CURRENT RESEARCH PROJECT: Chemical
Mechanical Planarization (CMP) and Role of "planarization" in semiconductor device fabrication: The performance of a miniaturized semiconductor device is governed by its signal processing speed, which in turn is determined by the gate and interconnect delay times. For devise features scaled down below 0.5 micron, the interconnect "RC delay" dominates the delay in signal processing, where R is the resistance of the wiring metal and C is the capacitance of the interlayer dielectric (ILD) used in the device. From simple considerations, it can be shown that [1,2] RC = [(rkl2)/(yd)]. Here r, l and d represent the resistivity, length and thickness of the wiring line, respectively; k and y represent the dielectric function and thickness of the ILD, respectively. Thus, decreasing the value of l and/or that of k can decrease the delay time. Decreasing the value of l is generally accomplished through the technique of multilevel metallization (MLM). In MLM, however, uneven topography is introduced as different levels are deposited, and such non-planar topography can be a serious problem (due to depth of focus limitations at short wavelength exposures) in the lithography used for integrated circuit fabrication. To remedy this problem, some type of surface planarization becomes necessary at each level of metallization in the MLM scheme. . Chemical mechanical planarization (CMP): CMP is currently the most commonly used planarization technique [1-3]. As its name indicates, CMP combines chemical surface reactions with mechanical planarization. CMP of metals includes one or all of the following chemical steps: (i) Chemical dissolution of the surface layer(s); (ii) Oxidation of the surface, with subsequent mechanical abrasion of the porous (mechanically unstable) oxide layers; (ii) Formation of soluble surface complexes that can be dissolved in the polishing solution, or can be removed with minimal mechanical abrasion. To support these specific functions, the CMP slurries usually contain a pH-adjusted aqueous background solution, an oxidizer, and a complexing agent and/or a "corrosion inhibitor". Slurry-stabilizers like surfactants are also used to prevent coagulation of abrasive particles in the electrolyte, which helps to minimize "scratches" and other defects caused by large particle aggregates. Mechanical performance of CMP is largely determined by the polishing parameters (down-pressure, rotation speed), pads and abrasives used. Chemical performance of CMP is governed by selective and collective reactions of different chemical ingredients of the slurry with the sample surface, as well as by interactions of abrasive particles with the sample. Our CMP research focuses primarily on the chemical surface reactions that control the chemical efficiency of CMP (material removal rates, uniformity across the wafer and defect-free quality of the polished surface). We use both D.C. and A.C. electrochemical techniques to study these surface reactions in CMP of various materials. Currently, our work in this area is centered on CMP of metals. Illustrative results of our work involving CMP of Cu, Ta and Ag have been published recently [see publications from our laboratory on CMP and ECMP listed below]. Our electrochemical CMP studies are designed to understand the origins of the necessary surface reactions for CMP, and to eventually develop chemically efficient abrasives as well as abrasive-free solutions for planarization of certain metals that are commonly used (or are potentially important for future applications) in the fabrication of integrated circuit microchips. Our general strategy for this research is schematically outlined in the diagram below. Electrochemical investigation of slurry chemistry in CMP of metals:
Planarization criteria for structures containing low-k materials: As noted above, decreasing the value of k is important for achieving fast processing time. Integration of low-k ILD can also reduce cross talks and power dissipation in the device. Most of the such low-k ILD materials (k < 2.5), however, are often porous and mechanically fragile. Therefore, to avoid damages to the ILD during IC fabrication, planarization of the ILD materials and their overlying structures must be performed at a low applied down-pressure (usually at < 1 psi). Electrochemical mechanical planarization (ECMP): The introduction of porous low-k materials in microchip devices has made "low down-pressure" (low-P) planarization an extremely important factor. Although low-P operation is difficult to incorporate in the currently available framework of CMP, it is possible to combine electrochemically controlled material removal with low-P mechanical polishing where the main role of the latter step is to provide uniform planarization across the sample surface (that is, to facilitate selective material removal from protrusions over recess regions of the surface). This approach [4-7], introduced by Applied Materials for industrial applications [6], is referred to as electrochemical mechanical planarization (ECMP), and can potentially lead to a more efficient planarization technology than the currently practiced chemical mechanical planarization (CMP). In most applications of ECMP, the mechanical abrasion step involves removal of thin passive films of oxides and/or surface complexes (not of the underlying bulk metal), and hence can be performed at a low down pressure (<1 psi). Eectrochemical techniques are often used only as a "probe" of CMP mechanisms to analyze corrosion/erosion behaviors of various CMP systems, but these techniques are not frequently applied to the actual CMP process. In ECMP, electrochemical techniques can be used to both activate and understand the mechanism(s) of material removal. In addition to its low-P processing capability, another major feature of ECMP is that it can be performed using electrolytes with no or very low concentrations of abrasive particles. This helps to eliminate several disadvantages of CMP that are associated with the use of slurries containing high concentrations of abrasive particles, such as lack of within-wafer uniformity, particle coagulation, slurry-handling and waste disposal. The task of endpoint detection is relatively straightforward in ECMP where simply controlling the applied voltage or current can accurately control the extent of planarization. Often it might also be possible to eliminate the need for certain expensive, unstable and/or side-reacting chemicals (oxidizers, surfactants, etc.) in ECMP. Our current work in the field of ECMP focuses on cetain fundamental aspects of ECMP of a number of metals (and low-k dielectrics). The electrolytes used for metal ECMP in our laboratory contain different combinations of nonspecifically adsorbing anions and complexing agents both with and without oxidizers. The activation voltages for material removal are designed as trains of repeated anodic voltage pulses of rectangular, triangular or staircase-type shapes. The voltage programs are designed to activate (system specific) anodic reactions that lead to direct metal dissolution and/or electrochemical generation of soluble or structurally unstable "soft" surface films. These films can be removed through low-P (< 1 psi) polishing. The integrated charge of the resulting current is used as a measure of electrochemically induced material removal. This quantity is also calibrated in terms of the thickness of the processed surface layer. The electrochemical control variables are varied to obtain optimized removal rates and defect-free finished surfaces. The processed surface morphology is examined using AFM and SEM. Preliminary studies of ECMP of Ag and Cu have been recently reported from our laboratory [see below publications by our group]. Both our CMP and ECMP works are presently done in collaboration with Professor S.V. Babu's research group at the Center for Advanced Materials Processing of Clarkson University. References [1] P. B. Zantye, A. Kumar, A.K. Sikdar, Mater. Sci. Eng. R45 (2004)
89. Recent publications on CMP and ECMP from our laboratory 1. K. A. Assiongbon, S. B. Emery, C. M. Pettit, S. V. Babu and D. Roy,
2. J. Lu, J. E. Garland,
C. M. Pettit, S. V. Babu, and D. Roy, 3. S. B. Emery, J.
L. Hubbley, M. A. Darling and D. Roy, 4. V.R.K. Gorantla, S. B. Emery, S. Pandija, S.V. Babu, D. Roy, 5. K. A. Assiongbon, S. B. Emery, V.R.K. Gorantla, S.V. Babu, D. Roy,
6. V. R. K. Gorantla, K. A. Assiongbon, S. V. Babu, D. Roy, 7. P. C. Goonetilleke, D. Roy, 8.P.C. Goonetillake, S.V. Babu and D.Roy, Voltage initiated material removal for electrochemical mechanical planarization of copper in NO 3 - , glycine and H 2 O2 containing electrolytes: Electrochem. And Solid State Lett. 8, (2005) G190-193. 9. Y. Hong, D. Roy and S. V. Babu, Ammonium Dodecyl Sulfate as a Potential Corrosion Inhibitor Surfactant for Electrochemical Mechanical Planarization of Copper, Electrochemistry and Solid-State Letters 8, (2005) G297-G300. 10. C.M. Pettit and D.
Roy, Role of iodate ions in chemical mechanical and electrochemical
mechanical planarization of Ta investigated using time-resolved impedance
spectroscopy, Materials Letters 59, 11. Y. Hong, U. B. Patri, S. Ramakrishnan, D. Roy and S.V. Babu, 12. C.M. Pettit, P.C. Goonetilleke and D. Roy, "Measurement of differential capacitance for faradaic systems under potentiodynamic conditions: Considerations of Fourier transform and phase-selective techniques" Journal of Electroanalytical Chemistry 589 (2006) 219-231 13. K.A. Assiongbon, S.B. Emery, V.R.K. Gorantla, S.V. Babu and D. Roy,
"Electrochemical impedance characteristics of Ta/Cu contact regions in
polishing slurries used for chemical mechanical planarization of Ta and
Cu: considerations of galvanic corrosion", Corrosion Science, 48,
(2006) 372-388 14. P.C. Goonetilleke and D. Roy, 15. Y. Hong, V. K. Devarapalli, D. Roy, and S. V. Babu, 16. S. Pandija, D. Roy and S.V. Babu, 17. S. Ramakrishnan, S.V.S.B. Janjam, U.B. Patri, D. Roy, and S.V. Babu,
18. S. V. S. B. Janjam, C. V. V. S. Surisetty, S. Pandija, D. Roy, and S. V. Babu, "Oxalic-Acid-Based Slurries with Tunable Selectivity for Copper and Tantalum Removal in CMP", Electrochemistry and Solid-State Letters 11 (2008) H66-H69. 19. C. V. V. S. Surisetty, P. C. Goonetilleke, D. Roy, and S. V. Babu, 21. S. V. S. B. Janjam, C. V. V. S. Surisetty, S. Pandija, D. Roy, and S. V. Babu, 22. J. P. Zheng, B. K. Klug, and D. Roy, 23. P.C. Goonetilleke and D. Roy, |