Thin Silicon Substrate Research Articles

Use of Monte Carlo modeling for interpreting scanning electron microscope linewidth measurements

AbstractA scanning electron microscope (SEM) can be used to measure the dimensions of the microlithographic features of integrated circuits. However, without a good model of the electron‐beam/specimen interaction, accurate edge location cannot be obtained. A Monte Carlo code has been developed to model the interaction of an electron beam with one or two lines lithographically produced on a multilayer substrate. The purpose of the code is to enable one to extract the edge position of a line from SEM measurements. It is based on prior codes developed at the National Institute of Standards and Technology, but with a new formulation for the atomic scattering cross sections and the inclusion of a method to simulate edge roughness or rounding. The code is currently able to model the transmitted and backscattered electrons, and the results from the code have been applied to the analysis of electron transmission through gold lines on a thin silicon substrate, such as is used in an x‐ray lithographic mask. Significant reductions in backscattering occur because of the proximity of a neighboring line.

Gettering of Cu and Ni Impurities in SIMOX Wafers

The gettering of Cu and Ni impurities in intentionally contaminated SIMOX wafers have been studied by means of cross‐sectional transmission electron microscopy, nanoprobe energy dispersive x‐ray spectroscopy, secondary ion mass spectrometry, and selective etching. The wafers with Cu or Ni surface concentrations ranged from about were annealed at various temperatures followed by slow cooling to room temperature. Single and multistep thermal treatments were applied. It has been found that the buried oxide does not prevent the diffusion of both Cu and Ni contaminants from the top silicon layer into the bulk substrate at the whole investigated temperature range from 600 to 950°C. Moreover the effective gettering of Cu and Ni in the thin silicon substrate layer located just beneath the buried oxide has been observed and explained as being due to the heterogeneous impurity precipitation at stacking fault tetrahedra formed there during the SIMOX manufacturing. The gettering process has remained stable during the thermal simulation of CMOS device process of new generation ICs with 0.25 μm feature size.

Direct observation of rare-earth silicide epilayer formation by RHEED technique

The solid phase reaction of yttrium thin film and silicon substrate was investigated in situ with help of RHEED technique. For the first time, we have measured directly yttrium silicide formation temperature which can be as low as 120 °C for a thin (60 Å) metal layer deposited on a Si(111) substrate. For this purpose, we developed a new experimental technique. A study of the growth of thin (10–150 Å) silicon overlayers on yttrium and dysprosium silicide films epitaxially grown on Si(111) was also made. In this paper, an in situ reflection high energy electron diffraction (RHEED) pattern and also azimuthal plot investigation of Si/RE-silicide/Si double heterostructures grown by solid phase epitaxy and reactive deposition method are presented.

Silicide formation in the Co-Si system by rapid thermal annealing

Rapid thermal annealing was used to obtain additional information on the phases of the Co-Si system formed during the silicidation reaction between cobalt thin films and silicon substrates. The phase sequence, their stability and their coexistence were investigated by X-ray diffraction and sheet resistance. At an appropriate temperature all three phases are likely to coexist for short times, although the most common observation is the sequential formation of Co 2Si, CoSi and CoSi 2 as reported earlier. Cursory determination of the activation energy for CoSi 2 formation by sheet resistance measurements is 241.1 kJ mol -1.

Kinetic pattern formation of Gd-silicide films in lateral growth geometry

The solid phase reaction of gadolinium thin film and silicon substrate was investigated in lateral growth geometry with the help of periodic titanium protective stripes by optical microscopy. In the lateral reaction zone the shape of the interface between gadolinium and Gd silicide was very complicated and showed pattern formation. This silicide growth can be described as a kinetic process modified by the structure of the Gd film in contrast to the previously proposed simple nucleation.

Oxygen distribution in a thin epitaxial silicon layer

An empirical solution for the oxygen distribution in a thin epitaxial silicon layer and substrate wafers heavily doped with boron and antimony accounting for the prebake step is developed. The results from secondary ion mass spectroscopy analysis indicate that heavily doping with either boron or antimony has no effect on oxygen outdiffusion during epitaxial deposition. In contrast to slow-diffusing species such as boron and antimony, the plane at which the oxygen concentration is equal to one-half of the bulk oxygen concentration does not remain at the epitaxial layer/substrate interface, but is a function of the effective diffusion length of oxygen atoms for the total epitaxial process.

Passive microelectrode arrays for recording of neural signals: A simplified fabrication process

A new process for the fabrication of microelectrode arrays is developed that is fully compatible with standard integrated circuit fabrication schedules. The process involves two innovations: the use of initially thin silicon substrates (50 μm), and the use of sputtered SiO2 as a plasma etching mask. The finished electrode array consists of a very thin silicon substrate (<20 μm) that provides a mechanically stable carrier for the electrodes. The electrode metal conductors are sandwiched between two layers of polyimide films that adhere firmly to the substrate. Small electrode interface and contact windows are made with lithographically limited minimum feature sizes. Dry plasma etching is employed both to open the windows and to thin the substrate. For the former, we have determined the appropriate etching conditions and thicknesses necessary for SiO2 to be a viable mask in this step. For the latter step, the electrode profile is first etched to the final desired substrate thickness from the front (with the same oxide as the mask), and the wafer is then turned over and fully etched until the electrodes separate. Using initially thin wafers simplifies considerably this final etch.

Low temperature reactions of thin layers of Mn with Si

Reactions between manganese thin films and silicon substrates, annealed at relatively low temperatures (<430 °C), have been systematically investigated. Three phases, i.e., Mn3Si, Mn5Si3, and MnSi, were formed through a layered growth process. The formation sequence for these silicides was Mn3Si, MnSi, and then Mn5Si3. The unusual phenomena of coexistence of these three phases and simultaneous growth of two phases (MnSi and Mn5Si3) were also observed. A model has been proposed to explain the growth behavior. The formation sequence of Mn3Si, MnSi, and then Mn5Si3, and simultaneous growth of MnSi and Mn5Si3, can be explained by considering both thermodynamic and kinetic effects. Only those reactions that are thermodynamically allowed and kinetically preferred can take place. Kinetic preference is determined by the composition in the reaction region, which is controlled by the diffusion flux of the moving reactant. The proposed model is also compared with existing models.

Photopolymers exhibiting a large difference between glass transition and curing temperatures

When the glass transition temperature T g of a completely polymerized system is considerably higher than the temperature of polymerization, the reaction usually stops before completion. With step reactions the difference between T g and T cure rarely exceeds 35K. We have found that in the chain crosslinking photopolymerization of 1,6-hexanediol diacrylate a difference of more than 70K can be observed. Photopolymerization continues in the glassy state, albeit at a self-decelerating rate. Possible causes of this continuation are (i) the presence of a significant amount of unreacted monomer up to high conversions and (ii) the persistence of a localized mobility of radical sites in the glassy state through tertiary hydrogen abstraction. It is proposed that the continuation of the reaction is not limited by vitrification but by freezing of secondary relaxations. The initial, fast part of the photopolymerization process was monitored with differential scanning calorimetry (d.s.c.), the slow continuation in the glassy state with Fourier-transform infra-red spectroscopy ( FTi.r.). By using thin silicon substrates the same samples could be used with both techniques. In the range of conversions accessible with d.s.c., a perfect correlation with FTi.r. results was obtained. This proves that Beer's law applies to the i.r. data.

Direct STEM ADF imaging of gold on silicon (111)

Many studies of thin metal film growth and the formation of metal-semiconductor contacts have been performed using a wide range of experimental methods. STEM annular dark field imaging could be an important complement since it may allow direct imaging of a single heavy atom on a thin silicon substrate. This would enable studies of the local atomic arrangements and defects in the initial stage of metal silicide formation.Preliminary experiments were performed in an ultra-high vacuum VG HB501A STEM with a base pressure of 1 × 10-10 mbar. An antechamber directly attached to the microscope for specimen preparation has a base pressure of 2×l0-10 mbar. A thin single crystal membrane was fabricated by anodic etching and subsequent reactive etching. The specimen was cleaned by the Shiraki method and had a very thin oxide layer left on the surface. 5 Å of gold was deposited on the specimen at room temperature from a tungsten filament coil monitored by a quartz crystal monitor.

Ion-induced reaction between thin chromium layer and silicon substrates

Ion-induced increment ΔAS in the intensity of the ion channeling surface peak of a Si(111) surface uncovered and covered with a thin Cr layer (Cr-coated Si) has been investigated between room temperature and 345°C. It is found that the t1S-fluence relation for Cr-coated Si at each temperature includes a term in addition to the saturation curve having the same growth rate as that observed for the uncovered Si surface. This additional term is also a saturation curve and ascribed to the Cr-Si reaction. The Cr-Si reaction rate is found to increase with increasing temperature, while the growth rate of ΔS in the uncoated Si surface is found to decrease with increasing temperature.

Epitaxial Growth of FeSi2 on (111)Si

ABSTRACTIsothermal annealing, two step annealing and ion beam mixing were performed to induce interfacial reactions between iron thin films and silicon substrate. Both orthorhombic and tetragonal FeSi2 were found to grow epitaxially on (lll)Si with orthorhombic FeSi2 being the predominant phase. No epitaxial growth of FeSi2 on (001)Si was detected. Epitaxial islands as large as 40 μm in size were formed by a scheme combining ion beam mixing and two step annealing.

Aluminum contact to nickel silicide using a thin tungsten barrier

An aluminum film in contact with NiSi is not stable in the temperature range around 450 °C usually applied for aluminum contact sintering. We used thin (about 2 kÅ) self-supported silicon substrates to investigate the interaction of aluminum films with NiSi by mega-electronvolt 4He + backscattering spectrometry. The thin substrate enables us to distinguish between the aluminum and the silicon signals, to isolate them, and to analyze the reaction. It is found that the aluminum reacts with the silicide and forms an NiAl 3 layer in direct contact with the silicon substrate. Simultaneously, a rise in the Schottky barrier height of the contact is observed. A thin layer (250 Å) of tungsten placed as a barrier between the aluminum and the silicide is shown to inhibit the aluminum-silicide reaction. A process is described to prepare a reliable aluminum contact to NiSi on a silicon substrate in a single annealing step.

MIS and SIS solar cells on polycrystalline silicon

MIS- and SIS-structured solar cells are receiving much attention in the photovoltaic community. Apparently these cells could be a viable alternative to thermally diffused p-n junctions for use on thin film polycrystalline silicon substrates. In this paper we review MIS- and SIS-structured solar cells and the possible advantages of these structures for use with thin film polycrystalline silicon. The results of efficiency calculations are presented. We also consider the lifetime stability and fabrication techniques amenable to large-scale production. Finally the relative advantages and disadvantages of these cells and the results obtained are presented.

Ion-beam-induced silicide formation

Rutherford-backscattering spectrometry and x-ray-diffraction analysis have been used to investigate intermixing between thin metal films (Pt, Ni, and Hf) and silicon substrates as a result of inert-gas ion bombardment. Silicide phases (Pt2Si, Ni2Si, and HfSi) were observed near the interface as long as the ion range exceeds the film thickness. For a fixed dose, the silicide thickness increases with the atomic mass of both ion and metal and is greater for Pt2Si than HfSi. The growth of Pt2Si showed a square-root dependence on ion dose for Ar, Kr, and Xe ions. The phenomenon of ion-induced silicide formation is similar to formation resulting from thermal anneal until the whole metal film is consumed in the reaction, at which point a progressive intermixing to redistribute the metal into deeper regions of the sample occurred along with the disappearance of the structure diffraction patterns.

Direct observations of the growth of sputtered silver and gold films on (111) silicon substrates

The growth of ion-beam-sputtered silver and gold films on (111) silicon substrates has been studied using the in situ electron microscope technique. The thin silicon substrates are obtained by jet chemical etching followed by ion-micromilling techniques. The substrate temperatures during growth vary from 300 to 600 °C. A two-layer phenomenon occurs in the growth of silver films. Both layers grow with a (111) orientation parallel to the substrate. Liquidlike coalescence which is observed in the second layer, however, is prohibited in the growth of first-layer islands. Gold films deposited under the same conditions result in the formation of discrete liquid islands even at substrate temperatures as low as 300 °C. Occurrence of solid polycrystalline gold films is first observed at 260 °C by lowering the substrate temperature after deposition. Deposition at lower temperatures results in the formation of polycrystalline and amorphous structures. Both cases can be attributed to the interface interaction between the overgrowth and the single-crystal silicon substrates.

Preparation of Thin Single Crystal Silicon Substrates for in situ Electron Microscope Studies

The in-situ electron microscope technique has been shown to be a powerful method for understanding the nucleation and growth of thin films formed both by vacuum vapor deposition and ion beam sputter-deposition. Single crystal silicon which has only been chosen as substrate for thin film deposition outside the electron microscope has now been prepared in a form suitable for in-situ deposition.The method of the preparation of thin silicon substrate is a combination of jet chemical etching and modified ion beam thinning. A specimen of thickness roughly 0.010 inch is first etched from both sides by the jet etching technique. After jet etching, it is transferred to the Commonwealth Scientific ion micro-milling instrument and bombarded from both sides with Argon ion beam. A pin hole occurs in the center of the specimen after about 30 minutes of ion bombardment.

Direct Observations of the Epitaxial Growth of Sputtered Ag on Si

The growth of Ag films deposited on various substrate materials such as MoS2, mica, graphite, and MgO has been investigated extensively using the in situ electron microscopy technique. The three stages of film growth, namely, the nucleation, growth of islands followed by liquid-like coalescence have been observed in both the vacuum vapor deposited and ion beam sputtered thin films. The mechanisms of nucleation and growth of silver films formed by ion beam sputtering on the (111) plane of silicon comprise the subject of this paper. A novel mode of epitaxial growth is observed to that seen previously.The experimental arrangement for the present study is the same as previous experiments, and the preparation procedure for obtaining thin silicon substrate is presented in a separate paper.