The Wright Etch is a preferential etch for revealing defects in (100) and (111) oriented, p- and n-type silicon. It was developed by Margaret Wright Jenkins in 1976 while working in research and development at Motorola Inc. It was published in 1977 [1]. This etchant reveals clearly defined oxidation-induced stacking faults, dislocations, swirls and striations with minimum surface roughness or extraneous pitting. These defects are known causes of shorts and current leakage in finished semiconductor devices (such as transistors) should they fall across isolated junctions. A relatively slow etch rate (~1 micrometre per minute) at room temperature provides etch control. The long shelf life of this etchant allows the solution to be stored in large quantities. The present report summarizes the highlights of the Wright etch. For a detailed discussion of observation and findings, please consult the original publication [2].
Contents |
The composition of the Wright etch is as follows:
60 ml conc. HF (hydrofluoric acid)
30 ml conc. HNO3 (nitric acid)
30 ml of 5 mole CrO3 (mix 1 gram of chromium trioxide per 2 ml of water)
2 grams Cu (NO3)2 . 3H2O (copper nitrate)
60 ml conc. CH3COOH (acetic acid)
60 ml H2O (deionized water)
In mixing the solution, the best results are obtained by first dissolving the copper nitrate in the given amount of water; otherwise the order of mixing is not critical.
The Wright etch consistently produces well-defined etch figures of common defects on silicon surfaces. This attribute is attributed to the interactions of the selected chemicals in the formula. Robbins and Schwartz [3][4][5] described chemical etching of silicon in detail using an HF, HNO3 and H2O system; and an HF, HNO3, H2O and HC2H3O2 (Acetic acid) system. Briefly, the etching of silicon is a two-step process. First, the top surface of the silicon is converted into a soluble oxide by a suitable oxidizing agent(s). Then the resulting oxide layer is removed from the surface by dissolution in a suitable solvent, usually HF. This is a continuous process during the etch cycle. In order to delineate a crystal defect, the defect area must be oxidized at a slower or faster rate than the surrounding area thereby forming a mound or pit.
In the present system, the silicon is oxidized with HNO3, CrO3 solution (Cr2O7=) and Cu (NO3)2. The Cr2O7=, a strong oxidizing agent, is considered to be the principal oxidizing agent. The ratio of HNO3 to CrO3 solution stated in the formula produces a superior etched surface. Other ratios produce less desirable finishes. With the addition of a small amount of Cu (NO3)2, the definition of the defect was enhanced. Therefore it is believed that the Cu (NO3)2 affects the localized differential oxidation rate at the defect site. The addition of the acetic acid gave the background surface of the etched silicon a smooth finish. It is theorized that this effect is attributed to the wetting action of the acetic acid which prevents the formation of bubbles during etching.
This etch process is a quick and reliable method of determining the integrity of pre-processed polished silicon wafers or to reveal defects that may be induced at any point during wafer processing. It has been demonstrated that Wright etch is superior in revealing stacking faults and dislocation etch figures when compared with those revealed by Sirtl [6] and Secco [7] etchings. Please see references for comparison micrographs. This etch is widely used in failure analysis of electrical devices at various wafer processing stages as seen in articles: "Pipeline Defects in Flash Devices Associated with Rings OSF" * [1] and "Defect Etching in Silicon" * [2] . In these publications, by comparison, the Wright etch was the preferred etchant to reveal defects in silicon crystals.