Karl Dietz 0000-00-00 00:00:00
The Tech Talk column has addressed dry film photoresist development issues over the years (see Ref. 1-8) but it has not specifically looked into how to measure and control the chemistry, nor has it explained the significance of controlling parameters such as “active carbonate,” “total carbonate,” and “resist loading.” So, here is the long overdue discussion on developer chemistry control. The Basics of Aqueous Development When sodium carbonate is dissolved in water, it comes to equilibrium forming sodium hydroxide and sodium bicarbonate (Figure 1, Equation 1) , creating a basic solution. The reaction does not proceed very far to the right because hydroxide (OH-) competes more successfully than carbonate (CO3 2-) for the hydrogen ions. The pH of the fresh carbonate solution will vary with the amount of total carbonate added to the water, higher carbonate concentrations resulting in higher pH. The binders in the resist, and occasionally the monomers, contain carboxylic acid groups for aqueous processing. When the unexposed resist enters the developer and comes in contact with the carbonate solution, the base and carboxylic acid react to form the sodium carboxylate salt (Figure 1, Equation 2). Water molecules cluster around the sodium, solvating the carboxylate salt. This action swells the binder and facilitates solubilization. The use of sprays in a conveyorized developer assists the speed of development by washing away the top layer of dissolved binders and monomers and brings fresh carbonate solution to the underlying layer of unexposed resist. During development of unexposed resist, sodium hydroxide is consumed in the carboxylate salt formation. This drives the reaction of Equation 1 to the right, resulting in a depletion of sodium carbonate and a build-up of sodium bicarbonate. Since the pH of the solution goes down as the concentration of the hydroxide goes down, pH makes a convenient indicator of the equilibrium condition of this reaction. Eventually, even though development of unexposed resist is still possible, the rate of development slows down and fresh carbonate needs to be added to maintain the development speed. Sodium carbonate or potassium carbonate (Na2CO3, K2CO3) are also referred to as “active carbonate” since the carbonate drives the development process. The bicarbonate (NaHCO3 or KHCO3) is the reaction product and it is not “active”. “Total carbonate” is the term used for amount or concentration of both, carbonate and bicarbonate. “Resist loading” refers to the volume of dry film resist per volume unit of developer solution. English and metric units are in use. The English unit is typically milft2/ gal whereby the “mil” indicates the thickness of the resist, the ft2 s the area of the developed resist in the one unit of developer solution, and “gal” is the unit for the liquid volume. The metric equivalent is micron-m2/liter. For the discussion of developer chemistry control two more terms need to be introduced and defined: “time-to-clean” and “breakpoint.” Time-to-Clean (Tc ): The time required, in seconds, to remove unpolymerized resist from the substrate. “Removal” is judged by visual inspection, using one of the methods described below. Time-to-clean is an important characteristic of the photoresist because it relates to productivity. Time-to-clean is not an absolute constant for a given resist but can only be stated in the context of defined process variables and equipment features. Breakpoint (BP), also referred to as wash-off point: After a panel enters the development chamber, there will be a point where unexposed resist appears to be removed from the copper surface. Breakpoint is the term used to refer to the location of removal in the development chamber. It is measured from the entrance of the developer chamber and reported as a percentage of the “usable” developer chamber length. For example, if a panel enters the development chamber and the unexposed resist is completely removed at a point half way between the entrance and exit of the chamber the breakpoint will be expressed as BP = 50 percent. The significance of controlling active carbonate, total carbonate, and resist loading Aqueous development needs to be controlled to avoid underdevelopment (late breakpoint) and overdevelopment (early breakpoint). This is achieved by setting the active carbonate concentration and the conveyor speed so that the recommended breakpoint (e.g. 50 percent) is maintained. In addition, resist loading needs to be controlled within certain limits because loading also affects time-toclean and the breakpoint (Figure 2). While the interrelationships shown in Figure 2 hold true for all aqueous resists, the specific data of Figure 2 are only valid for the resist used in this test, and other resists are likely to show slightly shifted curves due to differences in the formulations of the photoresists, e.g. the acid number of the binder. High resist loading can have another undesirable effect: the rinse may not be sufficient to remove the dissolved resist completely. Total carbonate needs to be controlled because exceedingly high levels affect the development quality by causing poorly defined resist sidewalls and a resist foot. Analytical methods to determine active carbonate, total carbonate, and resist loading Titration with Hcl to a methyl orange end point (pH 3.2) The most common and simplest method of analysis is to titrate the solution with acid to an end point using methyl orange. This analyzes for both the sodium carbonate and sodium bicarbonate as gtotal carbonate. Na2CO3 + Hcl NaHCO3 + NaCl NaHCO3 + Hcl NaCl + H2O + CO2 This method cannot distinguish between a fresh developing solution with high carbonate - low bicarbonate concentrations and a used developing solution with low carbonate - high bicarbonate concentrations. Titration with Hcl to a phenolphthalein end point (pH 8.2). This determines only the sodium carbonate. The titration stops when all carbonate has been converted to bicarbonate. Na2CO3 + Hcl NaHCO3 + NaCl This method has the advantage of determining the “active” carbonate and gives a good indication of the activity of the developing solution. The method gives no indication of the total carbonate concentration and the possible problems associated with total carbonate concentrations that are out of range. An indication of resist loading is only possible by comparing the active and total carbonate concentrations. Combined analysis method using phenolphthalein & methyl orange end points This method combines both end points in one analysis procedure and enables both “total” and “active” carbonate to be determined. Na2CO3 + Hcl NaHCO3 + NaCl (Phenolphthalein end point) NaHCO3 + Hcl NaCl + H2O + CO2 (Methyl Orange end point0 By calculating the active carbonate as a percentage of the total carbonate it is also possible to obtain a good indication of the resist loading of the solution. The percent active carbonate is inversely proportional to the resist loading. By controlling both the total carbonate concentration and the percent active carbonate it is possible to obtain optimum control of the developing results and resist sidewalls. The total carbonate concentration is primarily determined by the concentration of the solution used for replenishment. The active carbonate concentration is determined by the replenishment rate and how well it matches the amount of dry film developed. It should be controlled by increasing or decreasing the volume of replenishment. Analysis of a fresh developing solution may indicate an active carbonate content significantly lower than 100 percent. This may be due to sodium bicarbonate as an impurity in the sodium carbonate. Some sodium carbonate may also be consumed in softening the make up water. Titration Method Pipette a 10 ml. Sample of developing solution into a conical fl ask and add ca. 100 ml of deionized water. Add 4 drops of phenolphthalein indicator and titrate with 0.1 N hydrochloric acid to a colorless endpoint.( with a developing solution that is highly loaded with resist, the end-point may be purple to blue.) Record this titre as T1 Add 8 drops of methyl orange indicator and continue titration of the solution to a yellow to red end-point. (Do not refill the burette) Record the titre as T2 Calculation Total Carbonate concentration = T2 x 0.53 g/l Na2CO3 This equation is derived as follows: g/l Na2CO3 = T2x normality N of Hcl (0.1) x the formula weight for Na2CO3 (106) /sample volume (10 ml) x 2 (the normality of Na2CO3, i.e. there are 2 sodium ions to be converted to hydrogen ions. (Note: FW for sodium carbonate monohydrate is 124, FW for potassium carbonate is 138) Percent active carbonate = T1x 200/T2 % Active Carbonate Concentration = T1 x 1.06 g/l “active” Na2CO3 Resist loading = (T2/2 – T1) x 3.85 g/l (this is an empirical, best fit formula) Typical control limits are: Total Carbonate concentration: 7.0 - 9.0 g /l ( 0.7- 0.9 % ) Active carbonate concentration: 65 - 77.5 % of total carbonate Resist Loading: 7.5 g/l + 1.5 g/l for standard production < 5 g/l for fine line images ( <100ìm) Figure 3 shows a spreadsheet that can be used to calculate these parameters. Instead of using pH indicator reagents, one can use a pH probe and create a titration curve as shown in Figure 4. The two endpoints are marked by a rapid drop in pH. References 1. Fine Lines in High Yields, Part XIII: “The Balancing Act of Controlling Aqueous Development”, Karl H. Dietz, CircuiTree Magazine, August 1996, pg. 108 2. Fine Lines in High Yields, (Part XXVI): Replenishment Chemistry and Process Control for Aqueous Development, Karl H. Dietz, CircuiTree Magazine, October 1997, pg. 32 3. Fine Lines in High Yields, (Part LXII): Checking the “Break- Point” in Aqueous Development of Dry Film Resist, Karl H. Dietz, CircuiTree Magazine, November 2000, pg. 32 4. Fine Lines in High Yields, (Part LXIV): Developer Maintenance, Karl H. Dietz, CircuiTree Magazine, January 2001, pg. 42 5. Fine Lines in High Yields, (Part LXXXIV): Trends and Observations in Aqueous Development of resists, Karl H. Dietz, CircuiTree Magazine, September 2002, pg. 46 6. Fine Lines in High Yields, (Part LXXXV): Spray Nozzles, Karl H. Dietz, CircuiTree Magazine, October 2002, pg. 46 7. Fine Lines in High Yields, (Part CXXIII) : Developer Rinse, Karl H. Dietz, CircuiTree Magazine, December 2005, pg. 89 8. Fine Lines in High Yields, (Part CLVI) : Defect Problems and Defect Prevention in Development, Karl H. Dietz, CircuiTree Magazine, September 2008, pg. 22 Acknowledgment Valuable contributions to this article by my colleague John Raine are gratefully acknowledged.
Published by CircuiTree. View All Articles.
This page can be found at http://digital.bnpmedia.com/article/Tech+Talk/470042/44520/article.html.