SC-1处理后的硅晶片的TXRF分析

Silicon wafer cleaning is the most frequently applied processing step in the integrated circuit manufacturing sequence.  This process is intended to remove several different types of contaminants, among them particles, metallics, and organics.  It has been estimated, however, that over fifty percent of yield losses in integrated circuit manufacturing are caused by  contamination remaining on the surface of silicon wafers after cleaning. It is the object of this article to document the  effects of using improved, ultrahigh purity chemicals on silicon wafer surfaces as measured by total reflection x-ray  fluorescence, TXRF. During this study, silicon samples were cleaned with standard grade chemicals and ultrahigh purity  chemicals, and metallic impurities were then measured with TXRF. It was found that the use of ultrahigh purity chemicals  substantially reduced the amount of surface contamination present on wafer surfaces after cleaning.

Silicon wafer cleaning has remained an integral part of  semiconductor device fabrication since the 1950s, and in  fact, is the most frequently applied processing step in the  integrated circuit (IC) manufacturing sequence.The objective of wafer cleaning is the removal of contamination  from the silicon surface without degrading its texture. The importance of adequate cleaning cannot be underestimated  in that contamination remaining on the substrate surface is  known to degrade device performance, reliability, and  yield. It has been estimated that over fifty percent of yield  losses in IC manufacturing are caused by microcontamination.Wafer cleaning will continue to remain an important  process step in device manufacturing especially as device  geometries approach sub-half micron dimensions.

Contamination remaining on a semiconductor surface  can cause a variety of adverse effects during subsequent  processing that will depend on the nature of the impurity.  Particles can cause blocking or masking of various processing operations, for example during etching or photolithography. Particles present during film growth or deposition  can lead to pinholes and microvoids, and if sufficiently  large and conductive, will cause shorting between conductor lines. Metallic impurities will also contaminate the silicon wafer surface at almost any step in device processing.  Such contamination will contribute to the increase of current leakage at the p-n junction, decrease of oxide breakdown voltage, and deterioration of minority carrier lifetime.Many metals are capable of introducing localized  energy states close to the middle of the silicon bandgap  (1.12 eV), thereby creating efficient generation-recombination centers that degrade minority carrier lifetime. These  centers are otherwise known as traps. For example, by introducing bulk traps at 0.40 and 0.55 eV above the silicon  valence bandedge, iron may cause leakage currents that  prevent dynamic random access memory (DRAM) devices  from meeting refresh specifications. 

In TXRF, monochromatic x-rays impinge on an optically  flat sample surface at an angle below the angle for total  external reflection and excite only the top few atomic layers (about 3-8 nm in depth). The fluorescence x-rays from  these top few monolayers emit in many directions, and a  detector located perpendicular and close to the sample surface collects the emitted fluorescence x-rays and analyzes  them according to energy. It is this energy analysis which  yields elemental identification. 

Fig1

During the study; chemicals (H₂0₂ and NH₄OH) were  supplied to the Silicon Products Department in Sherman  by the Chemical Operations Department in Dallas. Initially, the materials were of the standard grade but were  later switched to ultrahigh purity. Specifications (highest  acceptable tolerances of an impurity) for the four metals  that were tracked during the study (Zn, Cu, Ni, and Fe) are  listed in Table I. As can be seen, the switch to the higher  purity grade chemicals resulted in tolerances that are  tighter by 30-100 times.

There is a definite correlation between impurity levels in  process chemicals and metal deposition on wafer surfaces.  It was further shown in this study that TXRF is an appropriate tool for the detection and measurement of surface  wafer impurities.