CMP of PC, PMMA and SU-8 Polymers

Chemical mechanical polishing of polymeric materials for MEMS applications

This article discusses advanced developments and applications of chemical mechanical polishing (CMP) published recently in the selected papers indexed by Web of Science. The topics covered are advances in slurry and abrasives, pads and conditioning, CMP for semiconductor device manufacturing, CMP for other applications, modeling and simulations, and CMP with ultrasonic vibrations, lasers, photocatalysts, or UV lamps. Nonspherical abrasive particles have been developed for CMP, resulting in increased material removal rates (MRRs). Advanced conditioning methods have been proposed to uniformly generate pad surface shapes. Fixed abrasive CMP has advantages with higher MRRs. New models for designing the pad and conditioner have been proposed, and more uniform pad shapes can be obtained. Integrated advanced process control improves the wafer-to-wafer variation. Dental implants treated by CMP perform equally or better than the baseline-machined implants and the biphasic calcium phosphate-treated implants. The slurry distribution and the abrasive behavior can be simulated by means of multiphase modeling. Molecular dynamics simulations can explore the mechanism of CMP. CMP of wafers can be simulated using an atomic force microscope with its tapping mode. Theoretical models have been developed to calculate removal depths, study the chemical action in CMP, and explore the crystal orientation effects. CMP assisted by ultrasonic vibrations increases CMP MRRs and lowers the roughness of polished surfaces. CMP assisted by using UV lamp power, femtosecond lasers, or photocatalysts enhances CMP MRRs.

Glass disk substrates are used in a wide range of portable devices because of their relatively high resistance to heat and shocks compared with aluminum substrates. Chemical mechanical polishing (CMP) is an available method to eliminate residual surface defects induced by grinding/lapping to meet final product requirements. In this work, CMP experiments with a low-cost fabric cloth pad were performed to study the effects of four key process factors on the material removal rate (MRR), flatness, and surface finish of polished glass disk substrates, and to explore the feasibility to achieve high machining efficiency, sub-nanometric surface roughness, and micrometric flatness of polished glass disk substrates simultaneously under optimized CMP conditions. The experimental results and analyses reveal that the interaction of pad rotational speed and polish head rotational speed had significant effect on the roughness (Ra), down force, and the interaction of down force and pad rotational speed had significant effects on the flatness, and down force and pad rotational speed had significant effects on the MRR. After compromise and prioritization analyses are performed, a combination of the four variables, which is the combination of 50 g/cm2 for down force, 40 rpm for pad rotational speed, 20 rpm for polish head rotational speed, and 100 ml/min for slurry supply velocity, is obtained to achieve high machining efficiency, sub-nanometric surface roughness and micrometric flatness of polished glass disk substrates simultaneously.

In this study, the effect of the friction and wear of a polishing pad on the material removal rate of a silicon oxide wafer was investigated during chemical mechanical polishing (CMP) with ceria slurry. Further, the effect of surface properties of the polishing pad, such as surface roughness and hardness, on the variation in the material removal rate was examined. From a tribological viewpoint, the in-situ friction force was monitored during the CMP process, and wear of the polishing pad was controlled by different types of conditioners. After CMP, the pad surface roughness was measured by optical profiling and scanning electron microscopy. Experimental results showed that the material removal rate was almost linearly proportional to the friction force between the pad and the wafer surface, irrespective of the properties of the pad. Experiments on the dependency of the pad wear rate on the material removal rate showed that the material removal rate increased with a decrease in the pad wear rate. Experiments and pad characterization confirmed that such a correlation was attributed to the pad surface roughness and the friction force.

With the rapid change of materials systems and decreased feature size, thin film microstructure and mechanical properties have become critical parameters for microelectronics reliability. An example of a major driver of this new technology is the data storage community who is pushing for 1 Terabit/square inch on its magnetic disk hard drives. This requires inherent knowledge of the mechanical properties of materials and in depth understanding of the tribological phenomena involved in the manufacturing process. Chemical mechanical polishing (CMP) is a semi-conductor manufacturing process used to remove or planarize ultra-thin metallic, dielectric, or barrier films (copper) on silicon wafers. The material removal rate (MRR), which ultimately effects the surface topography, corresponding to CMP is given by the standard Preston Equation, which contains the load applied, the velocity of the pad, the Preston coefficient which includes chemical dependencies, and the hardness of the material. Typically the hardness, a bulk material constant, is taken as a constant throughout the wafer and thereby included in the Preston coefficient. Through metallurgy studies, on the micro and nano scale, it has been proven that the hardness is dependent upon grain size and orientation. This research served to first relate the crystallographic orientation to a specific hardness value and secondly use the hardness variation in the previously developed particle augmented mixed-lubrication (PAML) model to simulate the surface topography and MRR during CMP. Recent test and results show that currently there is no empirical formula to relate the crystallographic orientation and thereby a critically resolved shear stress (CRSS) to a specific hardness value. The second part of this investigation utilized the variation in hardness values from the initial study and incorporated these results into the PAML numerical model that incorporates all the physics of chemical mechanical polishing (CMP). Incorporation of the variation of hardness resulted in a surface topography with a difference in roughness (Ra) from the bulk constant hardness value of 60 nm. The material removal rate (MRR) of the process differs by 2.17 μm3/s.

Epoxy resins are widely used in the coating, adhesive and electronics industries. However, it is a difficult-to-machine polymer owing to its nature of thermal softening and soft-plastic and corrosion resistance characteristics. To overcome these difficulties, surface roughness (Sa) is usually set at more than 4 nm, and material removal rate (MRR) is normally set lower than 2 μm h-1 during traditional chemical mechanical polishing (CMP). Nevertheless, atomic level surface roughness is desired by high-performance devices of epoxy resins. This poses a challenge in conventional CMP. To address this challenge, herein, a novel CMP was developed, and the CMP slurry included ceria, dichloromethane, sodium carboxymethyl cellulose and deionized water. After CMP, an atomic level surface roughness was achieved with an Sa of 0.198 nm, and the MRR was 5.46 μm h-1. To the best of our knowledge, both the surface roughness and MRR are the best values reported in the literature. An atom-molecular dual-scale model was proposed to investigate the dynamically interactive mechanism between dichloromethane and the epoxy resin. Fourier transform infrared and Raman spectroscopies revealed that the conformation of epoxy resin varied and surface swelling occurred during the diffusion of dichloromethane. Hydrogen atoms of dichloromethane were selectively adsorbed onto the oxygen atoms of ether bonds in the epoxy resin, reducing the energy of the chains. The adsorption energy was high, reaching up to -67.3 kcal mol-1, and this conformation was the most stable. During adsorption, 0.012 eV of energy was transferred to dichloromethane from the small crosslinked molecules. Molecular dynamics simulations and density functional theory were used to perform these calculations. Dichloromethane relaxed the resin surface, speeding up the migration rate of chains and reducing their energy of breaking. The combined effect of mechanical shear force, compressive force and relaxation of swelling led to the breakage of C-N bonds with the lowest energy in the chains. The broken short chains were wrapped in dichloromethane, separating from the surface of the resin with the slurry flow. Our results provide new insights into acquiring an atomic level epoxy resin surface with soft-plastic, thermal softening and corrosion resistance characteristics.

Chemical mechanical polishing (CMP) has become a critical planarization technique in the manufacture of advanced integrated circuit devices. It can effectively remove the topography of thin film on wafer, and achieves the good planarization either in local or global region, and also the polished object will have the excellent surface quality. Recently, new materials and architectures have been introduced in the manufacturing process of IC production, so some consumables still require further study and development, mainly due to the stricter requirements. In this dissertation, the improvement in CMP performance by various pad conditioner designs and fixed nanodiamond pad was investigated experimentally. The main consumables under research, which included the polishing pad and the pad conditioner, in CMP performance were evaluated, and the polishing planarization was also analyzed and improved. The fixed nanodiamond pad was the developed consumable, and through a series of experiments in silicon CMP to obtain the proper pad design and verify the pad’s validity. In the comparison of the performance of the experimental fixed nanodiamond pad, several commercial slurries were adopted. In citing the findings from the experimental results, the value and practicality of this study was discussed. The surface roughness and thickness of the polishing pad influenced the material removal rate directly, and the uniformity of both the surface roughness and thickness of the polishing pad were significant factors of the with-in wafer non-uniformity in CMP process. The compressibility and the friction force could be used to evaluate the polishing pad. Using a corrective conditioning profile, the non-uniformity of material removal rate was significantly improved, and also the service time of the polishing pad was lengthened. The design of pad conditioner was critical as it determined the efficiency of the grooving action. The pad conditioner chose the polycrystalline diamond as its substrate, which could increase the wear resistance. The diamond cutting tips on pad conditioner had the same size, shape, and protrusion height, which was beneficial for the dressing behavior and stability. The effect of various fixed abrasive pad designs on polishing characteristics during silicon wafer polishing was investigated. Due to the different manufacturing method for these fixed nanodiamond pads, the abilities of controlling the nanodiamond particles were different. Consequently, the surface roughness of the polished silicon wafer was remarkably improved by the fixed nanodiamond pad, and result also revealed that the pad hardness was a key factor in keeping the abrasives effective in the pad-wafer interface. In comparison with the performance of slurry process in silicon CMP, although the fixed nanodiamond pad process had a relatively low polishing efficiency that depended simply on the mechanical abrasion. For the same reason, a relatively thick damaged layer was formed in the polished silicon subsurface, as the fixed nanodiamond pad was used. However, it was comparable with the slurry process in terms of surface topography of polished silicon wafer, due to its extremely small and nearly spherical particles.

CMP characteristics and optical property of ITO thin film by using silica slurry with a variety of process parameters

This paper aims to study the chemical mechanical polishing (CMP) of glass wafer with effect of colloidal silica (SiO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> ) abrasive in different CMP process parameters. Conventional glass polishing usually adopts ceria (CeO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> ) abrasive due to its high material removal rate (MRR). However, with the development of 3D stacking integrated circuits (3DS-IC), the demand of glass wafer for trench-glass-via (TGV) instead of trench-silicon-via (TSV) has very tight specification of surface roughness and near-free-defects by CMP for next step of lithography process. Thus the ceria abrasive may not satisfy the surface quality as required in applications of glass wafer or TGV. The colloidal silica of SiO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> abrasive is considered because of its uniform particle size and familiar application in CMP for silicon wafer and thin film devices. In this study, viscosity and pH value of SiO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> slurry have been tested to characterize slurry properties. Some 40 × 40mm sliced substrates from 200mm (8") glass wafers have been used for polishing characteristics and evaluation of CMP process. Relationship between the process parameters, including down pressure, platen speed and slurry flow rate have been investigated. Performance of CMP tests including MRR, surface roughness and non-uniformity have been obtained for concluding the slurry performance for CMP of glass wafers.

The processing parameters directly affect the polishing quality of planar optics in chemical mechanical polishing (CMP). In this paper, a novel multiparameter optimization method based on the fuzzy theory was proposed to evaluate the manufacturing efficiency and accuracy of flat optical elements with CMP. The evaluation criteria and optimized parameters for CMP optimization were selected. The optimized objective function—combined fuzzy index of CMP (CFICMP)—was introduced to combine different criteria. A L9(34) orthogonal experimental design was implemented to improve the efficiency of experiments and analyze the effect of three parameters (polishing pressure, rotational speed of polishing pad, and slurry flow rate) on CMP. A half-normal distribution was applied to fit the trend of CFICMP. To verify the validity and feasibility of CFICMP, the material removal rate and the material removal uniformity were set as the optimized functions in verification experiments. The experimental results demonstrated that the optimal parameters of CFICMP could achieve better manufacturing efficiency and accuracy of planar optics simultaneously. The proposed method provides a feasible and effective way to select optimized processing parameters for CMP.

Chemical mechanical polishing (CMP) has become an essential process in the manufacturing of advanced microelectronic devices. More recently, CMP has also been applied to the process of other advanced materials such as optical crystals and thin films. Typically, a CMP slurry is formulated as an aqueous dispersion which may contain abrasive particles, activating agent, passivating agent, surfactant, etc. Due to its sensitivity to water, hygroscopic crystals must not be processed with aqueous based slurry. In this study, a new abrasive-free system based on water-in-oil microemulsion was investigated to address this challenge. More specifically, a dispersion made of dodecanol, Triton X-100, and water was studied for its potential application in KH2PO4(KDP) crystal processing. In this unique polishing system, water molecules are caged into micelles so the reaction between KDP and water is controlled. As a result, the static etch rate of the substrate surface is minimized. During polishing process, the frictional action between crystal surface and pad leads to the release of reactive water molecules. The material removal is, thus, enhanced. In this paper, the techniques used to characterize such abrasive-free system were first introduced. The water-in-oil structures were characterized and confirmed by conductivity, dynamic lighting scattering and dynamic nuclear magnetic resonance (NMR) measurements. The performance of this system on the process of KDP crystals was then discussed. The static etch rate and the material removal rate in polishing process were measured under various conditions in order to elucidate the polishing mechanism. Finally, the potential application of such a novel nonaqueous polishing system in CMP beyond KDP crystals is discussed.

Cobalt (Co)‐doped colloidal silica abrasives were synthesised by seed‐induced growth method. Time‐of‐flight secondary ion mass spectroscopy was used to characterise the composition of the obtained abrasives. The morphology of the abrasives was measured by using scanning electron microscopy. The chemical mechanical polishing (CMP) performances of the Co‐doped colloidal silica abrasives on sapphire substrates were investigated. Experiment results indicate that the Co‐doped colloidal silica abrasives exhibit lower surface roughness and higher material removal rate (MRR) than that of pure colloidal silica abrasive under the same testing conditions. Furthermore, the inductively coupled plasma‐atomic emission spectrometry and X‐ray photoelectron spectroscopy were also used to investigate the acting mechanism of the Co‐doped colloidal silica in sapphire CMP. Analysis results show that cobalt aluminate appears after polishing, implying the tribochemistry reaction occurs during CMP. The chemical reaction between element cobalt and sapphire surface during CMP can promote the chemical effect in CMP and lead to the increasing of MRR.

The chemical mechanical polishing (CMP) of tungsten has been recognized as an enabling technology for sub-half-micron multi-level interconnect. There are many technical challenges to be resolved when incorporating tungsten CMP into the manufacturing line. In this paper, key issues related to the tungsten CMP process--oxide erosion, plug recess, oxide thinning, surface roughness, microscratches, contamination, photolithographic alignment, metal line bridging, via resistance and device yield--were studied for three different types of slurries. The variables studied include pattern density, glue layer thickness, overpolish time and underlying oxide type. Atomic force microscopy, scanning electron microscope and long-scan surface profiler were used. Oxide thinning and erosion were found to depend strongly on slurry type and all the variables studied. With optimized polishing parameters and integration scheme, the tungsten CMP process has enough processing window for manufacturing.

Material removal rate model for chemical–mechanical polishing of single-crystal SiC substrates using agglomerated diamond abrasive

Real-time prediction of material removal rate for advanced process control of chemical mechanical polishing