Temperature distribution near the interface in sapphire crystals grown by EFG and GES methods

Abstract

This paper presents results of experiments on in situ temperature measurements during sapphire shaped crystal growth. The temperature distribution difference between the crystals grown by EFG (edge-defined, film-fed growth) and GES (growth from an element of shape) methods is considered.

Full text

Journal of Crystal Growth 198/199 (1999) 210 —214
Temperature distribution near the interface in sapphire crystals
grown by EFG and GES methods
V.M. Krymov*, V.N. Kurlov, P.I. Antonov, F. Theodore, J. Delepine
A.F. Ioffe Physical—Technical Institute, 26 Politekhnicheskaya, St. Petersburg 194021, Russia
Institute of Solid State Physics RAS, Chernogolovka, Moscow District 142432, Russia
DTA/CEREM/DEM/SPCM, Commisariat a%l+Energie Atomique, 17 rue des Martyrs, F-38054 Grenoble Cedex 9, France
Abstract
This paper presents results of experiments on in situ temperature measurements during sapphire shaped crystal
growth. The temperature distribution difference between the crystals grown by EFG (edge-defined, film-fed growth) and
GES (growth from an element of shape) methods is considered. 1999 Elsevier Science B.V. All rights reserved.
PACS: 81.05.!t; 81.10.!h; 81.10.Fq
Keywords: Shaped crystal growth; Sapphire; EFG/Stepanov method; GES method
1. Introduction
The favorable combination of excellent optical
and mechanical properties of sapphire com-
plemented with high chemical durability makes it
an attractive material for high-technology applica-
tions. By now the growth of sapphire single crystals
of various shapes is well developed. But the quality
of these crystals still remains a serious problem
limiting their application in optics.
It is very important to know the thermal history
of the growing single crystal, in order to improve
*Corresponding author. Fax:#7 812 247 8924; e-mail: anto-
nov@crystal.ioffe.rssi.ru.
the crystal quality and process yield. The temper-
ature distribution in the growing crystal has
a dominating influence on the formation of thermal
stresses. The thermal stresses in turn can lead to
plastic deformations and initiation of defects of
structure (dislocations, slip lines and grain bound-
aries). By adjusting the temperature in the crystal
by means of thermal zone modification involving
heater, die or shields, these defects can be control-
led more effectively [1—3].
The direct temperature measurement in a sap-
phire single crystal is very difficult because of its
high melting point and semitransparency to ther-
mal radiation. The temperature distribution along
a growing sapphire tubular crystal was measured
with a special IR pyrometer [2], and with the
0022-0248/99/$ —see front matter 1999 Elsevier Science B.V. All rights reserved.
PII: S 0 0 2 2 - 0 2 4 8 ( 9 8 ) 0 1 1 2 9 - 4

ingrowth of thermocouples [3]. The maximum cur-
vature of temperature distribution just near to crys-
tallization front was demonstrated.
This paper reports the direct temperature meas-
urement and comparison of temperature fields near
to the interface in sapphire crystals grown by the
EFG and GES methods. Sapphire ribbons were
grown as a model using both methods.
2. Experimental procedure
The experiments were done in an 8 kHz induc-
tion heated graphite susceptor/molybdenum cru-
cible setup held within a growth chamber. The
special design of this installation allows translation
of the pulling shaft in vertical and horizontal direc-
tions simultaneously. Sapphire ribbons were grown
from the melt by the EFG and GES methods. The
techniques utilize capillary rise from a melt source
to the top surface of a wetted die. For the EFG
technique the outer edges of the die determine the
shape of the meniscus, and thus of the growing
crystal [4]. The GES method has been developed
on the basis of the Stepanov method [5]. The
approach of the GES method consists of pulling
a shaped crystal from a melt meniscus which is only
a small element of the whole transverse cross sec-
tion of the growing crystal [6]. The crystal grows
layer by layer while traversing in the horizontal and
vertical directions.
Fig. 1. The scheme of the temperature measurement using the
ingrowth of thermocouples: (a) the EFG technique; (b) the GES
technique. 1 —die; 2 —meniscus; 3 —thermocouple; 4 —crystal.
Fig. 2. Photographs of the growth process made by TV camera.
EFG technique (a). The successive stages of GES technique: The
beginning growth from seed (b), the crystal ribbon growth (c),
the view of the ribbon after the quick lifting from the die (d).
Marking is the same as in Fig. 1.
V.M. Krymovet al. /Journal of Crystal Growth 198/199 (1999) 210–214 211

Fig. 3. The axial temperature distribution in sapphire ribbons grown by: 1 —GES technique (»"0.05 mm/min, »"6.3 mm/min);
2—EFG technique (»"1.0 mm/min); 3 —EFG technique (»"0.1 mm/min).
An initial charge was crushed sapphire Verneuil
boules. The atmosphere was high purity argon. The
molybdenum dies were 2.5;24 mmin cross sec-
tion for the EFG method (Fig. 1a), and 2.5;3mm

in cross section for the GES method (Fig. 1b). The
width of the capillary channels was 0.3 mm for both
variants. The temperature distribution was mea-
sured using WR 5/20 thermocouples, diameter of
the wire was 0.1 mm. The thermocouple seal was
located on the lower end of the seed ribbon initially
grown parallel to the crystallographic c-axis. The
sapphire thin tubes were used for insulating the
wires. At the beginning of the process the seed
ribbon with thermocouple was put down until it
made contact with the die and a column of the melt
was formed (Fig. 2). After the re-melting of the
crystal—melt region, the seal of the thermocouple
was on the crystal—melt phase boundary. Then the
crystal with ingrown seal of thermocouple was
pulled with a constant rate »"0.1 or 1 mm/min
(the EFG variant, Fig. 1a). For the GES variant the
crystal was pulled in vertical direction with rate
»"0.05 mm/min and reverse translated in hori-
zontal direction with rate »"6.3 mm/min both
together (Fig. 1b). Variation of the thermo-
electromotive force in time was registered by the
“Sefram-8400”recorder.
3. Results and discussion
3.1. EFG variant
The axial temperature distribution in sap-
phire ribbons grown by the EFG technique is
shown in Fig. 3. The curve 2 corresponds to the
pulling rate »"1.0 mm/min, the curve 3 corres-
ponds to the rate »"0.1 mm/min. The first dis-
tinguishing characteristic of the temperature
distribution is a sharp decrease of temperature
near the crystallization front. At the distance of
0—5 mm from the crystallization front the tem-
perature falls according to the exponential law
(exponent 1.2 mm\). With distance from the
crystallization front the smooth fall of temper-
ature is observed. The second distinguishing
characteristic is the variation of temperature
distribution with the pulling rate ». This cool-
ing rate increases when decreasing the growth
rate. This effect is connected to the efficiency
of heat radiation out of the crystal that is
better the slower the pulling is. Some experi-
mental observations suggest that this effect can
also be explained by the crystal quality: the trans-
parency is increased with decreasing the growth
rate.
212 V.M. Krymovet al. /Journal of Crystal Growth 198/199 (1999) 210–214

Fig. 4. The temperature—time measurement date for the GES method. Marking is the same as in Fig. 1. A —passing of the thermocouple
seal above the die at a movement of a ribbon in one direction; C —in the opposite direction; B and D —reversal points.
Fig. 5. The two-dimentional temperature distribution in sapphire ribbon near to crystallization front for the GES technique.
X-coordinate along ribbon width (0—20 is equal to 16 mm), ½-coordinate along ribbon length (0—20 is equal to 4 mm), Z—temperature
coordinate.
3.2. GES variant
The sapphire ribbons with 2.5;16 mmin cross
section have been grown. The average thickness of
each growing layer is 100 lm.
Fig. 4 shows the temperature measurement data
for the GES method. The point A corresponds to
a maximum of temperature at passing of the thermo-
couple seal above the die at a movement in one
direction, point C —in the opposite direction. The
points B and D correspond to a moment of reversal
of a direction of horizontal translation of a seed
holder. In this moment the thermocouple seal is in
the extreme position outside of a zone of the die.
V.M. Krymovet al. /Journal of Crystal Growth 198/199 (1999) 210–214 213

Fig. 4 shows that the temperature maximum
(points A, C) decreases with increasing distance
from the crystallization front, and disappears abso-
lutely on distance of 4—5 mm. Also the temperature
measurements have shown that in extreme points
B and D the temperatures difference is 15°C. This is
because the heat zone has the radial temperature
gradient, approximately equal to 1°/mm. The high-
resolution TV camera enables observation of the
meniscus shape during the crystal growth. A con-
cave crystal—melt phase boundary was established
(Fig. 2). After the measurements, the crystal was
quickly lifted from the die to preserve the shaped-
phase boundary in the moment of crystal-die con-
tact (Fig. 2d). The re-melting of crystal above the
die is 0.2—0.6 mm, that is, several earlier crystallized
layers. Fig. 3 (curve 1) illustrates the axial temper-
ature distribution in the ribbon grown by the GES
method. As in the case of EFG method the temper-
ature falls according to the exponential law. The
two-dimensional temperature distribution near the
crystallization front is shown for the GES method
(Fig. 5). This distribution was obtained on the
base of time-dependence of temperature distri-
bution in Fig. 4 for the die under the middle part of
the ribbon situation (A and C). The die has a pro-
nounced effect on the temperature distribution. In
parallel with the axial nonlinear temperature distri-
bution in a crystal near to interface there also has
been a strong distortion of the temperature field in
the radial direction. Analysis of the temperature
distribution and the second derivative of the tem-
perature suggests the high level of thermoelastic
stresses in the grown GES crystals.
References
[1] P.I. Antonov, S.I. Bakholdin, E.A. Tropp, V.S. Yuferev,
J. Crystal Growth 50 (1980) 62.
[2] A.L. Alishoev, L.M. Zatulovsky, Yu.K. Lingard, D.L. Shur,
Bull. Acad. Sci. USSR Phys. Ser. 52 (1988) 99.
[3] P.I. Antonov, S.I. Bakholdin, M.G. Vasilev, V.M. Krymov,
A.V. Moskalev, V.S. Yuferev, Bull. Rus. Acad. Sci. Phys. 58
(1994) 72.
[4] H.E. LaBelle, Mater. Res. Bull. 6 (1971) 581.
[5] A.V. Stepanov, The Future of Metalworking, Lenizdat,
Leningrad, 1963 (in Russian).
[6] P.I. Antonov, Yu.G. Nosov, S.P. Nikanorov, Bull. Acad.
Sci. USSR Phys. Ser. 49 (1985) 6.
214 V.M. Krymovet al. /Journal of Crystal Growth 198/199 (1999) 210–214