Nuclear Instruments and Methods in Physics Research B 237 (2005) 213–216 www.elsevier.com/locate/nimb
Thermal stability enhancement of silicides by using N2 and Ar implantation Tuung Luoh *, Maggie Liou, Hung-Wei Liu, Chin-Ta Su, Yung-Tai Hung, Ling-Wuu Yang, Chi-Tung Huang, Kuang-Chao Chen, Henry Chung, Joseph Ku, Chih-Yuan Lu Technology Development Center, Macronix International Co. Ltd., No. 16, Li-Hsin Road, Science Park, Hsinchu, Taiwan, ROC Available online 16 June 2005
Abstract Thermal stability of titanium and cobalt silicides were enhanced by implementing nitrogen or argon implantation prior to silicide formation. Silicide formed on N2 or Ar implanted blanket wafers, P-type, or N-type poly-silicon had better thermal stability as compared with non-implant ones. Furthermore, Ar implanted approach demonstrated superior thermal stability characteristics even in a RTP test of 1000 C/180 s. With the aid of N2 or Ar implantation, the grain growth of the silicide under high temperature was suppressed and thus it prohibited further diffusion and redistribution of the metal. 2005 Elsevier B.V. All rights reserved. PACS: 81.05.Je; 68.60.Dv; 85.40.Ry; 68.55.Ac; 73.40.Cg Keywords: Cobalt silicide; Thermal stability; Nitrogen implantation; Argon implantation; Grain growth; Resistance
1. Introduction Silicide is widely used as low-resistance gate electrodes and local interconnection in ULSI technology. In order to reduce parasitic resistance, sal-
*
Corresponding author. Tel.: +886 3 5786688x78173; fax: +886 3 5789087. E-mail address:
[email protected] (T. Luoh).
icide (self-aligned silicide) on gate and source/ drain regions is implemented for reducing sheet resistance. In addition to low resistance, good process compatibility with Si, e.g. ability to withstand high temperature, oxidizing ambient, various chemical cleans used during processing, little electro-migration, easy to be dry etched and good contacts to other materials make silicide process be implemented into many silicon integrated circuits. However, a serious limitation to the
0168-583X/$ - see front matter 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.nimb.2005.04.103
214
T. Luoh et al. / Nucl. Instr. and Meth. in Phys. Res. B 237 (2005) 213–216
application of silicide in the actual device is the thermal stability of silicide on long and narrow poly-silicon lines or in narrow contacts on Si substrate [1,2]. Cobalt silicide can be an attractive alternative to titanium silicide for CMOS process of 0.25 lm and beyond, because of its superior characteristics with low resistance, excellent chemical stability (up to 700–900 C), and little or no resistance degradation on narrow lines/gates [2]. However, it is well known that the degradation of CoSix after high temperature annealing is strongly affected by the salicide thickness [3]. Lateral silicide formation, non-planar silicide formation or silicide grain growth during high temperature annealing can result in silicide protrusions into junctions and thus junction leakage. For salicide on poly-silicon gates, grain growth problems can also lead to threshold voltage shifts or gate oxide leakage [4]. Ravesi et al. [5,6] found that Ar ion beam with beam energy of 1 keV made CoSix no increase in resistance up to 1000 C. The purpose of this paper is to investigate the titanium and cobalt silicide thermal stability by using N2 and Ar implantation on bare Si substrate, P-type and N-type poly-silicon for low-resistance interconnect application.
2. Experimental Thermal stability of titanium and cobalt silicide on blanket wafer, P-type, and N-type poly-silicon was investigated. Nitrogen or argon implantation prior to silicide formation was employed for thermal stability. Four-point probe sheet resistance, Rs, measurement was employed for pre- and post-RTP thermal annealing. The thermal stability was identified as the Rs deviation. The process flow is shown in Fig. 1 and described briefly as follows. Bare wafers, P+Poly-Si, or N+Poly-Si ! Sacrificial oxide growth ! N2 or Ar implantation ! Remove sacrificial oxide ! Ti or Co deposition ˚ , 80–140 A ˚ ) ! TiN deposition (50– (150–250 A ˚ 250 A) ! RTA1 ! Selective etch ! RTA2 ! Rs measurement ! RTP thermal annealing at 950 C or 1000 C ! Rs measurement.
Fig. 1. Process flow of CoSix, TiSix formation with N2 or Ar implantation.
3. Results and discussion 3.1. Thermal stability improvement on TiSix formation by nitrogen and argon implantation N2 and Ar implanted Si substrate were implemented prior to TiSix formation for thermal stability improvement. In Ti salicide formation, two-step annealing is used to get lower resistance and stable phase. It will form C49 phase and block lateral diffusion of Si in first step. In the second step, the higher annealing temperature transforms C49 phase to nucleation-limited C54 phase with low resistance. After silicidation, the N2 and Ar implantation indeed improve the thermal stability of TiSix formation, as shown in Figs. 2(a) and (b). As compared with directly TiSix formation, N2 and Ar implanted TiSix shows more stable Rs after enhanced 950 C RTP test. For the TEM cross-sectional images of N2 and Ar implanted TiSix, Ar implanted TiSix has smoother surface than N2 implanted one, as shown in Fig. 3. 3.2. Thermal stability improvement on CoSix formation by nitrogen and argon implantation CoSix formation is similar to TiSix process. It needs two-step annealing to achieve low-resistance silicide phase. After the second annealing step, non-implant-based CoSix demonstrates the lowest resistance among the CoSix with various N2 and Ar implantation conditions on blanket wafer. However, the thermal stability of non-implant-
T. Luoh et al. / Nucl. Instr. and Meth. in Phys. Res. B 237 (2005) 213–216 500
450
TiSix formation without Implant
450
400
N2 Implant 15KeV 2E15 N2 Implant 15KeV 6E15
400
Rs (Ohm/cm2)
Rs (Ohm/cm2)
500
350 300 250 200
TiSix Formation withoutImplant Ar Implant 35KeV 2E15 Ar Implant 35KeV 6E15
350 300 250 200
150
150
100
100
50
50 0
0
a
215
-50
0
50
100
150
200
250
300
350
400
-50
b
0
50
RTP Process Time (sec)
100
150
200
250
300
350
400
RTP Process Time (sec)
Fig. 2. Thermal stability of TiSix. by implantation of (a) N2 and (b) Ar.
Fig. 3. TiSix formation after RTA2 annealing: (a) N2 implantation of 15 keV/6 · 1015 and (b) Ar implantation of 35 keV/6 · 1015.
26
N2 Implant 15KeV 2E15
24
N2 Implant 15KeV 6E15
22
Ar Implant 35KeV 2E15
20
CoSix formation without Implant
1000C
18 2
Rs (Ohm/cm )
based CoSix becomes worse after using the enhanced 950 C RTP test for 60 s. On the contrary, CoSix with N2 or Ar implantation still shows their thermal stability even after 950 C RTP for 360 s, as shown in Fig. 4. For comparison between N2 and Ar implanted CoSix, Ar implanted condition exhibits superior thermal stability characteristics even enhanced RTP test up to 1000 C for 180 s. Eighty Angstrom Co deposition was implemented into CoSix formation on P-type and N-type poly-silicon with various N2 implanted conditions to investigate their thermal stability, as shown in Figs. 5(a) and (b). It shows that wafers with N2 implantation of 40 keV and lower dosage of 2 · 1015 (atoms/cm3) have stable and lower resistance. In addition, non-implanted N-type poly-silicon reveals unstable thermal stability as
950C
16 14 12 10 8 6 4 -50
0
50
100 150 200 250 300 350 400 450 500 550 600 RTP Process Time (sec)
Fig. 4. Thermal stability comparison among Ar implanted, N2 implanted and non-implanted CoSix formation.
216
T. Luoh et al. / Nucl. Instr. and Meth. in Phys. Res. B 237 (2005) 213–216 26
26
22
N2 Implant 15KeV 6E15
22
N2 Implant 15KeV 6E15
20
N2 Implant 40KeV 2E15
20
N2 Implant 40KeV 2E15
18
N2 Implant 40KeV 6E15
18
N2 Implant 40KeV 6E15
Rs (Ohm/cm2)
Rs (Ohm/cm2)
24
950C 1000C
16 14 12
1000C
16
950C
14 12
10
10
8
8
6
6
4
4 0
a
Non-Implant on P+POLY N2 Implant 15KeV 2E15
Non-Implant on P+POLY N2 Implant 15KeV 2E15
24
100
200
300
400
500
600
RTP ProcessTime (sec)
0
100
b
200
300
400
500
600
RTP ProcessTime (sec)
˚ Co deposition formed on N2 implanted and (a) P-type poly-silicon (b) N-type poly-silicon. Fig. 5. Thermal stability of CoSix with 80 A
compared with non-implanted P-type poly-silicon and other conditions even it has the lowest resistance before enhanced RTP test. Furthermore, N2 implanted condition with 40 keV and 6 · 1015 (atoms/cm3) has higher resistance as compared to others implantation conditions. 3.3. Thermal stability enhancement mechanism of CoSix formation on N2 and Ar implanted Si and poly-silicon The agglomeration starts with grain boundary grooving in the silicide. As a consequence, the grain separation and formation of silicide islands are observed and thus cause high-resistance interconnects and electrical leakage [1,2]. They depend on the ratio of the grain size and as-deposited film thickness. Ravesi [5] proposed that the grain size under a critical grain size might prevent agglomeration of silicide. With the aid of N2 and Ar implantation, silicide grain growth under high temperature annealing was suppressed. The grain boundaries are pinned by N or Ar atoms. Hence, the dissociation of Si in silicide, transport of Si atom in silicide, precipitation and epitaxial growth of Si, and deformation of silicide processes are retarded. It prohibits metal further diffusion and redistribution. The Ar ions demonstrate better pin-
ning capability than N2 and thus prohibit silicide further agglomeration. 4. Summary Silicide agglomeration has been suppressed successfully by N2 and Ar implantation. The thermal stability of TiSix and CoSix are enhanced. The stabilization of silicide layer with N2 or Ar implantation is attributed to the reduced atomic diffusion from grain boundaries by changing the energy balance at the grain surface. N2 or Ar implantations suppress the grain boundary grooving and retard the silicide agglomeration. References [1] E.G. Colgan, J.P. Gambino, Q.Z. Hong, Mater. Sci. Eng. R 16 (1996) 43. [2] J.P. Gambino, E.G. Colgan, Mater. Chem. Phys. 52 (1998) 99. [3] T.P. Nolan, R. Sinclair, R. Beyers, J. Appl. Phys. 71 (1992) 720. [4] W.M. Chen, S.K. Banerjee, J.C. Lee, Appl. Phys. Lett. 64 (1994) 1505. [5] S. Ravesi, F. La Via, V. Raineri, C. Spinella, Appl. Surf. Sci. 91 (1995) 19. [6] A. Alberti, F. La Via, S. Ravesi, S. Pannitteri, C. Bongiorno, Microelectron. Eng. 64 (2002) 151.