Search

Megasonic cleaning is one of the promising technologies to remove the particles during semiconductor processing. Acoustic bubble cavitation plays a key role in removing the particles. In this work, the effect of an anionic surfactant sodium dodecyl sulfate (SDS) on a bubble in the presence of hydrogen dissolved DIW water was studied. The bubble dynamics were observed using a high-speed camera. It was found that with the increase of surfactant the bubble characteristics were changed very significantly. Several parameters affecting the bubble dynamics were investigated.

Due to the emergence of sub-7 nm technologies, next generation CMP slurry formulations have continued to increase in additive (nanoparticle and chemistry) complexity to meet stringent device specifications. Therefore, it is essential to probe the molecular level interactions at the nanoparticle/slurry chemistry/substrate interface and in turn correlate them to key performance metrics such as removal rate, post CMP defects, and planarization efficiency. This work will address key interactions through a series of case studies focusing on the role of supramolecular structure and cleaning method (i.e. contact vs. non-contact) during STI post-CMP cleaning, probing the impact of supramolecular structure and mode of cleaning relevant to Cu post-CMP, and development of a biomimetic matrix with chemical activity to act as a brush in STI post-CMP cleaning processes. Results show in both BEOL and FEOL post-CMP cleaning there is a strong correlation to the delivery and “soft” nature of the chemistry to allow for effective particle removal at low mechanical force and prevent further defect formation. Furthermore, this work shows a clear correlation between supramolecular structure and particle removal efficiency under both contact and non-contact post-CMP processes.

Visualization experiments were performed to study the relation between free-surface motion and bubble translation in a 1-MHz ultrasonic cleaning bath. From the visualization with a video camera, the characteristic frequencies of the free-surface oscillation (under the acoustic radiation force) and the translational velocity of cavitation bubbles (trapped via the primary Bjerknes force) were extracted, showing that there is a strong correlation between the free-surface oscillation and bubble translation. From the context of megasonic cleaning, such free-surface oscillation is expected to contribute to uniform cleaning performance with cavitation bubbles.

The p-GaAs/n-InP heterojunction was fabricated by direct wafer bonding technology. The optimized atomic level contact between GaAs and InP is critical for getting good ohmic contact and removing the bubbles or voids at the interface, which is helpful to enhance the efficiency of wafer bonded multi-junction solar cells. Through the surface megasonic cleaning and the plasma treatment, we have achieved the high quality bonding interface without bubbles or voids and with interface resistivity of about 0.1 ohms/cm2. A GaInP/GaAs//InGaAsP/InGaAs 4-junction solar cell was prepared with the high efficiency of 34.4% (AM0) at 1 sun.

Acoustic cavitation is used for megasonic cleaning in the semiconductor industry, especially of wafers with fragile pattern structures. Control of transient cavitation is necessary to achieve high particle removal efficiency (PRE) and low pattern damage (PD). In this study, the cleaning performance of solutions with different concentrations of dissolved gas (H2) and anionic surfactant (sodium dodecyl sulfate, SDS) in DIW (DI water) on silicon (Si) wafers was evaluated in terms of PRE and PD. When only DIW was used, PRE was low and PD was high. An increase in dissolved H2 gas concentration in DIW increased PRE; however, PD also increased accordingly. Thus, we investigated the megasonic cleaning performance of DIW and H2-DIW solutions with various concentrations of the anionic surfactant, SDS. At 20 ppm SDS in DIW, PRE reached a maximum value and then decreased with increasing concentration of SDS. PRE decreased slightly with increasing concentrations of SDS surfactant when dissolved in H2-DIW. Furthermore, PD decreased significantly with increasing concentrations of SDS surfactant in both DIW and H2-DIW cases. A high-speed camera setup was introduced to analyze bubble dynamics under a 0.96 MHz ultrasonic field. Coalescence, agglomeration, and the population of multi-bubbles affected the PRE and PD of silicon wafers differently in the presence of SDS surfactant. We developed a hypothesis to explain the change in bubble characteristics under different chemical environmental conditions.

Under extreme ultraviolet (EUV) exposure, the surfaces of EUV mask and optics are exposed to hydrocarbon contaminant, which reduces the imaging performance and lifetime of EUV mask. Thus, an efficient method is required to remove hydrocarbon contaminant from the mask surface. In this work, a mixture of alkaline chemicals and organic solvents has been proposed for the removal of EUV exposed hydrocarbon contaminant from Ru (Ruthenium) capping layer of EUV mask. A hydrocarbon film with similar properties to EUV hydrocarbon contaminant was prepared by PECVD (plasma enhanced chemical vapor deposition). The hydrocarbon removal efficiency was evaluated by using four kinds of quaternary ammonium hydroxides as alkaline chemicals, and DMSO (dimethyl sulfoxide) and THF (tetrahydrofuran) as polar aprotic organic solvents. Among the alkaline chemicals, tetrabutyl ammonium hydroxide, together with an organic solvent, produced effective removal of hydrocarbon from Ru. Removal of hydrocarbon film using a higher polar organic DMSO solvent required a higher concentration and longer time than THF. The combination of megasonic power and cleaning chemistry enhanced the hydrocarbon removal efficiency significantly.

This study investigates the effect of the adhesion and removal of particles such as silica (SiO2), tantalum (Ta) and polystyrene latex (PSL) from three different kinds of EUV (extreme ultra violet) mask surfaces, specifically silicon (Si), tantalum nitride (TaN), and ruthenium (Ru). The results show that inorganic particles such as silica and Ta deposited on Si surface are harder to remove from its surface than from TaN and Ru surfaces. This is due not only to the presence of van der Waals forces but also the hydrogen bonding force between them. On the other hand, an organic contaminant such as PSL particles deposited on Si is easily removed from the surface. This is due to the presence of van der Waals forces between the surface and particles. Inorganic particles deposited on a Ru surface are easily removed from its surface compared with a Si surface. However, PSL particles deposited on a Ru surface are hard to remove due to the presence of pi bonding between the Ru surface and the PSL particles. To weaken the chemical bonding (pi bond), a tetramethylammonium hydroxide (TMAH) cleaning solution is used to improve the removal efficiency of PSL particles on the Ru surface.

Clarifying the mechanism of particle removal by megasonic cleaning and multiple-bubble dynamics in megasonic fields is essential for removing contaminant particles during nanodevice cleaning without pattern damage. In particular, the effect of the interaction of multiple bubbles on bubble-collapse behaviour and impulsive pressure induced by bubble collapse should also be discussed. In this study, a compressible locally homogeneous model of a gas–liquid two-phase medium is used to numerically analyse the multiple-bubble behaviour in a megasonic field. The numerical results indicate that, for bubbles with the same equilibrium radius, the natural frequency of the bubble decreases, and bubbles with smaller equilibrium radii resonate with the megasonic wave as the number of bubbles increases. Therefore, the equilibrium radius of bubbles showing maximum wall pressure decreases with an increasing number of bubbles. The increase in bubble number also results in chain collapse, inducing high wall pressure. The effect of the configuration of bubbles is discussed, and the bubble–bubble interaction in the concentric distribution makes a greater contribution to the decrease in the natural frequency of bubbles than the interaction in the straight distribution.

A novel transducer for megasonic cleaning of photomasks presents an approach that differs from previous configurations, and appears to have unique features for cleaning while minimizing damage. As the cleaning and damage processes are determined by the presence of cavitation, a thorough acoustic analysis was performed on the device, by using a calibrated hydrophone scanned at the photomask location, and a quartz photomask with embedded sensors.

In this study, usually, arrays of square or rectangular piezo devices bonded to a substrate and spaced as close together as possible for transduction for efficient megasonic cleaning, megasonic cleaning is applied to removing sidewall particles during post-silicon etch cleaning. A system was developed containing different measures of ultrasonic and megasonic processes, generated by adjusting the power in NH4OH. Megasonic waves in the gap between the wafer and the megasonic transducer ensure uniform sonic energy across the whole wafer. The generation of radicals is promoted by megasonic energy, such that residue removal occurs in dilute solution. The feature damage for the megasonic cleaning decreased in the NH4OH solution compared to ultrasonic. Moreover, the mechanical force of the bubble cavitation generated by megasonic cleaning also improves residue removal and enhances the mass transfer rate. Extreme cavitation is beneficial for removing particles but can damage wafer features. The damage caused by ultrasonic cleaning is significantly larger than megasonic cleaning. Compared with typical methods of silicon oxide residues removal, this megasonic process has a low material loss and a high residue removal efficiency for via, with a high aspect ratio. In addition, it minimizes defects to structures. These results suggest that megasonic cleaning can be applied to the nanoparticle cleaning process. Megasonic cleaning is quickly becoming an efficient method for post process cleaning in IC fabrication processes.

In the absence of EUV pellicles, EUV masks need to be cleaned frequently. Even after the implementation of the pellicles, EUV masks need to be cleaned multiple times. When pellicle studs need to be removed, aggressive cleaning recipes are required to remove glue from the surface mask. Therefore, surfaces in contact with chemicals during mask cleaning should be stable against aggressive, acidic and basic chemicals. In addition to, chemical durability, EUV masks are repeatedly exposed to EUV and UV radiation. In particular, with increasing the EUV source power, it is expected that EUV mask surface be exposed to EUV light with energy densities > 5 W/cm2 as well as out-of-band radiation in UV region. Such high energy radiation can oxidize Ru cap layer or promote Ruthenium silicide formation under the capping layer. Such mechanism can result in Ruthenium peel –off by multiple cleaning of EUV masks. When acidic chemistries are used to remove particles from pattern EUV masks, absorber layer can be etched and as a result mask CD will change by multiple cleanings. During the chucking process in an EUV scanner, there might be dents form in the backside conductive layer (e.g. CrN) which results in thinning backside conductive layer in the certain areas. Meantime, more aggressive megasonic cleaning is required to remove micron size particles from the backside of the EUV masks. Combination of multiple chucking, dent formation and aggressive cleaning may result in damage in the backside film. Finally, in the EUV masks with OPC sub-resolution-assist-feature (SRAF) can be easily damaged by megasonic during cleaning. This paper will discuss cleaning durability challenges for sub 10 nm half pitch nodes when high power EUV sources expect to be used. In particular, we present our latest results of multiple cleaning of Applied Materials blanks with conventional cleaning chemistries. The change in the EUV reflectivity of Ru cap multilayers by 100x cleaning will be presented. Impact of cleaning on conventional absorber (TaN) and newly developed thin absorber films in Applied Materials will be discussed. The impact of the cleaning processes on the substrate and CrN backside conductive film will be presented. Multiple surface characterization techniques will be used for study of cleaning impact on different films.

Recently, panel size has become increasingly large reaching the 8th generation (2200mm × 2500mm) and there is a trend that is changing from a-Si TFT (amorphous-Si Thin Film Transistor) to oxide TFT because of a demand for high resolution display and an expansion of OLED (Organic Light Emitting Diodes) industry. In previous panel manufacturing processes, cleaning process was not so important. However, nowadays pixel size is becoming smaller and the number of pixels per unit area is gradually increasing due to the high resolution. Consequently, an influence of submicron sized contaminated particles is increasing, so the importance of cleaning process has increased. In the oxide TFT, to achieve the high resolution, Cu based gate structure having a low resistance characteristic is necessary and in order to prevent a leakage current through the active layer, SiO2 deposition is needed on a gate insulator layer and ITO (Indium Tin Oxide) layer is used as a transparent electrode. For a large area and high resolution panel with a good production yield, a research on an effective cleaning process which can remove the submicron contaminated particles deposited on the Cu, SiO2and ITO surfaces during a TFT manufacturing process, is required. Generally, physical cleaning processes are used to remove the submicron particles in FPD (Flat Panel Display) process. Among them, a research on an eco-friendly physical cleaning processes such as DI water megasonic, mixing-jet, brush and electrolyzed ionized water (EIW) is actively being carried out. In this study, we have optimized these cleaning processes to remove the submicron particles from the substrate surface. The primary objective is to establish an optimal cleaning process by applying above processes for the Cu, SiO2 and ITO substrates. And a mechanism of cleaning process and a theoretical understanding of interaction force such as van der Waals force, zeta potential and electrostatic force between the particle and substrate for the optimization of particle cleaning are studied. In the course of the experiment, the particles (glass and aluminum) which were produced in TFT manufacturing process were intentionally deposited on the Cu, SiO2 and ITO substrate surface. And then, the cleaning test using megasonic (1 MHz and 250 W), mixing-jet (570 lpm CDA and 4 lpm DI water), brush (diamond shape) and EIW (pH 10.2 and -790 mV ORP) processes was conducted. After that, we evaluated the cleaning efficiency by counting the number of particles on the surface before and after the cleaning test using an optical microscope. As a result, we achieved a high cleaning efficiency of above 95% in the electrolyzed ionized water cleaning process by applying the megasonic cleaning on all substrates. Figure 1

Extreme ultraviolet lithography (EUVL) is the most promising lithography technique for the 22 nm half-pitch and beyond. The EUV patterned masks work in a reflective mode and it has 40 to 50 pairs of MoSi multilayer to reflect EUV radiation. This multilayer is capped by a 2.5 nm Ru thin film to protect the MoSi multilayer. Since EUV mask does not have a pellicle which protects mask surface from the contaminants. Therefore, EUV mask is more vulnerable to particle contamination during lithography processing which is a major yield loss in device fabrication. No study has been reported on the effect of particle size on CD variation if they exist on EUV mask. The critical defect size, which can cause 10 % CD error, has to be removed from the mask surface. When the defect is located between line and space patterns on the mask, the reflectivity will be reduced due to defect absorption. Simulation predicted that 30 nm sized silica particles can cause 10 % CD error on 16 nm line and space pattern. Thus, these contaminant particles have to be removed from EUV mask surface for preventing CD variation. In order to evaluate the 30 nm sized silica particle removal test, we adopted 30 nm standard spherical silica particles. Actually, it is a challenge to evaluate 30 nm particles removal test without inspection tools. So, we used atomic micro scope (AFM) for the evaluation of these particles. The silica particles were intentionally deposited on Ru surface using spinning method. Cleaning test was performed with twin type megasnoic. 30 nm silica particles removal efficiency was drastically decreased as compared with 137 nm silica particles at 1 MHz megasonic cleaning. Because the megasonic cleaning force was reduced to ultra-fine particles due to boundary layer thickness. Therefore, higher megasonic cleaning force is required to remove ultra-fine particles to overcome boundary layer thickness. It is well known that the large cleaning force is induced, not only particle removal efficiency but also pattern damage will be increased. Thus, megasonic cleaning process without causing pattern damage has to be optimized. In this study, we shall discuss the effect of megasonic frequency on the removal of ultra-fine particles and pattern damage. Also, gas dissolved water and dilute alkali cleaning solution are adopted to high frequency megasonic cleaning process to achieve ultra-fine particles removal without pattern damage effectively. Figure 1

Removal of particulate contaminants from surfaces is a critical area in the processing of wafers to build integrated circuits and is typically done using wet chemical formulations. Strategies for particle removal require an understanding of particle deposition mechanisms. In the first part of this review paper, colloid chemical principles underlying particle deposition on surfaces is discussed. Specifically, the importance of van der Waals and electrical double layer interaction energy on particle–surface interaction is reviewed. This is followed by a description of commonly used particle removal methods in wafer processing, viz., megasonic cleaning, brush cleaning, and nanospray cleaning. The importance of particle deposition conditions used in research on particle removal efficiency is highlighted through a study on the effect of aging.

Display Omitted Identified conditions in the range of 1-3MHz to achieve damage-free cleaning.Characterized cavitation using microelectrode and hydrophone measurements.OH concentration measured using fluorescence spectroscopy.Transient cavitation dominant at 1MHz and stable cavitation prevalent at 3MHz. Megasonic cleaning is routinely used for removal of particulate contaminants from various surfaces in integrated circuit industry. One of the drawbacks of megasonic cleaning is that although it can achieve good particle removal efficiencies at high power densities, it also causes feature damage. The current paradigm is that damage is primarily caused by transient cavitation whereas cleaning is affected by streaming and stable cavitation. In order to develop a damage-free and effective megasonic cleaning process, it is essential to understand the acoustic bubble behavior and identify conditions that generate significant stable cavitation without any transient cavitation. In the current work, microelectrode based chronoamperometry, pressure measurements using a hydrophone and fluorescence spectroscopy studies were conducted under different acoustic frequencies (1-3MHz) and power densities (2-8W/cm2) to fundamentally investigate the type of cavitation produced under these conditions and also establish a correlation to the generation of hydroxyl radicals for characterization of transient cavitation.

In semiconductor fabrication industries, photomask cleaning with no pattern damage is a critical issue, especially for advanced technology nodes. Thus, optimization of cleaning techniques adopted for photomask cleaning or development of new cleaning technique is always of great interest to meet process requirements. In this book chapter, the sources of contaminants on photomask and their impact on imprinted images are first discussed. Then a critical review is presented on the wide spectrum of cleaning techniques and strategies that are adopted for the removal of particulate contaminants and organics from the photomask. The applicability of each technique and its limitations are also examined. Because extreme ultraviolet lithography (EUVL) is considered to be a next-generation lithography, a separate section is devoted to EUVL mask-cleaning techniques.

In this article, the megasonic cleaning system for cleaning micro/nano particles from flat panel display (FPD) surfaces was developed. A piezoelectric actuator and a waveguide were designed by finite element method (FEM) analysis. The calculated peak frequency value of the quartz waveguide was 1002 ㎑, which agreed well with the measured value of 1003 ㎑. The average acoustic pressure of the megasonic cleaning system was 43.1 ㎪, which is three times greater than that of the conventional type of 13.9 ㎪. Particle removal efficiency (PRE) tests were performed, and the cleaning efficiency of the developed system was proven to be 99%. The power consumption of the developed system was 64% lower than that of the commercial system. These results show that the developed megasonic cleaning system can be an effective solution in particle removing from FPD substrate with higher energy efficiency and lower chemical and ultra pure water (UPW) consumption.

Display Omitted Investigated OH generation in various chemical formulations at ~1MHz.Generation rate of OH decreases in the presence of ammonia.Carbonates are better scavengers than bicarbonates.OH generation rate increases with temperature. Chemical formulations used for megasonic cleaning typically contain hydroxides, peroxides, carbonates, and others which can affect particle removal efficiency and feature damage. The current study is focused on investigating the role of carbonates in modulating the oxidation power of megasonic irradiated alkaline solutions through the scavenging of hydroxyl radicals as determined by fluorescence spectroscopy. Although studies exist in literature that report the scavenging of OH by carbonates, these studies have been conducted for waste water treatment applications using advanced oxidation processes. This work focuses on understanding the role of alkaline chemical formulations containing varying levels of carbonates and bicarbonates and solution temperatures on net generation of hydroxyl radicals for applications in semiconductor industry. Investigations were carried out at an acoustic frequency of ~1MHz and different power densities (2, 4 and 8W/cm2).

Megasonic cleaning is traditionally used for removal of particles from wafer surfaces in semiconductor industry. With the advancement of technology node, the major challenge associated with megasonic cleaning is to be able to achieve high cleaning efficiency without causing damage to fragile features. In this paper, a method based on electrochemistry has been developed that allows controlled formation and growth of a hydrogen bubbles close to a solid surface immersed in an aqueous solution irradiated with ∼1 MHz sound field. It has been shown that significant microstreaming from resonating size bubble can be induced by proper choice of transducer duty cycle. This method has the potential to significantly improve the performance of megasonic cleaning technology through generation of local microstreaming, interfacial and pressure gradient forces in close vicinity of conductive surfaces on wafers without affecting the transient cavitation responsible for feature damage.

Emerging ultrasonic and megasonic cleaning demands in various applications (solar cell, storage devices, wafer and mask cleaning, etc.) dictate the need to understand the acoustic cavitation under different operating conditions to optimize efficiency of cleaning and reduce damage. Major parameters that affect cavitation include frequency of the sound field, operating power of the transducer and the cleaning chemistry. Previous studies have reported the use of common techniques such as multi-bubble sonoluminescence [1] and sono-electrochemistry [2] to understand acoustic cavitation. The disadvantage with sonoluminescence technique is that it characterizes cavitation mainly in the bulk of the solution, which may not be pertinent to wafer cleaning applications where the interest is in understanding cavitation phenomena close to the wafer surface. Although, sono-electrochemical techniques employing microelectrode are capable of measuring cavitation in the vicinity of a solid surface, they are limited to measurements on an extremely small area due to the miniscule size (5-25 μm) of the electrode. In this context, hydrophone measurements offer significant benefit as they can be taken near a solid surface as well as on a relative large area (1-2 mm diameter) of the pressure sensitive tip.

Megasonic irradiation of cleaning solutions removes particulate contaminants from wafer surfaces through the mechanisms of acoustic streaming and acoustic cavitation. Uncontrolled cavitation, however, also damages fragile wafer features. It has been hypothesized that damage results primarily from violent transient cavitation while cleaning results from the gentler, stable cavitation and is assisted by substrate etching at alkaline pH. A central challenge in the megasonic cleaning field is to find physical and chemical conditions that would suppress violent cavitation without adversely affecting cleaning efficiency. We have previously reported the strong ability of aqueous CO"2 in suppressing sonoluminescence, a measure of transient cavitation, and wafer damage. However, dissolution of CO"2 also renders solutions acidic and lowers their ability to remove particles. Here we report the development of two new systems, NH"4HCO"3/NH"4OH and NH"4OH/CO"2(g) designed to suppress pattern damage and enhance cleaning efficiency. Using megasonic irradiation of bare and line-space patterned wafers in a single wafer spin cleaning tool, MegPie(R), we measured the efficiencies of these systems as well as NH"4OH, at pH 8.2, in removing SiO"2 particles from oxide Si wafers and suppressing damage to wafer features. All systems were found to achieve high particle removal efficiencies but extent of damage was strongly reduced in the newly designed NH"4HCO"3/NH"4OH and NH"4OH/CO"2(g) compared to NH"4OH. This study establishes a means for well controlled generation of CO"2(aq) over an extended pH range (4.0-8.5) and its application in strongly reducing wafer damage without compromising megasonic cleaning efficiency.

In the silicon dioxide chemical–mechanical polishing (CMP) process, one of the most challenging issues is the formation of defects such as scratches. In this study, scratch formation behavior and CMP performance were evaluated on high-density plasma oxide (HDP), plasma-enhanced tetraethylorthosilicate (PETEOS), and borophosphosilicate glass (BPSG) wafers. To evaluate the number of scratches after the CMP process, contaminated abrasive particles were removed using an optimized post-CMP cleaning process consisting of scrubber cleaning and dilute SC1 megasonic cleaning. The oxide wafers were then treated with dilute hydrogen fluoride (HF) in order to improve the visibility of the generated scratches. The number and shape of the scratches were investigated as a function of oxide film hardness. The results show that a decrease in film hardness correlates with an increase in the number of scratches. Three different types of scratches (chatter marks, line, and rolling) were observed on the oxide surfaces. The do...

A pre-bond cleaning process was developed utilizing a unique, radially uniform, large area proximity type Megasonic transducer. In prior work this new cleaning method was investigated for PRE (particle removal efficiency) as well as particle neutrality. These tests yielded higher values than those achieved with the processes of record. Subsequently, this process was integrated into an industrial volume low temperature fusion bonding process and enabled higher bonding yields. In the above process flow the process fluid was dispensed to fill the gap between the Megasonic transducer surface and the substrate using an atmospheric free flow stream applied to the substrate. Current work describes development, testing and operational verification of a process fluid management device used in conjunction with the wide area proximity Megasonic transducer. The goals of this development were reduction of process fluid amount required, increase the operating substrate rotation speed, and provide better control of process fluid parameters. The design criteria and process flow as well as test results demonstrating the benefits of the new system are presented.

The clarification of the mechanism and the control of a megasonic cleaning are necessary for the megasonic cleaning without pattern damage. In the present study, the numerical analysis of the bubble behavior in a megasonic field is performed to clarify the mechanism of the particle removal of megasonic cleaning. The bubble behavior in the standing wave of a megasonic wave is analysed using a compressible locally homogeneous model of a gas-liquid two-phase medium. The results show that the present numerical method is found to can simulate the characteristic of the bubble translational motion in a standing wave that the direction of the movement due to the primary Bjerknes force changes at the resonant radius. And the calculated wall pressure is higher than the exposed megasonic pressure, and the cavitation is confirmed to influence the particle removal of megasonic cleaning.

An investigation of particle removal efficiency and feature damage has been conducted in NH4OH/NH4HCO3 cleaning solutions irradiated with megasonic energy. By adjusting the pH of the solution in the range of 8.2 - 8.5, high particle removal efficiency (PRE) was achieved while feature damage was reduced significantly. The sonoluminescence data collected from NH4OH and NH4OH/NH4HCO3 solutions indicate significant suppression of transient cavitation in alkaline solutions containing aqueous CO2.

Megasonic cleaning has been one of the most successful techniques for Cu/low-k interconnects post-CMP cleaning. The structural deformation and stress of Cu and low-k materials in megasonic cleaning are examined with finite element method (FEM). The maximum stress is concentrated in the binding area between Cu and low-k. With decrease of Cu line width, the maximum stress increases and the max value exceeds the yield strength of Cu which results in the plastic deformation. The increasing frequency will change the bubble collision times. Therefore the fatigue is potential. The maximum displacement moves from center to the sides of top surface with increase of line width. When the line width is 25nm, the deformation is the largest.

Megasonic cleaning has been one of the most successful techniques for nanoparticle cleaning in semiconductor industry. However, the megasonic energy often causes a pattern collapse and some damage to very small structure. In this study, the distribution of sonic pressure in a wet cleaning bath with different arrangement of megasonic transducers is simulated. Megasonic wave transmits directional and will be disturbed. Thus the wafer surfaces should be arranged parallel to the sound wave direction of propagation. Convergence gain in the whole cleaning space is limited but results in great variance in local areas. Thus the wafer moving, rotating and oscillating is necessary.

The application of megasonic energy in semiconductor cleaning solutions has been shown to be very effective in improving the particle removal efficiency (PRE). There has been a significant interest in understanding the phenomena of cavitation and acoustic streaming, which are known to play an important role in particle removal during megasonic cleaning. In the present work, transient cavitation in acoustically (~1MHz frequency) irradiated aqueous solutions containing different dissolved gases (Ar, N"2 and CO"2) has been characterized with a 25@mm diameter microelectrode using high resolution cyclic voltammetry. Specifically, using ferricyanide as an electrochemical probe, current transients are obtained as a function of time. A simple mathematical analysis based on diffusion is used to correlate the collapse characteristics of a transient cavity to the magnitude of current peaks and the time scale of rise and fall in current.