Enhancement of Oxygen Precipitation in Quenched Czochralski Silicon Crystals
- PDF / 260,740 Bytes
- 4 Pages / 420.48 x 639 pts Page_size
- 17 Downloads / 250 Views
ENHANCEMENT OF OXYGEN PRECIPITATION IN QUENCHED CZOCHRALSKI SILICON CRYSTALS A. HARA, T. FUKUDA, I. HIRAI, AND A. OHSAWA Fujitsu Laboratories Ltd., Kamikodanaka, Nakahara-ku, Japan
Kawasaki,
ABSTRACT We studied oxygen precipitation in quenched Czochralski silicon crystals. The larger the cooling rate and the higher the quenching temperature, the more oxygen precipitated. The defects enhancing oxygen precipitation are eliminated by annealing above 900 0 C. The defects are formed and removed repeatedly by quenching and slow cooling. We think that the aggregation of vacancies is related to those defects. INTRODUCTION Intrinsic gettering can improve yield in device manufacturing, but the amount of oxygen precipitation must be controlled. The amount of oxygen precipitation depends on the thermal history of crystal growth [1,2]. There has been much research on oxygen precipitation in slowly cooled crystals to simulate device fabrication process and to study thermal history of crystal growth [1 I], but not much research has been done in oxygen precipitation in quenched Czochralski silicon crystals
[3].
EXPERIMENTAL To study oxygen precipitation in such crystals, we annealed samples above 1270 C in nitrogen or dry-oxygen to eliminate thermal history during crystal growth and to obtain initially homogeneous crystals and then quenched them. We then annealed samples to precipitate oxygen at 700 0 C for 360 h or 0 at 700 C for 20 h and then at 1O0 C-2 for 10 h. We used a conversion coefficient of 4.81 x 10 cm RESULTS Figure 1 shows relationship between the amount of oxygen precipitation and the cooling rate. Solution annealing was done at 1290 0 C for 40 min in nitrogen. We changed the cooling rate from 1700°C/min to 8 0 C/min. Annealing to precipitates oxygen was at 700 0 C for 360 h. The larger the cooling rate, the more oxygen precipitated. The largest was about 4 times as big as the smallest. This shows that the faster the cooling, the more unknown defects are introduced and cause oxygen to precipitate. Figure 2 shows how the amount of oxygen precipitated is affected by the temperature at the start of quenching. We annealed samples at 1270 0 C for 25 min and slowly cooled them 0 at 9 C/min to the quenching temperature (QT). After 15 min at. this temperature, we quenched them faster than 1000°C/min. Onej was quenched directlg from 1270 0 C. We changed quenching 0 temperature from 1270u Cto 1100 C. Annealing to precipitate: Mat. Res. Soc. Symp. Proc. Vol. 163.
1990 Materials Research Society
970
C~* -15
'E
ro6
0
E
S19 40 min; N2
PA: 700 ( 20 h + 1000% 10 h 1270 -25 min;N 2 ,0 2 fgC/min 15min OT
0 E RT
•4
Time
RRT
L_
L1
Time
N
•2 0
o
1
2
3i
104
103 102 10 Cooling rate (°C/min)
100
0 N2
(-D
PA: 700tC 360 h
00
between Fig.1. Relationship amount of oxygen precipitation and pqoling rate. [0.] = 1.78 x 10 cm-'. P(B) = 1• ohm-cm. RT: room temperature. PA: precipitation annealing.
0 02
0
1300 1200 1000 1100 Quenching temperature CC)
Fig.2. Amount of oxygen at different precipitation quenching g
Data Loading...
