MRS volume 697: Surface Engineering 2001 -Fundamental and Applications

 

Effect of tribological properties of undoped and florine-doped silicon di-oxide FIMS ON chemical mechanical planarization process

A.K. Sikder, S. Thagella., U.C. Bandugilla and  Ashok Kumar1

Center for Microelectronics Research

1Also with Department of Mechanical Engineering

University of South Florida, Tampa, FL 33620

 

Abstract

Chemical mechanical planarization (CMP) occurs at an atomic level at the slurry/wafer interface and hence slurries and the interaction of the films and polishing pads play a critical role in the successful implementation of this process. Understanding the tribological properties of a dielectric layer in the CMP process is critical for successful evaluation and implementation of the materials.  In this paper, we present the effect of tribological properties of undoped and florine doped silicon dioxide films on their CMP process. A micro-CMP tester was used to study the fundamental aspects of CMP process. We have studied the CMP process of oxides on polyurethane pads (IC1000-B4/SubaIV) with colloidal silica slurry at different conditions. The coefficient of friction (COF) and acoustic emission signal was monitored during process. The COF was measured during the process and was found to varies differently for different samples and with down force and platen rotation. The effects of machine’s parameters on the polishing performance and correlation of physical phenomena with the process has been discussed.

 

I. INTRODUCTION

         Continued miniaturization of the device dimensions and the related need to interconnect an increasing number of devices on a chip have led to building multilevel interconnection on planarized levels. In chemical mechanical planarization (CMP) very thin materials (≤5 μm) have to be removed very precisely while maintaining the precise control on the remaining thickness. CMP is a tribochemical process involving few basic components. Mechanical wear was accelerated by a chemical reaction, later induced by friction [1]. Planarization of the wafer results from the synergistic action of the mechanical shear forces and the chemical action of the slurry [2]. Many issues, e.g. polishing conditions, pad properties, slurry chemistry and hydrodynamics, and polishing head design, have been addressed by several researchers [3,4-6]. However, a poor understanding of complicated polishing phenomena makes it difficult to achieve local and global uniformity.

         Different CMP processes attempt to achieve a balance between removal rate and global/local planarization through a combination of solution chemistry, speed, applied pressure, and pad properties [7,8]. Often a change in slurry or operating conditions lead to conflicting performance.  There is a great need for a better understanding of all the complex tribological interactions between slurry, polishing pad, carrier film, wafer, polishing head, pad conditioner, etc. Fundamental tribological studies will allow to optimize the pad design, material selection, process pressure, orbital and linear speed, chemical solution, within wafer uniformity, and local planarization [9-16]. Solving these evolving challenges, while maintaining cost-effectiveness of the multi-faceted CMP process requires extensive experiments that involve a complex interplay between flow, thermal, and chemical processes, with particulate and mechanical contact and deformation phenomena.  It is important to understand the fundamental nature of surface planarization and identify the main mechanisms to control the CMP process.

         Although there is universal acceptance that device performance will improve by using lower-k (~ 2.5 or lower) dielectric films and Cu metallization, there are no of issues have to be solved in CMP point of view [17-19]. It is important to develop novel slurry/abrasive systems or to use chemically functionalized particle/slurry combinations to resolve these issues. It is necessary to develop a fundamental understanding of the relationship between the CMP process variables, slurry parameters and the characteristics of the low-k films.

         In this study, we discussed the tribological properties of undoped and florine doped silicon dioxide CMP process using CMP tester. It is often difficult to study in-situ the fundamental polishing properties in a real CMP polisher. Conventional studies mostly restricted themselves to the optimization of process parameters without going into the understanding of basic tribological properties. A CMP tester, used in this study, has several sensors (Force sensor, acoustic emission (AE) sensor and electrical sensor), which are very useful for the in-situ monitoring and optimizing the CMP process. Co-efficient of friction (COF) is measured using force sensor during polishing and it is found that COF has marked effects on the removal rate and acoustic emission. Correlation of mechanical properties along with the friction behavior of the films with the wear behavior has been discussed. Validity of Preston’s equation has also been discussed.

 

II. EXPERIMENTAL

The CMP process was performed on a CMP tester (CETR, Inc, CA) with variety of process parameters. Detalis of the CMP tester and its optimization has been discussed earlier [20]. Figure 1 shows a schematic of the CMP process. The lower platen can hold a pad up to 6” in diameter. The upper carriage can hold sample coupons up to 1.5” X 1.5”. The upper carriage can oscillate radially on the pad during planarization process. A strain gauge force sensor (0-100 N) can record both vertical and friction forces and hence co-efficient of friction (COF) is monitored during the process.  The system is also equipped with a high frequency acoustic emission sensor, which can detect the delamination, endpoint, and debris during polishing. Plasma enhanced chemical vapor deposited (PECVD) undoped silicon dioxide (U-SiO2) and high density plasma CVD florine doped SiO2 (SiOF) were polished with a wide range of process conditions using Klebesol 1501 colloidal silica slurry (pH ~10.5). Details of the samples are shown in Table 1. Mechanical properties of the films were measured using nanoindentation method. The CMP pads used in this study consisted of a dual layers (IC 1000-B4/Suba IV) (Rodel, Inc., Spcific Gravity 745). Pad was subjected to break-in for 20 min with flowing deionized water. CMP process conditions have been shown in Table 2.

 

Table 1. Details of the samples used for CMP studies.

 

Sample

Thickness (Ĺ)

Grown by

Hardness

(GPa)

Modulus

 (GPa)

Refractive Index

SiO2

3850

PECVD

6.3 ± 0.4

68.1 ±1.2

1.474

SiOF

1700

HDPCVD

4.3 ± 0.7

 

42.40±3.35

1.430

Fig. 1. Schematic of the CMP process

 

Table 2. Testing parameters and materials for measuring wear behavior of oxides

Normal Pressure

Variable (1-6 psi)

Platen Rotation

Variable (0.2--1.2 m/s or 42.2--269.8 RPM)

Slider Movement

45 mm with offset ± 5mm and velocity 10 mm/sec.

Slurry

Oxide slurry (Klebesol 1501), (100 ml/min.)

Pad

IC1000-B4/SubaIV

Time

20-80 Sec

Upper Specimen

Undoped and F-doped Silicon oxide

 

 

III. RESULTS AND DISCUSSION

           Typical polyurethane pads, either perforated or grooved IC 1000, consist of pores or voids of an average diameter of about 30 μm; voids account for approximately 30% of the volume of the pad. Fig. 2 (a) shows the cross section of a typical IC 1000/Suba IV pad and (b) shows the surface of an IC 1000 pad. The key element that affects polishing performance is pad surface roughness and this roughness is maintained with the   application of a conditioner, diamond grit embedded in a metal matrix. It can be seen that the pores are uniformly distributed throughout the pad, as pores can be seen even at the cross section.

        


Fig. 2 SEM micrographs of (a) cross section of IC 1000/Suba IV pad and (b) surface of IC 1000 pad

 
 
 

 

     (a)                                                             (b)

                                                                                          (c)                                                               (d)

 

Fig. 3.  Coefficient of friction with (a) RPM for U-SiO2 films,(b) with RPM for SiOF films, (c) with PSI for U-SiO2 films and (d) with PSI for SiOF films.

 

         A series of oxide samples (1” X 1”) were tested at different combinations of down force and platen speed while the slider was oscillating in the radial direction with a linear velocity of 10 mm/sec at a radial distance 45 mm ± 5 mm. COF is an important tribological property of films and pad, and was recorded during all the tests. Fig. 3 (a,b) and (c,d) show the COF vs. RMP and PSI during polishing of U-SiO2 and SiOF films respectively. COF was calculated with taking average of 10-20 sec. data during the polishing of total 20-80 sec. for different samples. With higher RPM, there is a slight increase of COF could be seen with increasing RPM for both U-SiO2 and SiOF films. With different PSI also, COF has a trend of a slight increase could be seen for both the films. Value of COF is slightly higher for the SiOF films. COF has marked effect on removal rate, local and global uniformity. Influence of COF in polishing performance has been discussed in the next section.

 

   

                                               (a)                                                                                    (b)

  

                                       (c)                                                                     (d)                  

 

Fig. 4.  Average removal rate with RPM and PSI for (a,b) for U-SiO2 and (c,d) for SiOF films.

 

         For removal rate calculations, thickness of the oxide layer was measured by 9-point thickness measurement after polishing using ellipsometer. Before polishing thickness of the films were also measured at several points. After subtracting all 9-points from the initial thickness, removal rate (RR) was then calculated by taking the average of all 9 points. Fig. 4 (a,b) and 4 (c,d) show the variation of  removal rate with the function of RPM and PSI for U-SiO­2 and SiOF films respectively. Removal rate increases with both RPM and PSI for both the samples. Removal rate decreases slightly at platen rotation 250 RPM for U-SiO­2, which has been noticed earlier also [20]. If this is due to inadequate slurry-film interactions during higher platen rotation, similar effect could have been seen for SiOF films also. This may be because of hydroplaning effect at higher rotation. It can be noticed from Fig. 4 that overall RR is higher for U-SiO2 than the SiOF films. There is no significant change in COF for these two films. Again nanoindentation results showed significant change in their hardness and modulus values (Table 1). Now there should be higher RR for softer films for a mechanical polishing which might not be true for chemical mechanical process. Higher RR for U-SiO2 films is may be due to higher chemical interaction of slurry with the film’s surface. Validity of Preston’s equation has also been tested. Fig. 5 (a) and (b) show the average RR vs. RPM*PSI for two kind of films and the linear relation indicates that polishing of these films follows Preston’s equation [19]. It can be seen that data are more scattered for U-SiO2 films than SiOF films. This may be because of higher mechanical polishing for SiOF films than chemical polishing. Fig. 5 (b) and (d) the variation of average RR with respect to RPM*PSI*COF for the two films respectively. Interestingly data is slightly more scattered for U-SiO2 films, whereas, data are more closer to the average line for SiOF films. More experiments are underway for understanding the effect of tribological properties of the films and pads on the CMP process.   

 

            

                                                (a)                                                                                                 (b)                   

            

                                             (c)                                                                                                   (d)               


   

Fig.5. Average removal rates plotted with RPM*PSI. Linear relation indicates that polishing follows Preston’s equation

           

              

IV. CONCLUSIONS

 

            The following conclusions can be made from this study:

·       Polyurethane pads consist of pores, or voids. These pores or voids are distributed uniformly in the bulk

·       Florine doped silicon oxide has lower hardness and modulus than undoped oxide.

·       Increasing trend of COF have been found for both the films with higher RPM and PSI

·       COF is an important parameter for the polishing process

·       Removal rate increases with platen velocity and down pressure.

·       Removal rate decreases slightly at platen rotation 250 RPM for the U-SiO2 films. This may be due to hydroplaning

·       Removal rate follows Preston’s equation.

 

 

V. ACKNOWLEDGEMENTS

          Part of the research was supported by SEMATECH grant # 2112-139LO. One of the authors (Ashok Kumar) would like to acknowledge NSF CAREER Grant No. 9983535 for support of this research.

 

      REFERENCES

 

  1. T. Fischer, Ann. Rev. Mater. Sci. 18 303 (1988).
  2. J.M.Steigerwald, S.P. Murarka, R.J. Gutmann, Chemical Mechanical Planarization of Microelectronic Materials, Wiley-Interscience, New York, 1997.
  3. Dipto G. Thakurta, Christopher L.Borst, Donald W.Schwendeman, Ronald J. Gutmann, William N.Gill, Thin Solid Films, 366, 181(2000).
  4. D.Wang, J.Lee, K. Holland, T. Bibby, T.Cale, J.Electrochem.Soc.,141,2843(1997).
  5. O.G. Chekina, L.M. Keer, H. Liang. J.Electrochem.Soc,145, 2100(1998).
  6. S. Sundararajan, D.G. Thakurta, D.W. Schwendeman, S.P. Murarka, W.N. Gill, J. Electrochem.Soc. 146, 761(1999).
  7. A.E. Braun,  Semiconductor International, 65-74, (1998).

8.     Timothy Marbeiter, Timothy Cleary, and Karen Sutter, Semiconductor International, 76-80 (1998).

9.     M. Bhushan, R. Rouse, and J. E. Lukens, Journal of Electrochemical Society, 142,  3845 (1995).

10.  R. Jairath et. al.,  Solid State Technology, pp. 71 (1994).

11.  R. DeJule, Semiconductor International, pp. 54-60 (1997).

12.  S. R. Runnels, L. M. Eyman, “Tribology Analysis of Chemical Mechanical Polishing,” Journal of Electrochemical Society, 141, 1900 (1994).

13.  James Reynolds, A. L. Swecker, Andrzej J. Strojwas, Ady Levy, and Bobby Bell, Solid State Technology, 39-44 (1998).

14.  S. Runnels, Journal of Electronic Materials, 25, 1574, (1996).

15.  D. P. Thakurta, C. L. Borst, D. W. Schwendman, R. J. Gutman, and W. N. Grill, Thin Solid Films, 336, 181, (2000).

16.  A.K. Sikder, I. M. Irfan, Ashok Kumar, A. Belyaev, S. Ostapenko M. Calves, J. P. Harmon and J. M. Anthony, Mat. Res. Soc. Symp. Proc., 671, M1.8.1-7 (2001).

17.  P. Wrschka, J. Hernandez, G. S. Oehrlein and J. King, J. Electrochem. Soc., 147, 706 (2000).

18.  T. Park, T. Tugbawa and D. Boning, Proc. CMP-MIC,  Santa Clara, pp.196 (2000).

19.  Y. Morand, Microelectronic Engineering, 50, 391 (2000).

20.  A.K. Sikder, Frank Giglio, John Wood, Ashok kumar and J.M. Anthony, J. Elec. Mater., 30(12) 2001.

21.  F. Preston, J. Soc. Glass Technol., 11, 214 (1927).