Copper electropolishing in phosphoric acid has been characterized using electroanalytical methods, primarily potential transient techniques. An uncommon voltage response, consisting of two distinct steps, was noted when the current was stepped to the limiting current. A slow ( ∼100–300s , depending on agitation) and relatively small (∼50mV) initial potential increase was followed by a fast (∼5s) and large (∼1.5V) potential rise. The latter always reached the oxygen evolution potential (∼1.6V) , irrespective of the process conditions. While the first, slow potential transient can be correlated in terms of a diffusion process, the second, rapid potential rise suggests the buildup of a highly resistive component, most likely a surface film. The nearly instantaneous potential relaxation upon current interruption further supports the resistive film model rather than a transport-related process. A two-stage mechanism is proposed and analytically modeled. Accordingly, the cupric ion concentration at the anode increases during the initial stage of the dissolution process due to transport limitations, until a saturation level is reached and a resistive surface film forms. During the second stage, the continuing imbalance between the rate of cupric ion formation and transport into the bulk leads to increasing film thickness and, consequently, to a rapid buildup in resistance. The model quantitatively correlates the experimentally measured transients and is consistent with all other observations relating to the copper electropolishing process.
Recent studies of Cu electropolish indicate planarization is not possible due to a large diffusion layer thickness relative to a small post-electroplate step height and small Cu overburden available for electropolish. Assuming an integration scheme that includes a CMP step followed by electropolish, the subsequent challenge of maintaining surface roughness over 300 m range while electropolishing thin Cu films is addressed by optimizing applied current. For a typical electropolish solution of 6.4 M H 3 PO 4 , 5.4 M glycerin and 17.5 M H 2 O, an rms value of 38 Å, comparable to the incoming post-CMP wafers, is demonstrated for 2500 Å Cu removal.12 Potential application of electropolish as a chemical mechanical polish ͑CMP͒ replacement for Cu/low-k interlevel dielectric ͑ILD͒ integration has been discussed previously. 1-4 However, it has not been demonstrated yet, because of low planarization efficiency and pattern sensitivity of the electropolish process. Planarization of the small aspect ratio features by conventional electropolishing is fundamentally limited by the minimum mass transport boundary layer ͑BL͒ thickness. 5,6 For planarization to occur, the step height needs to be on the order of the BL thickness. According to Suni 5 and West, 6 the typical BL thickness in phosphoric acid based electrolytes is in the range of 5-30 m, which is approximately two orders of magnitude greater than the post-electroplate topographical step height. This BL thickness depends on physical parameters such as fluid viscosity and wafer rotation rate and cannot be easily reduced to an appropriate thickness to enable planarization.Use of a CMP step prior to electropolish for initial topography planarization and surface smoothing has been considered. 7 In this scenario, CMP is not likely to damage the low-k dielectric, because of the protective action of a remaining thin blanket layer of Cu. 5 Considering a realistic Cu thickness of 5000 Å and assuming CMP processes consume half to achieve planarity targets, electropolish conditions must be established and surface roughness targets met in a budget of only 2500 Å.It is generally accepted that electropolishing occurs when metal dissolution is controlled by the mass transfer of the limiting species or products to or from the electrode surface. The process is independent of crystallographic orientation. 8 Copper electropolishing in phosphoric acid has been extensively studied over the years. A number of theories have been proposed to explain mass-transfer limitations, including salt-film mechanism 9 and the water acceptor mechanism 10 clearly favored by the most publications. 4,11 Regardless, a key factor of detailed mechanisms in establishing electropolish conditions is the time to reach the mass-transfer limitation and amount of copper dissolved during the transient time.More recently, Padhi et al. 4 studied anodic galvanostatic transients and demonstrated that time to form a resistive BL is inversely proportional to the square of the applied current density. However, their optimum con...
This paper demonstrates the application of a modified Levich equation for chemical systems with varying viscosity. A commonly used technique to analyze rotating disc electrode (RDE) experiments is to fit the data to the Levich equation assuming a constant effective diffusion coefficient which may be valid for conditions where the viscosity does not vary significantly (less than an order of magnitude). However, most diffusion coefficient models (e.g. Stokes-Einstein) show an inverse relationship with viscosity which consequently indicates that a constant effective diffusion coefficient may result in poorer model-to-data agreement. Here, data are presented for a series of RDE experiments for the electrodissolution of Cu in phosphoric acid, water and glycerin based baths. Viscosity changes of greater than one order of magnitude allow for testing the assumption of a constant effective diffusion coefficient. The collected data, as well as data published elsewhere, can be explained by a modified Levich equation which takes into account the viscosity dependence of the diffusion coefficient. List of SymbolsC A Concentration of A in solution (mol l -1 ) D Diffusion coefficient (m 2 s -1 ) D A Effective diffusion coefficient of A (m 2 s -1 ) D AB Mutual diffusivity at infinite dilution of A in B (m 2 s -1 ) D AB Mutual diffusivity (m 2 s -1 ) F Faraday's constant (C mol -1 ) I lim Limiting current per unit area (A m -2 ) k Boltzman's constant (J K -1 ) m Molality of solute (mol (kg of solvent) -1 ) n Ionic charge r Effective radius (m) s Solvent coordination number T Absolute temperature (K) V Molar volume (m 3 mol -1 ) c ± Mean ionic activity coefficient of solute l Absolute viscosity (cP) m Kinematic viscosity (m 2 s -1 ) u B Association factor for solvent B for Wilke-Chang equation w B Parachor parameter for component B for Tyn-Calus equation xRotational speed (rad s -1 )
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.