Dopant Deactivation Research Articles

Effect of buried Si∕SiO2 interface on dopant and defect evolution in preamorphizing implant ultrashallow junction

P -type ultrashallow junctions are widely fabricated using Ge preamorphization prior to ultralow-energy boron implantation. However, for future technology nodes, issues arise when bulk silicon is supplanted by silicon-on-insulator (SOI). An understanding of the effect of the buried Si∕SiO2 interface on defect evolution, electrical activation, and diffusion is needed in order to optimize the preamorphization technique. In the present study, boron has been implanted in germanium preamorphized silicon and SOI wafers with different preamorphizing implant conditions. Subsequent to implantation an isothermal annealing study of the samples was carried out. Electrical and structural properties were measured by Hall-effect and secondary-ion-mass spectroscopy techniques. The results show a variety of interesting effects. For the case where the Ge preamorphization end-of-range defects are close to the buried oxide interface, there is less dopant deactivation and less transient-enhanced diffusion, due to a lower interstitial gradient towards the surface.

Defects and electrical behavior in 1MeV Si+-ion-irradiated 4H–SiC Schottky diodes

In this paper, the formation and evolution of defects induced by ion irradiation with 1MeV Si+ ions in Ni2Si∕4H–SiC Schottky diodes were studied and correlated with the electrical properties of the contacts. The current-voltage characteristics of the contacts monitored before and after irradiation showed an increase of the Schottky barrier height, of the series resistance, and of the leakage current with increasing irradiation fluence. The changes in the barrier height and in the series resistance values could be attributed to the dopant deactivation in the near-interface region, while the increase of the leakage current was associated with the formation of irradiation-induced defects. These defects showed an evolution with increasing irradiation fluence. Moreover, a combination of deep-level transient spectroscopy and current-voltage measurements of the diodes allowed us to demonstrate that the Z1∕Z2 center of 4H–SiC has the major influence on the increase of the diodes leakage current in the irradiated material.

Amorphous layer depth dependence on implant parameters during Si self-implantation

Preamorphization followed by low temperature solid phase epitaxial regrowth has been proved to provide a high activation of the dopants with minimal diffusion. However, the end of range damage present after regrowth beyond the initial amorphous/crystalline interface causes diffusion and deactivation of dopants during subsequent annealing. In this paper, we study the influence of implant conditions on the depth of the amorphous layer during Si self-implantation. We compare experimental data with our simulation results obtained using an atomistic amorphization–recrystallization model recently developed. We show that the amorphous/crystalline interface depth initially increases with dose but saturates at high doses. Beam current and wafer temperature also alter the depth of the amorphous layer and the amount of residual damage by affecting the dynamic annealing of the damage. These parameters are not always well controlled or specified in experiments and can explain differences observed in dopant profiles.

Physical atomistic kinetic Monte Carlo modeling of Fermi-level effects of species diffusing in silicon

An accurate physically based Fermi-level modeling approach, amenable to be implemented in an atomistic process simulator, is reported. The atomistic kinetic Monte Carlo method is used for point and extended defects, in conjunction with a quasiatomistic, continuum approach treatment for carrier densities. The model implements charge reactions and electric bias according to the local Fermi level, pairing and break-up reactions between particles, clustering-related dopant deactivation, and Fermi-level-dependent solubility. We derive expressions that can be used as a bridge between the continuum and the atomistic frameworks. We present the implementation of two common dopants, boron and arsenic, using parameters that are in agreement with both ab initio calculations and experimental results.

Ion irradiation of inhomogeneous Schottky barriers on silicon carbide

In this paper, the effects of ion irradiation on Schottky barriers formed on silicon carbide are discussed. After Si-ion irradiation at the near-interface region in Ti/4H-SiC contacts an increase of the Schottky barrier height from 1.05 to 1.21 eV was observed, accompanied by a lowering of the reverse leakage current. The combination of several methods allowed us to determine the physical properties of the Schottky barrier and to explain the mechanism responsible for the barrier height changes. In particular, the structural and electrical modifications of the interfacial region, both of Ti layer and SiC (i.e., different orientation of the Ti layer, irradiation-induced defects in the epilayer, dopant deactivation, and the consequent reduction of the surface electric field) are responsible for the increase of the Schottky barrier height and the reduction of the leakage current. The electrical characterization of the contacts at different temperatures also suggested that ion irradiation induced modifications in the inhomogeneous nature of the Ti Schottky barrier.

Ion-Beam Induced Modifications of Titanium Schottky Barrier on 4H-SiC

The effects of ion irradiation on the Ti/4H-SiC Schottky barrier are discussed. The Ti/SiC interfacial region of test Schottky diodes was irradiated with 8 MeV Si+4 ions at fluences between 1×109 and 1×1012 ions/cm2. By increasing ion fluence, an increase of the Schottky barrier FB was observed, from FB=1.05 eV in the non-irradiated sample to FB =1.21 eV after irradiation at the highest fluence, accompanied by a decrease of the leakage current. Using different experimental analytical techniques enabled us to correlate the modification of the interfacial region with the contacts electrical behaviour. In particular, the structural and electrical modifications in the nearinterface region (different orientation of the Ti film, defects in the epilayer, dopant deactivation) were responsible for the change of the Ti barrier after irradiation.

Materials Characterization and Modeling of SiC in Europe - From the Viewpoint of a Theorist

SiC technology is presently still burdened by a number of problems caused by process- or operation-induced defects. Experimental materials characterization in cooperation with atomistic modeling can be helpful in designing strategies against them. In recent years, considerable theoretical effort has been devoted to clarify the dynamics of defect creation and the mechanisms of dopant (de)activation. The investigation of epitaxial growth and of thermal oxidation has also begun. Here an attempt is made to survey the most important theoretical results of the past four years from Europe.

Physical modeling of Fermi-level effects for decanano device process simulations

We report on a physically based Fermi-level modeling approach designed to be accurate and yet amenable to be implemented in a device-size process simulator. We use an atomistic kinetic Monte Carlo method in conjunction with a continuum treatment for carrier densities. The model includes: (i) charge reactions and electric bias according to the local Fermi-level; (ii) pairing and break-up reactions involving charged particles; (iii) clustering-related dopant deactivation; and (iv) Fermi level-dependent solubility. Degenerated statistics, band-gap narrowing, and damage-induced electrical compensation are also included. The parameters used for charged particles are in agreement with ab initio calculations and experimental results. This modeling scheme has proved to be very computationally efficient for realistic device-dimension process simulations. We present an illustrative set of simulation results for two common dopants, boron and arsenic, and discuss the potential of this approach for accurate process simulation of decanano CMOS devices.

Novel polysilicon gate engineering with a laser thermal process for sub-40 nm CMOS devices

We demonstrate a novel poly-Si gate engineering by laser thermal process (LTP) to suppress gate depletion for high performance sub-40 nm CMOS devices. To show the full advantage of LTP, an optimized BEOL process has been employed to suppress dopant deactivation and achieve a reduction of electrical inversion gate oxide thickness by 0.1/0.06 nm (PMOS/NMOS), which improves the I on current by 11/7% (PMOS/NMOS). Furthermore, a novel source-drain rapid thermal annealing (SD-RTA) has been developed to suppress dopant dose loss to achieve a 5% improvement of the I on current for NMOS.

Effects of contamination on selective epitaxial growth

As MOSFET dimensions scale down in size, it has become increasingly difficult to maintain high drive current while suppressing the off-state leakage current. One method of avoiding short-channel effects is to scale the source/drain (S/D) junction depths proportionally with the gate length. Unfortunately, this increases the S/D resistance, which slows the circuit. To keep the S/D junction shallow without affecting the S/D resistance, a raised S/D (RSD) structure is required. Integrating RSD can be difficult. Selective epitaxial growth (SEG) is the process used to incorporate RSD. This process requires a relatively clean surface to initiate the growth. Insertion of SEG earlier in the process flow facilitates selective epitaxial growth. Insertion of SEG later in the process flow results in higher levels of contamination at the interface of the Si substrate and the RSD structure. In this paper, we identify some mechanisms that determine the quality of the selective epitaxial film. Results indicate that Si defects are not a dominant mechanism in SEG film quality. Instead, results suggest that higher levels of contamination increased the surface roughness of the epitaxial film. PMOS regions were found to have higher levels of contamination and rougher epitaxial films than NMOS regions. Hydrogen bake as high as 900 °C was required to lower the surface contamination and provide excellent epitaxial morphology. Unfortunately, this high temperature causes enhanced dopant diffusion and deactivation of the device. Previous work [H. van Meer, K. De Meyer, Symposium on VLSI Technology Digest of Technical Papers, 2002, p. 170.] identified an alternative integration that provides excellent quality selective epitaxy, without dopant diffusion and deactivation.

Effects of contamination on selective epitaxial growth

As MOSFET dimensions scale down in size, it has become increasingly difficult to maintain high drive current while suppressing the off-state leakage current. One method of avoiding short-channel effects is to scale the source/drain (S/D) junction depths proportionally with the gate length. Unfortunately, this increases the S/D resistance, which slows the circuit. To keep the S/D junction shallow without affecting the S/D resistance, a raised S/D (RSD) structure is required. Integrating RSD can be difficult. Selective epitaxial growth (SEG) is the process used to incorporate RSD. This process requires a relatively clean surface to initiate the growth. Insertion of SEG earlier in the process flow facilitates selective epitaxial growth. Insertion of SEG later in the process flow results in higher levels of contamination at the interface of the Si substrate and the RSD structure. In this paper, we identify some mechanisms that determine the quality of the selective epitaxial film. Results indicate that Si defects are not a dominant mechanism in SEG film quality. Instead, results suggest that higher levels of contamination increased the surface roughness of the epitaxial film. PMOS regions were found to have higher levels of contamination and rougher epitaxial films than NMOS regions. Hydrogen bake as high as 900 °C was required to lower the surface contamination and provide excellent epitaxial morphology. Unfortunately, this high temperature causes enhanced dopant diffusion and deactivation of the device. Previous work [H. van Meer, K. De Meyer, Symposium on VLSI Technology Digest of Technical Papers, 2002, p. 170.] identified an alternative integration that provides excellent quality selective epitaxy, without dopant diffusion and deactivation.

Statistical modeling of radiation-induced proton transport in silicon: deactivation of dopant acceptors in bipolar devices

We show that radiation-induced dopant deactivation in MOS capacitors that simulate the base oxides of silicon bipolar transistors is due primarily to direct neutralization by protons. The strong dependence of the deactivation process on electric field is related to the transport of H/sup +/ in the depletion region. The probability of acceptor neutralization near the Si surface is higher for small irradiation biases when protons diffuse through the depletion layer rather than drift through it. The observed dependence of the neutralized-acceptor concentration on the irradiation bias is explained by analytical modeling and statistical Monte Carlo simulations. The analytically estimated and numerically calculated density of passivated acceptors near the Si-SiO/sub 2/ interface is in very good agreement with experimental data. Neutralization of protons in the Si is not necessary to describe the observed dopant deactivation.

Radiation-induced acceptor deactivation in bipolar devices: Effects of electric field

We show that radiation-induced boron acceptor deactivation in silicon bipolar transistors is due primarily to direct neutralization by protons. The strong dependence of the deactivation process on electric field is due mainly to the transport of H+ in the depletion region. The observed dependence of the neutralized-acceptor concentration on the irradiation bias is explained by analytical modeling and Monte Carlo simulations. Neutralization of protons in the Si is not necessary to describe the observed dopant deactivation.

Dopant deactivation in heavily Sb doped Si (001): A high-resolution x-ray diffraction and transmission electron microscopy study

Laser annealing is being studied as an alternative dopant activation technique to form the ultrashallow, low resistivity junctions required in future generations of integrated circuits. This method benefits from the ability to create uniform, box-shaped dopant profiles with concentrations that can exceed equilibrium solubility values. These supersaturated dopant concentrations, however, exist in a metastable state and deactivate upon subsequent thermal processing. Electrical measurements show that antimony deactivation becomes increasingly severe as the concentration increases. High-resolution x-ray diffraction (HR-XRD), secondary ion mass spectroscopy and transmission electron microscopy were used to study the deactivation process. Under most conditions, Sb deactivation occurs through the precipitate mechanism in which the inactive dopant forms precipitates, leading to a decrease in the substitutional dopant concentration. However, the HR-XRD data reveal that at high concentrations above 6.0×1020 cm−3 during low temperature anneals, the inactive dopant forms a type of inactive structure that resides on a substitutional site within the Si matrix.

Thermodynamic and kinetic studies of laser thermal processing of heavily boron-doped amorphous silicon using molecular dynamics

Laser thermal processing (LTP) has been proposed as a means to avoid unwanted transient enhanced diffusion and deactivation of dopants, especially boron and arsenic, during the formation of ultrashallow junctions. Although experimental studies have been carried out to determine the efficacy of LTP for pure Si and lightly B-doped junctions, the effects of high concentrations of dopants (above 2% B) on the thermodynamic and kinetic properties of the regrown film are unknown. In this study, a classical interatomic potential model [Stillinger–Weber (SW)] is used with a nonequilibrium molecular dynamics computer simulation technique to study the laser thermal processing of heavily B-doped Si in the range 2–10 at. % B. We observe only a small effect of boron concentration on the congruent melting temperature of the B:Si alloy, and thus the narrowing of the “process window” for LTP is predicted to be small. No significant tendency for boron to segregate was observed at either the regrowth front or the buried c-Si interface during fast regrowth. The B-doped region regrew as defect-free crystal with full activation of the boron atoms at low boron concentrations (2%), in good agreement with experiments. As the concentration of boron increased, the number of intrinsic Si defects and boron interstitials in the regrown materials increased, with a minor amount of boron atoms in clusters (<2%). An instability limit for crystal regrowth was observed at around 8%–10% boron atoms during fast regrowth; systems with 10% B showed partial amorphization during regrowth. Comparison with tight-binding quantum mechanical calculations showed that the SW model gives similar diffusivities in the liquid and tendency to cluster, but the lifetimes of the SW clusters are considerably too long (>150 ps, compared to 5 ps in tight binding). The importance of adequate system size is discussed.

Physical processes associated with the deactivation of dopants in laser annealed silicon

Laser annealing is being investigated as an alternative method to activate dopants and repair the lattice damage from ion implantation. The unique properties of the laser annealing process allow for active dopant concentrations that exceed equilibrium solubility limits. However, these super-saturated dopant concentrations exist in a metastable state and deactivate upon subsequent thermal processing. Previously, this group compared the electrical characteristics of the deactivation behavior of common dopants (P, B, and Sb) across a range of concentrations and annealing conditions. Boron and antimony were shown to be stable species against deactivation while P and As deactivate quickly at temperatures as low as 500 °C. In this work, we present additional data to understand the underlying physical mechanisms involved in the deactivation process. It is proposed that As and P deactivate through the formation of small dopant—defect clusters while B and Sb deactivate through precipitation.

Study of reverse annealing behaviors of p+/n ultrashallow junction formed using solid phase epitaxial annealing

Solid phase epitaxial (SPE) annealing at low temperature has the advantage of high dopant activation and very little dopant diffusion. However, due to the low thermal budget engaged in SPE, a large amount of defects can exist in the area beyond the original interface of the crystal and the pre-amorphized layer. These defects may cause severe junction leakage. They may also cause dopant diffusion and deactivation in a following higher temperature process. This work studies the reverse annealing behaviors during a second annealing step for SPE-formed p+/n junction using either 1 keV B+ or 5 keV BF2+ implants. Four-point probe, secondary-ion-mass spectroscopy, and transmission electron microscopy are used in this study. The results show that the boron deactivation after second-step annealing is not only correlated with the transmission electron diffraction (TED), but also correlated with the end of range defect evolution. The results also show that BF2 implanted wafer has slower boron deactivation, less TED and slower defect evolution than the B implanted wafer. In addition, the BF2 implanted wafer shows a much slower SPE regrowth rate at 550 °C than the B implanted wafer.

Electrical and Structural Characterization of Boron Implanted Silicon Following Laser Thermal Processing

AbstractOne alternative to conventional rapid thermal annealing (RTA) of implants for ultra-shallow junction formation is that of laser annealing. Laser thermal processing (LTP) incorporates an excimer pulsed laser capable of melting the near surface region of the silicon (Si) substrate. The melt depth is dependent upon the energy density supplied by the irradiation source and the melting temperature of the substrate surface. A process window associated with this technique is able to produce similar junction depths over a range of energy densities due to the melting temperature depression established with pre-amorphization of the substrate surface prior to dopant incorporation. The process window of germanium (Ge) preamorphized, boron (B) doped Si was investigated. 200 mm (100) n-type Si wafers were preamorphized via 18 keV Ge+ implantation to 1x1015/cm2 and subsequently implanted with 1 keV B+ to doses of 1x1015/cm2, 3x1015/cm2, 6x1015/cm2, and 9x1015/cm2. The wafers were laser annealed from 0.50 J/cm2 to 0.88 J/cm2 using a 308 nm XeCl excimer irradiation source. Transmission electron microscopy (TEM) was used to determine the process window for each implant condition, and correlations between process window translation and impurity concentration were made. Four-point probe quantified dopant activation and subsequent deactivation upon post-LTP furnace annealing.

Deactivation kinetics of supersaturated boron:silicon alloys

The effect of laser annealing on the electrical activity of boron-doped silicon wafers has been investigated as a function of boron concentration, annealing time, and annealing temperature (from 600 °C to 1050 °C). Metastable supersaturated alloys were produced by the laser annealing of ion-implanted Si 〈100〉 wafers using an excimer laser with a pulse duration of 30 ns. The extent of dopant activation, deactivation, and tendency towards precipitation were subsequently studied following rapid thermal annealing in an argon ambient using a four-point probe of the sample resistance. Sheet resistances as low as 15 Ω/□ were achieved in 200 nm layers. Following laser anneals, boron atoms remained active at concentrations of 7.5×1020/cm3 up to 800 °C for 210 s. A two-mode relaxation model including defect association and precipitation was proposed to describe the annealing behavior. These results show that laser processing can produce metastable B-doping levels, stable to moderate thermal processing, at concentrations adequate for all anticipated device structures.

The influence of polysilicon gate morphology on dopant activation and deactivation kinetics in deep-submicron CMOS transistors

In this paper, the impact of gate microstructure on the activation and deactivation kinetics of ion-implanted dopants is discussed. A comparison is made between large-grained polysilicon that was obtained by recrystallizing deposited amorphous silicon, and fine-grained polysilicon. Very good gate activation was achieved for both As- and B-implanted fine-grained polysilicon gates using post-implant rapid thermal annealing for 20 s at 980–1010°C, leading to gate depletion levels below 5%. Similar levels of gate activation are found for spike annealing of 1 s at 1100°C. The kinetics of dopant deactivation in large- and fine-grained polysilicon have been measured in the temperature range from 700 to 800°C. From these measurements, the loss in the gate activation due to extra annealing steps in a 0.13 μm CMOS process flow has been evaluated. Silicidation is the most harmful process step leading to an increase in the gate depletion level to 6% for a p-type fine-grained polysilicon gate, representing the worst case situation. Finally, TSUPREM-4 gate depletion simulations for future CMOS generations (≤0.10 μm) have revealed the need of high dopant concentration (4–5×10 20 cm −3) which questions the possibility of polysilicon use as a gate material.