G. J. van Gurp

Cobalt silicide layers on Si. I. Structure and growth

Cobalt silicide layers have been grown by electron‐beam vacuum deposition of Co onto Si wafers, and subsequent thermal treatment above 400 °C. The compounds Co2Si, CoSi, and CoSi2 were found by x‐ray diffraction. The compounds grow as successive layers with a thickness ratio of about 3:1:0.1, as was found from comparison with powder diagrams. The silicide growth proceeds faster along grain boundaries. After prolonged heating the Co is consumed and the Co2Si thickness then decreases. Subsequently the Co2Si is consumed, the CoSi thickness decreases, and a growing CoSi2 layer is left. No significant preferred orientation was found in the three silicides, except in CoSi2, which exhibits a preferred 〈100〉 orientation, when grown on 〈100〉 Si for more than 3 h at 550 °C. Both from x‐ray diffraction intensities and from thickness measurements it is shown that the growth of the Co2Si and CoSi layers, which is initially approximately linear with time, proceeds with a t1/2 dependence with an activation energy V of 1...

Interactions in the Co/Si thin‐film system. II. Diffusion‐marker experiments

The Kirkendall effect in thin‐film Co/Si couples is investigated by He backscattering using diffusion markers. The couples contain the compounds Co2Si and CoSi. The markers used are Xe implanted into Si, C, or Co2Si and a discontinuous W film between Si and Co. In the case of Xe implantation into Co, the experiments can be explained by dragging of Xe bubbles by the moving Co/Co2Si phase boundary. Xe implanted into Si is located at the Si/CoSi interface after the reaction and W is found at the Co/Co2Si interface after the reaction. The same applies to Xe implanted through Co into Co2Si. The results are evidence for predominant Co diffusion in Co2Si growth and Si diffusion in CoSi growth. Grain‐size measurements suggest that grain‐boundary diffusion plays a role in the growth of the silicide layers. The results of a Xe implantation into Co2Si without Co suggests that at the Co2Si/CoSi interface both Si and Co diffusion currents flow.

Interstitial and substitutional Zn in InP and InGaAsP

Closed‐ampoule Zn diffusion in InP results in a net acceptor concentration that is much smaller than the Zn concentration. After subsequent annealing of InP in an atmosphere without Zn, the Zn and net acceptor concentrations have become almost identical, due to a decreased Zn concentration and an increased net acceptor concentration. The annealing treatment gives rise to a decreased contact resistivity. The diffusion depth has not changed. Annealing with a SiN cap on the InP surface does not have this effect on the concentrations. These annealing effects also take place in InGaAsP on InP layers. The results can be explained quantitatively by assuming that Zn is incorporated as both substitutional acceptors and interstitial donors and that only the interstitial Zn is driven out by the annealing, owing to its large diffusion coefficient. Profiles calculated with this interstitial‐substitutional model can be fitted to experimental profiles assuming Zn to diffuse as singly ionized interstitial donors. This mo...

Diffusion‐limited Si precipitation in evaporated Al/Si films

The precipitation behavior of evaporated Al/Si films with up to 1.8% Si is studied by microscopy as well as by resistance measurements after isochronal and isothermal aging. After deposition of the film most of the Si is precipitated. Temperature treatments above 300°C cause the solution of Si into Al. Aging of a supersaturated solution results in precipitation on grain boundaries at a rate that is much greater than in bulk Al/Si. The precipitation is described as a two‐dimensional diffusion‐limited process. A theoretical expression for the rate constant permits the determination of the diffusion coefficient of Si in Al. This is much greater than in bulk Al/Si and has an activation energy of about 0.85 eV. In view of the small vacancy concentration in the film, these results are explained by diffusion along dislocations, in agreement with electron micrographs revealing a dislocation density of about 1010 cm−2.

Aluminum‐silicide reactions. I. Diffusion, compound formation, and microstructure

Material reactions as a result of thermal treatment were studied on thin‐film Al/silicide/Si systems with CoSi2, PtxNi1−xSi, and MoSi2 for the silicide. Auger electron spectroscopy and Rutherford backscattering analysis showed the transport of Si and the metal released from the silicide into the Al and transport of Al into the silicide. X‐ray diffraction showed the formation of Co2Al9 at 400 °C, PtAl2, NiAl3, and PtNiAl2 at 275 °C, of which the latter disappeared above 450 °C, and MoAl12 at 535 °C, as well as free Si. The Co2Al9 formation followed a linear time dependence with an activation energy of 2.3 eV. The MoAl12 formation followed a parabolic time dependence with an activation energy of 3.6 eV. A thin tungsten layer between Al and the silicide proved to be effective as a diffusion barrier below 500 °C, at which temperature WAl12 was formed. The microstructure was studied by scanning and transmission electron microscopy, electron microprobe analysis, and scanning Auger electron spectroscopy. The rea...

Cellular structure and silicide formation in laser‐irradiated metal‐silicon systems

Laser irradiation of thin Co, Mo, and Pd films on single‐crystalline silicon using Q‐switched Nd‐YAG laser pulses was shown by He backscattering to result in deep metal penetration into the Si. Evidence of the silicide formation was obtained by x‐ray diffraction. Transmission electron microscopy showed the simultaneous occurrence of two types of cells with metal‐rich walls: small cells of about 0.1‐μm diameter, attributed to rapid solidification from a supercooled melt, and larger cells of about 1‐μm diameter, attributed to convection in the melt (Benard cells).

Zinc diffusion in n-type indium phosphide

Profiles of Zn in n‐type InP〈100〉 wafers after ampoule diffusion were measured by secondary‐ion mass spectrometry, Auger electron spectrometry, differential Hall‐effect measurements, capacitance measurements, and scanning electron microscopy. The results can be explained by an interstitial‐substitutional mechanism, in which Zn diffuses as a singly ionized interstitial and is incorporated in the In sublattice as an electrically active substitutional acceptor or as an electrically inactive complex. At Zn concentrations lower than the background donor concentration the profile is cut off, as interstitial diffusion breaks down. The acceptor solubility increases with background donor concentration. Activation energies for diffusion and solubility were found to be 1.40 and 1.0 eV, respectively.

Tungsten as a marker in thin‐film diffusion studies

Vacuum‐deposited tungsten of about 30‐A thickness has been used as a diffusion marker in the reaction between a thin Co film and Si. In order to produce a discontinuous W film, it is necessary first to deposit a Sn film with island structure and then W, after which the Sn is dissolved. It was found by MeV He+ backscattering analysis that the W is located at the Co‐Co2Si interface, which means that Co is the dominant diffuser in Co2Si.

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 barrie...

Electromigration in Al Films Containing Si

Electromigration‐induced failure in Al films is retarded by the addition of Si. It is then governed by surface diffusion instead of by grain‐boundary diffusion as it is in pure Al films. This is evidenced by activation energies, which were found to be 0.3 eV in Al/Si films and 0.55 eV in Al films. Failure in the Al/Si films is due to void formation in a temperature gradient near the cathode, whereas in Al films the failure mode is void formation at structural inhomogeneities.