- Email: [email protected]
Characteristics of large-sized ruby crystal grown by temperature gradient technique
Optical Materials 27 (2005) 699–703 www.elsevier.com/locate/optmat
Characteristics of large-sized ruby crystal grown by temperature gradient technique Ci Song *, Yin Hang, Changtai Xia, Jun Xu, Guoqing Zhou Shanghai Institute of Optics and Fine Mechanics, P.O. Box 800211, 390 Qinghe Road, Jiading, Shanghai, People’s Republic of China Received 22 May 2004; accepted 14 September 2004 Available online 27 October 2004
Abstract Large ruby with the size of B75 · 45 mm was grown by temperature gradient technique for the first time. Absorption spectrum was carried out in the range of 190–800 nm by spectrophotometer, and the concentration spatial distribution of Cr3+ in ruby was calculated from the absorption coefficient that based on the Beer–LambertÕs Law. Cr3+ ions gradually increase along both the growth axis and the radial direction. The shape and ingredient of the inclusions were measured by means of Leitz wide field microscopy and scanning electron microscopy. Laue photos and X-ray omega scan show the good quality of as grown ruby. The optimized growth conditions were pointed out based on the observation. 2004 Elsevier B.V. All rights reserved. PACS: 61.72.Q; 81.10; 61.16.B Keywords: Concentration distribution; TGT method; Scanning electron microscopy; Leitz wide field microscopy; Laue photos
1. Introduction Ruby is a very important crystal not only for its applications in optics, but also for its nobleness in gemstones. It can act as a laser medium and make visiblelight laser, which possesses many good properties such as narrow linewidth, long fluorescent lifetime, large quantum efficiency and very wide absorption band. Besides, the host crystal corundum makes its physical– chemical properties excellent. As a kind of gemstone, it has charming color that people are very fond of it. Thus, many techniques have been undertaken to artificially create ruby, such as flame fusion method [1,2], seeded solution growth and Czochralski method [3], floating zone method [4] and hydrothermal method [5]. In flame fusion method, the ruby ‘‘boule’’ usually splits to prevent cracking while cooling down, and they are of-
ten of undependable optical quality. In seeded solution growth, Czochralski and floating zone methods, the process is costly and takes considerable time to produce ruby large enough to cut crystals with good quality and the gas bubbles are inclined to be induced in ruby [4]. The hydrothermal methods cannot obtain large sized ruby. Previously our group have grown large sized Ti:Al2O3 [6], sapphire [7,8] and other crystals [9–12] successfully by temperature gradient technique (TGT). Currently, we have successfully grown large ruby with the size of B75 · 45 mm and with good quality. The regularity of Cr3+ distribution and the inclusions in the crystals have been investigated.
2. Experimental 2.1. Growth of ruby by TGT
*
Corresponding author. Tel.: +86 021 69918550; fax: +86 021 69918485. E-mail address: [email protected] (C. Song). 0925-3467/$ - see front matter 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.optmat.2004.09.012
The TGT method is a simple directional solidification technique that adapts for growing high temperature
700
C. Song et al. / Optical Materials 27 (2005) 699–703
crystals. The setup with the proper thermal field for growing ruby is similar to that described elsewhere [7]. The stable thermal field with proper temperature is achieved by cylindrical graphite heating elements, the Mo crucible, the Mo shields in the furnace and the cooling water system. The typical dimensions of the tapered molybdenum crucible used here are OD = 80 mm, ID = 76 mm and cylindrical part height of 70 mm. The oxide powders of Al2O3 and Cr2O3 were ignited for several hours at 1300 C to remove moisture, then the powders were mixed (Cr2O3) and pressed to the form of block with diameter closed to the inner diameter of the Mo crucible. The block was then sintered at 1400 C for 48 h in air and loaded into the crucible with the cylindrical [0 0 0 1]-orientation Cr:Al2O3 (Cr2O3 = 0.05 wt%) seed in the bottom. Then the furnace was vacuumized to 105 Torr. After the materials totally melted and keeping the molten mixture for several hours, the crystallization of ruby was started and driven by slow cooling (0.5–1 C) with high precision temperature controller. The whole crystallization process was completed automatically. When the crystallization was complete, the crystal was then annealed in situ and cooled down to room temperature at the desired cooling rate (30–50 C/h). The sanguine-colored ruby crystal boule with the diameter of 75 mm and the cylindrical partÕs height of 45 mm removed from the Mo crucible is shown in Fig. 1(a). 2.2. The absorption spectrum and distribution of Cr3+ in TGT-ruby Several pieces of slabs were cut from as grown crystals that parallel to or perpendicular to the [0 0 0 1] direction, then they were finely polished for observation. Absorption measurements were taken by V-570 UV/ VIS/NIR spectrophotometer at RT on the slab with a spectral resolution of 0.5 nm in the 190–800 nm spectral region. From the absorption coefficient of the different part of the slab, which was cut from the center of as ruby that parallel to the [0 0 0 1] direction, the concentration spatial distribution of Cr3+ in ruby was calculated based on the Beer–Lambert law.
2.3. Observation of macroscopic growth defects in TGT-ruby Considering the symmetrical characteristic we observe different regions of the slab that cut from the center of as ruby that parallel to the [0 0 0 1] direction (Fig. 1(b)) to find the regularity of the inclusion distribution. The shape of the inclusions was observed too, and the photos were selected and recorded on the PC via a CCD camera coupled with Leitz wide field microscopy, then analysis of the impurity elements was performed by energy dispersive X-ray (EDX) in JSM-6360LA scanning electron microscopy (SEM). 2.4. X-ray diffraction analysis The crystallographic orientation was determined by using a homemade back-reflection Laue X-ray camera. An X-ray diffraction omega scan to the (0 0 0 1) slab was taken by Philips XÕpert MRD diffractometer.
3. Results and discussion 3.1. Absorption spectrum and spatial distribution of Cr3+ in TGT ruby The absorption spectrums parallel to the [0 0 0 1] direction is shown in Fig. 2. It shows the typical absorption peaks of Cr3+ in ruby near 410 nm and 550 nm. They are corresponding to the energy transition 4 A2 ! 4F1 and 4A2 ! 4F2 of Cr3+. The sharp and weak absorption peak R1 and R2 near 694 nm were detected too. Intense main absorption peak in the ultraviolet region at 206 nm and two small humps can be seen at the slightly low energy of 225 nm and 255 nm. They are related to the F-center in as grown crystal which is due to the oxygen vacancies during the growth process. This phenomenon is similar to Al2O3 single crystals that was described elsewhere [8,13], The formation of oxygen vacancies is caused by the reductive condition. The absorption spectrum indicate that there is no Cr2+, that is because the fine vacuum impede the C transformed into CO during the growth process, and the solid–liquid
Fig. 1. Large ruby crystal grown by TGT. (a) 1075 g (B75 · 45 mm) ruby as removed from TGT Crucible. (b) Slab cut from the crystal parallel to the [0 0 0 1] direction.
C. Song et al. / Optical Materials 27 (2005) 699–703
701
in the radial direction is higher than along the growth direction. These may be associated with two factors, the one is the speed of the solid–liquid interface excluding the chromium faster than it entering into the solid phase during the growth process, the other is that the interface was kept convex during the entire course, so the chromium ions may be rejected to the edge of the boule. 3.2. Growth defects in TGT-ruby
Fig. 2. Absorption coefficient at RT as the function of wavelength for E ? C of TGT ruby.
interface is submerged beneath the surface also avoid the possible reductive condition which will make the Cr3+ transform into Cr2+. The absorption coefficient a was calculated by Eq. (1) based on Beer–LambertÕs law: 2:303LogðI=I 0 Þ ð1Þ L where L represents the thickness of sample I. The concentration of Cr3+ was calculated based on the equation
a¼
C ¼ a=r ? abs; 550 nm
ð2Þ 20
2
where a?abs, 550 nm (=22.8 · 10 cm ) refers to absorption cross-section for E ? C at 550 nm obtained from Ref. [14]. The Cr3+ spatial concentration distribution is shown in Fig. 3. It can be seen that Cr3+ concentration rise both along the growth direction and along the radial direction, and the lines also showed that the rate of rise
Fig. 3. Distribution of Cr3+ concentration along the growth axis and radial direction.
The diagrammatic sketch of growth striations represent the convex shape of the solid–melt interface is shown in Fig. 4. It is similar to other reports [6,15], where the same TGT growth is used. While investigating slab that cut from the center of as ruby that parallel to the [0 0 0 1] direction we found that the inclusions distributed like a trumpet (part a in Fig. 4). The inclusions are mainly in the cone-shaped sector and periphery of the crystal. This phenomenon can be explained as follows. When solid–liquid surface advance slower than mass transport, the foreign impurity may easily be trapped in the melt and be enveloped by the solid phase. The fluctuations of the thermal field during initial stage of the crystal growth lead to the relatively more inclusions near to the seed crystal. The thermal stability increased from the periphery to the center, so there are fewer inclusions in the center. The photos obtained from Leitz wide field microscope are shown in Fig. 5. The specimens were all cut from the cone shaped area. Some inclusions are tubular substances with bright center along the growth axis, which are considered to be the gas bubbles trapped during the melting process and kept in the saturated liquid. Others are regular hexagonal with the hexagon face be perpendicular to the [0 0 0 1] direction or triangular grain, which was testified to be graphite grains by the energy spectra analysis. Anyway,
Fig. 4. Schematic drawing of as grown ruby. The dash pitch arc represent the striations of the crystal. The area a that near to periphery of the crystal and the near the seed has more inclusions and the area b is of good quality.
702
C. Song et al. / Optical Materials 27 (2005) 699–703
Fig. 5. The inclusions observed by transmission optical microscope.
of the whole crystal, as the part b of Fig. 4 shown, the center and the end sector nearly have no inclusions. The size of crystal cut from the bulk with good quality is about B60 · 30 mm. From this phenomenon, it comes out that we have made progress in controlling the thermal field in the crystal growth in TGT method. 3.3. Study of the inclusions by SEM and energy dispersive X-ray Further study of the inclusions has been taken. Photos taken from SEM are shown in Fig. 6. The pit in the surface is just corresponding to the bubbles and other inclusions observed by above microscope. The EDX analysis indicated that the main impurity is carbon. The constituents of the impurity are illustrated in Table 1. The hexagon inclusion which Fig. 6(a) show contains 97.57 at.% carbon. To the erose pit of Fig. 6(b), it contains a majority of carbon (63.38 at.%) and a little ferrum (2.06 at.%). The carbon must have come from the graphite heater. The ferrum may from the steel disk during the sampleÕs polishing process. Another kind of inclusions exhibited farinose shape (Fig. 6(c)). Except for the mainly composition of aluminum and oxide, a small amount of chlorine (1.33 at.%), potassium
(1.59 at.%), phosphorus (0.53 at.%) and sodium (0.35 at.%) are included. From the shape and the ingredient, we can make a conclusion that the impure oxide powder was introduced by the uneven mixing of the stuff had been trapped in the melt. Attribute to the better controlling of the environmental pollution, we did not find Mo in the EDX analysis although we have used the Mo crucible and the Mo shields in the furnace.
3.4. Crystallinity of as grown ruby Fig. 7 illustrates the back-reflection Laue X-ray photographs and Fig. 8 showed the X-ray diffraction pattern in omega scan of the (0 0 0 1) slab cut from area b. The sharpness and perfect symmetry of the patterns of the Laue spots are evidence of the high degree of crystalline perfection of the slab. The full width at half maximum (FWHM) of the (0 0 0 1) reflection is only 2800 . It should point out that the divergent angle of the beam of Philips XÕpert MRD diffractometer is 1200 . The line broadening of the peak was mostly instrumental resulting from the ‘‘imperfectly parallel beam’’, and the mechanical polish of the slab may broaden the peak either, thus, these results indicate the high quality.
Fig. 6. The SEM micrographs of the inclusion of ruby. (a) The hexagonal inclusion, (b) the erose pit and (c) the farinose shape inclusion.
C. Song et al. / Optical Materials 27 (2005) 699–703
703
Table 1 The constituents of the impurity from the cone shaped area Sample
a b c
Impurity (atomic%) C
O
Al
Cl
97.59 63.38
2.28 20.74 53.84
0.13 13.82 42.36
1.33
Fe
K
P
Na
1.59
0.53
0.35
2.06
tion. From our previous work and present study, the inclusions are bubbles and foreign impurities. The inclusion of crystals grown by TGT mainly in the cone shaped sector and the periphery of as grown crystal. To decrease the amount of inclusions and grow larger ruby with good quality, thoroughly mixing is needed, high purity powders should been prepared, good vacuum are needed, and a clearly growth condition must be kept. Besides, proper cooling down rate and more stable thermal field need to be improved in our later experiments.
Fig. 7. The back-reflection Laue X-ray of the (0 0 0 1) slab.
Acknowledgement The authors gratefully acknowledge engineer Yongzhong Zhou and Quanyue Xu for the work of crystal growth. References
Fig. 8. The X-ray diffraction pattern in omega scan photograph.
4. Conclusions Large ruby with the size of B75 · 45 mm was grown by TGT method for the first time, in which the size of B60 · 30 mm bulk is of good quality. In TGT grown ruby, no Cr2+ and Cr4+ but only Cr3+ existed. Based on Beer–LambertÕs law and the absorption spectrum of the slab that parallel to [0 0 0 1] direction, we found that the amount of Cr3+ doping gradually increasing both along the growth axis and along the radial direc-
[1] R.L. Barns, in: G.E. Brock, (Ed.), Proceedings of Technical Conference on Metallurgy of Advanced Electronic Materials, Interscience, New York, 1962. [2] G.W. Duker, C.M. Kellington, M. Katzmann, J.G. Atwood, Appl. Opt. 4 (1965) 109. [3] C. Bobert, A. Linares, J. Phys. Chem. 26 (1965) 1817. [4] M. Saito, J. Cryst. Growth 74 (1986) 385. [5] A. Kuznedsov, A.A. Shternberg, Sov. Phys. Crystallagr. 12 (1967) 280. [6] Y. Zhou, H. Xia, D. Peizhen, et al., Solid State Lasers III, SPIE, 1627, 1992, p. 230. [7] J. Xu, Y. Zhou, J. Xu, et al., SPIE 3557 (1998) 11. [8] J. Xu, Y. Zhou, G. Zhou, et al., J. Cryst. Growth 193 (1998) 123. [9] Y. Zhou, J. Cryst. Growth 78 (1996) 31–35. [10] P. Deng, J. Qiao, Hubing, et al., J. Cryst. Growth 92 (1992) 276. [11] G. Zhao, X. Zeng, J. Xu, et al., J. Cryst. Growth 252 (2003) 355. [12] G. Zhao, X. Zeng, J. Xu, et al., J. Cryst. Growth 253 (2003) 290. [13] Lee, J.H. Crawford Jr., Electron centers in single-crystal Al2O3, Phys. Rev. B 15 (8) (1977) 4065. [14] D.C. Cronemeyer, J. Opt. Soc. Am. 56 (1966) 1703. [15] P. Deng, J. Qiao, Hubing, et al., J. Cryst. Growth 92 (1988) 276.
Characteristics of large-sized ruby crystal grown by temperature gradient technique Ci Song *, Yin Hang, Changtai Xia, Jun Xu, Guoqing Zhou Shanghai Institute of Optics and Fine Mechanics, P.O. Box 800211, 390 Qinghe Road, Jiading, Shanghai, People’s Republic of China Received 22 May 2004; accepted 14 September 2004 Available online 27 October 2004
Abstract Large ruby with the size of B75 · 45 mm was grown by temperature gradient technique for the first time. Absorption spectrum was carried out in the range of 190–800 nm by spectrophotometer, and the concentration spatial distribution of Cr3+ in ruby was calculated from the absorption coefficient that based on the Beer–LambertÕs Law. Cr3+ ions gradually increase along both the growth axis and the radial direction. The shape and ingredient of the inclusions were measured by means of Leitz wide field microscopy and scanning electron microscopy. Laue photos and X-ray omega scan show the good quality of as grown ruby. The optimized growth conditions were pointed out based on the observation. 2004 Elsevier B.V. All rights reserved. PACS: 61.72.Q; 81.10; 61.16.B Keywords: Concentration distribution; TGT method; Scanning electron microscopy; Leitz wide field microscopy; Laue photos
1. Introduction Ruby is a very important crystal not only for its applications in optics, but also for its nobleness in gemstones. It can act as a laser medium and make visiblelight laser, which possesses many good properties such as narrow linewidth, long fluorescent lifetime, large quantum efficiency and very wide absorption band. Besides, the host crystal corundum makes its physical– chemical properties excellent. As a kind of gemstone, it has charming color that people are very fond of it. Thus, many techniques have been undertaken to artificially create ruby, such as flame fusion method [1,2], seeded solution growth and Czochralski method [3], floating zone method [4] and hydrothermal method [5]. In flame fusion method, the ruby ‘‘boule’’ usually splits to prevent cracking while cooling down, and they are of-
ten of undependable optical quality. In seeded solution growth, Czochralski and floating zone methods, the process is costly and takes considerable time to produce ruby large enough to cut crystals with good quality and the gas bubbles are inclined to be induced in ruby [4]. The hydrothermal methods cannot obtain large sized ruby. Previously our group have grown large sized Ti:Al2O3 [6], sapphire [7,8] and other crystals [9–12] successfully by temperature gradient technique (TGT). Currently, we have successfully grown large ruby with the size of B75 · 45 mm and with good quality. The regularity of Cr3+ distribution and the inclusions in the crystals have been investigated.
2. Experimental 2.1. Growth of ruby by TGT
*
Corresponding author. Tel.: +86 021 69918550; fax: +86 021 69918485. E-mail address: [email protected] (C. Song). 0925-3467/$ - see front matter 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.optmat.2004.09.012
The TGT method is a simple directional solidification technique that adapts for growing high temperature
700
C. Song et al. / Optical Materials 27 (2005) 699–703
crystals. The setup with the proper thermal field for growing ruby is similar to that described elsewhere [7]. The stable thermal field with proper temperature is achieved by cylindrical graphite heating elements, the Mo crucible, the Mo shields in the furnace and the cooling water system. The typical dimensions of the tapered molybdenum crucible used here are OD = 80 mm, ID = 76 mm and cylindrical part height of 70 mm. The oxide powders of Al2O3 and Cr2O3 were ignited for several hours at 1300 C to remove moisture, then the powders were mixed (Cr2O3) and pressed to the form of block with diameter closed to the inner diameter of the Mo crucible. The block was then sintered at 1400 C for 48 h in air and loaded into the crucible with the cylindrical [0 0 0 1]-orientation Cr:Al2O3 (Cr2O3 = 0.05 wt%) seed in the bottom. Then the furnace was vacuumized to 105 Torr. After the materials totally melted and keeping the molten mixture for several hours, the crystallization of ruby was started and driven by slow cooling (0.5–1 C) with high precision temperature controller. The whole crystallization process was completed automatically. When the crystallization was complete, the crystal was then annealed in situ and cooled down to room temperature at the desired cooling rate (30–50 C/h). The sanguine-colored ruby crystal boule with the diameter of 75 mm and the cylindrical partÕs height of 45 mm removed from the Mo crucible is shown in Fig. 1(a). 2.2. The absorption spectrum and distribution of Cr3+ in TGT-ruby Several pieces of slabs were cut from as grown crystals that parallel to or perpendicular to the [0 0 0 1] direction, then they were finely polished for observation. Absorption measurements were taken by V-570 UV/ VIS/NIR spectrophotometer at RT on the slab with a spectral resolution of 0.5 nm in the 190–800 nm spectral region. From the absorption coefficient of the different part of the slab, which was cut from the center of as ruby that parallel to the [0 0 0 1] direction, the concentration spatial distribution of Cr3+ in ruby was calculated based on the Beer–Lambert law.
2.3. Observation of macroscopic growth defects in TGT-ruby Considering the symmetrical characteristic we observe different regions of the slab that cut from the center of as ruby that parallel to the [0 0 0 1] direction (Fig. 1(b)) to find the regularity of the inclusion distribution. The shape of the inclusions was observed too, and the photos were selected and recorded on the PC via a CCD camera coupled with Leitz wide field microscopy, then analysis of the impurity elements was performed by energy dispersive X-ray (EDX) in JSM-6360LA scanning electron microscopy (SEM). 2.4. X-ray diffraction analysis The crystallographic orientation was determined by using a homemade back-reflection Laue X-ray camera. An X-ray diffraction omega scan to the (0 0 0 1) slab was taken by Philips XÕpert MRD diffractometer.
3. Results and discussion 3.1. Absorption spectrum and spatial distribution of Cr3+ in TGT ruby The absorption spectrums parallel to the [0 0 0 1] direction is shown in Fig. 2. It shows the typical absorption peaks of Cr3+ in ruby near 410 nm and 550 nm. They are corresponding to the energy transition 4 A2 ! 4F1 and 4A2 ! 4F2 of Cr3+. The sharp and weak absorption peak R1 and R2 near 694 nm were detected too. Intense main absorption peak in the ultraviolet region at 206 nm and two small humps can be seen at the slightly low energy of 225 nm and 255 nm. They are related to the F-center in as grown crystal which is due to the oxygen vacancies during the growth process. This phenomenon is similar to Al2O3 single crystals that was described elsewhere [8,13], The formation of oxygen vacancies is caused by the reductive condition. The absorption spectrum indicate that there is no Cr2+, that is because the fine vacuum impede the C transformed into CO during the growth process, and the solid–liquid
Fig. 1. Large ruby crystal grown by TGT. (a) 1075 g (B75 · 45 mm) ruby as removed from TGT Crucible. (b) Slab cut from the crystal parallel to the [0 0 0 1] direction.
C. Song et al. / Optical Materials 27 (2005) 699–703
701
in the radial direction is higher than along the growth direction. These may be associated with two factors, the one is the speed of the solid–liquid interface excluding the chromium faster than it entering into the solid phase during the growth process, the other is that the interface was kept convex during the entire course, so the chromium ions may be rejected to the edge of the boule. 3.2. Growth defects in TGT-ruby
Fig. 2. Absorption coefficient at RT as the function of wavelength for E ? C of TGT ruby.
interface is submerged beneath the surface also avoid the possible reductive condition which will make the Cr3+ transform into Cr2+. The absorption coefficient a was calculated by Eq. (1) based on Beer–LambertÕs law: 2:303LogðI=I 0 Þ ð1Þ L where L represents the thickness of sample I. The concentration of Cr3+ was calculated based on the equation
a¼
C ¼ a=r ? abs; 550 nm
ð2Þ 20
2
where a?abs, 550 nm (=22.8 · 10 cm ) refers to absorption cross-section for E ? C at 550 nm obtained from Ref. [14]. The Cr3+ spatial concentration distribution is shown in Fig. 3. It can be seen that Cr3+ concentration rise both along the growth direction and along the radial direction, and the lines also showed that the rate of rise
Fig. 3. Distribution of Cr3+ concentration along the growth axis and radial direction.
The diagrammatic sketch of growth striations represent the convex shape of the solid–melt interface is shown in Fig. 4. It is similar to other reports [6,15], where the same TGT growth is used. While investigating slab that cut from the center of as ruby that parallel to the [0 0 0 1] direction we found that the inclusions distributed like a trumpet (part a in Fig. 4). The inclusions are mainly in the cone-shaped sector and periphery of the crystal. This phenomenon can be explained as follows. When solid–liquid surface advance slower than mass transport, the foreign impurity may easily be trapped in the melt and be enveloped by the solid phase. The fluctuations of the thermal field during initial stage of the crystal growth lead to the relatively more inclusions near to the seed crystal. The thermal stability increased from the periphery to the center, so there are fewer inclusions in the center. The photos obtained from Leitz wide field microscope are shown in Fig. 5. The specimens were all cut from the cone shaped area. Some inclusions are tubular substances with bright center along the growth axis, which are considered to be the gas bubbles trapped during the melting process and kept in the saturated liquid. Others are regular hexagonal with the hexagon face be perpendicular to the [0 0 0 1] direction or triangular grain, which was testified to be graphite grains by the energy spectra analysis. Anyway,
Fig. 4. Schematic drawing of as grown ruby. The dash pitch arc represent the striations of the crystal. The area a that near to periphery of the crystal and the near the seed has more inclusions and the area b is of good quality.
702
C. Song et al. / Optical Materials 27 (2005) 699–703
Fig. 5. The inclusions observed by transmission optical microscope.
of the whole crystal, as the part b of Fig. 4 shown, the center and the end sector nearly have no inclusions. The size of crystal cut from the bulk with good quality is about B60 · 30 mm. From this phenomenon, it comes out that we have made progress in controlling the thermal field in the crystal growth in TGT method. 3.3. Study of the inclusions by SEM and energy dispersive X-ray Further study of the inclusions has been taken. Photos taken from SEM are shown in Fig. 6. The pit in the surface is just corresponding to the bubbles and other inclusions observed by above microscope. The EDX analysis indicated that the main impurity is carbon. The constituents of the impurity are illustrated in Table 1. The hexagon inclusion which Fig. 6(a) show contains 97.57 at.% carbon. To the erose pit of Fig. 6(b), it contains a majority of carbon (63.38 at.%) and a little ferrum (2.06 at.%). The carbon must have come from the graphite heater. The ferrum may from the steel disk during the sampleÕs polishing process. Another kind of inclusions exhibited farinose shape (Fig. 6(c)). Except for the mainly composition of aluminum and oxide, a small amount of chlorine (1.33 at.%), potassium
(1.59 at.%), phosphorus (0.53 at.%) and sodium (0.35 at.%) are included. From the shape and the ingredient, we can make a conclusion that the impure oxide powder was introduced by the uneven mixing of the stuff had been trapped in the melt. Attribute to the better controlling of the environmental pollution, we did not find Mo in the EDX analysis although we have used the Mo crucible and the Mo shields in the furnace.
3.4. Crystallinity of as grown ruby Fig. 7 illustrates the back-reflection Laue X-ray photographs and Fig. 8 showed the X-ray diffraction pattern in omega scan of the (0 0 0 1) slab cut from area b. The sharpness and perfect symmetry of the patterns of the Laue spots are evidence of the high degree of crystalline perfection of the slab. The full width at half maximum (FWHM) of the (0 0 0 1) reflection is only 2800 . It should point out that the divergent angle of the beam of Philips XÕpert MRD diffractometer is 1200 . The line broadening of the peak was mostly instrumental resulting from the ‘‘imperfectly parallel beam’’, and the mechanical polish of the slab may broaden the peak either, thus, these results indicate the high quality.
Fig. 6. The SEM micrographs of the inclusion of ruby. (a) The hexagonal inclusion, (b) the erose pit and (c) the farinose shape inclusion.
C. Song et al. / Optical Materials 27 (2005) 699–703
703
Table 1 The constituents of the impurity from the cone shaped area Sample
a b c
Impurity (atomic%) C
O
Al
Cl
97.59 63.38
2.28 20.74 53.84
0.13 13.82 42.36
1.33
Fe
K
P
Na
1.59
0.53
0.35
2.06
tion. From our previous work and present study, the inclusions are bubbles and foreign impurities. The inclusion of crystals grown by TGT mainly in the cone shaped sector and the periphery of as grown crystal. To decrease the amount of inclusions and grow larger ruby with good quality, thoroughly mixing is needed, high purity powders should been prepared, good vacuum are needed, and a clearly growth condition must be kept. Besides, proper cooling down rate and more stable thermal field need to be improved in our later experiments.
Fig. 7. The back-reflection Laue X-ray of the (0 0 0 1) slab.
Acknowledgement The authors gratefully acknowledge engineer Yongzhong Zhou and Quanyue Xu for the work of crystal growth. References
Fig. 8. The X-ray diffraction pattern in omega scan photograph.
4. Conclusions Large ruby with the size of B75 · 45 mm was grown by TGT method for the first time, in which the size of B60 · 30 mm bulk is of good quality. In TGT grown ruby, no Cr2+ and Cr4+ but only Cr3+ existed. Based on Beer–LambertÕs law and the absorption spectrum of the slab that parallel to [0 0 0 1] direction, we found that the amount of Cr3+ doping gradually increasing both along the growth axis and along the radial direc-
[1] R.L. Barns, in: G.E. Brock, (Ed.), Proceedings of Technical Conference on Metallurgy of Advanced Electronic Materials, Interscience, New York, 1962. [2] G.W. Duker, C.M. Kellington, M. Katzmann, J.G. Atwood, Appl. Opt. 4 (1965) 109. [3] C. Bobert, A. Linares, J. Phys. Chem. 26 (1965) 1817. [4] M. Saito, J. Cryst. Growth 74 (1986) 385. [5] A. Kuznedsov, A.A. Shternberg, Sov. Phys. Crystallagr. 12 (1967) 280. [6] Y. Zhou, H. Xia, D. Peizhen, et al., Solid State Lasers III, SPIE, 1627, 1992, p. 230. [7] J. Xu, Y. Zhou, J. Xu, et al., SPIE 3557 (1998) 11. [8] J. Xu, Y. Zhou, G. Zhou, et al., J. Cryst. Growth 193 (1998) 123. [9] Y. Zhou, J. Cryst. Growth 78 (1996) 31–35. [10] P. Deng, J. Qiao, Hubing, et al., J. Cryst. Growth 92 (1992) 276. [11] G. Zhao, X. Zeng, J. Xu, et al., J. Cryst. Growth 252 (2003) 355. [12] G. Zhao, X. Zeng, J. Xu, et al., J. Cryst. Growth 253 (2003) 290. [13] Lee, J.H. Crawford Jr., Electron centers in single-crystal Al2O3, Phys. Rev. B 15 (8) (1977) 4065. [14] D.C. Cronemeyer, J. Opt. Soc. Am. 56 (1966) 1703. [15] P. Deng, J. Qiao, Hubing, et al., J. Cryst. Growth 92 (1988) 276.