Behaviour of implanted arsenic during rapid thermal annealing of Ti on Si

Ponpon, J.P.; Saulnier, A.; Stuck, R.

doi:10.1007/bf00616694

Cited by 7 publications

(3 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…XRD spectra of a Ti-Ta film on Si ͑111͒ RTA at 1123 K for the times indicated.techniques,8,[16][17][18][19][20][21]21 and the variation in the results were analyzed in Ref. 14 in favor of the value of 2.5-2.6 eV.…”

mentioning

confidence: 99%

A conductance model (approach) for kinetic studies: The Ti–Ta–Si system

Rubinovich

2000

Journal of Vacuum Science &Amp; Technology A: Vacuum, Surfaces, and Films

View full text Add to dashboard Cite

Electrical resistance measurements provide a convenient way to study the kinetics of transformations in isothermally annealed thin films. The measured resistance and the amount of a new phase formed during some sort of reaction in a transformation process are often related by a linear function. This simple relationship between the measured resistance and the fraction of the new phase formed is not always valid and the reliability of the results might be questionable. In this communication, a different approach is suggested, which is based on conductance. The validity of the model is tested by experimental data for the Ti-Ta-Si system. The suggested and the linear resistance models are compared, and it seems that it is more appropriate in certain cases to use the conductance approach for kinetic purposes. An effective activation energy can be derived from the results by defining the time needed at each temperature to obtain either a constant resistance, or alternatively, a certain fraction of the new phase. The activation energy derived for the end phase, i. e. ͑Ti,Ta͒Si 2 formation from the experimental data in the example ͑Ti-Ta-Si͒ considered is ϳ2.6 eV.

show abstract

mentioning

confidence: 99%

A conductance model (approach) for kinetic studies: The Ti–Ta–Si system

Rubinovich

2000

Journal of Vacuum Science &Amp; Technology A: Vacuum, Surfaces, and Films

View full text Add to dashboard Cite

show abstract

“…The redistribution of arsenic into the TiSi 2 was also observed in conventional SALICIDE processes. 7,8,21 By using TiSi 2 as an arsenic diffusion source, Privitera et al found that the segregation coefficient K s , where K s ϭ C As ͑TiSi 2 ͒/C As ͑Si͒, was 5 at 800°C and increased to 15 at 1000°C. The value of the segregation coefficient suggests that arsenic is more likely to reside in TiSi 2 instead of the Si substrate.…”

Section: Resultsmentioning

confidence: 99%

“…Extensive research has been carried out on this topic. [7][8][9] It was found that dopants in Si reduce the silicide growth rate, with arsenic having the strongest impact. Furthermore, during conventional TiSi 2 formation, dopants easily diffuse into the growing silicide without a ''snow plow'' effect.…”

mentioning

confidence: 99%

Arsenic Redistribution during Rapid Thermal Chemical Vapor Deposition of TiSi[sub 2] on Si

Fang

Öztürk

O’Neil

et al. 2001

J. Electrochem. Soc.

View full text Add to dashboard Cite

This paper studies the redistribution behavior of implanted arsenic during selective rapid thermal chemical vapor deposition of titanium disilicide (TiSi 2 ). The arsenic implant doses ranged from 3 ϫ 10 14 cm Ϫ2 to 5 ϫ 10 15 cm Ϫ2 . The TiSi 2 films were deposited either directly on arsenic-implanted silicon substrates or on epitaxial silicon buffer layers selectively deposited with varying thicknesses before TiSi 2 depositions. SiH 4 and TiCl 4 were used as precursors for TiSi 2 depositions and Si 2 H 6 and Cl 2 for selective silicon epitaxial growth. Experimental data revealed that the majority of the implanted arsenic was lost from the silicon substrate into the deposited TiSi 2 films when epitaxial silicon buffer layers were not employed. With the inclusion of the buffer layer, the arsenic loss could be reduced significantly. The loss of arsenic observed could not be explained by considering substrate consumption alone. In both cases, arsenic exhibited strongly enhanced out-diffusion from the silicon substrate into the TiSi 2 film. The injection of vacancies during the TiSi 2 depositions has been proposed as the reason for this enhanced out-diffusion. Monte Carlo simulations have been performed to verify the proposed model. Titanium disilicide (TiSi 2 ) is widely used to form low resistivity contacts to silicon devices. 1 Self-aligned silicide ͑SALICIDE͒ technology is used to form TiSi 2 on source/drain junctions and polysilicon gate electrodes of complementary metal-oxide-silicon ͑CMOS͒ devices. 2 As device dimensions shrink into the deep submicrometer regime, reducing the silicon substrate consumption and achieving the low resistivity C54 phase of TiSi 2 is becoming much more difficult. 3 Selective chemical vapor deposition ͑CVD͒ of TiSi 2 offers an alternative process solution with the potential to alleviate these problems. By introducing titanium and silicon from gaseous sources, TiSi 2 can be formed without substrate consumption. 4,5 Preliminary results have shown that formation of TiSi 2 on fine polysilicon lines can also be improved by CVD. 6 Because TiSi 2 is formed simultaneously on source/drain and polysilicon gate electrodes doped with different impurities, it is important to study the interactions between silicide formation and dopant redistribution in silicon. Extensive research has been carried out on this topic. 7-9 It was found that dopants in Si reduce the silicide growth rate, with arsenic having the strongest impact. Furthermore, during conventional TiSi 2 formation, dopants easily diffuse into the growing silicide without a ''snow plow'' effect. A recent study from this laboratory explored selective CVD of TiSi 2 on heavily arsenicdoped substrates. 10 It was found that arsenic introduces a barrier to TiSi 2 nucleation and results in enhanced substrate consumption. It was also shown that arsenic displays a strong tendency to diffuse into TiSi 2 during deposition. Arsenic loss into TiSi 2 during CVD processing was also observed by Southwell et al. 11 They attributed the effect to the injec...

show abstract