Documentos de Académico
Documentos de Profesional
Documentos de Cultura
Kazuki Morita
1
Contents
1. Introduction 2. Metallurgical refining for SOG-Si
- Thermodynamic properties of impurities in molten Si alloys - Optimized Process Combined with Leaching Treatment -
MG-Si
SOG -Si
Fe Ti
Al
Ti Al P B
Fig. Solar-cell efficiency versus impurity concentration for 4-cm p-base devices.
6
0
Gibbs energy of formation for oxides(kJ/mol O2)
+
4 Cu
O 2 Cu 2 = O2
= 2 Fe + O 2 2 FeO
-400
C + O2 = CO2
-800
-1200
e O4 2C+ 1/2 F 3 = O = O 2 + O e 2 n 2 CO M 2 3/2 F = O 2 n+ 2M O l 2O 3 = Si 2 2/3 A O 2 = + O 3 l + O2 Si-SiO2 Si O = 2/3 Ti 2 A 3 / 4 Mg-MgO i+ 2 4/3 T O Mg Ca-CaO 2 = Al-Al2O3 2 O g+ CaO 2M =2 O2 a+ C 2
Fe-FeO
Temperature(K)
P, B, C, etc.
T=T1 Si standard : solid
GM solid 0liquid
GM 0-
liquid solid
Si
Component X
Si
Component X
T1
S+L Si
T1
S+L S Si
Component X
Component X
Fig. Relationship between Gibbs free energy and phase diagram below melting point of silicon.
104 103 Fe Al 2 Ti 10 P B after first solidification 101 V Cr 100 Ni -1 after second 10 solidification -2 10 10-3 10-4 10-5 10-6 10-7 10-8 10-9 10-10 10-11 -6 10 10-5 10-4 10-3 10-2 10-1 100 segregation coefficient
9
and Iron
Silicone Plug
Mullite Tube
Ar+P4
Red Phosphorus
Ribbon Heater
Graphite Holder
398-461K 190cc/min
Ar+P4 (P2+P)
0.002 PP21/2(atm1/2)
0.001
1750
1800 T(K)
1850
Fig. Relationship between equilibrium phosphorous partial pressure and phosphorous concentration of silicon at 1823K.
Fig. Temperature dependence of free energy change of phosphorous dissolution into silicon.
12
Hertz-Knudsen equation
Mi dy 0 = i pi X i 2RT dt
1823K 1867K estimated Ikeda et al.
10-4 10-5 10-6 10-7 10-8 10-9 10-10 10-11 10-12 10-13 10-14 10-15 10-16 -6 10
(V/A)log([mass%P]0/[mass%P]t)(m)
0.03
PP ,PP2 (atm)
PP PP2
0.02
0.01
10-5
10-2
10-1
Fig. Relationship between equilibrium partial pressure of P, P2 and phosphorous content of silicon at 1823K.
2 3 4 Time (ks) Fig. Relationship between time for vacuum treatment and (V/A) log([mass%P]0/[mass%P]t). 13
Graphite Holder Graphite Lid Graphite Crucible Molten Si-M (M:Fe,Ti) Molten Lead Porous Alumina Block
Si
Pb
Ti(l) in Si=4.4810-4
Ti Ti in Si=3.97
-6
-1 -2 -3 -4 -5 -6 0
-2
-7
-8
lnXTi in Pb-lnXTi in Si-13.2XPb in Si -10.8XSi in Pb-1.74 = -7.71+3.97XTi in Si
-9 0
0.1
0.02
0.04
0.06
XFe in Si
15
Al
Al2O3-Al6Si2O13 Crucible
Ca
SiO2 Crucible Molten SiO2 satd. CaO-SiO2 Slag Molten Si-Ca Alloy SiO2 Crucible MgSiO3, SiO2 pellet
Mg
1823
T(K) 1773 Ca Al
Fig. Temperature dependence of the activity coefficient of aluminum, calcium, and magnesium in molten silicon relative to pure liquid.
18
20
Ca
2000
10
Al
1000
Mg
3 4 5 Time(ks) Fig. Relationship between vacuum time and impurity content of silicon at 1823K.
0 6
30
3000
mass ppm P
19
Vacuum melting by electron beam Removal of P, (Al, Ca) Directional solidification Removal of Fe,Ti,Al
Vacuum
Plasma melting with water vapor Removal of B, C Directional solidification Removal of Fe,Ti,Al
Argon
SOG-Si
Is there any way to reduce the times of solidification refining? If the treatment of MG-Si can remove 90% of Ti and Fe,
Ar
Graphite Crucible
Chemical analysis
22
Removal Fraction of Fe
23
Fig. Phase Diagram of the Si rich corner for the Ca-Fe-Si System [calculated by Thermo-Calc]
24
Si-3.26%Ca-0.424%Fe, 4.4K/min
Fig. (a) SEM micrograph of Si-Ca-Fe alloy. (b) Microstructure of Si-Ca-Fe alloy. (c) Concentration profile by EPMA line analysis.
25
Si-2.41%Ca-0.685%Fe, 4.4K/min
Fig. (a) SEM micrograph of Si-Ca-Fe alloy. (b) Microstructure of Si-Ca-Fe alloy. (c) Concentration profile by EPMA line analysis.
26
Acid leaching treatment with Ca addition NEDO SOG-Si Manufacturing Process Removal Fe and Ti Co.) (Operated inof JFE Steel
MG-Si
First step purification Second step purification
P : 30 < 0.1ppmw
Directional solidification Removal of Fe,Ti,Al
Vacuum
SOG-Si
27
Acid leaching treatment with Ca addition NEDO SOG-Si Manufacturing Process Removal of Ti, P and B (Operated inFe, JFE Steel Co.)
MG-Si
First step purification Second step purification
P : 30 < 0.1ppmw
Directional solidification Removal of Fe,Ti,Al
Vacuum
SOG-Si
28
Segregation ratio k i =
Si
X i in Si Al melt
Temp.
Tm,A T2
b)
Component B
c)
Solid Si
Solid Si
Experimental method
Temperature Gradient Zone Melting Method
Al and P contents in solid silicon EPMA analysis
5 4 6 7 8 9 10
Distribution of P (and Al) between solid Si and Si-Al-P alloy - Red phosphorus stuck Al foil Distribution of B (and Al) between solid Si and Si-Al-B alloy
2 1
11 12
Fig. 1. Schematic diagram of an experimental apparatus: 1-SiC heating element; 2-thermocouple connected to PID controller; 3-gas outlet tube; 4-stainless steel holder; 5-single crystalline silicon; 6aluminum foil; 7-mullite tube; 8-thermocouple for measuring temperature of molten zone; 9-porous alumina boat 10-gas inlet tube; 11-alumina plate; 12-Sponge titanium 33
Segregation ratio
400 Intensities of K radiation of aluminum and phosphorus P Al 300
200
(b) Migrated region (c) Si-Al zone (a) Single crystalline Si (d) Single crystalline Si
100
10mm
200 300 400 500 600 Distance ( m) Fig. Intensity profiles of Ka radiation of aluminum and phosphorus of the sample (accelerating voltage of 15kV, sample current of 200nA, sampling step of 10m).
34
100
Segregation ratio
1 0.5
0.1 0.05
Segregation ratios are smaller at low temperature. Low temperature refining is effective.
35
Derivation of segregation ratio of impurity elements between solid Si and Si-Al melt
Equality in chemical potential of impurity between solid Si and Si-Al melt
Gi + RT ln ai = Gi
l0 l
s0
+ RT ln ai
l0
1)
ln k i = ln
Xi Xi
s l
Gi i = + ln s 0 RT i
fus
2)
<
37
Impurity contents of Si after low temperature solidification refining with Si-Al melt at 1273K and ordinary Si solidification refining of MG-Si.
Contents after purification; Ci = ki Ciini
Contents of MG-Si Contents after solidification refining with Si-Al melt at 1273K Contents after ordinary solidification refining
103 102 101 100 10-1 10-2 10-3 10-4 10-5 10-6
Fe
Ti
38
Viewing the development of the solidification refining of Si with Si-Al melt at low temperature
Si-Al melt
40
Separation of solidified Si from Si-Al melt using electromagnetic force Induction furnace
3 1 4 5 6
Si-55.3at%Al alloy (liquidus, 1273K) was melted and held at 1323K in the induction furnace
8 9 10 11 12
Measuring the surface temp. of the melt by the infrared pyrometer Cooling and Solidifying the sample by lowering from the position of induction coil
41
Bottom 1cm
Solidification of Si at lower position Downward bulk flow Agglomeration of Si grains to the bottom of the sample
B Induced flow
Si-Al melt
Solidified Si
Experimental method
Prism Gas inlet tube
Synthesized MG-Si was alloyed to Si55.3at%Al(liquidus, 1273K) or Si-64.6 at%Al(liquidus, 1173K) with pure Al
Quartz tube
Cooling and solidifying the sample by lowering from the coil Crashing the Si agglomerated part (> 840m), then acid cleaning
Sample No. PrSi-1 Fe 4500 2160 Ti 691 248 Al 1280 1560 B 56 36 P 36 19
PrSi-2
44
Source
PrSi-1 PrSi-1 PrSi-2 PrSi-2 PrSi-1 PrSi-1 13 13 20 27 47 36
Fe
(99.7%) (99.7%) (99.1%) (98.8%) (99.0%) (99.2%)
Ti
5.2 (99.2%) 2.7 (99.6%) 2.8 (98.9%) 4.5 (98.2%) 7.7 (98.9%) 5.6 (99.2%)
B 0.81 (98.6%) 0.88 (98.4%) 0.71 (98.1%) 1.90 (94.8%) 0.98 (98.3%) 0.99 (98.2%)
P
0.93 (97.4%) 1.2 (96.7%) 0.72 (96.3%) 1.0 (95.1%) 0.42 (98.8%) 0.66 (98.2%)
Al
599 (53.1%) 534 (58.1%) 575 (63.1%) 602 (61.3%) 538 (57.8%) 453 (64.5%)
Fe, Ti Well reduced but not as well as predicted values B, P Effectively removed
SOG-Si
46
Acknowledgments
Present data are based on two Doctoral Dissertations by Takahiro Miki (1999), now a research associate of Tohoku University, and Takeshi Yoshikawa (2005), now an assistant professor of Osaka University.
48