An Examination of the Chemical Staining of Silicon

Abstract

Large-scale chemical lapping of diamond-sawed silicon wafers to remove residual mechanical damage results in the formation of wafer stains. Highly doped p-type substrates are particularly susceptible and have been a primary cause of low wafer yields during production. This paper describes an investigation into the cause of silicon staining and a practical solution to the problem. On the premise that the stain is a suboxide of silicon, various nitric-hydrofluoric acid mixtures, as well as mixtures of hydrofluoric acid with oxidizing agents such as Cr +6, Fe +~, Cu+2; and I ~ were used to synthesize silicon stains. Ion-scattering spectroscopy (ISS) and secondary ion mass spectrometry (SIMS) analyses established that all of the stains were suboxides of silicon (SiO<2). Their origin is the incomplete oxidation of silicon, Which is determined by the nature of the oxidant and its concentration. Although it is not practical to prevent silicon stains from forming, they can readily be removed by a 15 sec immersion in a 2:1 volume ratio of hydrofluoric acid: 0.038M potassium permanganate solution.

(100) Silicon Etch‐Rate Dependence on Boron Concentration in Ethylenediamine‐Pyrocatechol‐Water Solutions

“…The reactions in Eq. [11] and [12] are similar to those proposed by Turner (29) (29) found m ~ 2 and k ~ 4; we also believe that k ~ 4, which fits our etch-rate dependence results (Fig. 6), but that m ~ 0.…”

“…Silicon staining correlation.--Finally, the correlation between the etch-ratio decrease and the silicon staining phenomenon (12,36,37) is further evidence for the proposed oxidation-reduction reactions. Both effects are predominant for heavily boron-doped (100) silicon for concentrations greater than ,~ 1019 cm-~.…”

“…The ~1000A thick boron glass formed on the wafer was removed b y dilute H F ( H F : H 2 0 = 1:20) for 1 min in the dark. T h e etching was done in the d a r k to prevent stain formation on the heavily doped p -t y p e silicon surface, which occurs especially in dilute H F solutions in bright light (12). No stains were observed, especially after a subsequent RCA-1 clean designed to remove them.…”

“…Furthermore, in a way analogous to that by which insufficient oxidizing agent can cause staining (12), it is possible that it is also caused by electron concentration deficiency.…”

Experimentally Derived Catalytic Etching Kinetics for Defect-Utilized Dual-Porous Silicon Formation

“…Thus, the APS growth can be described by the oxygen diffusion into silicon depending on the applied etching time. 21,22,24,[35][36][37][38]55 To investigate the APS dissolution by HF, dipping the etched samples in a 49 wt % HF solution was conducted for 5 min. Figure 4 displays the atomic ratio of the HF-treated APS of 90 s etched sample as a representative.…”