A plasticity-based model of material removal in chemical-mechanicalpolishing (CMP)

It is well known that the chemical reaction between an oxide layer and a water-based slurry produces a softer hydroxylated interface layer. During chemical-mechanical polishing (CMP), it is assumed that material removal occurs by the plastic deformation of this interface layer. In this paper, the behavior of the hydroxylated layer is modeled as a perfectly plastic, material, and a mechanistic model for material removal rate (MRR) in CMP is developed. The deformation profile of the soft pad is approximated as the bending of a thin elastic beam. In addition to the dependence of MRR on pressure and relative velocity, the proposed plasticity-based model is also capable of delineating the effects of pad and slurry properties. The plasticity-based model is utilized to explore the effects of various design parameters (e.g., abrasive shape, size and concentration, and pad stiffness) on the MRR. Model predictions are compared with existing experimental observations from glass polishing, lapping, and CMP     

A model for wafer scale variation of material removal rate in chemical mechanical polishing based on viscoelastic pad deformation

It is well known that within-wafer nonuniformity (WIWNU) due to the variation in material removal rate (MRR) in chemical mechanical polishing (CMP) significantly affects the yield of good dies. The process control for a batch CMP operation is further complicated by wafer-to-wafer nonuniformity (WTWNU) caused by MRR decay when a number of wafers are polished with the same unconditioned pad. Accordingly, the present work focuses on modeling the WIWNU and WTWNU in CMP processes. Various material removal models suggest that the MRR is strongly influenced by the interface pressure. It is also well known that the viscoelastic properties of the pad play an important role in CMP. In the present work, an analytical expression for pressure distribution (and its associated MRR) at the wafer-pad interface for a viscoelastic pad is developed. It is observed that under constant load, which is typical during main polishing in CMP, the spatial distribution of the interface pressure profile may change with time from edge-slow to edge-fast, depending on the combination of wafer curvature, down pressure, and pad properties. For constant displacement operations, the pressure profile retains its edge-slow or edge-fast characteristics over time. The analytical model predictions of MRR based on viscoelastic pad properties also correlate very well to existing experimental observations of MRR decay when an unconditioned pad is used to polish a number of wafers. Based on these observations, it may be conjectured that the viscoelastic material properties of the pad play a primary role in causing the observed MRR decay. The analytical results obtained in the present work can also provide an estimation of evolution of thickness removal distribution over the entire wafer. This may be used for determining the optimum thickness of the overburden material and its polishing time, and for effective control of CMP processes.     

Material removal mechanism in chemical mechanical polishing: theoryand modeling

The abrasion mechanism in solid-solid contact mode of the chemical mechanical polishing (CMP) process is investigated in detail. Based on assumptions of plastic contact over wafer-abrasive and pad-abrasive interfaces, the normal distribution of abrasive size and an assumed periodic roughness of pad surface, a novel model is developed for material removal in CMP. The basic model is MRR=ρ_wNVol _removed, where ρ_w is the density of wafer N the number of active abrasives, and Vol_removed the volume of material removed by a single abrasive. The model proposed integrates process parameters including pressure and velocity and other important input parameters including the wafer hardness, pad hardness, pad roughness, abrasive size, and abrasive geometry into the same formulation to predict the material removal rate (MRR). An interface between the chemical effect and mechanical effect has been constructed through a fitting parameter H_w a “dynamical” hardness value of the wafer surface, in the model. It reflects the influences of chemicals on the mechanical material removal. The fluid effect in the current model is attributed to the number of active abrasives. It is found that the nonlinear down pressure dependence of material removal rate is related to a probability density function of the abrasive size and the elastic deformation of the pad. Compared with experimental results, the model accurately predicts MRR. With further verification of the model, a better understanding of the fundamental mechanism involved in material removal in the CMP process, particularly different roles played by the consumables and their interactions, can be obtained     

Inverse analysis of material removal data using a multiscale CMP model

This paper describes a mechanical model for a representative dual axis rotational chemical mechanical planarization (CMP) tool. The model is three-dimensional, multiscale and includes sub-models for bulk pad deformation, asperity deformation, lubrication based slurry flow, carrier film deformation, wafer compliance and material removal by abrasive particles in the slurry. With the model, material removal rate (MRR) can be determined as a function of stress applied to the wafer, relative sliding speed, and material and geometric parameters of the pad and slurry. Experimental material removal rate profiles obtained from Cu polishing experiments performed on a wafer without rotation are analyzed as an inverse problem. We use MRR data to predict local CMP conditions such as fluid film thickness, fluid pressure and contact pressure. The results are consistent with available experimental and analytical information. This inverse technique offers promise as an improved method of CMP model verification.     

A model for wafer scale variation of removal rate in chemical mechanical polishing based on elastic pad deformation

It is well known that within wafer non-uniformity (WIWNU), due to the variation in material, removal rate (MRR) in the whole wafer plays an important role in determining the quality of a wafer planarized by CMP. Various material removal models also suggest that the MRR is strongly influenced by the interface pressure. In the present work, an analytical expression for pressure distribution at the wafer and pad interface is developed. It is observed that depending on the wafer curvature and polishing conditions, the interface pressure may exhibit significant variation. The analytical model predictions are first verified against finite element method (FEM) simulations. The predicted analytical pressure profiles are then utilized in Preston's equation to estimate the MRR, and these MRR predictions are also compared to experimental observations. The analytical results suggest, that for a specified wafer curvature there exists a certain polishing condition (and vice versa) that will enable holding the WIWNU within a specified tolerance band. The proposed model facilitates the design space exploration for such optimal polishing conditions.     

Current-carrying capacity of anisotropic-conductive film joints for the flip chip on flex applications

The effect of the substrate-pad physical properties (surface roughness and hardness) on the current-carrying capacity of anisotropic-conductive film (ACF) joints is investigated in this work. Flip chips with Au bumps were bonded to the flexible substrates with Au/Cu and Au/Ni/Cu pads using different bonding pressure. It was found that the current-carrying capacity of ACF joints increased to a maximum value with the rise of the bonding pressure; then, it reduced if the bonding pressure continually increased. The maximum average value per unit area of Au/Ni/Cu pad and Au/Cu pad ACF joints is about 93 μA/μm² and 118 μA/μm², respectively, at 100-MPa bonding pressure. The variation trend of connection resistance is the opposite of current-carrying capacity. The variation of current-carrying capacity (or connection resistance) of Au/Cu pad joints is larger than that of Au/Ni/Cu pad joints. The current-carrying capacity is related to the variation of the resistance of ACF joints. The connection resistance of ACF joints depends primarily on the particle constriction resistance (R_coi), R_coi ∞ 1/a, where “a” is the radius of contact spot. A smaller contact area results in larger joule heat generation per unit volume (Q_g), Q_g ∞ 1/a⁴, which preferentially elevates the temperature of the constriction. The raised temperature increases the resistance because of the temperature-dependent coefficient of the metal resistivity. The theory of tribology is used to explain the difference between Au/Cu pad and Au/Ni/Cu pad ACF joints. For the Au/Cu pad ACF joints, the deformation of the particles’ upper and bottom sides is nearly symmetrical; the contact between conductive particles and pad has the character of “sliding contact,” especially under high pressure. For the Au/Ni/Cu pad ACF joint, the contact between particles and pad determined the conduction characteristics of ACF joints. It has the character of “static contact.” Thus, the current-carrying capacity (or connection resistance) of Au/Cu pad joints is more sensitive to the bonding pressure.     

Effect of polishing pad with holes in electro-chemical mechanical planarization

Electro-chemical mechanical planarization (ECMP) process dissolves copper ions electrochemically by applying an anodic potential on the copper surface in an aqueous electrolyte, and then removes a copper (Cu) complex layer by the mechanical abrasion of a polishing pad or abrasives in the electrolyte. The ECMP process is a low pressure polishing method for metals such as copper, aluminium (Al) and tungsten (W) on dielectric materials such as silicon dioxide, low-k (LK) and ultra low-k (ULK) dielectrics, comparing to the amount of defects by the traditional Cu chemical mechanical planarization (CMP). The polishing pad used in the ECMP process is a conventional closed cell type pad (IC 1400 K-groove pad) with holes. It supplies the aqueous electrolyte to the copper surface and removes the copper complex layer. The material removal rate (MRR) and MRR profile were simulated and tested according to the changes of the wafer overhang distance (WOD) from the platen and the electric contact area (ECA). In order to derive the design rule of the system, the experimental results are compared with the simulation results. After the ECMP process, it was verified that the within wafer non-uniformity (WIWNU) was lower than 2% using the relatively uniform ECA pad (C-type) under the smallest WOD condition. The experimental results well matched the simulated results.     

Effect of abrasive particle concentration on preliminary chemical mechanical polishing of glass substrate

Effect of abrasive particle concentration on material removal rate (MRR), MRR per particle and the surface quality in the preliminary chemical mechanical polishing (CMP) of rough glass substrate was investigated. Experimental results showed that the MRR increases linearly with the increase of abrasive concentration and reaches to the maximum when the abrasive concentration is 20 wt.%, and then tends to be stable. When the abrasive concentration increases from 2 to 5 wt.%, the MRR per particle increases greatly and reaches a peak. Then the MRR per particle decreases almost linearly with the increase of the abrasive concentration. The root mean squares (RMS) roughness almost decreases with increasing particle concentration. In addition, in situ coefficient of friction (COF) was also conducted during the polishing process and the zeta potentials of abrasive particles in slurry with different solid concentration were also characterized. Results show that COF value is not related to zeta potential but be sensitive to glass surface conditions in terms of rough peaks in preliminary polishing of glass substrate.     

Study of Intermetallic Growth and Kinetics in Fine-Pitch Lead-Free Solder Bumps for Next-Generation Flip-Chip Assemblies

With continued advances in microelectronics, it is anticipated that next-generation microelectronic assemblies will require a reduction of the flip-chip solder bump pitch to 100 μm or less from the current industrial practice of 130 μm to 150 μm. With this reduction in pitch size, and thus in bump height and diameter, the interaction between die pad metallurgy and substrate pad metallurgy becomes more critical due to the shorter diffusion path and greater stress. Existing literature has not addressed such metallurgical interaction in actual fine-pitch flip-chip assemblies. This work studies intermetallic growth and kinetics in fine-pitch lead-free solder bumps through thermal aging of flip-chip assemblies. Based on this study, it is seen that Ni from the die pad diffuses to the substrate pad region and Cu from the substrate pad diffuses to the die pad region, thus the resulting intermetallic compounds at the die and substrate pad regions are influenced by the other pad as well. Such cross-pad interaction is much stronger in fine-pitch solder bumps with smaller standoff height. It is seen that the die pad region contains Ni₃P and (Cu,Ni)₆Sn₅ after thermal aging, while the substrate pad region contains Cu₃Sn and (Cu,Ni)₆Sn₅. By digitally measuring the thickness of the interfacial phases, the kinetics parameters and the activation energy were calculated for the growth of (Cu,Ni)₆Sn₅ on the substrate side. The Cu diffusion coefficient through the intermetallic compound (IMC) layer was found to be 0.03370 μm²/h, 0.1423 μm²/h, and 0.4463 μm²/h at 100°C, 125°C, and 150°C, respectively, and the apparent activation energy for the growth of compound layers was 67.89 kJ/mol.     

固结磨料抛光K9光学玻璃的工艺实验研究

采用一种亲水性固结磨料抛光垫(FAP),通过单因素实验法,系统地研究了抛光K9光学玻璃过程中抛光时间、偏心距、压力、转速、抛光液流量及pH值等工艺参数对材料去除速率(MRR)和表面粗糙度的影响规律,并对实验结果进行了解释。结果表明:随着抛光时间的延长,K9光学玻璃的MRR逐渐呈下降趋势;在抛光20min时,MRR达最大值310nm/min,且表面粗糙度降至最低值为2.73nm;选择较大的偏心距和碱性抛光液环境均有利于提高MRR;随着抛光盘转速的升高,MRR将显著增大。而在一定范围内,抛光压力和抛光液流量对MRR的影响不大。     

Failure analysis of pad-height effects in the fine-pitch interconnection of the anisotropic conductive films

This study analyzes the effect of the upper-to-lower pad-height ratio on the global failure probability of IC/substrate assemblies packaged using Anisotropic Conductive Film (ACF). In modeling the failure of the IC/substrate package, the probability of an opening failure in the vertical gap between the pads is calculated using a Poisson function, while the probability of a bridging failure between the pads in the pitch direction is computed using a modified box model. The opening and bridging probabilities are then combined using probability theory to establish an overall failure prediction model for the IC/substrate assembly. The results show that the failure probability increases as the sum of the lower pad height and upper pad height increases, or as the ratio R_h of the upper pad height to the lower pad height increases. Furthermore, for a given gap size between the IC device and the substrate, the minimum failure probability is obtained when the ratio of the upper pad height to the lower pad height has a value of R_h = 1. Overall, the results suggest that the reliability of ACF-packaged IC/substrate assemblies can be improved by reducing the total height of the two pad arrays or by utilizing pad arrays with an equivalent height.     

Mathematical modeling of CMP conditioning process

Up to now, the conditioning model with an oscillating conditioner wheel has not been studied. In this paper, kinematic analysis of the conditioning process and mathematical modeling of pad wear while the conditioner wheel oscillates is studied and the results show how the various parameters of the conditioning process influence the pad shape. The conditioning of the polishing pad is one of the most important processes associated with the CMP (Chemical Mechanical Polishing). As the wafer is polished, the surface of the pad can be deteriorated with a reduced polishing rate and reduced planarity due to wear and glazing of the pad. Thus, the polishing pad needs to be conditioned to maintain its effectiveness. In general, the conditioning process is used to regenerate the pad surface by breaking the glazed area of the pad and increase the MRR (Material Removal Rate) and give us longer pad life. However, as the conditioning process continues, the pad shape becomes more and more concave over the whole pad while the conditioner wheel oscillates (Y.Y. Zhou, E.C. Davis, Mat. Sci. Eng. B. 68 (1999), 91-98). It has been shown that the concavity of the polishing pad increases with conditioning time - longer conditioning induces a higher incidence of concavity of the polishing pad. Therefore, the conditioning process is related to the WIWNU (Within Wafer Non-Uniformity). Through this conditioning model, thickness variation of the polishing pad can be predicted.     

Size effect of nanoparticles in chemical mechanical polishing - a transient model

When a workpiece to be polished is placed on the carrier of a polishing machine, it is pressed down to the polishing pad. Large abrasives make contact between the pad and the workpiece before the smaller ones. The larger abrasives are pressed into the pad and indented into the workpiece. These particles are the active abrasives and participate in material removal. The abrasives with a size less than the gap between the pad and the workpiece move freely in the valleys/voids of the pad and are inactive. As the gap decreases during the polishing process, smaller abrasives trapped between the pad and the workpiece become active in polishing. Thus, the process of chemical-mechanical polishing is dynamic, while all previous modeling is static. This paper establishes a dynamic model for the abrasives. The modeling considers the transient motion of the workpiece/particle/pad in the vertical direction and the change of the roughness of the workpiece. A study of the transient motion shows an increasing number of active particles and a changing polishing rate in the first 2 min. It also demonstrates that the viscoelastic properties of the pad and the workpiece surface roughness are important factors in determining the polishing rate. This paper also shows that when the average particle size is smaller than an optimum size, the polishing rate increases with increasing particle size for the same particle density or same wt% abrasives. Yet, if the average particle size is larger than the optimum size, the polishing rate decreases with increasing particle size.     

Abrasive-Free Chemical Mechanical Polishing of Hard Disk Substrate with Cumene Hydroperoxide-H2O2 Slurry

With magnetic heads operating closer to hard disks, the hard disks must be ultra-smooth. The abrasive-free polishing (AFP) performance of cumene hydroperoxide (CHP) as the initiator in H₂O₂-based slurry for hard disk substrate was investigated in our work, and the results showed that the slurry including CHP could improve the material removal rate (MRR) and also reduce surface roughness. Electron spin-resonance spectroscopy (EPR), electrochemical measurement and Auger electron spectroscopy (AES) were conducted to investigate the acting mechanism with CHP during the polishing process. Compared with the H₂O₂ slurry, the EPR analysis shows that the CHP–H₂O₂ slurry provides a higher concentration of the HO^O free radical. In addition, the AES analysis shows the oxidization reaction occurs in the external layer of the substrate surface. Furthermore, electrochemical measurements reveal that CHP can promote the electrochemical effect in AFP and lead to the increase of MRR.     

Detectivity studies of SMD-packaged ZnSSe and ZnMgS UV detectors

The 1/f noise and shot noise studies were carried out on molecular beam epitaxy (MBE)-grown ZnS_0.85Se_0.15 and Zn_0.43Mg_0.57S ultraviolet (UV) detectors packaged in surface mount lead frames. Three ZnS_0.85Se_0.15 detectors with different thicknesses of the active layer and the top electrode pad were used. The highest onset of reverse bias for the appearance of 1/f noise is ?27.5 V, and the highest dark resistances at zero bias is R₀=3.7×10¹³ Θ. The observed difference in their noise performance implies that the increase of the thicknesses of both the active layer and the top electrode pad can significantly lower the noise levels and in turn lead to higher detectivity. The best detectivity achieved is 8.75×10¹³ cm Hz^1/2 W^?1 in a ZnS_0.85Se_0.15 detector with an active layer of 5000 Å and a top electrode pad of Cr (50 Å)/Au (8000 Å). The typical detectivity achieved on Zn_0.43Mg_0.57S devices that have an active layer thickness just exceeds the critical thickness of phase transition is 1.54×10¹² cm Hz^1/2 W^?1.     

Electromigration study in SnAg3.8Cu0.7 solder joints on Ti/Cr-Cu/Cu under-bump metallization

This paper investigates the electromigration-induced failures of SnAg3.8Cu0.7 flip-chip solder joints. An under-bump metallization (UBM) of a Ti/Cr-Cu/Cu trilayer was deposited on the chip side, and a Cu/Ni(P)/Au pad was deposited on the BT board side. Electromigration damages were observed in the bumps under a current density of 2×10⁴ A/cm² and 1×10⁴ A/cm² at 100°C and 150°C. The failures were found to be at the cathode/chip side, and the current crowding effect played an important role in the failures. Copper atoms were found to move in the direction of the electron flow to form intermetallic compounds (IMCs) at the interface of solder and pad metallization as a result of current stressing.     

A copper/polyimide Metal-base packaging technology

A unique substrate MCPM (Mitsubishi Copper Polyimide Metal-base) technology has been developed by applying our basic copper/polyimide technology.¹ This new substrate technology MCPM is suited for a high-density, multi-layer, multi-chip, high-power module/package, such as used for a computer. The new MCPM was processed using a copper metal base (110 × 110 mm), full copper system (all layers) with 50-μm fine lines. As for pad metallizations for the IC assembly, we evaluated both Ni/Au for chip and wire ICs and solder for TAB ICs. The total number of assembled ICs is 25. To improve the thermal dispersion, copper thermal vias are simultaneously formed by electro-plating. This thermal via is located between the IC chip and copper metal base, and promotes heat dispersion. We employed one large thermal via (4.5 mm^?) and four small vias (1.0 mm^?) for each IC pad. The effect of thermal vias and/or base metal is simulated by a computer analysis and compared with an alumina base substrate. The results show that the thermal vias are effective at lowering the temperature difference between the IC and base substrate, and also lowering the temperature rise of the IC chip. We also evaluated the substrate’s reliability by adhesion test, pressure cooker test, etc.     

On the Wafer/Pad Friction of Chemical–Mechanical Planarization (CMP) Processes—Part II: Experiments and Applications

This paper presents the experimental validation and some application examples of the proposed wafer/pad friction models for linear chemical–mechanical planarization (CMP) processes in the companion paper. An experimental setup of a linear CMP polisher is first presented and some polishing processes are then designed for validation of the wafer/pad friction modeling and analysis. The friction torques of both the polisher spindle and roller systems are used to monitor variations of the friction coefficient in situ . Verification of the friction model under various process parameters is presented. Effects of pad conditioning and the wafer film topography on wafer/pad friction are experimentally demonstrated. Finally, several application examples are presented showing the use of the roller motor current measurement for real-time process monitoring and control.     

Effect of misalignment on electrical characteristics of ACF joints for flip chip on flex applications

The effect of misalignment on the electrical properties of anisotropic conductive film (ACF) joints is investigated in this work. It is found that along with the increase of misalignment, the connection resistance of ACF joints increases. When the misalignment in x-direction is less than 5 μm, the increase rate of connection resistance is quite large. Then, along with the severity of misalignment, the increase rate becomes smaller. Finally, when the misalignment is close to 20 μm, the increase rate rises again. The Holm's electric contact theory is used for understanding the connection resistance variation. On the other hand, with the increase of misalignment in x-direction, the insulation resistance between ACF joints decreases. If the misalignment exceeded 10 μm, the decrease is prominent for the Ni particle ACF joints. This phenomenon can be explained by the effect of dielectric damage of the epoxy.Computer programs are also developed to calculate the variation of the probability of open and shorting after misalignment and predicate the maximum misalignment tolerance. The results show that the open and shorting probability increase abruptly after misalignment. On the view of pad parameters, the open probability is mainly related to the pad area, while the pads gap is critical to the shorting probability. Large pads gap (small pad width) can reduce the shorting probability obviously. On the other hand, enlarging the pad area by increasing pad length decreases the open probability significantly. So comparing to square shape pad, rectangle shape pad can reduce the failure probability greatly.     

Analysis of flow between a wafer and pad during CMP processes

In this paper, we summarize the development of a numerical model for the chemical mechanical planarization (CMP) process and experimentally investigate the effects of pad conditioning on slurry transport and mixing. A simplified two-dimensional numerical model of slurry flow beneath a stationary wafer was developed to determine the pressure and shear stress beneath a wafer. The initial results indicate that in the hydrodynamic regime a positive upward pressure is exerted on the wafer. We also examined three cases to study pad effects on slurry transport; polishing with an Embossed Politex pad, an unconditioned IC1000 pad, and a conditioned IC1000 pad. Cab-O-Sperse SC1 slurry was used in a 1:1.5 dilution with water. Mixing data show that conditioning has a negligible effect on the rate of slurry entrainment and mixing; however, conditioning has a large effect on the thickness of the slurry layer between the wafer and pad. Conditioning was found to increase the slurry thickness by a factor of two. In addition the gradients in slurry age beneath the wafer were compared among the three cases. The IC1000 pads supported a gradient in the inner third of the wafer only, while the Embossed Politex pad showed a linear gradient across the wafer implying it retains pockets of unmixed slurry in the embossed topography.