Transient Enhanced Diffusion (TED) is one of the biggest modeling challenges present in predicting scaled technologies. Prior work has concentrated on TED resulting from sub 200 KeV implants. However, the results applicable to the low energy regime cannot be extrapolated per se into the high energy regime. Furnace anneals have been performed and through simulations it has been observed that there is a differential diffusivity enhancement between the part of the profile before the profile peak and that beyond the profile peak. Longer saturation times compared to the low energy implants have also been observed.

Boron and antimony doped superlattices (DSLs) were implanted with arsenic at 40 keV to doses of 2×1014 cm−2, 5×1015 cm−2 and 2×1016 cm−2. Increasing the arsenic dose above 5×1015 cm−2 resulted in a reduction in the extent of arsenic transient enhanced diffusion (TED) following annealing at 700°C, 16 hr. Concurrent with this reduction in TED was a reduction in the number of free interstitials beyond the end-of-range, as measured by the boron diffusion enhancement in the doped superlattices. No enhancement in antimony diffusivity was observed in this region, indicating that vacancies play no direct role in the diffusion of arsenic in this region, although an indirect role for vacancies as recombination centers for mobile interstitials is not precluded by these experiments. We conclude that interstitials dominate arsenic diffusion in the end-of-range region and beyond. Interpretation of the DSL data in the projected range region is complicated by Fermi level and segregation effects and no definitive conclusion can be reached about the point defect populations in this region.

We discuss the sample preparation technique for the two-dimensional dopant concentration profiling in silicon based on chemical staining and transmission electron microscopy analysis. The capability of characterizing ULSI silicon devices is demonstrated by showing an example of failure analysis performed on a high speed bipolar transistor. The high spatial resolution of the technique allows to reveal many details which are not detectable by other techniques. The limitations of this method concern the impossibility to simultaneously delineate n and p-type regions on the same sample, and the need to have a p+n junction for delineating a boron profile with an acceptable sensitivity. We demonstrate that these limitations can be overcome by developing a novel electrochemical staining procedure based on anodic oxidation of silicon.

Very shallow p+/n junctions (lower than 60 nm) have been fabricated by implanting Sb and subsequently BF2, at a higher dose, in a n-type Si substrate. The preamorphisation with Sb avoids the B channeling and increases the n-type doping in the junction region, thus confining the depth of the p layer. Furthermore, both the transient enhanced diffusion, being the B implanted in a preamorphized layer, and the standard diffusion, due to the pairing between donors and acceptors, are strongly reduced. This procedure allows us to obtain very shallow junctions even after annealings with relatively high thermal budget, like 800 C/8h, or 900 C/lh, or 950 C/15min or 1000 C/60s. Dopant diffusion is strongly affected by the direct donor-acceptor interaction. Good agreement between experimental and simulation results can only be obtained using a simulation code which takes into account the formation of neutral, near immobile, Sb-B pairs.

We use a kinetic Monte Carlo model to simulate the implantation of low energy Boron in Silicon, from 0.5 to 1 keV, at high doses, 1015 ions/cm2. The damage produced by each ion is calculated using UT-Marlowe, based on a binary collision approximation. During implantation at room temperature,, silicon self-interstitials, vacancies and boron interstitials are allowed to migrate and interact. The diffusion kinetics of these defects and dopants has been obtained by ab initio calculations as well as Stillinger Weber molecular dynamics. Clustering of both self-interstitials, vacancies and boron atoms is included. We also model the diffusion of the implanted dopants after a high temperature annealing in order to understand the transient enhanced diffusion (TED) phenomenon. We observe two different stages of TED During the first stage vacancies are present in the lattice together with interstitials and the diffusion enhancement is small. The second stage starts after all the vacancies disappear and gives rise to most of the final TED.

An atomistic model for B implantation, diffusion and clustering is presented. The model embodies the usual mechanism of Si self-interstitial diffusion and B kick-out and also includes the formation of immobile precursors of B clusters prior to the onset of transient enhanced diffusion. These immobile complexes, such as BI2 (a B atom with two Si self-interstitials) form during implantation or in the very early stages of annealing, when the Si interstitial concentration is very high. They then act as nucleation centers for the formation of B-rich clusters during annealing. This model explains and predicts the behavior of B under a wide variety of implantation and annealing conditions.

The transient enhanced diffusion behavior of B after ion implantation above the amorphization threshold is investigated. The experimental structure uses a layer of epitaxially grown Si, uniformly doped with B to act as a diffusion monitor. Wafers using this structure are implanted with amorphizing doses of Si, As, or P and annealed for various times at various temperatures. The experimental results show that upon annealing after Si implantation, there is a large amount of B pile-up that occurs at the amorphous/crystalline (A/C) interface while B is depleted from the region just beyond the A/C interface. This pile-up/depletion phenomenon can be attributed to the dislocation loops that form at the A/C interface. These loops act as sinks for interstitial point defects. There is also B pile-up/depletion behavior for As and P implants as well. However, this behavior may be explained by an electric field enhancement effect. While dislocation loops are known to form at the A/C interface for all of the investigated implant conditions, it appears that while they are necessary to simulate for Si amorphizing implants, they may not be necessary to simulate for As and P amorphizing implants.

The lattice Monte Carlo method with parameters from recent first-principle calculations1,2 are used to investigate dopant diffusion in silicon. In the simulations, vacancy hopping on a silicon lattice is biased by changes in system energy, including interactions up to the sixth-nearest neighbor. We find that vacancy-mediated diffusivity increases dramatically above 1020 cm−3, in agreement with experimental observations3 and previous calculations.4 However, for very long simulation times, arsenic diffusivity is reduced due to formation of AsxV complexes, with clustering more pronounced at high doping levels. As suggested by Ramamoorthy and Pantelides,5 we find that As2V complexes are mobile, and although they diffuse much more slowly than AsV pairs, they appear likely to have a significant role in high concentration diffusion due to their much higher numbers. We also investigated dopant fluxes in a vacancy gradient. For dopants like As for which pair diffusion is limited by the dissociation to third-nearest neighbor distances, the dopant flux is less than that predicted by pair diffusion models, with greater difference at higher temperatures. In contrast, for phosphorus/vacancy pairs, whose diffusion is limited by dopant/vacancy exchange, the dopant flux is close to the predictions of pair diffusion.

Prediction of transient enhanced diffusion (TED) requires modeling of extended defects of many types, such as {311} defects, dislocation loops, boron-interstitial clusters, arsenic precipitates, etc. These extended defects not only form individually, but they also interact with each other through changes in point defect and solute concentrations. We have developed a fundamental model which can account for the behavior of a broad range of extended defects, as well as their interactions with each other. We have successfully applied and parameterized our model to a range of systems and conditions, some of which are presented in this paper.

The phenomenon of transient diffusion of boron in crystalline silicon is considered. The local equilibrium assumption from the law of mass action is not used in order to estimate the length of the transient regime. For an interstitial diffuser such as boron, the pair of diffusion equations, one each for the mobile and the immobile species, can be interpreted as a generalized Fick's Law in the total concentration with a position- and time-dependent diffusion coefficient whose concentration-weighted average, called the global diffusion coefficient, has an analytic solution for an appropriately defined time-lag. At 850°C for boron, the transient in the global diffusion coefficient is negligible, showing that the local equilibrium assumption is justified. An estimate is given for conditions under which local equilibrium may not be valid.

An arsenic diffusion model was proposed with the emphasis on a new deactivation process which accounts for excess interstitial silicon generation. Appropriate parameter set for the binding energies of arsenic-point defect pairs were obtained for reproducing various arsenic activation levels and interaction with boron in the case of co-diffusion. Such parameters were extracted from the data of carefully performed secondary ion mass spectroscopy with lowering the primary beam energy.

Transient Enhanced Diffusion (TED) is one of the biggest modeling challenges present in predicting scaled technologies. Damage from implantation of dopant ions changes the diffusivities of the dopants and precipitates to form complex extended defects. Developing a quantitative model for the defect behavior during short time, low temperature anneals is a key to explaining TED. The surface can play a defining role in the removal of point defects from the bulk, but there is a lot of controversy over the role and strength of the surface sink for point defects. The controversy will be reviewed, and new experimental results will be presented that investigate the role of the surface on TED.

(100) silicon wafers were implanted at elevated temperatures up to 600°C with l×lO15-5×1015 Ge+/cm2 using 120 keV. The wafers were tilted by 5–7° and rotated by 5–15°. The implanted germanium profile was monitored as a function of implant temperature using RBS-channelling. Considerable profile broadening was seen together with apparent mass germanium migration away from the surface in samples implanted at 300°C and above. Control implants into hot and cold samples simultaneously rule out excess loss of germanium from the heated wafers. Channelling data indicate that while room temperature implants lead to amorphisation, with hot implants good quality layers are obtained in which the germanium atoms occupy substitutional sites. Hot implants into a sample previously implanted with germanium at room temperature does not lead to any redistribution in the original germanium profile. This result indicates that the apparent enhanced indiffusion of germanium is not a radiation assisted phenomenon and could be explained by a considerable channelling of the implanted germaniumatoms along the <100> direction.

P+NN+ silicon rectifiers have been irradiated by protons and alpha particles. DLTS and reverse recovery time (the so called TRR) measurements were performed in both as-irradiated and annealed (400°C) samples. The evolution of implantation induced-centers with the annealing is not the same when using the two particles. For example, the concentration of the well known A center increases (decreases) with the annealing when protons (alpha particles) are used. On the other hand, the measurements of the TRR show that the rectifier speed decreases with the annealing in all cases studied.

(001) CZ silicon wafers were implanted with As at lOOkeV to a dose of 1×1015/cm2. The implant was amorphizing in nature and the peak As concentration was below the As clustering threshold. Subsequently, a second As+or Ge+ implant at 30keV at doses of 2×1015/cm2, 5×1015/cm2 and 1×1016/cm2 were performed, respectively, into the as-implanted samples. The samples with a double arsenic implant induce As clustering at the lower doses and As precipitation at the highest dose at the projected range. However, the samples with the Ge do not induce clustering or precipitation. The samples were annealed at 700°C for various times to regrow the amorphous layer and to cause enhanced arsenic diffusion beyond the end-of range region. These samples wereanalyzed by SIMS and TEM. The difference in the defect evolution at the end-of-range region and TED beyond the end-of-range region between the As and Ge implanted samples was used to isolate the effects of As clustering and precipitation.

We have studied the lattice sites of Er in CZ Si single crystals by using conversion electron emission channeling from the isotope 167mEr (2.27 s) which is the decay product of radioactive Tm 167m(9.25 d). Following 60 keV implantation of 167Tm at a dose of 4×1013 cm−2 and annealing at 600°C, more than 90% of 167mEr is found close to tetrahedral insterstitial(T) sites. The tetra-hedral fraction of 167mEr decreases considerably after 10min annealing at 800°C and above. We attribute this to the onset of diffusion of the parent 167Tm and its trapping at other defects, presumably oxygen atoms or clusters of Tm/Er.

In this paper, we report Current-Voltage (I-V), Capacitance-Voltage (C-V) and Deep Level Transient Spectroscopy (DLTS) measurements on ultra-shallow (p+/n) junctions obtained by boron implantation into crystalline and Ge preamorphized Si-substrates. Germanium implantations were carried out at the energies of 30, 60 and 150 keV at a dose of 1015 cm−2. Boron was then implanted at an energy of 3 keV at a dose of 1015 cm−2. Dopant activation was obtained by Rapid Thermal Annealing (RTA) performed at 950°C for 15s in a nitrogen ambient. The aim of this work is to study the relation that exists between a population of End-Of-Range (EOR) defects, measured by TEM, and electrical properties of the Ge preamorphized diodes. The electrical measurements allow us to give the conduction mechanism which dominates in the diodes. Moreover, DLTS measurements showed the presence of two majority-carrier traps in direct relation with the EOR defects, measured at Ec-0,22eV and Ec-0,47eV. The presence of energy-distributed G-R centers independent upon the preamorphization stage is also discussed. Nevertheless, high quality (p+/n) junctions can be obtained by this technology with a judicious choice of Ge implantation energy.

Using a double-paddle oscillator, we have studied the internal friction of thin Si layers which have been disordered by ion implantation. We have identified the temperature-independent internal friction common to all amorphous solids below 10K, and also a narrow relaxation peak at ∼ 48K. This peak has been shown to be caused by divacancies in the damaged crystalline region underneath the amorphous layer. Thermal, rapid thermal, and laser-flash annealing have been compared. The minimum disorder is left after rapid thermal annealing.

We studied effects of nitrogen on reduction of boron diffusion in BF2-implanted source/drain regions of a p-channel MOSFET. Evolution of extended defects induced by nitrogen implantation in silicon and its dependence on implant energy were investigated. Characterization with TEM and SIMS led to validation of models of nitrogen as interstitial traps and strain-induced gettering. Electrical measurement of the PMOS threshold voltage shows a consistent trend of the nitrogen effects on boron redistribution in the S/D regions.

Plan-view and cross-sectional transmission electron microscopy have been used to study the microstructural characterization of the nucleation and growth behavior of {113} rodlike defects, as well as their correlation with {111} dislocation loops in silicon amorphized with 50 keV, 36×1014 Si/cm2, 8.0 mAand annealed by rapid thermal anneals at temperatures from 500 °C to 1100 °C for various times. We found that the nucleations of the {113} rodlike defects and {111} dislocation loops are two separate processes. At the beginning of anneals, excess interstitials accumulate and form circular interstitial clusters at the preamorphous/crystalline interface at as low as 600 °C for 1 s. Then these interstitial clusters grow along the <110> direction to form {113} rodlike defects. Later, while the {113} defects have begun to grow and/or dissolve into matrix, the {111} faulted Frank dislocation loops start to form. We also found that the initial interstitial clusters prefer to grow along the <110>directions inclined to the implantation surface.