Purged gas purification for contamination control of DUV stepper lenses
09/01/2001
Serena Barzaghi, Alberto Pilenga, STMicroelectronics, Agrate Brianza, Milan, Italy
Giorgio Vergani, Sara Guadagnuolo, Saes Getters, Lainate, Milan, Italy
overview
The measurement and control of optical lens contamination by airborne molecular contaminants has not received much attention. To control and enhance light throughput, the universal solution is to purge optical parts with ultra-high-purity gases [1]. But as geometries shrink and wavelengths decrease from 193nm to 157nm, hydrocarbon species control becomes an absolute must, because a very wide range of airborne molecular organic contaminants, especially hydrocarbons, absorb light at the corresponding wavelengths [2-4]. This article addresses a solution.
DUV lithographic process sensitivity toward airborne molecular contamination has been recognized as one of the most difficult challenges of next-generation submicron geometries[5]. Condensation on the lens and the presence of contamination between the lens and the substrate can result in poor optical transmission, not to mention the cost and downtime associated with lens cleaning [6, 7].
 |
Clean dry air (CDA) is commonly used as a purge gas for stepper lens purging and cooling. CDA, for this specific application, has to strictly comply with specifications given by the stepper manufacturer for humidity, hydrocarbons, and some inorganic ion contents. For future-generation steppers, 157nm in particular, N2 gas is expected to replace CDA as the main purge gas. Even if high-purity N2 contains, in principle, fewer impurities than CDA, it is thought that local purification of N2 purge will still be required.
 |
At STMicroelectronics Central R&D Fab in Agrate Brianza, Milan, Italy, the common practice has always been to use synthetic air from bottles for steppers that specify CDA as a purge gas. At the end of 1999, during a fab expansion, a new generation of DUV steppers for 0.18µm geometries and smaller were installed. For these machines, the CDA flow necessary was much higher than previously used, making bottled air inconvenient. A decision was made to investigate the use of centralized pneumatic CDA coupled with point-of-use (POU) purifiers to ensure the required flow and quality specifications. SAES Getters and the facility technology group collaborated to study and implement purifiers capable of guaranteeing the appropriate specifications.
Requirements
A lithography equipment manufacturer has recommended that the purge gas CDA at the stepper lens barrel meet the following specifications:
- dew point: -60°C,
- temperature: 19-26°C,
- particle content: <3500 particles (f µm)/m2, and
- contaminant concentration limits shown in Table 1.
CDA purging is in fact limited to the lens compartment with a ~2slpm required flow for each stepper.
 |
From Sept. 1999 to Jan. 2000, analytical measurements were performed to evaluate the quality of the fab's centralized CDA. Analytical sampling was performed with absorbing traps in the line after the compressor. Traps collected sample gas for 24-hr intervals, so the reported results represent an average value over 24 hr. Analyses were performed by an external third party, SINAL, an accredited analytical laboratory. Sampling and measurements were performed following registered standard procedures. Measurements of species of interest were performed with the standard analytical techniques. C6-C30 organic compounds were analyzed following two different analytical methodologies. The first samples were collected in tenax traps and analyzed with a gas chromatograph coupled with a mass spectrometer (GC/MS). This was done to qualitatively identify contaminant species. Sample volume was thermally desorbed into the analyzer.
 |
Exact quantification of C6-C30 organic compounds was performed with a second analytical methodology (M.U. 565) using a flame ionization detector gas chromatograph (0.5µg/m2 limit of detection). Sample volumes were obtained desorbing active carbon traps with carbon sulfide. Determination of sulfuric acid (SO42-) and nitric acid (NO3-) concentrations was performed according to the Niosh 7903 methodology collecting gas contaminants on silica gel traps and analyzing samples by ion chromatography (detection limit of 1µg/Nm2). Sample volumes were obtained by desorbing traps with distilled water. Further, ammonia (NH4+) was determined by collecting sample gas volumes with silica gel traps and quantified by spectrographic UV-VIS methods (Niosh 6015) having a detection limit of 1µg/m2. Sample volumes were obtained desorbing traps with distilled water.
For a correct evaluation of the data presented, note that detection limits reported for the methods mentioned above refer solely to the analytical equipment. These, in fact, do not take into account the intrinsic sampling measurement errors, which are related to sample collection procedures. Sample collection analytical errors could, in principle, contribute to increasing the overall analytical methodology detection limits; nonetheless, in the specific case, low concentration measurements were found to be acceptable and within the desired specifications.
The measurements indicated that CDA compressed with an on-site compressor, filtered with 0.01µm filters, then dried to a dew point of -70°C, was in accordance with the required dew point specifications. A -70°C dew point, in fact, corresponds to a moisture concentration of 2.55ppm and -60°C corresponds to 10.6ppm. However, the quality of CDA in terms of organic and inorganic pollutants was found to be unacceptable for the lens purging and clearly variable in time. Data analysis (see Table 2) shows that the average concentration value relative to C6-C30 organic compounds was quite high. Also very high was the standard deviation of the collected data, showing the intrinsic variability of the contaminant species. Furthermore, actual chromatograms show that the lighter hydrocarbon species give the undesired contamination. Attention must also be paid to ion contaminants.
POU purifier decision
The team decided to study a customized purifier that could guarantee the stepper manufacturer's CDA specifications. Space management and time constraints did not allow studying a centralized system for the entire lithography area, though this would have had economic benefits and space advantages if designed at the outset.
On the basis of the cleanroom layout, the best compromise was a 10slpm POU purifier system connected to the supply gas of four steppers.
The use of a POU purification system can guarantee a reproducible and consistent gas quality delivery. This is a cost-effective solution to an intrinsically variable source gas where the quality is naturally fluctuating.
Purifier development
To meet the goal of removing heavy hydrocarbons (C6 through C30 organic compounds) from the CDA for purging the lens compartment, a convenient converter material was identified, developed, and tested for efficiency. The converter material converts hydrocarbons (HCs) to H2O and CO2. Gas specifications related to product species H2O and CO2 were in the range of several ppm or negligible. The identified converter material met specifications. Even in a worst case scenario where concentration estimates were based on the conversion of hydrocarbons at maximum inlet levels, H2O product concentrations were in the range of 5-10% of the maximum acceptable specification level.
In general, if HC inlet expectations do not make it possible to meet moisture specifications reliably, it is possible to add absorber material at the outlet of the converter material. The additional absorber material will trap the converted species, H2O and CO2, and deliver a gas consistently up to specifications. At ST, the converter material solution was sufficient to guarantee high-quality CDA needed to purge the stepper lens apparatus.
After initial testing, a vessel containing the identified material was dimensioned to maximize efficiency in the specific flow-rate range and at the specified impurity load conditions. It was then installed in ST for beta site testing and qualification. The test unit was rated for a maximum of 10slpm and dimensioned to convert at least 1.5mg/m2 of inlet C6 through C30 HCs. A qualification test phase was scheduled to validate the product for the specific application and eventually extend the flow-rate capability of the system.
 Figure 1. Comparison of purifier inlet and outlet C6-C30 organic compound concentration values. |
The purifier material is operated at high temperature, ~350-400°C. One technical concern regarding delivery of gas to the stepper lens compartment was that the outlet gas be at room temperature at the stepper machine inlet. This concern was valid given the fact that a higher-temperature gas would modify the expected refractive index of the lens barrel volume. The outlet gas temperature was monitored at the maximum specified flow rate meeting the requirement, between 19-26°C. To assure the purifier's most efficient conversion conditions, two thermocouples monitored the column temperature at the same point (two for redundancy) and a control system with customer-specified alarm and pre-alarm set points on the temperature measurement monitored temperature and thermocouple reliability.
 Figure 2. Ion contaminant species concentration values at purifier inlet. |
The converter has a virtually indefinite lifetime; conversion of the impurities occurs without the consumption of the media. This is not true for the volume of optional absorber material for H2O and CO2 removal capability. However, such units have commonly been requested for bulk applications. High flow justifies the use of regenerable units with a double absorber column. In this case, while one column is in an operating, purifying condition, the other is being regenerated off-line, getting ready to take over the next purification cycle. This happens in an automatic mode on a cycle time that can be tuned both by customer requests and total inlet gas impurity load.
 Figure 3. Ion contaminant concentration values at purifier outlet. |
CDA can be used according to manufacturer's specifications in two different parts of the stepper machine — the lens compartment and the illuminator compartment. Gas specifications, the maximum acceptable impurity concentrations, in both these areas are compatible, making the same purifier system appropriate for each compartment. Flow rates to the lens are commonly much lower than to the illuminator, the latter compartment requiring several normal cubic meters/hour (Nm2/hr). In general, the dimensions of the converter and absorber unit can be scaled to meet bulk delivery requirements.
Test results
The prototype purifier was installed in ST's facility for validation at the beginning of 2000. Analyses were performed to compare purifier inlet and outlet airborne molecular contamination (AMC). Scope of the test program was to confirm purifier efficiency on site and determine background concentration levels.
Tables 3 and 4 report the data analysis performed during the purifier beta test to compare inlet and outlet purity levels. All contaminant concentrations at the purifier outlet are consistently below the required specifications and thus prove the purifier's excellent efficiency in removing the fluctuations and high inlet contamination present in the inlet gas stream. The concentration species measurements reported in the tables are individually represented in Figures 1, 2, and 3 to show the beneficial effect of the purifier in-line.
Figure 1 represents the totality of C6-C30 organic compound concentration measurements obtained before and after the purifier, and also prior to its installation. Purifier inlet (Fig. 2) and outlet values (Fig. 3) for the other specified ionic impurities also indicate the efficiency of the purifying media.
The figures demonstrate purifier efficiency in consistently meeting stringent gas quality specifications independently to the gas source's natural variability, indicating the success of the test program.
Conclusion
STMicroelectronics Agrate R&D facility technology group was driven to look for a reliable and technically viable solution for the purge gas (CDA) flow and quality optimization of recently installed new-generation 0.18µm geometry stepper machines. In a joint effort between STMicroelectronics and SAES Getters, a purifier for CDA was developed capable of guaranteeing the demanding specs received from the stepper manufacturer.
During the development, a converter material was identified and positively tested for efficiency and reliability. A complete set of purifiers is now installed to serve such tools at the ST facility, and the cooperation process has served to develop a new product to meet industry needs. This purification approach can also be applied to N2 purged systems other than CDA systems.
References
- C. Wagner, W. Kaiser, J. Mulkens, D.G. Flagello "The Technical Considerations of Extending Optical Lithography," Solid State Technology, pp. 97-108, Sept. 2000.
- "Airborne Molecular Contamination Control for DUV Lithography," Cleanroom Technology, pp. 31-33, June 2000.
- A.J. Dallas et al., "Protecting DUV Process and Optimizing Optical Transmission," Metrology, Inspection and Process Control for Microlithography Conference, SPIE Proceedings 2000, pp. 14-26.
- J. Mulkens et al., "Challenges and Opportunities for 157nm Mask Technology," Proceedings of SPIE, Vol. 3873, pp. 372-384.
- The International Technology Roadmap for Semiconductors, SIA, 2000.
- R.R. Kunz, V. Liberman, D.K. Downs, "Photo-Induced Organic Contamination of Lithographic Optics," Microlithography World, pp. 2-8, Winter 2000.
- R.R. Kunz, V. Liberman, D.K. Downs, "Experimentation and Modeling of Organic Photo-Contamination on Lithographic Optics," Optical Microlithography XIII, Proceedings of SPIE, Vol. 4000, pp. 474-487, 2000.
- Serena Barzaghi received her degree in chemical engineering from the Polytechnic University of Milan, Italy. In 1988, she joined STMicroelectronics, where she is a project engineer for process tools installation in Italy. STMicroelectronics, Via C. Olivetti, 2, I-20041 Agrate Brianza, Italy; ph 39/039-603-1, fax 39/039-603-5700.
Alberto Pilenga joined STMicroelectronics in 1989 as a lithography engineer. Since 1996, he has worked on pure gas distribution systems as R&D manager in Italy.
Giorgio Vergani graduated with a degree in industrial chemistry at the University of Milan. He joined SAES Getters SpA in 1995 to develop new materials for gas purification, and has contributed to the development of a number of new products for ultra-high gas purity. He is currently R&D project coordinator for gas purification and related fields.
Sara Guadagnuolo graduated in physics from the University of Milan in 1994. She joined SAES Getters SpA in 1996, and works in the corporate R&D labs in the field of gas purification and analysis. She has contributed to the development of new equipment in both analysis and purification.