AN ATOM-PROBE COMPOSITIONAL STUDY OF Pd-Si INTERFACES

HAL Id: jpa-00225950

https://hal.archives-ouvertes.fr/jpa-00225950

Submitted on 1 Jan 1986

HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

AN ATOM-PROBE COMPOSITIONAL STUDY OF Pd-Si INTERFACES

T. Sakata, Y. Hasegawa, A. Kobayashi, T. Sakurai

To cite this version:

T. Sakata, Y. Hasegawa, A. Kobayashi, T. Sakurai. AN ATOM-PROBE COMPOSITIONAL STUDY OF Pd-Si INTERFACES. Journal de Physique Colloques, 1986, 47 (C7), pp.C7-321-C7-326.

�10.1051/jphyscol:1986755�. �jpa-00225950�

JOURNAL DE PHYSIQUE

Colloque C7, supplément au n o Il, Tome 47, Novembre 1986

AN ATOM-PROBE COMPOSITIONAL STUDY OF Pd-Si INTERFACES

T. SAKATA*, Y. HASEGAWA, A. KOBAYASHI and T. SAKURAI

The Institute for Solid State Physics (ISSP), The University of Tokyo, Minato-ku, Tokyo 106, Japan

' ~ o l l e g e of Integrated Arts and Science, University of Osaka Prefecture, Sakai, Osaka 591, Japan

Abstract

The compositional analysis of thin films of Pd-silicides to the Si substrate has been performed for the first time using an atom-probe field ion microscope. Pd-silicides were grown on Si emitter surfaces at 50 K and 823 K (550 C). It was found that the outermost layer of the films was of pure Si in both cases and that it was followed by Si rich Si-Pd mixed region, PdSi, Pd rich mixed region, PdnSi and the substrate. The interface reactions were discussed in terms of a substitution mode1 of Si and Pd atoms at the reaction front of the interface.

1. Introduction

Metal-semiconductor interfaces have been studied extensively [1,2]

due to their importance in the electronic device applications. Pd-Si is one of the most interested and studied systems among the

metal-semiconductor systems. It has been known that only one silicide phase, PdzSi, exists as a stable phase upon heating up to 700 C.

Atom-probe field ion microscopy is a unique technique for investigating surfaces and interfaces [ 3 ] and has been successfully applied for the micro analysis of metal-semiconductor interfaces C4,51. In the present work, a compositional analysis of the Pd-Si system has been performed by using the focusing-type ToF atom-probe field ion microscope at ISSP [61.

2, Experimental

The atom-probe field ion microscope used in the present work were described elsewhere in detail [5,6,71 and will not be discussed here.

The Si specimen tips were fabricated by cutting the commercially available Si wafers in a square rod of 0.5 x 0.5 mmz with the 11001 direction along the tip axis[8]. They were mounted on a 0.1 mm

diameter Mo U loop and chemically etched into sharp needles. The Si tip was field-evaporated in the presence of 10-5 Torr hydrogen in the main atom-probe chamber to obtain an atomically clean surface.

Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphyscol:1986755

C7-322 JOURNAL D E PHYSIQUE

A Pd evaporator is a Pd wire of 0.05 mm diameter wound on a W filament of 0.1 mm diameter. It was placed 2 cm away from the Si tip in the main chamber. Pd was evaporated in-situ on the Si tip by resistively heating the filament. Before the Pd deposition, the hydrogen imaging gas was pumped out and Pd was deposited on the Si tip surface in the base pressure below 1x10-10 Torr. After the

deposition, the Si tip was heat treated in UHV at various temperatures. The tip temperature was controlled by resistively heating the Mo loop and was monitored by a Au-0.07%Fe/chromel thermocouple and an optical pyrometer.

Compositional analysis was performed in the region of the interface by a slow evaporation rate of 0.003 atom/pulse in the presence of approximately 10-8 Torr hydrogen. Various pulse

fractions were used in a range between 0.05 to 0.4. A difficulty was encountered when we attempted smooth pulse evaporation crossing the interfacial boundary of the PdzSi phase-Si substrate.

3. Results and discussion

It was reported that the Pd silicide was formed at the substrate temperature as low as 180 K basically in the same way as at room

temperature 181. In the present work, we have found that the silicide could be formed even at 50 K. In Fig. 1 is shown a typical mass histogram obtained from a sample which was kept at 50 K during and after deposition. It is clearly seen that Si was detected as'Si+, Siz+ and their hydride ions and Pd as P d + , Pdz+ and their

hydride ions. The complex ions of Si and Pd atoms were not detected.

In Fig. 2 is shown the cumulative number of Si or Pd ions from the same sample against the total number of signals. From Fig. 2, it is seen that the outermost layers consist of pure Si atoms.

It was previously reported by Oura et a1.[91 that pure Si layer with about 3 A thickness was formed on Pd2Si when annealed at 250 -

600 C. However, it is interesting to note that the pure Si layers were formed even at such a low temperature as 50 K. This fact may suggest that the diffusion process is not a limiting step for the formation of the Pd silicide. Instead, a spontaneous formation of the silicide is plausible through some other mechanisms, which will be discussed later. In order to analyze the composition variation in more detail, Fig. 2 was devided into several segments for enlargement, each of which included approximately 70 ion signals. This enlargement makes possible to observe the composition of each segment in detail.

HASS TO CHARGE R A T I O (m/n)

Pig.1 Mass histogram from Pd deposited

on the Si tip at 50 K.

Fig.2 Cumulative numbers of individua Si and Pd ions detected are

plotted against the total number of signals.

80

e u

Fig.3 Enlarged segments of Fig. 2. A , B and C correspond to A , B and C in Fig. 2, respectively.

- Si A ^u _e

k

VI LL O

c x

3

O

Some examples are shown in Fig. 3. The compositional variations thus determined are summarized schematically in Fig. 4. It can be seen from these figures that the composition varies continuously from the outermost layer to the interface. The silicide phases identified from these analyses were PdSi, PdzSi and PdSi4. This finding is

2

---

V>

Pd

LI O W OL

m

PdS i

O

O NUMBER OF Si+Pd 120

- si i

--- _Pd

/,d,,i

s p

O NUMBER OF S i+Pd 120

C7-324 JOURNAL DE PHYSIQUE

interesting because it has so far been believed that only one Pd-Si phase, PdzSi, exists as a stable bulk phase upon annealing up to 700 C. However, it has been noted recently that excess Si atoms existed in the Pd2Si layer and decreased with distance from the interface [IO].

Depth from the surface +

Fig.4 Schematic diagram of concentration gradient determined from Figs. 1-3.

MASS TO CHARGE R A T I O ( m / ~ )

n

g ²⁰⁰

W k W

150

V1 Z O

....

; 100

Fig.5 Mass histogram from the Pd-Si interface formed by annealing at 823 K for 3 min.

Another interesting finding is that there existed a monolayer of PdSi4 right at the interface (see Fig. 3). The existence of the PdSir was foreséhn by Freeouf [Ill to explain the change of the work function at the first monolayer of Pd on Si(ll1). We take our observation here as the first evidence for the existence of PdSi4 at the interface.

Another set of experiments was carried out by annealing the Si tip at 823 K for 3 min. after the deposition of Pd. The mass

histogram obtained is shown in Fig. 5, which is similar to that shown in Fig. 1 obtained without annealing. In Fig. 6 plotted are the cumulative number of Si or Pd ions against the total number of

signals. In Fig. 7 are several enlarged segments of the data in Fig.

6.

O CL

@ 50 P d2' Pd'

3 Z

50 LOO

S ^i2'

Si'

Fig.6 Cumulative numbers of individual Si and Pd ions detected from the sample used in Fig. 5.

LL O w

w m

r / ^{PdS i}

3 Z

/&---

" A- ^__..--

Fig.7 Enlarged segments of Fig. 6. A and B correspond to A and B in Fig. 6, respectively.

Fig. 8 is the schematic diagram of the compositional depth profile in this case. The depth profiles shown in Fig. 8 are essentially the same as that shown in Fig. 4 up to the beginning of the Si substrate.

As noted earlier, the Si tips fractured almost al1 the cases after the detection of the PdrSi phase when the Pd-Si was annealed at 823 K.

The cause for this tip fracture is not clear at present, but it is speculated that the evaporation field of the PdzSi phase may be significantly higher than that of the Si substrate. This can lead to the tip fracture due to the increased stress by the high electric field needed to field evaporate the interface layer. A similar difficulty was noticed in the case of Ni segregation observed in the course of surface segregation study of Ni-Cu binary alloys [12].

Assuming that the tip fractured at the sharp boundary between the Pa-silicide phase and the Si substrate, we can infer the depth profile after the tip fracture in dashed lines shown in Fig. 8.

We propose a following model for the mechanism of Pd silicide formation to explain the experimental results presented here. At the reaction front of the interface, Si and Pd atoms interdiffuse

simultaneously by substituting each other. At the region near the interface, the stable PdzSi phase is formed and the excess Si atoms diffuse to the vacuum surface. This model would explain the existence of pure Si layer at the surface and the observed concentration

gradient in Figs. 4 and 8.

JOURNAL DE PHYSIQUE

k p t h fram the surface

Fig.8 Schematic diagram of concentration gradient determined from Figs. 5-7.

Finally, we note that the pulsed laser atom probe field ion

microscopy Cl31 may be helpful for the investigation of the interface formed at high temperature to avoid the tip fracture.

References

1 ) Thin Films-Interdiffusion and Reactions, edited by J. M. Poate, K. N. Tu and J. W. Mayer (Wiley, NY,1978)

2) L. J. Brillson, Surf. Sci. Reports 2 (1982) 123.

3) T.Sakurai, T.Hashizume, A.Jimbo and T.Sakata, J.Phys.Colloq.

(1984) Cg-453.

4) S. Nakamura, T. Hashizume, Y. Hasegawa and T. Sakurai, Surf. Sci.

Lett. (1986) in press.

5) A.Kobayashi, T.Sakurai, T..Hashizume and T.Sakata, J. Appl. Phys.

59 (1986) 3448

6) K ~ a k u r a i , T.Hashizume and A.Jimbo, Appl. Phys, Lett. (1984) 38.

7) T. Sakurai and T, Hashizume, Rev. Sci. Instrum. a (1986) 236.

8 ) T. Sakurai, T. Sakata and A. Jimbo, Jpn. J. Appl. Phys. 22 (1983) L775.

9 ) G. W. Rubloff, P. S. Ho, J. F. Freeouf and J. E. Lewis, Phys. Rev.

B23 (1981) 4183.

10) K. Oura, S. Okada and H. Hanawa, Appl. Phys. Lett. (1979) 705.

11) J. L. Freeouf, Solid State Comm. 33 (1980) 1059.

12) T. Sakurai, T. Hashizume, A. Jimbo, A. Sakai and S. Hyodo, Phys.

Rev. Lett. (1985) 514.

13) A.Jimbo, T.Hashizume, T.Sakata and T.Sakurai, J. de Phys. 46

(1985) Cl0