Method Measures Lithography Gas Purity to ppt Levels

Optics are lithography's main driver, fueling progress toward smaller features. As CDs continue to shrink, molecular contamination becomes a major problem for DUV lithography in the area of purge gases. The SIA roadmap states that, for linewidths below 0.10 µm, the total contamination for lithography processes must not exceed 100 ppt. In the case of 157 nm processes, hydrocarbon limits have been set at 10 ppt. Contamination must be detected well before lens damage occurs through high-energy gas phase reactions depositing carbon, SiO₂, and other inorganic salt monolayers on the optical surfaces. These strongly adhesive coatings on CaF₂ surfaces can alter the lenses' optical properties, resulting in transmission degradation. This troublesome situation is intensified at 193 and 157 nm wavelengths.

Contaminants primarily originate from elastomeric materials — such as O-rings that are part of the lithography platform — that outgas hydrocarbons within the tool, and from the purge gases used during processing. Gases with ppt purity levels are needed to remove these contaminants, requiring quantitative measurements for hydrocarbons at ppb and low ppt levels, to monitor contaminants in the fragile lithography environment. Appropriate analytical capabilities are necessary to meet these requirements.

Direct analysis of hydrocarbon impurities using a gas chromatograph (GC) with flame ionization detector (FID) is capable of yielding a 3 ppb detection limit. For ppt levels, Aeronex (San Diego) has developed an indirect measuring method that uses a thermal desorption tube (TDT) with molecular sieves to concentrate volatile hydrocarbons from sample gases, enabling a <25 ppt detection limit. If sample collection optimization is used, a detection capability of 1 ppt is deemed attainable. In this case, because most commercial hydrocarbon gas standards are only obtainable at sub-ppm levels, a double dilution process was used to create sub-ppt standards. The system achieved accurate dilution of a known gas standard down to a 1:100,000 ratio.

	Accurate dilution of a known gas standard down to 1:100,000 ratio (<ppt standards) can be achieved by using a double dilution system. (Source: Aeronex)

A single dilution process is applied to reduce a known hydrocarbon concentration from a ppm to a ppb level (1:1000), by mixing it with ultrahigh-purity (UHP) air prior to mixing (Figure). The sought-after hydrocarbon concentrations are obtained by adding a calculated flow rate of each gas. MFC1 and MFC2 control the flow rate of purified air and hydrocarbon standard, respectively. By applying a second dilution process, the hydrocarbon standard from the first dilution is further diluted with UHP air using MFC3 and MFC4, respectively. The excess gas from the first dilution is vented through a back-pressure regulator (BPR1), thus ensuring that a constant operating pressure (P5) is maintained.

When the sample reaches ppt levels, it is then sent through a TDT to capture hydrocarbons. This collection process is continued for a specific time at a constant flow rate and pressure. Again, the excess gas is vented through BPR2. After sample collection, the TDT is closed (valves V7 and V8) and heated to 300°C. This procedure revolatizes captured hydrocarbons. The gas then goes to the helium carrier gas in a GC, using an FID to carry out the measurement of organic compounds.

Light (butane) and heavy (toluene) molecular weight hydrocarbons were investigated for measurements ranging from ppm to ppt. The ppb calibration data was generated by diluting known, established standards of butane and toluene. These samples were then injected into the GC via a six-port gas sampling valve with a 1 mL sample volume at 27 psig. Calibration points between 3 and 75 ppb were graphed and a least-squares fit formula was then derived to determine ppb butane and toluene levels from the peak area. A detectable signal was correlated with the lower detection limits.

Ppt calibration was carried out by sending known concentrations of hydrocarbons through the TDT. Butane was sent through the TDT at a 200 sccm flow rate and 30 psig pressure for 100 min. Toluene was collected at 4 slm and 30 psig pressure for 5 min. The TDT was then heated to 300°C to revolatize the hydrocarbon and passed to the GC for hydrocarbon measurement. TDT testing was done at 0, 25, 50, 95 and 500 ppt of butane. For toluene, the TDT was collected with 25, 30, 75, 100 and 500 ppt. It was determined that the TDT can measure butane and toluene to <25 ppt.

Finally, the lens gas purifier's efficiency was TDT-tested for butane and toluene removal individually. The gas (60 ppb butane and 60 ppb toluene) flowed through the TDT. From the initial 60 ppb of butane and toluene upstream of the purifier, no butane or toluene was detected downstream via the concentrated method, showing removal to <25 ppt.

For additional information on inspection, measurement and test, go to www.semiconductor.net/imt