Co-silicides were prepared with several techniques, such as annealing of evaporated. Co-layers on a Si-substrate (silicide surface layers) and annealing of ...
Hyperfine Interactions 57 (1990) 2133-2140
2133
C o - S I L I C I D E L A Y E R S S T U D I E D BY M ( ) S S B A U E R S P E C T R O S C O P Y AND RUTHERFORD BACKSCATYERING SPECTROSCOPY A. V A N T O M M E a, M . F . W U 2 I. DI~ZSI 3, p.Q. Z H A N G a n d G. L A N G O U C H E
4
Instituut voor Kern- en Stralingsfvsika, Universitv of Leuven, B-3030 Leuven, Belgium
Co-silicides were prepared with several techniques, such as annealing of evaporated Co-layers on a Si-substrate (silicide surface layers) and annealing of Co-implanted Si (buried silicide layers). By adding some 5VCo to the stable S9Co, the formation of the various Co-silicides could clearly be followed as a function of annealing temperature by means of MiSssbauer spectroscopy. In the case of surface silicide layers, Co2Si, CoSi and CoSi2 were formed subsequently. In the case of buried layers however, CoSi~ was the only crystalline phase that could be observed. In both cases, the CoSi 2 spectra showed an anomalous side resonance. Moreover, it was found that when 5VFe was implanted (instead of 57Co), a drastic increase in the intensity of this side resonance could be detected by CEMS.
I. Introduction T r a n s i t i o n m e t a l silicides (like CoSi2) h a v e b e e n u n d e r v e r y i n t e n s i v e investigation d u r i n g r e c e n t years. Both s u r f a c e a n d b u r i e d layers of CoSi 2 are of g r e a t interest for a p p l i c a t i o n s in m i c r o e l e c t r o n i c s [1-3]. C o S i 2 is c h o s e n as a m e t a l l a y e r b e c a u s e of its low resistivity a n d its g o o d t h e r m a l stability. M o r e o v e r , its crystal s t r u c t u r e is c o m m e n s u r a t e with Si (with a small lattice m i s m a t c h of o n l y 1.2%). In this c o n f e r e n c e p a p e r , we w a n t to give a review of the m a j o r results of o u r Co-silicide studies, a n d we refer to a n u m b e r of e x t e n s i v e r e p o r t s [ 4 - 7 ] for m o r e details.
2. Experimental S u r f a c e Co-silicide layers were f o r m e d b y f u r n a c e a n n e a l i n g ( f o r 1 hr u n d e r H , ) of C o - c o v e r e d Si single crystals. A n 80 n m thick C o - l a y e r was m a g n e t r o n s p u t t e r e d o n t o a S i ( 1 0 0 ) s u b s t r a t e . 57Co w a s t h e n i m p l a n t e d w i t h a t o t a l d o s e of 1 2 3 4
Research Assistant, N.F.W.O. (National Fund for Scientific Research, Belgium). On leave from the Peking University, Beijing, China. On leave from the Central Research Institute for Physics, Budapest, Hungary. Present address: Institute of High Energy Physics, Beijing, China.
9 J.C. Baltzer A.G. Scientific Publishing Company
2134
A. Vantomme et al. / Co-silicide layers studied bv M.S. and backscattering
3 x 10 TM a t o m s / c m 2 and an energy of 80 keV. This gives an implantation range of about 24 nm, with a width of 16.5 nm [8]. The M/Sssbauer spectra are therefore representative for the behavior of the Co-layer. Buried silicide layers were obtained by high dose implantation of Co into Si(111) with energies of 50, 70 and 160 keV at a substrate temperature between 310~ and 350~ The ion beam consisted of 59Co, which was alternated with small fractions of SVCo M/Sssbauer probes. These samples were subsequently annealed (for 1 hr at 600~ and 0.5 hr at 1000 ~ under a hydrogen atmosphere. All M6ssbauer spectra were measured at room temperature, using a single line absorber of 0.5 or 1.0 m g / c m 2 57Fe in sodium ferrocyanide (SFC : Naa[Fe(CN)6].10H20). The isomer shift values were converted to absorber isomer shifts relative to ~-iron. Rutherford Backscattering Spectroscopy (RBS) measurements were simultaneously done and compared with the M/Sssbauer data. Finally, highly pure MBE-grown CoSi 2 was implanted with 3 • 1015 5VFe a t o m s / c m 2 at 40 keV with the target kept at 300 ~ A target from the same MBE-grown sample was implanted under the same conditions with 57Co with 1015 a t o m s / c m 2. We compared the transition M6ssbauer spectrum of the 5VCo-source with the CEMS spectrum (using a 57Co in Cu single line source) of the 57Fe target.
3. Discussion
3.1. Co-SILICIDE SURFACE LAYERS After each annealing step of the Co-covered Si-sample, we compared the obtained M~Sssbauer spectrum with the known [9] spectra of the various Co-silicides. Figure 1 compiles the results of all our spectra. It shows that during annealing of this system up to 900~ Co2Si, CoSi and CoSi 2 are formed subsequently. It is a remarkable fact that two different Co-silicides can grow simultaneously: between 350 ~ and 425 ~ Co2Si and CoSi coexist, and between 475 ~ and 575 ~ CoSi and CoSi 2 are formed together. This phenomenon has only been discovered for two transition metal silicides, namely for Co and Pt [10]. RBS measurements were done simultaneously and confirmed fully our temperature dependent M/Sssbauer study. These RBS-data further show that the various silicide layers are formed in parallel layers on top of the Si-substrate. A more detailed study of our surface layer experiments is given in Ref. [4]. A surprising observation is that, although CoSi 2 has a cubic lattice (CaF 2, [11]) and thus is expected to give rise to a single line M/Sssbauer spectrum, we always find a broad side resonance next to the dominating single line at relative velocity v = - 0 . 0 9 m m / s (fig. 2), the origin of which is not straightforward. Some possible interpretations will be discussed further (3.3).
A. Vantomme et al. / Co-silicide layers studied by M.S. and backscattering 0
I-- 0.80 < J LLI r'r" 0.75 0.70
-ts'
-1.o '
-o.5 '
O. o
o.5 '
lo '. VELOCI
t5 '
TY[m m/s)
Fig. 2. The 57Fe M~Sssbauer spectrum of a 57Co-doped CoSi 2 surface layer on Si ( M B E deposited).
2136
A. Vantomme et al. / Co-silicide layers studied by M.S. and backscattering Energy 25
x:J
1.0
20
(MeV)
1.2
1.4
1.6
I
I
!
o~
a
o a
>7D (lJ N
QQ
~
15
O O O
Q
Q
O 10
E L.-
Q
o Z
Q
5
9 -- ~"~'~d.~
0
~
I
2oo
250
m I~
300 Chennel
Fig. 3. RBS spectra of a buried CoSi 2 layer after ]000 ~ Co/cm
2 at 70 keV). o, random:
I
350
4oo
almealing ( i m p l a n t a t i o n w i t h 0.9 x ] 0 iv zx, a l i g n e d .
at positive velocity side show the existence of other local phases as well (CoaSi, CoSi, interstitial Co . . . . ). The RBS spectrum shows a Co-implantation profile with a nearly Gaussian shape, and a Co-peak concentration of only 22%, which is not enough to form a continuous (macroscopic) CoSi 2 layer. The rather high m i n i m u m yield of the aligned spectrum (Xmin = 50%) indicates that the sample is still heavily damaged. Combining these channeling results with those of the M6ssbauer measurements, we can conclude that after implantation, microscopic precipitates of various silicides are formed, the largest part being CoSi 2. The spectrum taken after annealing at 6 0 0 ~ is the same as in the case of surface CoSi2-1ayers on Si (fig. 2), which means that we have the same kind of anomaly as in the former case. The RBS spectrum confirms the occurrence of a buried continuous CoSi 2 layer. However, when we measure in channeling geometry, it is found that the minimum yield Xmin = 25%, indicating that a good quality single crystal is not yet obtained. Only after annealing at 1000 o C, the Xmin drops to 5% (fig. 3). Considering this enormous improvement of the crystalline quality, it is a surprising fact that the M~Sssbauer spectrum doesn't change at all: the same type of spectrum as in fig. 2 is obtained. A cross sectional T E M picture (fig. 4) corroborates that, finally, we obtain a monocrystalline and continuous buried layer with sharp interfaces (the interface roughness is estimated to be only about 15 A). More details about the formation of buried Co-silicides can be found in ref. [6].
A. Vantomme et al. / Co-silicide layers studied by M.S. and backscattering
*------
Si
2137
Surface
Buried
CoSi2
Layer
Si S u b s t r a t e
[
2oo
Fig. 4. Cross sectional T E M picture of a buried CoSi2 layer (implanted at 50 keV with a dose of 0.6 • 10 Iv C o / c m 2, and annealed at 1000 ~ C).
3.3. C O M P A R I S O N O F 57Fe IN CoSi 2 A F T E R D I R E C T ION I M P L A N T A T I O N A N D A F T E R ION I M P L A N T A T I O N O F A R A D I O A C T I V E P A R E N T
Figures 2 and 5 show the M6ssbauer spectra of the 5VCo and 57Fe implanted MBE-grown samples, respectively. Both spectra can be fitted with the same set of parameters. The two observed resonances have isomer shifts of 61 = -0.09(5) m m / s and 32 = +0.41(5) m m / s respectively. The striking observation is that, whereas almost all SVFe atoms after 5VFe implantation are found with isomer shift 32, almost all 57Fe atoms after 5VCo implantation are found with isomer shift 61. Z O
W 1.06 LIJ I",< _J 1.04
or1.02
1.00 |
I
-1.s
-I.O
I
-o.5
I
I
I
I
o.o
os
I~
1.s
VELOCITY [ram/s) Fig. 5. The 57Fe M6ssbauer spectrum of a 57Fe-doped CoSi 2 surface layer on Si ( M B E deposited).
2138
A. Vantomme et al. / Co-silicide layers studied by M.S. and backscattering
In the case of 57Co(CoSi2), the dominating M~Sssbauer resonance (fig. 2) with isomer shift 8~ is considered to be due to 57Fe in the decay of 57Co on regular Co lattice sites in cubic CoSi 2 [4-6]. The origin of the second resonance is much harder to account for. Its isomer shift is close to the value measured for CoSi, but careful X-ray analysis could not detect any crystalline trace of this phase. It cannot even be associated with the implantation process as it also appears after diffusion in bulk CoSi 2. We can tentatively propose two interpretations for this anomalous resonance: (1) It is associated with SVFe atoms in the decay of 57Co atoms in a lattice configuration which is different from the normal substitutional Co atoms in CoSi2. As the electron density at the 5VFe nucleus is close to the one observed in CoSi, a local deviation from the correct stoichiometry towards a more Co-rich one can be responsible for this. (2) The 57Co parent atoms are found in a correct CoSi 2 stoichiometry, but an anomalous charge state is stabilized at least during the 200 nanosecond lifetime of the excited M~Sssbauer state of 5VFe. When the 57Fe isotope is implanted rather than SVCo, it is a surprising observation to discover (fig. 5) that the large marjority of 57Fe is found with the same hyperfine interaction parameters as the anomalous resonance of the 57C0 implantations. So far we do not have extensive data on different types of CoSi~ implanted with 57Fe, so that we cannot make a full comparison under differeni conditions. The present observations can somehow fit in both proposed interpretations for 5VCo: (1) During the final relaxation phase of the collision cascade, it can not be excluded that an impurity atom like Fe ends up in a somewhat different local stoichiometric surrounding than regular CoSi2. What we then observe is that the local (Co-rich?) configuration which is preferred by the majority of the Fe atoms is the same as the one preferred by a minority of the Co atoms. (2) If the two isomer shifts are due to different electronic configurations in the same stoichiometric CoSi 2, we have to conclude that both are very stable, and that Fe atoms that have been implanted and came to rest long before the M/Sssbauer experiment prefer a different electronic configuration than the Fe atoms that came into existence only nanoseconds before the M/Sssbauer experiments.
4. Conclusion
MiSssbauer spectroscopy, complemented with RBS and channeling spectroscopy, has shown that after annealing of Co surface layers on Si, C%Si, CoSi and CoSi 2 are formed subsequently. When Co is implanted in Si, buried CoSi 2 can be formed after 600 ~ C annealing. The best crystalline quality is reached after
A. Vantomrne et al. / Co-sificide layers studied by M.S. and backscattering
2139
1000~ annealing, as confirmed also by TEM. Finally, our study shows that different hyperfine interaction parameters are observed at Fe atoms that were directly implanted into CoSi 2, and Fe atoms that were formed due to the radioactive decay of implanted 57Co atoms.
Acknowledgements This work was supported by the Belgian I.I.K.W. and G.O.A. (Government Concerted Action) projects. The authors would also like to thank J.M. Phillips for providing the MBE-samples.
References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11]
J.C. Hensel, A.F.J. Levi, R.T. Tung and J.M. Gibson, Appl. Phys. Lett. 47 (1985) 151. J.C. Hensel, MRS Symp. Proc. 54 (1986) 499. S.P. Muraka, in: Silicidesfor VLSI Applications (Academic Press, 1983). A. Vantomme, I. D6zsi and G. Langouche, Hyp. Int. 41 (1988) 725. A. Vantomme, I. D4zsi and G. Langouche, Nucl. Instr. Methods B39 (1989) 284. A. Vantomme, M.F. Wu, I. D4zsi, G. Langouche, K. Maex and J. Vanhellemont, to be published in E-MRS Syrup. Proc. (1989). A. Vantomme, I. D6zsi and G. Langouche, to be published in Nucl. Instr. Methods. J.F. Ziegler, J.P. Biersack and U. Littmark, in: The Stopping, and Range of Ions in Matter, ed. J.F. Ziegler (Pergamon, New York, 1985) p. 202. I. D6zsi, H. Engelman, U. Gonser and G. Langouse, Hyp. Int. 33 (1987) 161. G.J. van Gurp and C. Langereis, J. Appl. Phys. 46 (1975) 4301. M.-A. Nicolet and S.S. Lau, in: VLSI Electronics: Microstructure Science, Vol. 6 (Academic Press, 1983) p. 402.