Al Si Contact Research Articles

Initial stages of the Schottky-barrier formation for abrupt covalent interfaces

The initial stages of the Schottky-barrier formation for an AlSi contact are theoretically analyzed. Four different geometrical configurations are discussed, and their charge neutrality level, interface Fermi level and local density of states are calculated. We conclude that the Schottky barrier is formed with a monolayer and that the charge neutrality level at the junction is a function of the metal-semiconductor interaction.

Initial stages of the Schottky-barrier formation for abrupt covalent interfaces

Abstract The initial stages of the Schottky-barrier formation for an AlSi contact are theoretically analyzed. Four different geometrical configurations are discussed, and their charge neutrality level, interface Fermi level and local density of states are calculated. We conclude that the Schottky barrier is formed with a monolayer and that the charge neutrality level at the junction is a function of the metal-semiconductor interaction.

Performance of titanium nitride diffusion barriers in aluminum–titanium metallization schemes for integrated circuits

Three different contact schemes Al–Si, Al/Ti–Si, and Al/TiN/Ti–Si have been studied for their electrical characteristics and thermal stability on shallow n+p-junctions. Contact resistance and leakage current measurements indicate that the Al/Ti structure remains stable up to 450 °C and the Al/TiN/Ti structure up to 500 °C for 15 min, whereas Al–Si contacts degrade already at 400 °C. The stability of the Al/Ti contacts is limited by the thin film reaction between Al and Ti while the failure of the Al/TiN/Ti structure is attributed to local defects such as pinholes in the TiN barrier.

Effects of ion mixing on Al-Si contacts

Abstract Al-Si contacts irradiated with arsenic ions and annealed at 200–500°C have been investigated for interfacial reactions and variations in electrical contact resistance. For arsenic doses above the amorphization threshold of silicon (about 2×10 14 ions cm -2 ) the crystallographic reaction pits become irregular in shape and seem to be confined to the amorphous layer. The contact resistance on n + -Si is almost one order of magnitude higher in irradiated samples as compared with unirradiated samples after annealing at 425°C for 10 min. Nuclear reaction analysis indicates the presence of a thin interfacial oxide layer in Al-n + -Si samples. Differences in surface morphologies of ion-irradiated Al-n + -Si and Al-p + -Si samples after long-term annealing at 200°C suggest that the dispersion of the interfacial oxide is very limited for the ion doses used in this study.

Failure temperature of amorphous Cu-Ta alloys as diffusion barriers in Al-Si contacts

Self-supporting films of amorphous Cu20Ta80 alloys have a crystallization temperature as high as 800 °C, yet they fail as diffusion barriers in aluminum-silicon contacts at much lower temperature because of compound formation. On Si(100) the a-Cu20Ta80 films react above the temperature at which TaSi2 forms (650 °C). In contact with aluminum, Al3Ta starts to form at 500 °C, which is also the failure temperature of a-Cu20T80 as diffusion barrier between Al and Si(100). The reaction characteristics of the barrier constituents with the surrounding elements as well as the crystallization temperature determine the thermal stability of amorphous alloy diffusion barriers.

Al metallization by ionized-cluster beam deposition and epitaxy

From the standpoint of metallization for semiconductor devices, Al film deposition and epitaxy by ionized-cluster-beam (ICB) are studied. Film properties related to the stability of Al-Si contact, electromigration resistance, step coverage, surface flatness, etc. are examined. Because of the effects caused by kinetic energy and the electric charge of ICB, such as in situ cleaning during deposition and enhancement of adatom migration, problems of Al metallization are remarkably diminished, and it is concluded that the Al film deposition and epitaxy by ICB meet the increasingly more stringent demands on device metallization.

Reduction of silicon-aluminum interdiffusion by improved semiconductor surface ordering

Aluminum overlayers on highly ordered single-crystal silicon (100) and (111) surfaces in ultrahigh vacuum exhibit characteristic interface widths less than tens of angstroms at room temperature and hundreds of angstroms at 400 °C—orders of magnitude more abrupt than conventionally prepared Al-Si contacts. We demonstrate that surface disorder plays a critical role in promoting Si diffusion into the Al overlayer.

Al-Si contacts formed by ion irradiation and post-annealing

Al-Si contacts have been formed by implantation of As ions through Al-Si interfaces followed by heat treatment at 400–500 °C for 10 min. The erosion of Si proceeds uniformly in contact areas at the sintering temperatures. Diodes using Al-Si contacts produced by this technique have been fabricated on thin n+ layers in p-type Si substrates with junction depths of 0.35 μm and contact areas of 5 μm2. The average leakage current per diode (with an approximate junction area of 14×26 μm2) is about 10−8 A, as compared to the leakage current of 10−5 A for diodes with Al-Si contacts prepared by sintering of Al on Si at 420 °C. We attribute the improvement to the uniformity of Si erosion after the interfacial oxide has been dispersed by ion irradiation.

Contact resistance behavior of the Pd2 Si-AlCuSi system

Aluminum to silicon contact resistance, Rc , is influenced by the choice of intermediate contact and interconnection metallurgies, as well as anneal cycles associated with metallization and post-metallization passivation processes in single and multi-level metal VLSI technologies. This is particularly evident for the metallurgical system comprised of palladium silicide, Pd2 Si, and AlCuSi. This system has been investigated to ascertain the relationships between R , anneal time, and anneal temperature for various Pd Si thicknesses of AlCuSi contacts to N+ single-crystal silicon and polysilicon. The evaluations revealed that Rc and Rc distribution are inversely proportional to Pd2 Si thickness, and increase with anneal time and temperature. The results are compared to the known physical chemical interactions of Pd Si and Al. The studies demonstrate that by proper selection of process parameters and contact structure design, stable Al-Si contact resistance can be achieved for semiconductor device applications when Pd is used as an intermediate contact metallurgy.

Modelling electrical behaviour of nonuniform AlSi Schottky diodes

The heat-treated AlSi contacts show large nonuniformities and cannot be described by a one-dimensional theory. The authors partially succeed in describing circular Al/ n Si contacts by a crude model: two diodes in parallel (the first central and circular, the second peripheral and annular). It was shown that only a part of the contact area is effective for samples treated at lower temperatures, due to the interfacial oxide layer. However, the contacts treated at 550°C seem to have more or less uniform electrical properties, despite large nonuniformities in the contact geometry. Moreover, the effective energy barrier of these contacts is little influenced by the preparation conditions, including the amount of Al available for AlSi interaction and the rate of cooling. These results may modify the present conception of the formation of the Al-doped layer at the AlSi interface.

Contact resistance response surface of sintered Al films on (100) silicon

Al–Si contact resistance for (100) p-type Si was studied as a function of sintering gas ambient, temperature, and time. Variables were N2 vs forming gas (10%H2, 90%N2); 540°C, 500°C, 450°C; 10, 20, 30, and 40 minutes. Results showed that forming gas yields higher contact resistance values than does N2. Away from the Al–Si eutectic point, time and temperature did not appear to be significant variables. The quantum mechanical electron tunneling model for metal–semiconductor contact resistance developed by Yu in 1970 is briefly described. This model predicts very well the range of contact resistance values measured in this experiment. The effects of forming gas can be understood by the addition of a space charge due to H2 diffusion into the Al film.

Effect of vacuum ambience on Al–Si contacts

Aluminum–silicon contacts are extensively used in integrated circuits. Contact resistance of Al–Si contacts is sensitive to silicon surface preparation, vacuum ambience of aluminum film deposition, and contact sintering conditions. Controlled contamination experiments have been carried out using H2O, O2, and CO during the aluminum film deposition to evaluate the effect on the Al–Si contact resistance. A known gas was admitted to the vacuum chamber by a controlled leak valve and the partial pressure of the gas was monitored with a UTI 100 C quadrupole mass analyzer. The substrates consisted of n‐type (111) silicon slices completely processed through boron base and phosphorous emitter diffusions. Al–Si contacts were sintered in the 400°–550°C temperature range in N2 ambience. Al surfaces and Al–Si interfaces were Auger characterized. These results along with the contact resistance data (determined with an appropriate test structure) have been examined and it is concluded that gaseous contaminants such as H2O, O2, and CO in the vacuum chamber up to partial pressures of 5×10−7 Torr (6.7×10−5 Pa) do not adversely affect the aluminum silicon interface contact resistance.

2.4 Formation of various metal semiconductor junctions by means of ion evaporation

Abstract The paper deals with the result of some preliminary experiments of applying ion-evaporated metal layers to silicon and heat treatment of the samples. Ion evaporation is based on the electron beam evaporation, namely, the ions arising from the 180° deflection type e-gun are accelerated to cause implantation effects during the condensation of the neutral particles. The accelerating voltage is typically 20 kV, the ion current 5–20 mA. Four metals have been chosen for the experiments: Al, Ni, Cu and Mo. The I–V-characteristics of the contacts deposited by ion evaporation were ohmlc. The curves appreciably changed after heat treatment of the samples at 400, 500 and 580°C. The AlSi contacts had very high resistivity after the treatment at 400°C, but then Al alloyed diode type contacts were achieved at 500°C. Ni and Mo formed alloyed type contacts at 580°C, and similar results were obtained in case of Cu. Very interesting results were obtained for Ni and Mo at 500°C, where we obtained I–V-characteristics of the Schottky-diode type, but with the reverse polarity.

Cobalt silicide layers on Si. II. Schottky barrier height and contact resistivity

Cobalt silicide layers have been grown by electron-beam vacuum deposition of Co onto Si wafers and subsequent thermal treatment. The barrier height φB and contact resistivity ρc of the Co-Si contacts were determined from current-voltage characteristics. The barrier height for Co on P-type Si decreases on annealing and reaches a value of 0.38 eV after annealing at 450 °C. The barrier height on N-type Si is then 0.68 eV. Presumably these are the values for a CoSi-Si contact. Annealing at higher temperatures results in a higher barrier (0.40 eV) on P-type Si and a lower barrier on N-type Si, probably by the formation of CoSi2. The values for the contact resistivity on P-type Si were determined using a transmission-line model and found to be comparable to those for Al-Si contacts. Annealing above 400 °C results in a decrease of the contact resistivity, mainly due to an increase in the boron concentration. By annealing above 500 °C the contact resistivity increases again, in agreement with the increased barrier height.

Electromigration Failure at Aluminum-Silicon Contacts

Previous experiments to measure electromigration properties of Al–Si contacts were conducted at high current densities. Failures resulted from a combination of electromigration damage and diffusion of silicon into the aluminum lands. At low current densities, the silicon diffusion is significantly reduced and the median time to failure is inversely related to the current density. In addition, contacts ranging from 10×5 μ to 25×15 μ were tested to determine the dependence of median time to failure on contact size.