Substitutional carbon is known to locally reduce silicon self-interstitial concentrations and act as a barrier to self-interstitial migration through the carbon rich regions. A silicon spacer between two carbon rich SiGe layers is fabricated in this work to examine self-interstitial generation in a region that is isolated from self-interstitial formation at the surface or in the silicon bulk. Boron marker layers above, below and in between two SiGeC layers are used to monitor the self-interstitial concentration between the substitutional carbon. No evidence of self- interstitial depletion in the silicon spacer is observed, despite annealing in conditions believed sufficient to allow the self-interstitials to reach and react with surrounding substitutional carbon. Simulations of the self-interstitial and carbon indicate that the silicon self interstitial concentration in the spacer layer can be sustained in part due to a silicon self-interstitial recycling process through a reverse “kick-out” reaction.

The vibrational spectra of N-pairs and nitrogen-vacancy-oxygen defects in nitrogen doped Czochralski silicon have been investigated using density functional theory calculations. We found that 771 cm−1 and 967 cm−1 lines measured by FTIR are fingerprints for N-pairs in interstitial position. These confirm that nitrogen atoms are paired and bonded to Si atoms. Calculated local vibration modes of N2On complexes provide the best matching with observed FTIR frequency of N-O complexes. Nonetheless, VmN2On (m,n =1,2) can develop during crystal cooling or wafer processing, as revealed by local vibrational modes falling around, 806 and 815 cm−1 FTIR frequencies.

Silicon on insulator (SOI - Smartcut®) wafers were implanted with 1MeV and 300keV silicon ions to doses of 3.8x1015 cm−2 and 3x1014 cm−2, respectively, in order to modify the vacancy concentration in a controlled way. Boron was then implanted at 2keV to a dose of 1×1015 cm−2 into the near-surface part of the vacancy-engineered region. Atomic profiles were determined using SIMS and electrical profiles were measured using a novel Differential Hall Effect (DHE) technique, which enables profiling of electrically active dopants with a nanometer depth resolution. The electrical profiles provide pairs of carrier concentration and mobility values as a function of depth. The buried oxide (BOX) is proven to restrict the back diffusing interstitials positioned below the BOX from entering the silicon top layer and interacting with the boron profile. Also an increase of ∼50% in boron activation is achieved when a co-implant is used. However, SOI shows a reduced degree of activation when compared to bulk silicon, with or without a co-implant.

We have investigated the role of fluorine in the reduction of transient enhanced diffusion (TED) and thermal diffusion (TD) of B in preamorphized Si layers implanted with F. For this purpose, we have employed B delta-doped layers, grown by molecular beam epitaxy (MBE), as markers for silicon self-interstitials (Is). We have shown that boron TED decreases with increasing amount of incorporated F up to the complete TED suppression. Furthermore, we have clearly demonstrated that the physical mechanism that suppresses the boron TED is not a B-F chemical bonding, but a strong interaction between F atoms and Is. In addition, we have seen that fluorine strongly reduces B diffusion also under Is thermal equilibrium concentration. Our results clearly show that the presence of F lowers the Is density very effectively, reducing the boron TED as well as the dopant diffusion under equilibrium conditions.

We have developed a model for high concentration fluorine diffusion and fluorine diffusion in amorphous silicon. In this context, we define high concentration fluorine to mean fluorine doses above the threshold of amorphization for implantation into silicon, which is approximately 1×1015/cm2 dose. We pre-amorphized with silicon to create a continuous amorphous region, and samples were subsequently implanted with the fluorine conditions of 16keV at 2×1015cm−2 dose, 30keV at 2×1014 cm−2 dose, or 16keV at 8×1015cm−2 dose. Samples were annealed by either conventional furnace or RTA with an N2 ambient for various times at temperatures of 550-750 °C. SIMS was used for depth profiling, and TEM images were also taken of the samples to check for defects and amorphous depth. We then created the model for the data by extending the fluorine model presented in our previous work, and it models the profile motion and the time dependence well. The model is also still capable of describing our previous work and fits it very well.

The susceptibility of mill-annealed (MA) and gas tungsten arc welded (GTAW) Alloy 22 to Pb-assisted stress corrosion cracking (PbSCC) was evaluated in supersaturated PbCl2 and Pb(NO3)2 solutions. The solution chemistries and pH were varied to evaluate the effects of specific Pb-bearing species on the SCC resistance of GTAW Alloy 22 double U-bend samples. Anodic polarization tests in acidic and basic supersaturated Pb(NO3)2 solutions indicated that the passive current densities generally increased with applied potential. These results were then compared to previous anodic polarizations results for supersaturated acidic PbCl2 solutions. Constant deformation SCC tests indicate that MA and GTAW Alloy 22 were resistant to PbSCC in supersaturated PbCl2 (16,000ppm Pb, 95°C, pH 0.5, and Eapplied = -100, -50, 0, and 50mVSCE) and Pb(NO3)2 (16,000ppm Pb, 95°C, pH 3.9 and 10.7 and Eapplied = 50 and 140mVSCE, respectively). The overall results from this PbSCC investigation suggest that MA and GTAW Alloy 22 are resistant to SCC in supersaturated PbCl2 and Pb(NO3)2 solutions at 95°C. Provided that these high Pb concentrations are not attainable in the anticipated repository environments, Alloy 22 is unlikely to be susceptible to SCC by the presence of Pb.

The main challenges for PMOS ultra shallow junction formation remain the transient enhanced diffusion (TED) and the solid solubility limit of boron in silicon. It has been demonstrated that low energy boron implantation and spike annealing are key in meeting the 90 nm technology node ITRS requirements. To meet the 65 nm technology requirements many studies have used fluorine co-implantation with boron and Si+ or Ge+ pre-amorphization (PAI) and spike annealing. Although using BF+2 can be attractive for its high throughput, self-amorphization and the presence of fluorine, many studies have shown that for the fluorine to successfully reduce TED its energy needs to be well optimized with respect to the boron's, therefore BF+2 does not present the right fluorine/boron energy ratio for the optimum junction formation. In this work we optimize the fluorine energy when a deep or shallow PAI is used. We also demonstrate that the fluorine dose needs to be carefully optimized otherwise a reverse effect can be observed. We will also show that the optimized junction depends less on the Ge+ energies between 2 keV and 20 keV and when HF etch is implemented after Ge+ PAI, improvements in both the junction depth and the sheet resistance are observed.

The feasibility of the SPER junction process as a reasonable alternative to the spike anneal junction is proved in this work. Good control of the SCE and performance competitive results as compared to the spike junction are obtained. An analysis of the interaction between the halo dopant and the SPER junctions has been carried out; it is shown that the performance degrades with increasing halo dose as a consequence of an overlap resistance problem.

The germanosilicidation of Ni on strained (001) Si0.8Ge0.2, particularly Ge segregation, grain boundary grooving, and surface morphology, during rapid thermal annealing (RTA) was studied. High-resolution cross-sectional transmission electron microscopy (HRXTEM) suggested that Ge-rich Si1−zGez segregation takes place preferentially at the germanosilicide/Si1−xGex interface, more specifically at the triple junctions between two adjacent NiSi1−uGeu grains and the underlying epi Si1−xGex, and it is accompanied with thermal grooving process. The segregation process accelerates the thermal grooving of NiSi1−uGeu grain boundaries at the interface. The segregation-accelerated grain boundary grooving has significant effect on the surface morphology of NiSi1−uGeu films in Ni-SiGe system.

The stable growth conditions of erbium-silicide on silicon-on-insulator (SOI) are investigated considering annealing temperature, SOI and sputtered erbium thickness. From the sheet resistance measurement, X-ray diffraction and Auger electron spectroscopy analysis, the optimum annealing temperature is determined as 500°C. Also, for the stable growth of erbium- silicide on SOI, the sputtered erbium thickness should be less than 1.5 times of SOI thickness. As the SOI thickness decreases below this critical thickness, erbium-rich region is formed at the erbium-silicide and buried-oxide interface. By applying the optimized erbium-silicide growth conditions, 50-nm-gate-length n-type SB-MOSFET is manufactured, which shows the possible usage of erbium-silicide as the source and drain material in the n-type Schottky barrier MOSFETs for decananometer regime applications.

In every wafer processing step wafer stress management is extremely important for advanced device manufacturing. Thermally induced stress on device wafers has a large impact on lithography and affects device yield. Thermally induced stress during rapid thermal annealing (RTA) steps in high density 512MB DRAM device fabrication was investigated using a lamp-based (cold wall) RTA system and compared to results using a furnace-based (hot wall) single wafer RTA system. Compared to the lamp-based (cold wall) system, RTA in a furnace-based (hot wall) system was found to be very effective in suppressing thermally induced stress and increasing device yield due to superior pattern transfer characteristics in lithography.

The use of flash lamp annealing for processing semiconductor materials is outlined. Specific applications include ultra-shallow junction formation and heteroepitaxial growth of improved quality thin films of cubic silicon carbide. It is demonstrated that flash lamp annealing holds great promise as a technique for fabricating novel devices.

We present experimental results of impurity and self-diffusion in an isotopically controlled silicon heterostructure extrinsically doped with phosphorus. As a consequence of extrinsic doping, the concentration of singly negatively charged native defects is enhanced and the role of these native defect charge states in the simultaneous phosphorus and Si self-diffusion can be determined. Multilayers of isotopically controlled 28Si and natural silicon enable simultaneous analysis of 30Si self-diffusion into the 28Si enriched layers and phosphorus diffusion throughout the multilayer structure. An amorphous 260 nm thick Si cap layer was deposited on top of the Si isotope heterostructure. The phosphorus ions were implanted to a depth such that all the radiation damage resided inside this amorphous cap layer, preventing the generation of excess native defects and enabling the determination of the Si self-diffusion coefficient and the phosphorus diffusivity under equilibrium conditions. These samples were annealed at temperatures between 950 and 1100°C to study the diffusion. Detailed analysis of the diffusion process was performed on the basis of a P diffusion model which involves neutral and positively charged mobile P species and neutral and singly negatively charged self-interstitial.

An accurate method to measure the four point probe (4PP) sheet resistance (Rs) of USJ Source-Drain structures is described. The new method utilizes Elastic Material probes (EM- probe) to form non-penetrating contacts to the silicon surface. The probe design is kinematic and the force is controlled to ensure elastic deformation of the probe material. The probe material is selected so that large direct tunneling currents can flow through the native oxide thereby forming a low impedance contact. Sheet resistance measurements on USJ implanted P+/N structures with SIMS junction depths as shallow as 15 nm have been measured. The sheet resistance values obtained with the new EM-probe 4PP method were found to be consistent with expectations. In this paper, the method will be demonstrated on a variety of implanted USJ structures.

The local structural information around the germanium atom in boron doped SiGe alloys is important in understanding the dopant diffusion mechanisms. Epitaxial SiGe test structures with B and C markers were grown on Si substrates by using rapid thermal chemical vapor deposition (RTCVD). The local structure around the Ge atom was probed using Ge K-edge x-ray absorption fine structure spectroscopy (XAFS) to determine the effects of the B and C on the Ge sites. The concentration profiles obtained from secondary ion mass spectroscopy are correlated with the Ge XAFS results. The modifications on the local structure around the Ge atoms are revealed from the multiple scattering analyses on the Ge near-neighbors. First and second shell XAFS fits to the B doped SiGe samples indicate a direct evidence of the Ge trapping of the B atoms whereas the C is randomly distributed to the Si lattice sites.

Factors limiting the activation of indium in silicon are examined via the ab initio pseudopotential technique. The role of carbon in the enhancement/retardation of activation/diffusion respectively is clarified. It is found that (1) adjacent substitutional indium atoms are deactivated. Only second neighbour sites of indium are activated unlike the case of boron, where all substitutional sites remain activated. (2) Silicon self-interstitials deactivate indium by trapping them on substitutional sites. Carbon, on the other hand, traps such self- interstitials with higher binding energy and prevents them from deactivating indium. (3) Since both indium and carbon diffusion is interstitial mediated, carbon reduces indium diffusion on account of its higher binding energy with the self-interstitial. Moreover, the release of the carbon interstitial is more favourable than the release of the indium interstitial from a carbon-indium pair. Therefore, carbon minimises indium interstitial diffusion. (4) Arsenic enhances de- activation of indium by neutralisation and by strong binding on adjacent substitutional sites. Furthermore since the release of the indium interstitial is more favourable in comparison to the release of the arsenic interstitial from the indium-arsenic pair, indium diffusion is enhanced in the presence of arsenic.

Phase formation and interfacial microstructure evolution of nickel germanides formed by rapid thermal annealing in a 15-nm Ni/Ge (100) system have been studied. Coexistence of a NiGe layer and Ni-rich germanide particles was detected at 250°C. Highly textured NiGe film with a smooth interface with Ge was observed. Annealing at higher temperatures resulted in grain growth and severe grooving of the NiGe film at the substrate side, followed by serious agglomeration above 500°C. Fairly low sheet resistance was achieved in 250-500°C where the NiGe film continuity was uninterrupted.

The ability to activate greater amounts of dopants at lower temperatures is a persistent contingency in the continual drive for device scaling in Si microelectronics. We report on the effect of incorporating atomic hydrogen on the activation of implanted boron in shallow junctions. Hydrogen incorporation into the sample was carried out by exposure to an electron cyclotron resonance (ECR) hydrogen plasma. Enhanced activation was observed in hydrogenated samples for post-implantation annealing temperatures of 450°C and below, as measured by spreading resistance profilometry, and confirmed by identical boron atomic profile in hydrogenated and unhydrogenated samples. The enhancement in boron activation at lower temperature is attributed to the creation of vacancies in the boron-implanted region, the lattice-relaxation effect by the presence of atomic hydrogen, and the effect of atomic hydrogen on boron-interstitial cluster formation.

During high current ion implantation into photoresist-covered substrates, evolution of gaseous by-products of photoresist breakdown occurs that can affect the dose control of the process as well as diffusion and activation of the implanted dopants in silicon. The dosimetry effects are well understood and accounted for in modern ion implanter design. In this work, we report on the effect of photoresist outgassing on the distribution of boron concentration in silicon during the subsequent annealing process, boron electrical activation and sheet resistance measurements. Experiments were performed on mono- crystalline and pre-amorphized silicon and oxide covered wafers that help elucidate the mechanism of boron accumulation and diffusion. Higher carbon concentration on the wafer surface due to photoresist outgassing results in boron clustering and reduction of transient enhanced diffusion, leading to higher sheet resistance at low anneal temperature. Implantation through a thin oxide layer can be used to reduce carbon surface contamination and minimize the effects on the boron profiles and sheet resistance.

Silicidation of small features of various geometries and sizes using Ni-silicide was studied. Effects of dopants, surface preparation and silicidation parameters on silicide morphology were investigated. It was found that Ni silicide thickness and quality of the silicide/silicon interface (presence of NiSi2 pyramids) are dependent on the area dimensions and geometry. NiSi formed on narrow lines is thicker compared to wide areas. We found that pyramids of epitaxial NiSi2 protruding into the Si substrate are formed during the silicidation of undoped, boron-doped and BF2-doped active areas. The presence of pyramids was not observed on As-doped silicon. Rough interface due to the pyramids could be the reason for a more pronounced linewidth dependence of leakage current in the case of p+/n diodes. A higher density of pyramids was found on narrow lines (below 0.2μm) and particularly islands structures. No pyramids were found when a Ge preamorphisation implant was used.