Throughout the semiconductor fabrication process, it is necessary to minimize the amount of contamination that comes into contact with the wafer and the wafer processing equipment. Contamination control has become a major factor in manufacturing yield and profitability in semiconductor processing. Contamination from the environment, factory personnel, processing equipment, and all materials used to fabricate the devices are major concern in the manufacturing environment. Water is used in the rinsing step at the end of almost every cleaning operation. The frequent use of water makes it imperative that contain minimum amounts of potentially harmful contaminants.

© 2007 Arizona Board of Regents for The University of Arizona

Types of contaminants that may exist in pure drinking water which cannot be tolerated in water used for microelectronics include: 1 Dissolved inorganic salts such as sodium and calcium salts. These are dissolved by the water as it flows through pipes, rocks and soil. 2 Dissolved organic compounds from industrial waste or living matter 3 particulate matter such as small silica particles from rock and soil. 4 Microbiological life that sustains itself on other contaminants. Water containing these types of impurities must be purified to obtain to reach the levels in table 1 Contamination of ions causes the largest problem in semiconductors. For example, sodium interferes with the normal operation of semiconductors devices by rapidly drifting through silicon dioxide towards a region with a negative voltage and then gives rise to excessive leakage of current. The process of ion exchange, or deionization was used almost exclusively in water purification prior to the last few years. Ion exchange is the removal of positive and negative ions using activated resins. A typical ion exchange contains the following elements: 1 Chemical treatment (often chlorination), kills organism present in feed water. 2 Sand filter; removes particles from the incoming water. 3 Activating charcoal filter. Removes free chlorine and traces of organic matter. 4 Diatomaceous earth filter. Retains additional contaminants. 5 Anion exchange. Removes strongly ionized acids such as sulfuric, hydrochloric, and nitric acid. 6 Mixed bed polisher. Contains both cations and anions resins, and removes any ions missed by previous exchange filters. 7 Sterilization. Controls growth of bacteria; often achieved by chlorination or ultraviolet light. 8 Filter. Removes any residual particles left in the wafer prior to its first use. The general methods of removing the ionic and nonionic impurities Table 2 Page 20-21 Some of the methods listed above for removal of impurities are still in use as alternatives to demineralization in certain applications. Also these processes are used as treatments prior to demineralization, either for reducing the ion exchange load on the resins, such as the lime pretreatment for the reduction of alkalinity, or for removing impurities that would foul and clog the resins, such as pretreatment and removal of color, turbidity, and organic matter. The old methods of removing ionic impurities have certain limitations. They can remove only some cations or anions, whereas demineralizers remove all of them. Nevertheless, they continue in use, because they are economical: The chemicals added are less expensive than the acid and caustic soda regenerants used for demineralization, and their doses are smaller. In many water purifying projects, the old methods are included as pretreatment, and the combinations results in lowest total operational cost.. This pretreatment allows the demineralizers to be used as a finisher and refiner. The hardness of water is due mainly to the presence of calcium and magnesium cations. The removal of hardness for the purpose of avoiding the harmful effects of it is termed water softening.The two general methods of softening water are precipitation with lime and soda ash (lime soda process) and ion exchange (sodium cation and hydrogen cation cycles). The process of softening water at ambient temperatures by adding chemicals, such as lime soda and ash was developed commercially about 100 years ago in Great Britain. The lime ( Ca(OH)2 )reacts with the bicarbonate hardness (Ca(HCO3)2 )to precipitate calcium carbonate (CaCO2) and magnesium hydroxide (Mg(OH)2 ).The soda ash reacts with the non carbonate hardness such as magnesium sulfates and chlorides. to form the same fairly insoluble products. These precipitates are allowed to settle out, and the settled water is usually clarified by filtration. Ca(HCO3)2 + Ca(OH02 č 2CaCO3 + 2 H2O Mg(HCO3)2 + 2Ca(OH02 č 2CaCO3 + Mg(OH)2 + 2 H2O The reactions between soda ash and noncarbonate sulfate hardness are as follows: CaSO4 + Na2CO3 č CaCO3 + Na2SO4 MgSO4 + Na2CO3 + Ca(OH)2 č CaCO3 + Mg(OH)2 + Na2SO4 Finally any free carbon dioxide must be also removed by adding enough lime to raise the pH value to optimum required for the process. CO2 + Ca(OH)2 č CaCO3 + H2O The previous equations appear to indicate complete precipitation of hardness; actually the reactions do not go to completion, and some residual hardness is always left in solution. The residuals can be decreased by doses of lime and soda ash in excess of those suggested by the stoichiometry of the reactions. The excess increases the amounts of dissolved carbonate and hydroxide anions in the softened water, providing a common ion effect that depresses the solubilities of calcium carbonate and magnesium hydroxide. (solubility constants) Problems calculating the amount of lime soda, ash, needed to soften water assuming that the reactions go to completion. Answers in PPM. The demineralization process The process of demineralizing water or solutions by ion exchange consists of the conversion of salts to their corresponding acids by hydrogen cation exchangers and the removal of these acids by anion exchangers. salts dissolved in water dissociate into positively charged(cations) and negativelt charged ions (anions), which allow the solution to conduct electricity (electrolytes). likewise ion exchanger contain positively charged cations and negatively charged anions in a condition of electro neutrality. The exchangers differ from solutions in that only one of the two ionic species is mobile (exchangeable). For example a typical sulfuric acid cation exchanger has inmobile ion exchange site consisting of SO3- radicals to wich are attached mobile cations, such as H= or Na+, that may be exchanged in an ion exchange reaction. An anion exchanger similarlllly has immobile cationic sites to which are attached mobile hydroxide anions. When ion exchange occurs, the cations or anions in the solution are interchanged for those in the exchanger, but both the solution and the exchanger remain in a condition of electro neutrality. In the case of the cation exchanger, for example, on calcium cation, which has 2 positive charges (Ca++), when it leaves the water must replace, in the exchanger, two hydrogen cations, which have a single positibe changer (H+). Ion exchange between the solid exchanger and the water containing the electrolytes takes place without structural changes in the solid material (solid does not go into solution). The ions in the solution rapidly diffuse into the molecular network of the exchanger, reaching the exchange sites, where interchange of the ions occurs. The ions in the exchanger similarly diffuse out of the exchanger into the solution. The ion exchange demineralization takes place with the equilibrium (reversible) reaction. They may be expressed in simple form by the following 2 equations: Cation exchange: Z *a(+) + b(+) čç Z*b(+) + a(+) Z is the matrix and anionic fixed site of the cation exchanger, and a(+) and b(+) are two cations. Anion exchange: A*c(-) + d(-) čç A*d(-) + c(-) A is the matrix and cationic fixed site of the anion exchanger, and c(-) and d(-) are two anions. In the previos equilibria equations, the reverse may take place in the exchanger, as during regeneration. The direction the reaction will take depends mainly on the affinity of the resin for the various ions in the water or solution. Ion exhange equilibria are often expressed in terms of “selectivity coefficients” K K = (concentration of a(+) in resin)* (concentration of b(+) in solution) (concentration of b(+) in resin) * ( concentration of a(+) in solution) This applies only to ions of equal valences. Problems having to do with mass of ions removed in parts per million, amount of water produced….