2002 Progress Report: Removal of Photoresist and Post-Plasma Etch Sidewall Films Using Superciritical and Subcritical CO2 with AdditivesEPA Grant Number: R829554
Title: Removal of Photoresist and Post-Plasma Etch Sidewall Films Using Superciritical and Subcritical CO2 with Additives
Investigators: Hess, Dennis W. , Levitin, Galit , Myneni, Satya
Institution: Georgia Institute of Technology
EPA Project Officer: Carleton, James N
Project Period: January 1, 2002 through December 31, 2004
Project Period Covered by this Report: January 1, 2002 through December 31, 2003
Project Amount: $325,000
RFA: Technology for a Sustainable Environment (2001) RFA Text | Recipients Lists
Research Category: Sustainability , Pollution Prevention/Sustainable Development
Surface cleaning and preparation comprises approximately 35 percent of the production steps performed in advanced integrated circuit manufacturing and photoresist (PR) stripping accounts for essentially half of these cleaning steps. Incomplete removal of residues leads to defects and the incorporation of impurities into devices, resulting in reduced device yield and reliability. Conventional stripping by plasma and wet processes has serious technological drawbacks. Plasma cleaning can alter the film structure and the dielectric constant (k) of insulators and may create particles. Wet stripping can lead to the oxidation of silicon and metal surfaces and result in difficulty in penetrating narrow trenches in modern devices (<100 nm). In addition, the use of wet chemistry, which is primarily based on hazardous solvents, bases, and acids, generally has a negative environmental impact. Moreover, large amounts of de-ionized water (DIW) and isopropanol (for drying) are required to complete the stripping process. Super- and subcritical CO2 is being considered as an environmentally benign alternative for PR removal as indicated by the International Technology Roadmap for Semiconductors (ITRS, Edition 2001, Interconnect, http://public.itrs.net Exit ). This approach can reduce the use of harmful chemicals and DIW and thus may serve as a replacement for the present chemically intensive approaches.
The primary objective of this research project is to develop a more environmentally benign process for postplasma etch PR/residue removal based on super- or subcritical CO2 mixtures. This requires characterization of the composition and bonding structure of the residues and an understanding of their interactions with CO2 mixtures at both elevated and ambient pressures. In addition, the phase state (e.g., number of phases, supercritical vs. subcritical) of the cleaning mixture may affect the efficacy of residue removal.
Supercritical fluids have densities approaching those of liquids and viscosities and diffusivities approaching those of gases. However, supercritical CO2 has little solvating power for PR and most inorganic materials. Thus, the addition of modifiers is required to enhance residue removal. In our work, tetramethylammonium hydroxide (TMAH) was chosen as the base additive (<5 wt percent in CO2) in the cosolvent mixture due to its ability to attack the fluorinated PR crust and to the presence of alkyl groups, which contribute to enhanced solubility in CO2. In addition, TMAH attacks Si-Si, Si-O, and perhaps Si-C bonds. Due to the polar nature of TMAH, its solubility in CO2 is expected to be small. Because methanol and other low-molecular weight alcohols have significant solubility in CO2, 25 percent TMAH in methanol solution was chosen as the cosolvent. Due to the acidic nature of CO2 gas, it participates in acid-base reactions with TMAH. We performed a comprehensive study of the reaction products and the effect of these reaction products on PR residue removal. In these studies, industrial samples containing residues from the plasma etching of a low-k film were studied. The degree of film and PR residue removal was evaluated by x-ray photoelectron spectroscopy (XPS). Additional information was also obtained from infrared spectroscopy (IR) and scanning electron microscopy (SEM).
Our results demonstrated that tetramethylammonium carbonate, formed from the reaction of TMAH with CO2, has a low solubility in methanol and hence, precipitates and deposits on the walls of the tubing and reactor. As a result, no residue removal was observed. However, the carbonate salt of TMAH was soluble when water (one part per four parts of the TMAH solution) was added to the solution. The addition of water thus ensured that the salt was soluble in the methanol-water mixture and allowed some of the mixture to partition into the CO2 phase. However, with an excess of CO2, the carbonate salt is converted to the bicarbonate salt. Because the bicarbonate salt has higher solubility in methanol than the carbonate salt, complete residue removal was obtained without intentional water addition.
The effect of time, temperature, and cosolvent concentration on the efficacy of residue removal also was probed. At a pressure of 20 MPa and temperature of 70°C, a treatment time of 25 minutes was required for complete removal of the residue.
The investigation of the phase behavior of TMAH and its bicarbonate salt with and without water addition in CO2 is being performed in a variable volume high-pressure cell. Knowledge of the mixture phase behavior will help to establish an understanding of the residue removal mechanism.
Results/Discussion. The average surface composition of the untreated etch residue was determined by XPS and is shown in Table 1. Surface bonding structure was determined by deconvoluting C1s peak using Gauss-Lorentzian curves, where each peak was identified based on literature values for binding energies. The bonds and the corresponding binding energies are shown in Figure 1. Fluorine is present on the surface of the film mainly due to incorporation by the fluorocarbon etch gas, while nitrogen is present because it was used as a diluent in the plasma atmosphere. The bulk PR, which is a poly(4-hydroxystyrene)-based polymer, is not modified as a result of the plasma etching. This is confirmed by transmission Fourier Transform Infrared Spectroscopy spectra obtained from a blanket PR exposed to the Coral™ etch plasma.
Figure 1. Peak Fit of High Resolution C1s Peak From Untreated Sample
Preliminary experiments were performed in a liquid bath (beaker) to identify solvents that are sufficiently strong to remove the residue either by attacking the crust or undercutting the capping layer. After treatment, the samples were analyzed using XPS. The results are summarized in Table 2. Propylene carbonate, an additive to CO2 reported previously, was not effective in removing the residue even in the liquid phase where the density and concentration of the chemical species are high. A 25 weight percent solution of TMAH in methanol (as purchased from Sigma) removed the PR and etch residue to the detectability of XPS. The crust, which is the highly cross-linked and fluorinated residue on top of the PR, cracked and lifted off the surface during the liquid treatment. A 0.3-M solution of tetrabutylammonium fluoride in acetonitrile also completely removed the residue. Here, the fluoride ion may be attacking the Si-O underneath the residue and lifting the residue from the capping layer.
Because CO2 is an acidic gas, it reacts with TMAH to form carbonate and bicarbonate salts of the tetramethylammonium ion. No more than two of these three components (Equation ) co-exist in appreciable amounts in solution because an acid-base reaction eliminates the third. The initial product of the TMAH reaction with CO2 is tetramethylammonium bicarbonate (Equation ). However, this intermediate product further reacts with excess hydroxide and shifts the equilibrium toward formation of tetramethylammonium carbonate (Equation ) according to LeChatelier's principle. Titration of the as-purchased solution of 25 weight percent TMAH in methanol using 0.1 N HCl revealed that only 86.55 ± 0.66 mole percent of TMA+ exists as the hydroxide and 13.45 ± 2.29 mole percent as the carbonate salt.
|(CH3)4NOH + CO2 (CH3)4NHCO3|||
|(CH3)4NOH + (CH3)4NHCO3 [(CH3)4N]2CO3 + H2O|||
|Mixture||Treatment time (min)||Temperature (°C)||Average Percent Si|
|25 Percent TMAH in MeOH||45||55||22.27|
|0.3M TBAF in Acetonitrile||45||55||25.20|
When excess CO2 is added, the equilibrium between carbonate and bicarbonate is shifted toward bicarbonate formation.
|[(CH3)4N]2CO3 + CO2 + H2O (CH3)4NHCO3|||
To probe the effect of interactions between TMAH and CO2 on PR residue removal efficiency at atmospheric pressure, CO2 was bubbled through 10 mL of the original TMAH solution. During bubbling, the temperature of the solution increased by 9.2°C due to the exothermic nature of the reaction between CO2 and TMAH. Titration of this solution containing CO2 revealed that the amount of TMA+ existing as a carbonate increased to 80.41 ± 0.4 mole percent. Formation of a white precipitate during bubbling indicated the existence of a solubility limit of tetramethylammonium carbonate in methanol. Separation of the precipitate formed during CO2 bubbling and titration of the filtrate revealed that approximately 24.64 ± 2.41 mole percent tetramethylammonium carbonate was soluble in methanol. The higher solubility of tetramethylammonium carbonate relative to the carbonates of alkaline metals in methanol can be explained by the presence of tetraalkyl groups.
Additional bubbling of CO2 into the hydroxide/carbonate mixture results in complete dissolution of the precipitate and formation of a homogeneous transparent solution. These observations are consistent with the regeneration of tetraalkylammonium hydroxide by continuous bubbling of CO2 through aqueous tetramethylammonium carbonate solutions, which leads to the formation of tetramethylammonium hydrogen carbonate (Equation 3).
The formation of tetramethylammonium hydrogen carbonate from tetramethylammonium carbonate is not favored in completely aprotic media. However, in our experiments, the original (as purchased) solution of TMAH contains about 5.5 percent water (as determined by Karl-Fisher titration), which can serve as the required proton source for bicarbonate formation. Indeed, titration of the above transparent solution confirmed that all tetramethylammonium ions exist as tetramethylammonium bicarbonate. The absence of a precipitate indicates high solubility of tetramethylammonium bicarbonate in methanol. These results differ substantially from the results reported previously for analogous bicarbonate salts of alkaline metals.
Similar experiments were performed with 4:1 volumetric mixtures of TMAH in methanol and DIW. The addition of water to a suspension solution of the carbonate/hydroxide mixture led to the complete dissolution of the precipitate.
PR and plasma etch residue removal experiments were performed in closed beakers to prevent the evaporation of methanol and CO2 at 50°C. The beakers contained TMAH in methanol solution, through which CO2 was bubbled for variable times resulting in an increasing percentage of TMA+ existing as the carbonate salt. The cleaning or removal time in all experiments was maintained at 45 minutes. Following PR and etch residue removal, the samples were rinsed with methanol and the surface composition was analyzed by XPS. XPS data provided the chemical composition and bonding configurations of the contaminants remaining on the treated surfaces. All mixtures (with and without the addition of water) resulted in complete PR residue removal. This indicates that there is a sufficient concentration of OH- ions in these mixtures to ensure complete residue removal.
CO2 Cosolvent Studies. A matrix of experiments was conducted with combinations of the components as cosolvents in CO2. The purpose of this set of experiments was to identify the role of each component in the mixture and results are summarized in Table 3. Clearly, the addition of both water and TMAH is necessary to obtain acceptable residue removal. When TMAH solution is introduced into CO2, reactions , , and  result in the formation of a carbonate, followed by a bicarbonate salt. Because the carbonate is less soluble in methanol, the salt precipitates and deposits on the walls of the tubing and the view cell. Methanol preferentially dissolves in CO2. The addition of water thus ensures that the salt is soluble in the methanol-water mixture and some of it partitions into the CO2 phase. Titration of the liquid collected at the entrance of the view cell and at the exit demonstrated that the TMAH completely reacted and formed a bicarbonate salt. Thus, the post-etch residue is exposed to a biphasic mixture containing TMA bicarbonate in a semi-aqueous solvent. In the absence of methanol, aqueous TMAH solution is relatively insoluble in CO2; therefore, effective cleaning is not achieved.
|Cosolvent||Average Percent Si|
|4:1 volumetric mixture of Methanol and DIW||0|
|25 Percent TMAH in Methanol||0|
|4:1 volumetric mixture of 25 percent TMAH in Methanol and DIW||26.49|
|TMA bicarbonate (from bubbling CO2 through 25 Percent TMAH in Methanol)||19.03|
The PR used for the current study is poly(4-hydroxy styrene) based and as such, dissolves in basic solutions. The dissolution mechanism of such polymers is deprotonation by OH- ions and dissolution of the resulting phenoxide chain in water or other polar solvent. In the present case, the addition of water enhances dissociation of the salt resulting in a significant concentration of OH- ions. Using pKa1 and pKa2 of H2CO3, one can estimate the concentration of OH- ions in aqueous salt solutions. For example, the concentration of OH- ions in a TMAHCO3 solution in water is approximately 5x10-5M at 70°C. These OH- ions also attack Si-O bonds in the capping layer. Because the cosolvent is the reactive component for film and residue removal, there is a minimum concentration of cosolvent needed to obtain complete cleaning in 45 minutes. One series of experiments was conducted in which the cosolvent flow rate was varied from 2 mL/hr to 24 mL/hr. This corresponds to approximately 1 to 12 percent by weight in the final mixture, including CO2. From Figure 2, it is evident that at least 6 percent by weight of cosolvent is required to obtain effective cleaning performance. Under these conditions, more than one phase exists, as the "cloudy" nature of the fluid suggests.
Figure 2. Effect of Cosolvent Flow Rate on the Extent of Residue Removal (Treatment for 45 Minutes at 3,000 psi and 70°C Using TMAH in Methanol + DIW as Cosolvent)
Mechanistic Considerations. The top surface of the PR is fluorinated and is highly crosslinked due to energetic particle bombardment during the low-k etching process; therefore, it is difficult to dissolve this material in CO2 or other more aggressive solvents. The mechanism-controlling residue removal appears to be dissolution of the underlying uncross-linked polymer in the CO2/cosolvent mixture and undercutting of the SiO2 capping layer by TMAH mixtures. A schematic of the mechanism is shown in Figure 3. Experiments performed in a liquid solution indicate that cosolvent in the liquid phase can etch the plasma-enhanced chemical vapor-deposited SiO2 capping layer at a rate of 13 nm/hr. When the sample is exposed to the two-phase mixture, small droplets of liquid containing TMA bicarbonate diffuse into the film along with CO2, which aids in film swelling and in diffusional transport of species to and from the film. Some of the "soft" resist film beneath the crust is dissolved as indicated by SEM studies of the wafer after exposure to the mixture. The OH- ions, generated due to dissociation of the salt, can slowly attack Si-Si and Si-O bonds between the capping layer and the PR layer. Subsequently, the loosened crust can be removed by pulsing the pressure between 3,000 and 1,200 psi. When incomplete film/residue removal occurs, a several-second postclean rinse in DIW or methanol effectively removes the remaining film.
Figure 3. Schematic Representation of the Mechanism of Etch Residue Removal by High-Pressure Biphasic CO2-Based Mixtures
Role of CO2. The mixture displaying effective cleaning is not a single-phase fluid; accordingly, it behaves differently from a liquid. In the present case, the liquid is in the form of droplets dispersed in CO2; the composition of these droplets and the concentration of the cosolvents in the CO2-rich phase are currently under investigation. The role of CO2 is to provide efficient transport of these droplets of active chemical species to the surface to be cleaned. This role is especially critical as features become smaller, and the surface tension and capillary forces limit transport of liquids into and out of small features and hence, inhibit the efficient removal of residues from the sidewalls and trench bottom. Angle Resolved-XPS (AR-XPS) was used to analyze the sidewalls of the features. AR-XPS uses differential shading of the sidewalls to detect the sidewall composition. Normally, the XPS measurements are performed with the angle between the normal and the detector at 45°. At higher takeoff angles, more photoelectrons will be detected from the sidewall. No fluorine or nitrogen was detected even at 75°C, confirming that most, if not all, of the sidewall residue is removed. The CO2 also swells the film and thus increases film stress, providing a driving force for the film to crack and/or peel from the surface. Indeed, extensive cracking of the film has been observed in these studies.
Phase-Behavior Studies. The above results indicate that knowledge of the phase
behavior could assist in the formulation of CO2 mixtures containing additives
to achieve greater cleaning performance. This was one incentive for investigating
the phase behavior of these systems with and without water addition. All phase-behavior
studies were performed in a specially designed variable volume cell. A sapphire
window was incorporated to allow visual observation of the phases in equilibrium
with a charge-coupled device (CCD) camera mounted on a borescope. The borescope
and video camera not only allowed safe observation of phase equilibrium, but
also provided a significant magnification of the viewable area.
The entire cell was placed in thermostatted air bath. After thermal equilibrium was attained, the contents of the cell were compressed into a one-phase region by moving the piston forward. The piston in the cell was then slowly retracted until a second phase appeared. The pressure at which a liquid phase appeared (dew point) and a vapor phase appeared (bubble point) defined the phase behavior of the mixture at a particular temperature. In our case, bubble points were measured with varying amounts of liquid cosolvent in the cell. The measurement was repeated at least three times at each data point. In all cases, the composition of the predominant phase equals the overall mixture composition because the mass present in the second phase is negligible. Once the data point was obtained, a known mass of CO2 was added to the cell, and the procedure was repeated at the new concentration.
Figure 4. Pressure-Composition Isotherm for CO2-TMAH System at 70°C
Figure 4 shows the pressure-composition isotherms for the CO2-TMAH system. As mentioned above, the reaction between the acidic CO2 and the strong base TMAH should be considered during the study of the system-phase behavior. Initial addition of CO2 to the TMAH at elevated pressure results in a sharp increase in solution temperature, with a rapid return to the initial temperature. This sharp increase in solution temperature indicates the exothermic nature of the TMAH-CO2 reaction.
The fact that subsequent addition of CO2 did not affect solution temperature and that the solution became clear (no precipitate) implies that the neutralization reaction was complete after the first CO2 addition, and that tetramethylammonium bicarbonate was formed. The x-axis in Figure 4 shows the "actual" CO2 composition in the cell (i.e., moles of CO2 remaining after subtracting the amount of CO2 consumed by reaction with the base). A sharp increase in bubble point pressure was observed with the addition of more than 30 mole percent CO2. In the presence of water, the required pressure to obtain a single phase is higher.
Figure 5. Pressure-Composition Isotherm for CO2-Tetramethylammonium Bicarbonate (Formed by Bubbling CO2 Through TMAH Solution ex situ) System at 70°C
Figure 5 shows the pressure-composition isotherms of tetramethylammonium bicarbonate formed by bubbling CO2 through the original TMAH solution ex situ. The curves are similar to those in Figure 4. This supports our assumption that the initial addition of CO2 to TMAH converts it to its bicarbonate salt.
It is interesting to note that the system tetramethylammonium bicarbonate CO2 provides more reproducible results than does the system TMAH-CO2. This means that the in situ neutralization reaction between TMAH and CO2 has an impact on the mixtures phase behavior, which could be attributed to the kinetic and thermodynamic (heat generated) aspects of the reaction. The mixture composition (12 percent TMAH/methanol/water mixture, 88 percent CO2) that provided efficient residue removal occurs in the biphasic region.
Compositions of high-pressure vapor and liquid methanol phases in equilibrium with CO2 mixtures will be determined. These phase behavior studies will include the development of a theoretical model to assist in the prediction of experimental results.
To allow for more fundamental studies of the residue removal rate as a function of process variables (concentration, mixture composition, temperature, and pressure), model residue films will be formulated. These films will be deposited in a fluorocarbon-based plasma and will be designed to have a similar surface composition and bonding structure to those of residue films formed during IC processing.
Journal Articles on this Report : 2 Displayed | Download in RIS Format
|Other project views:||All 14 publications||8 publications in selected types||All 7 journal articles|
||Levitin G, Myneni S, Hess DW. Reactions between CO2 and tetramethylammonium hydroxide in cleaning solutions. Electrochemical and Solid State Letters 2003;6(8):G101-G104.||
||Levitin G, Myneni S, Hess DW. Post plasma etch residue removal using CO2-TMAHCO(3) mixtures: Comparison of single-phase and two-phase mixtures. Journal of the Electrochemical Society 2004;151(6):G380-G386||