Aqueous Antireflective Coatings for Photoresist Compositions

A suspension of 3-aminophenol (47 g, 0.43 mol) in acetone (100 ml) was chilled to 0.degree. C. To this, a solution of methacrylic anhydride (66.2 g, 0.43 mol) in acetone (150 ml) was added dropwise. A white precipitate formed and the reaction mixture was stirred for 1 hour. The resultant precipitate was poured into crushed ice and the solid product collected by filtration, and recrystallized (1:1 EtOH (ethanol):H.sub.2 O). The melting point was 172.degree.-173.degree. C. and the yield was 86.6%.

A solution of N-(3-hydroxyphenylmethacrylamide) (5.32 g, 0.03 mol) from Example 1, in .gamma. butyrolactone (40 ml), was prepared by heating the mixture to 65.degree. C. while stirring. On complete dissolution, the solution was degassed by vigorously bubbling argon, via an inlet needle in a sealed rubber septum, through the solution for up to 2 hours. Previously degassed diethylene glycol monovinyl ether was then injected into the reaction mixture. After an additional 20 minutes, an aliquot from a solution of AIBN (azo-bis-isobutyro-nitrile) (49 mg, 1 mol % total monomers) in .gamma. butyrolactone (1 ml) was also injected. In total, 2 aliquots were added at intervals of 6 hours. Both inlet and outlet needles were then removed and the sealed vessel was allowed to stir at 65.degree. C. for 18 hours. The resultant solution was then precipitated into ethyl acetate (300 ml) as nonsolvent, and the solid collected by filtration, redissolved in ethanol (50 ml) and reprecipitated into ethyl acetate. Purification was further carried out by successive reprecipitations of polymer solution until no monomer was observed by TLC. The polymer formed as a white powder and was dried under reduced pressure overnight.

The diazonium salt of sulfanilic acid was prepared as follows. To a solution of tetramethyl ammonium hydroxide in water (25% solids) (3.6 ml, 0.01 mol) was added sulfanilic acid (1.73 g, 0.01 mol), followed by isobutyl nitrite (1.5 ml. 0.012 mol) and the resultant suspension temperature maintained below 10.degree. C. A solution of HCl (37.8 wt % in water) (1.62 ml, 0.02 mol) was added to water (5 ml) and the solution slowly added to the reaction mixture, forming the diazonium salt.

The copolymer from Example 2 (3.09 g, 0.01 mol equiv. phenylacrylamide) was dissolved in ethanol (30 ml), then on addition of tetramethyl ammonium hydroxide in water (25 wt % solids) (7.2 ml, 0.02 mol), the immediate precipitate that formed was gradually redissolved. Additional water (5 ml) and ethanol (10 ml) were then added and the solution chilled to 5.degree. C. The diazonium salt prepared from Example 3 was then added dropwise to the polymer solution, which immediately formed a suspension. Additional water was added throughout the course of the addition (20 ml total) and the mixture allowed to stir. Gradually the suspension was converted to an orange solution (4 hours) which was filtered and the subsequent solution precipitated into 2-propanol (700 ml) as nonsolvent. The product was collected by filtration and purified by reprecipitation.

11.96 g (0.0675 mol) of N-(3-hydroxyphenylmethacrylamide) from Example 1 was dissolved in .gamma. butyrolactone (75 ml) and DMF (dimethylformamide) (75 ml), The mixture was heated to 65.degree. C. while stirring. The solution was degassed by vigorously bubbling argon, via an inlet needle in sealed rubber septum, through the solution for 1 hour. 4.41 ml (0.0225 mol) of N-(hydroxymethyl)acrylamide and 6.48 ml (0.06 mol) methyl methacrylate were injected into the solution. The solution was degassed for 0.5 hour. An aliquot from a solution of AIBN (0.251 g, 1.5 mmol, 1 mol % total monomer) in .gamma. butyrolactone (2 ml) was injected and degassed for 0.5 hour. In total, 2 aliquots were added at intervals of 4 hours. Both the inlet and outlet needles were removed and the sealed vessel was stirred at 65.degree. C. for 19 hours.

4.68 g (0.03375 mol) of 4-aminobenzoic acid were dissolved in 6.68 ml (0.084375 mol) of concentrated hydrochloric acid and 70 ml of water, contained in a 250 ml three-neck round-bottom flask. A thermometer was placed in the solution and the flask immersed in a bath of crushed ice, cooled until the temperature of the solution fell below 2.degree. C. The solution contained a white suspension. Diazotization was done by the addition of 4.19 ml (0.0338 mol) tert-butyl nitrite. The diazonium solution was stirred in ice-water for about 1 hour. The suspension changed to a yellow solution.

80 ml (0.075 mol) of the polymer solution from Example 5 was placed in a 500 ml round bottomed flask. To this was added 320 ml of DMF and a solution of (25% in water) 42.52 ml (0.1185 mol) tetramethylammonium hydroxide while stirring. The solution was cooled in an ice-water bath to 10.degree. C. and the cold diazonium salt solution was added. The solution changed to orange color. The mixture was allowed to stir about 3 hours and warmed to room temperature. The resultant solution was precipitated into excess isopropanol, and the polymer collected by filtration and vacuum dried.

Several 4" wafers were coated with AZ.RTM. 7805 (available from Hoechst Celanese Corporation, 70 Meister Ave., Somerville, N.J. 08876) and baked using a temperature of 90.degree. C. for 90 seconds to give thickness from 0.5 .mu.m (micrometers) to 0.9 .mu.m (micrometers). These wafers were imagewise exposed with a NIKON.RTM. 0.54 NA i-line stepper using a clear quartz as the reticle and a program that directed the stepper to print a 11.times.11 exposure matrix with dose increments of 2 mJ/cm.sup.2. The exposed wafers were baked at 110.degree. C. for 60 seconds and puddle developed with AZ.RTM. 300 MIF developer (available from Hoechst Celanese Corporation, 70 Meister Ave., Somerville, N.J. 08876) for 35 seconds. The minimum dose required to clear the film was plotted as a function of the corresponding resist thickness, a sinusoidal curve was obtained, called the swing curve. The % Swing ratio was calculated empirically by the following equation:

where Emax and Emin corresponding to the dose-to-clear the resist film thickness at the maximum and minimum energy on a swing curve. The smaller the value of % Swing Ratio the lower is the impact of reflectivity and the better is linewidth control over reflective substrate or topography.

The % Swing Ratio for AZ.RTM. 7805 was 34.8%.

The polymer from Example 4 was dissolved in water to give a 5 weight % solution. The polymer solution was spin coated on several 4" silicon wafers, and baked on a hot plate at 220.degree. C. for 60 seconds to give a thickness of 0.121 micron. The wafers were then coated with with AZ.RTM. 7805 (available from Hoechst Celanese Corporation, 70 Meister Ave., Somerville, N.J. 08876) and baked using a temperature of 90.degree. C. for 90 seconds to give thickness from 0.5 .mu.m (micrometers) to 0.9 .mu.m (micrometers). These wafers were imagewise exposed with a NIKON.RTM. 0.54 NA i-line stepper using a clear quartz as the reticle and a program that directed the stepper to print a 11.times.11 exposure matrix with dose increments of 2 mJ/cm.sup.2. The exposed wafers were baked at 110.degree. C. for 60 seconds and puddle developed with AZ.RTM. 300 MIF developer for 35 seconds. The minimum dose required to clear the film was plotted as a function of the corresponding resist thickness, and a sinusoidal curve was obtained, called the swing curve. The % Swing Ratio was calculated as in Example 8.

The % Swing Ratio for AZ.RTM. 7805 with antireflective polymer coating of this Example was 22%, showing a reduction in the % Swing Ratio from the resist without the antireflective coating.

The polymer from Example 7 was dissolved in water to give a 5 weight % solution. The polymer solution was spin coated on several 4" silicon wafer, and baked on a hot plate at 220.degree. C. for 60 seconds to give a thickness of 0.198 micron. The wafers were then coated with with AZ.RTM. 7805 (available from Hoechst Celanese Corporation, 70 Meister Ave., Somerville, N.J. 08876) and baked using a temperature of 90.degree. C for 90 seconds to give thickness from 0.5 .mu.m (micrometers) to 0.9 .mu.m (micrometers). These wafers were imagewise exposed with a NIKON.RTM. 0.54 NA i-line stepper using a clear quartz as the reticle and a program that directed the stepper to print a 11.times.11 exposure matrix with dose increments of 2 mJ/cm.sup.2. The exposed wafers were baked at 110.degree. C. for 60 seconds and puddle developed with AZ.RTM. 300 MIF developer for 35 seconds. The minimum dose required to clear the film was plotted as a function of corresponding resist thickness, and a sinusoidal curve was obtained, called the swing curve. The % Swing Ratio was calculated as in Example 8.

The % Swing Ratio for AZ.RTM. 7805 with antireflective polymer coating of this Example was 2.1%, showing a significant reduction in the % Swing Ratio from the resist without the antireflective coating.