Journal of the Japanese Association for Crystal Growth
Online ISSN : 2187-8366
Print ISSN : 0385-6275
ISSN-L : 0385-6275
Silicon at the Crystal Growth Technology Edge of the Next Century : Silicon(<Special Issue>Bulk Crystals for Human Activity in the New Millennium)
Wilfried v. Ammon
Author information
JOURNAL FREE ACCESS

2000 Volume 27 Issue 2 Pages 8-13

Details
Abstract

At the beginning of this century, the main activities in Si crystal growth development are focused on the transition to the next larger diameter (300 mm) . As this transition entails huge investments in new fabs, -for silicon suppliers as well as device manufacturers-, a thorough cost analysis of the whole process sequence for wafer manufacturing is of utmost importance. In this respect, crystal growth receives special attention, as it is the most expensive and technologically most challenging process step in this sequence. While the complexity of the growth process increases with larger charge size and crystal diameter, the perfection of the growth process must significantly improve , as any disturbances that result in structure loss during growth cause massive material losses. The use of magnetic fields for the control of melt convection will probably become indispensable for a full scale production of 300 mm crystals. With regard to the future bulk quality, radical changes may be required as the design rule approaches the size of the prevailing grown-in crystal defect (vacancy aggregate) . One solution, which considerably reduces the defect size, is nitrogen doping in conjunction with a fast pull rate and, perhaps, an additional wafer heat treatment. Nitrogen doping also appears to be attractive for sufficient internal gettering capability in low thermal budget device processes due to the enhancement of oxygen precipitation. The other, more thorough approach is to grow defect free crystals, which, however, is a tedious task. It will be shown that neither the defect free crystal nor the high temperature treament approach are likely to meet the 300 mm cost of ownership expectations of the device industry, whereas epi wafers offer an economical compromise between high quality and price and, hence, will probably dominate the 300 mm era.

Content from these authors
© 2000 The Japanese Association for Crystal Growth
Previous article Next article
feedback
Top