It is well known that oxide dishing occurring in STI CMP leads to considerable sidewall and edge-parasitic conduction. Thus, a closed-form solution for quantitative prediction of oxide dishing is needed. A contact-mechanics-based approach to describe the steady-state oxide dishing occurring in STI CMP process is presented. The theory is validated through comparison with experimental data in the literature. Once validated, the model is used to quantify the effect of pattern geometry on oxide dishing. It is shown that the predictions of the model agree reasonably well with the experimental results measured in overpolishing time.

The interaction between Cu surface and abrasive particles in slurry solution was characterized. The adsorption behavior of the citrate ions was dependent on the pH of the slurry and the concentration of the citric acid. The adsorption of citrate ions generated a highly negative charge on the alumina surface and shifted isoelectric point (IEP) to lower pH values. The Cu removal rate of alumina slurry was higher than that of colloidal silica based slurry in the investigated pH ranges. Although lower friction forces of Cu were observed in alumina based slurry of pH 4, 6 and 8, a higher friction force was observed at pH 2. This high friction force was attributed to the positive zeta potential and greater adhesion force of particle. It indicates that the magnitudes of particle adhesions on Cu surfaces in slurries can be directly related to the frictional behavior during CMP process.

Efficient end point detection (EPD) helps prevent numerous Chemical Mechanical Polishing (CMP) process defects such as dishing, erosion, excessive over-polish etc. During the CMP of Shallow Trench Isolation (STI) structures, the process should be effectively stopped at the buried Si3N4 layer to prevent any/all of the aforementioned process defects. In this research, novel in-situ metrology technique that used the real time tracking of Coefficient of Friction (COF) data during polishing was employed for EPD during STI-CMP. The experiments were performed on the CETR CP-4 CMP Tester using 1“X 1” sample coupons having a 2 ‘mu;m pitch STI pattern. The COF signal during polishing is a characteristic signature of the given process parameters and conditions. The in-situ EPD was based on the principle that the COF values are strongly dependent upon the surface of the materials that are being polished. Extended polishing to polish the underlying layer after the process end point was reached, can give an estimate of the polishing slurry selectivity with respect to the buried layer. The ex-situ surface characterization of coupons was performed at different points of the process using Atomic Force Microscopy (AFM). The repeatability and reliability of this technique was evaluated after carrying out multiple tests at similar process conditions. The demonstrated methodology can also be implemented for characterization of polishing pads and slurries besides STI-CMP process optimization.

This paper proposes a universal Chemical Mechanical Planarization (CMP) process description language. We discuss the appropriate attributes needed to describe complex CMP processes and the capability for the standardization of CMP. We show an implementation of the language based on an open standard.

Abrasive Free Copper Chemical Mechanical Polishing (AF-CMP) was developed for Cu/low k materials. Blanket film Cu removal rate of 6000 Å/min with very less non-uniformity of 3% achieved for polishing pressure of 1.5 psi. CMP process window and lower critical pressure were identified with pattern wafers. Material removal mechanism was studied using surface morphology of Copper blanket wafers polished using different pressure, rotation rate and slurry flow rate. Material Removal Mechanism (wear mechanism) such as Chemical wear (etching) and mechanical wear (fatigue wear, particle adhesion wear and abrasion wear) have been found. The increase of slurry flow rate and relative velocity and the decrease of pressure give the dominance of chemical wear in material removal mechanism, and vice versa. Dishing control was achieved during Cu polish using different carrier/platen speed for Cu/SiLKTM patterned wafers. The cumulative distribution of the metal line-to-line leakage current measurements of wafers shows good performance and it is comparable to abrasive process.

One of the critical parameters in chemical mechanical polishing (CMP) technology is adhesion between abrasive particles and surfaces being polished. Accumulated up to date theoretical and experimental facts insist that development of advanced material removal and cleaning methods requires understanding and evaluation of particle adhesion in these processes. To this end, our research focuses on AFM studies of interaction between the slurry particles and different substrates treated by CMP processes. To conduct adhesion measurements, the particles present in a CMP system and model latex particle were first attached to the surface of silicon wafer covered with monolayer of polymer(s) having high affinity to particles and the silicon wafer. Glass/quartz sphere (10-60 microns) was attached to AFM cantilever with appropriate spring constant. (The sphere represented the surface of material being polished and might be covered with various materials during future investigations.) Next, AFM force volume mode, which utilizes the collection of the force-distance curves over selected surface areas, was used for the adhesion measurements.

Chemical mechanical planarization (CMP), which can globally planarize both silicon dioxide (the prevalent interlayer dielectric), and copper films, has become the key process in the damascene method used for producing integrated circuit (IC) devices with multilevel copper interconnects. Cu CMP is typically carried out with slurries containing oxidizing agents, complexing agents, and corrosion inhibitors as the principal chemical components. In such slurries, complexing agents enhance the solubility of copper and increase the dissolution rate of the abraded material in Cu CMP. They also assist achieving high copper removal rates during dynamic polishing conditions. The nature of the complexing agent used, the pH and the redox potential of the slurry system are among the main factors controlling the dissolution and passivation behaviors of copper during CMP. Consequently, these factors are intimately related to the key CMP performance metrics such as removal rate and planarity. In this paper, potentialpH diagrams of copper in aqueous systems containing a number of organic complexing agents including ethylenediaminetetraacetic acid (EDTA), nitrilotriacetic acid (NTA), oxalic acid and malonic acid are presented. The predominance regions of copper complexes under different copper and ligand activities and their implications on copper removal during CMP are discussed.

This paper reviews the effect of pattern symmetry on Chemical Mechanical Planarization (CMP) and proposes a methodology to reduce the computational time of the CMP process simulation based on the boundary element method (BEM). We focus on a unit field which generates the symmetry structures and we formulate the BEM field matrix by parallel translations and spin/flip operations. We also discuss the characteristics of the field matrix and demonstrate the applications of the methodology.

Chemical mechanical polishing (CMP) is widely used for local and global planarization of microelectronic devices. It has been demonstrated experimentally in the literature that the polishing performance is a result of the synergistic effect of both the chemicals and the particles involved in CMP. However, the fundamental mechanisms of material removal and the interactions of the chemical and mechanical effects are not well understood. A comprehensive model for CMP was developed taking into account both the chemical and mechanical effects for monodisperse slurries. The chemical aspect is attributed to the chemical modification of the surface layer due to slurry chemistry, whereas the mechanical aspect is introduced by indentation of particles into the modified layer and the substrate depending on the operating conditions. In this study, the model is extended to include the particle size and pad asperity distribution effects. The refined model not only predicts the overall removal rate but also the surface roughness of the polished wafer, which is an important factor in CMP. The predictions of the model show reasonable agreement with the experimental observations.

Material removal in CMP occurs during intervals of pad-wafer contact separated by intervals of non-contact. One predictable sequence of non-contact intervals for a fixed point on the wafer is the traverse of the pad grooves, during which the wafer surface is renewed with fresh chemistry and heat is conveyed away. It is well understood that good uniformity requires machine kinematics that expose all points on the wafer to the same total contact time, mean slurry concentration, and temperature. Less widely known is that coherent structures tens to hundreds of nanometers high and matching the pitch of the pad groove pattern may be formed on an otherwise planar wafer despite multiple rotary motions. This unexpected phenomena is of interest not only because it manifests the impact of grooves and transport at scales not easily studied, but also because shrinking device architectures will ultimately disqualify even nanoscale departures from planarity. Wafer polish experiments are conducted alternately using circular, Cartesian grid, and spiral groove patterns using specialized pad conditioning and CMP recipes to amplify groove-induced nanotopography. Polish results illustrate sharp patterns in finished wafers (visible to the naked eye) that should not survive dual-axis tool kinematics. Computational 3-D model results are then presented for transient slurry mixing in the pad-wafer gap of a 200-mm polisher using the same groove patterns. A direct correspondence is found between observed wafer nanotopography and predicted groove-scale slurry mixing dynamics. In particular, the surface structures are underpolished areas traceable to intervals of non-contact protracted by depleted polish chemistry that prevails in groove segments when oriented relative to the local pad and wafer motion in a way that suspends transverse mixing in the groove crosssection. The study conclusively defines the features required in a groove pattern and polish recipe to form coherent structures matching the groove pitch. As validation of the theory, a groove pattern expected to form no surface topography is defined, experimentally tested, and shown to perform as predicted. Findings are discussed in the context of next-generation pad grooving and texturing as required for progressively more demanding applications of CMP.

Obtaining local and global planarity is one of the prime criteria in dielectric and metal planarizations. Although Chemical Mechanical Planarization (CMP) helps us achieve this criterion in constant pattern density surfaces, the same is not true for variable pattern density surfaces this results in formation of global step heights across the die. This paper provides a pressure open loop control algorithms for obtaining planarity across a die containing variations in pattern densities. Based on the variation of pattern density and surface heights across the die, the surfaces are separated into zones and the pressure is varied spatially and/or temporally to obtain uniform surface heights, with enhanced step height uniformity. One of the algorithms looks ahead and recalculates/modifies the pressure values by identifying the step heights that could be formed after a specified time step. The final surface predictions have improved uniformity on the upper surface as well as on the step heights across the entire die. The simulation would help us track the polishing process for each time step and guide us with the optimized pressure values that can be applied in order to an uniform final surface evolution.

Wafer-level three dimensional (3D) IC technology offers the promise of decreasing RC delays by reducing long interconnect lines in high performance ICs. This paper focuses on a viafirst 3D IC platform, which utilizes a back-end-of-line (BEOL) compatible damascene-patterned layer of copper and Benzocyclobutene (BCB). This damascene-patterned copper/BCB serves as a redistribution layer between two fully fabricated wafer sets of ICs and offers the potential of high bonding strength and low contact resistance for inter-wafer interconnects between the wafer pair. The process would thus combine the electrical advantages of 3D technology using Cu-to-Cu bonding with the mechanical advantages of 3D technology using BCB-to-BCB bonding.

In this work, partially cured BCB has been evaluated for copper damascene patterning using commercially available CMP slurries as a key process step for a via-first 3D process flow. BCB is spin-cast on 200 mm wafers and cured at temperatures ranging from 190°C to 250°C, providing a wide range of crosslink percentage. These films are evaluated for CMP removal rate, surface damage (surface scratching and embedded abrasives), and planarity with commercially available copper CMP slurries. Under baseline process parameters, erosion, and roughness changes are presented for single-level damascene test patterns. After wafers are bonded under controlled temperature and pressure, the bonding interface is inspected optically using glass-to-silicon bonded wafers, and the bond strength is evaluated by a razor blade test.

A ceria-based in-situ conditioning process was successfully developed. A new single layered metal bonded conditioner was developed to solve problems encountered in implementing ceria based in-situ conditioning process. The planarity was significantly improved compared with that of conventional Ni-plated in-situ conditioning.

The major aim of CMP is not the removal of excess material but the planarization of the surface. Therefore the determination of the planarization length appears to be more important than the removal rate itself. It has been shown, that the planarization length is not a constant process parameter, but is related to the removal respectively to the polish time in a square root behaviour. Founded on models proposed by Boning, Ouma, et. al. we applied a sequential polish on a single quasi infinite step. The resulting profile could be simulated by a sequential convolution of the surface contour with a Gaussian transfer function.

To come closer to the situation on a chip pattern we investigated the planarization behaviour on a specific pattern of the MIT854AZ copper CMP test chip, where a large area of unpatterned surface touches a pattern with a specific constant density.

The 200 mm wafer samples consisted of RIE structured oxide films covered with 850 nm ECD copper. The polish was performed on a standard semiconductor manufacturing tool, using a commercial consumables set. The surface profiles were determined by a high resolution profiler within the polishing sequence. The densely patterned areas are removed within a certain polishing time while the transition point between the unpatterned and patterned area appears as a global step. The deposited copper thickness is sufficient to study the contour evolution in both phases, before and after removal of the dense pattern. The paper presents the experimental results on the contour evolution for the patterned fields as well as the global step.

We present an extension of the density-stepheight model for pattern effects in oxide CMP which accounts for the roughness of the polishing pad's surface. The model is compared to polishing data for processes using different pressure and speed. Agreement with the data is improved especially in the initial regime of polishing before the pad contacts the down areas.

Chemical Mechanical Planarization (CMP) is the process of choice for planarization of the constituent layers of the Multi Level Metallization schemes in modern Integrated Circuits. Besides having a lot of advantages, copper CMP process still needs significant process control to eliminate defects such as delamination, microscratches, dishing, erosion etc. In this research, effect of heat generated at the interface on the generation of CMP defects has been investigated. CMP of blanket and patterned samples has been carried out at two conditions of pressure x velocity values with varying slurry temperature. Post CMP metrology is carried out using Atomic force microscopy (AFM) in order to characterize the variation in scratch depth, dishing profile and non-uniformity in step coverage. Pictures of the patterned samples polished at different temperatures are captured using Optical Microscopy (OM) to study the dishing and dissolution of copper lines in greater detail. The primary goal of this study was to gain deeper understanding of the effect of heat generation and rise in temperature at the pad-wafer-slurry interface on CMP induced defectivity.

The friction behavior and adhesion of abrasive particles were experimentally investigated during Cu CMP process. The highest particle adhesion force was measured in alumina slurry without citric acid. However, the alumina slurry with addition of citric acid had the lowest particle adhesion due to the adsorption of citrate ions on the alumina surfaces. While citrate ions could be easily adsorbed on alumina particles, silica particle showed the least effect on adsorption in citric acid solutions. The magnitude of adsorptions of citrate ions on the particle surfaces had significant effect on frictional behavior as well as adhesion force. Higher particle adhesion force resulted in higher friction, particle contamination and scratches in CMP process. It indicates that the magnitudes of particle adhesions on wafer surfaces in slurries can be directly related to the frictional behavior during CMP process.

A family of novel low-shear and surface-engineered pads for copper chemical mechanical planarization (CMP) are designed and evaluated for polishing characteristics and process induced stress. Tribological studies are carried out on blanket copper films using these low-shear/surface-engineered pads and comparison is made with state of the art commercial pads (IC1000 and JSR). Results indicate that low-shear surface-engineered pads exhibit vastly improved tribological performance. These pads exhibit a 40% lower coefficient of friction (COF) over a larger range of Sommerfeld numbers (So) compared to commercial pads. A simplified XRD line profiling technique has been utilized to characterize stress within post-CMP copper films. The (222) Bragg peak position obtained from the XRD spectra of films polished using low-shear surface-engineered pads indicates virtually no shift compared to the peak position for the unpolished copper films. This result establishes that almost no stress is incorporated during bulk Copper removal using the newly designed pads.

In copper CMP, transitions from copper to barrier as well as barrier to dielectric layer are typically sensed in situ using an optical reflectance technique. Spectroscopic techniques such as Raman, which allow monitoring the vibrational modes of silicon and low-k layers, have interesting potential for detecting these transitions. In this paper the use of Raman spectroscopy in detecting in situ removal of barrier layers from CDO materials is reported. Intensities of Raman peaks characteristic of Si-Si vibrations from Si substrate and C-H vibrations from low-k materials have been used for monitoring CDO layer thickness and detecting removal of Ta overlayer. An abrasion cell is integrated with a Raman spectrometer to demonstrate the feasibility of Raman monitoring in-situ. Capabilities and limitations of the Raman spectroscopic method are discussed.