InGaAlP and of p-InGaP can be prepared by hydrogen plasma treatment at 250°C. Proton implantation with energy 100 keV is shown to create high-resistivity layers (10 4–10 5 Ω cm) both in n and in type InGaP and InGaAlP.

Real time in situ monitoring of surfaces during glow discharge processing: NH3 and H2 plasma passivation of GaAs

Numerous device applications of GaAs are hampered by poor electronic properties of GaAs surfaces and interfaces. Room temperature NH3 and H2 downstream plasma passivation of native oxide contaminated GaAs surfaces is investigated using attenuated-total-reflection (ATR) Fourier-transform-infrared spectroscopy (FTIR) and photoluminescence (PL). Using ATR FTIR concentrations of –As–O, –As–H, H–OH, and C–H bonds are monitored in situ and in real time during exposure of the GaAs surface to H (D) atoms from a microwave discharge through NH3 (ND3) and H2 (D2). Photoluminescence intensity from the GaAs is monitored simultaneously with the FTIR spectra and used as a measure of surface state reduction. At room temperature, H atoms produced from the discharge remove –As–O and C–H contaminants but not Ga2O3. The appearance of As–H bonds and corresponding increase in PL are delayed significantly from plasma initiation, an increase in –O–H concentration, and removal of –As–O bonds. Because the As antisite defects are concentrated at the GaAs–oxide interface, the rates of As–H formation and PL enhancement may be limited by diffusion of H through the oxide and water layer which forms on the surface. We find that the concentration of physisorbed H2O on the GaAs surface increases throughout passivation. There are two sources of water detected on the surface: (i) reduction of As–O bonds and (ii) reaction of H with the quartz reactor walls. Ga2O3 grows on the surface via oxidation of GaAs by the physisorbed water. We surmise that higher H concentration in NH3 plasmas results in faster water accumulation on the surface and trapping of the As–H species subsurface. In H2 plasma, water accumulation on the surface is slower and trapping of As–H is not observed.

High-performance p-channel poly-Si TFT's using electron cyclotron resonance hydrogen plasma passivation

This letter presents a summary of the first detailed investigation of electron cyclotron resonance (ECR) hydrogen plasma exposure treatments of p-channel poly-Si thin film transistors (TFT's). It is shown that ECR hydrogenation can be much more efficient than RF hydrogenation. Poly-Si p-channel TFT's fabricated at low temperatures (/spl les/625/spl deg/C) and passivated with the ECR hydrogenation treatment are shown to exhibit ON/OFF current ratios of 7.6/spl times/10/sup 7/, subthreshold swings of 0.62 V/decade, threshold voltages of -4.6 V, and hole mobilities over 18 cm/sup 2//V.s. >

The effects of fluorine passivation on polysilicon thin-film transistors

The fluorine implantation on polysilicon was found to improve the characteristics of polysilicon thin-film transistors (TFT's). The fluorine passivates the trap states within the polysilicon channel, as compared with the H/sub 2/-plasma passivation. The fluorine implantation passivates more uniformly both the band tail-states and midgap deep-state, while the H/sub 2/-plasma treatment is more effective to passivate deep states than tail states. A fluorine-implanted device can be further improved its performance if an H/sub 2/-plasma treatment is applied. In contrast to the H/sub 2/-plasma passivation, the fluorine passivation improves the device hot-carrier immunity. Combining the fluorine passivation and H/sub 2/-plasma passivation, a high performance TFT with a high hot-carrier immunity can be obtained. >

Real-time monitoring of surface chemistry during plasma processing

Abstract

Plasma Passivation of III-V Semiconductor Surfaces

Plasma Passivation of III-V Semiconductor Surfaces

Plasma passivation of gallium arsenide*

Improved electrical properties of the SiO2–GaAs interface have been obtained using in situ plasma surface treatments prior to film deposition. We present a comparison of several hydrogen based plasma passivation schemes: H2, H2/N2, and H2S. Hydrogen plasmas remove native oxides, while nitrogen or sulfur form passivating surface layers. Samples were characterized using metal–insulator–semiconductor C–V analysis, x-ray photoemission spectroscopy (XPS) and spectroscopic ellipsometry. The H2/N2 and H2S treated samples display improved C–V characteristics. XPS indicates the presence of nitrogen and sulfur respectively on the uncapped samples, although little evidence remains after SiO2 deposition. H2S plasmas offer the best results, providing a self-terminating process that prevents roughening of the GaAs surface by hydrogen plasma etching. However, surfaec doping effects were observed after exposure to high temperatures.

Real-time, in situ monitoring of surface reactions during plasma passivation of GaAs

Real-time, in situ observations of surface chemistry during the remote plasma passivation of GaAs is reported herein. Using attenuated total reflection Fourier transform infrared spectroscopy, the relative concentrations of -As-O, -As-H, -H2O, and -CH2 bonds are measured as a function of exposure to the effluent from a microwave discharge through NH3, ND3, H2, and D2. The photoluminescence intensity (PL) from the GaAs substrate is monitored simultaneously and used qualitatively to estimate the extent of surface state reduction. It was found that, while the -CHx(x = 2,3) and -As-O concentrations are reduced rapidly, the rates at which the -As-H concentration and the PL intensity increase are relatively slow. The concentration of -H2O on the GaAs surface increases throughout the process as surface arsenic oxides and the silica reactor walls are reduced by atomic hydrogen. These observations suggest that removal of elemental As by reaction with H at the GaAs–oxide interface limits the passivation rate.