Results and discussion
4.3 STI gap fill
Figure 4.20 Wet etch rate vary TEOS trend chart
Figure 4.20 shows that increases in TEOS flow, wet etch rate also increases. The important of these results are beneficially to design both nucleation deposition and gap fill deposition to complete trench void free as well.
4.3 STI gap fill
4.3.1 Trench gap fill study by partial deposition
For STI gap filling, it is found that the void free of STI trench after steam and dry annealing, ramp up conditions of TEOS also show void-free, seamless in trenches for all splits. SEM shows that there is a very well bottom up process for high aspect ratio trench, the solution is to for the first step uses a method called “TEOS ramp” with an initial start of low TEOS on a fractional flow per seconds. The low TEOS flow starves deposition process with an O3/TEOS ratio up to 1000mgm TEOS flow and results in a homogeneous nucleation layer and a less surface selective film growth (figures 4.21-22). The figures 4.23-24 show the gap filling after deposition 1 and 2 and it found there
TEOS flow Vs Etch rate
0 1000 2000 3000 4000 5000 6000 7000
TEOS
Etch rate
Etch rate (nm/min)
- 102 -
is a very well STI gap fill without furnace annealing, no seam or void observed after trench gap filling process, so it can be define the initial homogeneous nucleation is extremely important for trench gap fill with bottom up deposition, TEOS ramps to 2700mgm with same ozone concentration maintaining also provides a very well trenches filling capability and no interfaces interruption.
After bi-layers gap fill monitoring, trenches filling with the whole stack film monitoring show there is seamless and voids free for the whole stack film deposition after steam annealing (figures 4.25-4.27).
- 103 -
4.3.2 STI gap fill check with three steps recipe
The studies described here indicate a promising approach to improve HARP SA-USG for STI gap fill in technology of 70nm and beyond. Initial
- 104 -
experiments data also show that process with TEOS ramp-up during the first stage deposition can improve gap fill capability. The most important difficulty encountered during the first stage tests is how to obtain a more flexible and reliable TEOS ramp-up. Software and hardware work requests have been issued to meet the desired ramp situation.
Finally, we had some window splits for trenches filling effect on the initial TEOS flow and TEOS ramp up rate during transition process step, also, different anneal conditions will be tested to more efficiently heal the seams, of which we need to optimize annealing conditions for better understanding of the whole process and to set up the final process flow.
4.4 STI gap fill windows check
4.4.1 O3/TEOS ratio adjustment
A number of approaches were attempted to improve gap filling (DOE_1~5), optimization of the existing process in the critical variables, and a high ozone process in which the TEOS is introduced in a varying but controlled manner (TEOS ramp). From this work, it became apparent that the initial oxide layer is very critical in achieving a good gap-fill. The high quality oxide film essential for good gap-fill needs an extremely Ozone rich environment coupled with low TEOS concentration in the gas phase. But this leads to deposition rates that are unacceptably low. At the time it was conjectured that the later layers may tolerate increasingly poorer ozone environment or conversely increasing higher TEOS concentrations in the gas phase. This leads to the development of TEOS ramp process that introduces very low amounts of TEOS in ozone rich environment to get a high quality
- 105 -
initial oxide layer followed by a continuous but slow increase in TEOS concentration, so that the subsequent layer is deposited faster without adversely affecting the trench filling.
There were some TEOS ramp rate tests with same O3 concentration to monitor gap fill capability by using the same matrix wafers. TEOS ramp rate 5mgm/sec is the proper flow can achieve the gap filling with steam anneal, but, TEOS ramp rate at 15mgm/sec showed poor gap-fill result with thinner linear oxide of initial deposition.
For the critical structure, to achieve STI gap filling does not only need to use low TEOS ramp rate , but also to have enough liner oxide thickness to prevent bigger seam locating at the bottom of trench.
- 106 -
4.4.2 Anneal conditions
Steam annealing gives a very good gap fill in dense areas and very good conformality in opening area. The another challenge is reflow angle after furnace annealing, the less reflow angle means movable film is more sensitive to high temperature, the effect of steam annealing not only give a seamless and/or void free results, but also it provides a less reflow angle to beneficial on chemical mechanism polishing .
- 107 -
Comparing to steam annealing, using dry furnace annealing can not get a better gap fill result as what we respected. It found that voids found in trenches whatever in dense areas or opening area, the reflow angle is also steep comparing with steam annealing. Thus, it can be concluded that dry furnace annealing is beneficial on film densification, but high temperature may prohibit significant diffusion of movable USG film.
- 108 -
Furnace reflow is very important to HARP SA-USG, the experiments show that gap fill achieving is mainly by steam annealing, but dry anneal may densify the stack film to get the closely film properties likes HDPCVD film that which is beneficial on chemical mechanism polishing. For those STI structure wafers were with oxide liner, if just use high temperature of dry annealing that which causes silicon oxidation and gap filling can not be achieved as well as it could. The steam annealing by furnace results in a very well trench filling at 8000C, there is also no measurable trench oxidation found.
In order to have the compatible film density as well as HDPCVD, we approached two steps annealing to get the closely film density. The acceptably result of two steps annealing will be the finalized reflow in our test, steam reflow at 8000C for 30 minutes and a sequent dry anneal at 10000C for 30minutes. The approach met the criteria of gap filling and film density as well as HDPCVD without any other side effects.
- 109 -
Chapter 5 CONCLUSION
From this study, we have found that the first few HARP USG layers are critical to achieve a good gap-fill. The high quality oxide film essential for good gap-fill needs an extremely ozone rich environment coupled with a low TEOS concentration in the gas phase. Even though this TEOS ramp method provides a void-free gap filling and improved seam performance, it still cannot completely eliminate this weak region for more aggressive structures, especially after the film is subjected to a mandatory high temperature anneal.
A solution to this problem turns out to be in the modification of the annealing environment. Addition of the steam during anneal has shown a remarkable improvement in the healing of seams In order to meet the demand for void- free gap-fill for the 65nm generation STI using HARP process in the HARP SACVD tool, potential hardware concerns and optimization of process windows has to be properly addressed.
Finally, no plasma damage issue to the active areas and conformal films enable high aspect ratio trench is the biggest advantages of HARP SACVD process for the shallow trench isolation gap filling of advanced DRAM.
- 110 -
References
[1]. Ootsuka, F.; Wakahara, S.; Ichinose, K.; Honzawa, A.; Wada, S.; Sato, H.;
ando, T.; Ohta, H.; Watanabe, K.; Onai, T., “A highly dense, High-Performan -ce 130nm node CMOS Technology for large scale system-on-a-chip applicat- ions”, Electron Devices Meeting, IEDM Technical Digest. International, 575-578, (2000).
[2]. W. Kern and G. L. Schnable, “Low Pressure Chemical Vapor Deposition for VLSI Processing-A Review”, IEEE Trans. electron Dev. ED-26, 647 (1979).
[3]. P. Singer, “Techniques of Low Pressure CVD”, Semiconductor Intl.129, 72, (1984).
[4]. F. C. Eversteyn, “chemical-Reaction Engineering in the Semiconductor Industry”, Philips Res. Rep., 29, 45, (1974).
[5]. J. Li, J. P. McVittie, J. Ferziger, K. C. Saraswat , and J. Dong,
“Optimization of Intermetal Dielectric Deposition Module Using Simulation”, J. Vac. Sci. Tech. B, 13, 1867, (1995).
[6]. M. Hammond, “Introduction to Chemical Vapor Deposition”, Solid State Technology, 61 (1979).
[7]. A. C. Adams, “Dielectric and Polysilicon Film Deposition,” in VLSI Technology (S.M. Sze, ed.), Chap. 3, p.93. McGraw-Hill, New York, (1983).
[8]. B. Mattson, “CVD Films for Interlayer Dielectrics”, Solid State Technology, p.60, (1980).
[9]. A. A. Chernov, “Growth Kinetics and Capture of Impurities during Gas Phase Crystallization”, J. cryst. Growth, 42, 55, (1977).
[10]. R. D. Gretz, and J. P. Hirth, “Nucleation and Growth Processes in CVD”,
- 111 -
in Proc. Conf. chemical Vapor Deposition of Refractory Metals, alloys and Compounds,73, (1967).
[11]. J. M. Blocher, Jr., “The Role of Intrinsic Impurities in CVD”, the Gordon Research Conf. on Thin Films, New Hampton, N. H., (1966).
[12]. S. M. Gates and S. K. Kalkarn, “Kinetics of Surface-Reactions in Very Low Pressure Chemical Vapor Deposition of Si from SiH4”, Appl. Phys. Lett., 2963, (1991).
[13]. P. Sutardja and W. G. Oldham, “Modeling of Stress Effects in Silicon Oxidation”, IEEE Trans. Elec. Dev., 36, 2415, (1989).
[14]. J. A. Thorton, “Influence of Apparatus Geometry and Deposition
Conditions on the Structure and Topology of thick Sputtered Films”, J. Vac.
Sci. tech., 11, 666, (1974).
[15]. S. Dunham, J. Appl. Physics, “Interaction of silicon Point Defects with Silicon Oxide Films”, 71, 685, (1992).
[16]. W. Shockley and J. Last, “Statistics of the Charge Distribution for a localized Flaw in a Semiconductor”, Phys. Rev.,107,392, (1957).
[17]. W. Kern, “Deposited Dielectrics for VLSI”, Semiconductor International 8, 122, (1985).
[18]. L. Holland, “The Properties of Glass Surface”, p.5, John Willey, New York, (1964).
[19]. T. I. Kamins, “Structure and Stability of Low Pressure chemically Vapor-Deposited silicon Films”, J. Electrochem. Soc. 125, 927-932, (1978).
[20]. A. C. Adams and C. D. Capio, “Planarization of Phosphorus-Doped silicon Dioxide”, J. Electrochem. Soc.,128, (1981).
[21]. K. Denbigh, “The Principles of Chemical Equilibrium”, 4th ed., Cambridge University Press, Cambridge, UK, (1983).
- 112 -
[22]. F. C. Eversteyn, “Chemical-Reaction Engineering in the Semiconductor Industry”, Philips Res. Rept. 29, (1974).
[23]. Richard S. Rosler, “Low Pressure CVD Production Processes for poly, Nitride, and Oxide”, Solid state Technology, pp.63-70, (1970).
[24]. A. C. Adams, “Dielectric and Polysilicon Film Deposition”, in S. M. Sze, Ed., VLSI Technology, 2nd ed., McGraw-Hill, New York, p.233, (1988).
[25]. A. Learn, “ Modeling the Reaction of low Pressure Chemical Vapor Deposition of Silicon Dioxide”, J. Electrochem. Soc. 132, 390, (1985).
[26]. B. Gorowitz, T. B. Gorczyca, R. J. Saia, “Applications of PECVD in VLSI”, Solid State Technology, p.197, (1985).
[27]. B. E. Deal, M. Sklar, A. S. Grove, and E. H. Snow, “Characteristics of the Surface State Charge of Thermal Oxidized Silicon”, J. Electrochem. Soc., 114, 266, (1967).
[28]. E. H. Snow, A. S. Grove, B. E. Deal, and C. T. Sah, “ Ion Transport Phenomena in Insulating Films”, J. Appl. Phys., 36, p.1664, (1965).
[29]. S. Veprek, “Plasma-Induced and Plasma –Assisted Chemical Vapor De- position”, Thin Sol.film, pp.130-135, (1985).
[30]. K. Fujino, Y. Nishimoto, N. Tokumasu, and K. Meada, “Doped Silicon Oxide Deposition by Atmospheric Pressure and low Temperature Chemical Vapor Deposition Using Tetraethoxysilane and Ozone”, J. Electrochem, Soc., 138, 3019, (1991).
[31]. N. Tsubouchi, “Application of the High-Pressure Oxidation Process to the Fabrication of MOS LSI”, IEEE Trans. Elec. dev., 618, (1979).
[32]. M. Finetti, P. Negrini, S. Solmi, and D. Nobili, “Electrical Properties and Stability of Supersaturated Phosphorus-Doped silicon Layers”, J. Electrochem.
Soc., 128, 131, (1981).
- 113 -
[33]. M. M. Mandurah, “Phosphorus Doping of Low Pressure Chemically Vapor-Deposited Silicon Films”, J. Electrochem. Soc.126, 1019-1023, (1979).
[34]. Tsai, Y.H.; Hsu, S.L.; Tseng, F.C.; Yoo, and, C.S, “Investigation on the flow properties of BPSG films”, VLSI Technology, Systems and Applications, (1989).
[35]. P. Lee, M. Galiano and P. Keswic, “Sub-atmospheric Chemical Vapor Deposition of TEOS-Ozone USG and BPSG” in IEEE VMIC Conf., (1990).
[36]. P. Lee, M. Galiano and P. Keswic, “Sub-atmospheric Chemical Vapor Deposition of TEOS-Ozone USG and BPSG” in IEEE VMIC Conf., (1990).
[37]. Kern and G. L. Schnable, “CVD BPSG for Si Devices”, RCA Review, 423, (1982).
[38]. S O Kang, H Y Seo and J H Cha, “Producer SABPSG Film Properties and Gap Fill Performance on DRAM ILD Process in Production Line”, ET Conf., (2001).
[39]. A. Bhatnagar, R. Pan, “Characterization of Gap-fill Capability of SACVD Ozone Doped Glass Films Annealed at Low Thermal Budget Conditions for 0.18 um Technology and Beyond” ET Conf., (2000).
[40]. Y. M. Kimo, G. Q. Lo, H. Konoshita, and D. L. Kwong, “Roles of Extended Defect Evolution on the Anomalous Diffusion of Boron in Si during Rapid Thermal Annealing”, J. Electrochem. Soc. 138, 1122, (1991).
[41]. R. B. Fair, “Junction formation in silicon by Rapid thermal annealing”, in R. B. Fair, Ed., Rapid Thermal Processing: Science and technology, Academic Press, Boston, (1993).
[42]. H. F. sterling and R. G. G. Swann, “chemical Vapor Deposition Promoted by R. F. Discharge”, solid-state electronics 8, 653-654, (1965).
- 114 -
[43]. W. L. Johnson, “Design of Plasma Deposition Reactors”, Sol. state Technology, p.191, (1983).
[44]. T. I. Kamins and K. L. Chiang, “Properties of Plasma-Enhance CVD silicon Films”, J. electrochem. soc. 129, 2326-2331, (1982).
[45]. E. Jerry, N. Tong, “Process and Film Characterization of PECVD Borophosphosilicate Films for VLSI Applications”, solid State Technology, pp.
161-170, (1984).
[46]. A. C. Adams, F. B. Alexander, C. D. Capio, and T. E. Smith,
“Characterization of Plasma Deposited Silicon Dioxide”, J. Electrochem. Soc.
128, 1545, (1981).
[47]. J. E. Tong, K. Schertenleib, and R. A. Carpio, “Process and Characterization of PECVD BPSG films for VLSI applications”, Solid State Technology 27, p.161, (1984).
[48]. G.Y. Lee, “A low Re-deposition Rate High Density plasma CVD Process for High Aspect Ratio T175nm Technology and Beyond”, International Interconnect technology Conference, 152-154, (1999).
[49]. D. R. Cote, “Plasma-Assisted Chemical Vapor Deposition if Dielectric Thin Films for ULSI Semiconductor Circuit”, IBM Journal of Research and Development, 43, 1-2, (1999).
[50]. Bikram Kapoor, Katsunari Ozeki, Anchuan Wang, Zia Karim,
“Innovative HDP CVD process for 65nmSTI applications”, Applied Materials ET Conference, (2003).
[51]. 51. Kirthik Janakiraman, Edwin Suarez, Kirby Floyd, Nitin Ingle, Steven Gianoulakis, “HARP Process kits: Enabling Hardware for Next Generation
SACVD Films”, Applied Materials ET Conference, (2001).
- 115 -
[52]. W. Bang, “HARP Enabling Hardware: T4 Liquid Flow Meter”, Applied Materials ET Conference, (2001).
[53]. K. Harvey, V. Abramson, “Chamber Clean Endpoint for SACVD Films Using NDIR Optical Absorption”, Applied Materials ET Conference, (2002).