GaN single crystals grown under moderate nitrogen pressure by a new flux: Ca3N2

GaN single crystals grown under moderate nitrogen pressure by a new flux: Ca3N2

ARTICLE IN PRESS Journal of Crystal Growth 291 (2006) 72–76 www.elsevier.com/locate/jcrysgro GaN single crystals grown under moderate nitrogen press...

310KB Sizes 1 Downloads 15 Views

ARTICLE IN PRESS

Journal of Crystal Growth 291 (2006) 72–76 www.elsevier.com/locate/jcrysgro

GaN single crystals grown under moderate nitrogen pressure by a new flux: Ca3N2 J.K. Jiana,b, G. Wanga, Cong Wangb, W.X. Yuanc, X.L. Chena, a

Beijing National Laboratory for Condensed Matter Physics, Institute of Physics, Chinese Academy of Sciences, Nano-scale Physics and Device Laboratory, Beijing 100080, PR China b Center of Condensed Matter and Materials Physics, School of Sciences, Beihang University, Beijing 100083, PR China c Department of Chemistry, School of Applied Science, University of Science and Technology of Beijing, Beijing 100083, PR China Received 18 February 2006; received in revised form 17 March 2006; accepted 21 March 2006 Communicated by R. Fornari

Abstract Using a new flux, Ca3N2, bulk GaN crystals were grown from Ga melt at 900 1C under a nitrogen pressure of about 0.2 MPa. Optical observations indicated that the grown GaN crystals were transparent hexagonal prisms with length up to 1.5 mm. The morphology of these GaN crystals was characterized by scanning electron microscopy (SEM) and compared with that of GaN crystals grown by using Li and Na flux. Raman scattering examinations revealed that the GaN crystals grew along [0 0 0 1] direction. These results demonstrated that Ca3N2 was an effective new flux in the crystal growth of GaN besides the known fluxes of Li, Na and Na–Ca. r 2006 Elsevier B.V. All rights reserved. PACS: 81.05.Ea; 81.10.Dn Keywords: A2. Single crystal growth; B1. Gallium compounds; B2. Semiconducting III–V materials

1. Introduction GaN is of great importance for use in high-power, shortwavelength optoelectronic devices due to its unique optical, electronic and thermodynamical properties. Great achievements in GaN technology have been gained during the past decade. GaN-based light-emitting diodes (LEDs) have been commercially available and room-temperature blue GaN laser diode has been reported, too [1–3]. These GaNbased devices have been usually fabricated on sapphire and 6H-SiC substrates by epitaxial growth. However, there are many defects introduced into the devices because of the mismatch between GaN film and the substrates, which seriously affect the performance of the devices. Although there are some technologies adopted to improve the quality of the heteroepitaxial GaN films, such as the epitaxial Corresponding author. Tel.: +86 10 82649039.

E-mail addresses: [email protected] (J.K. Jian), [email protected] (X.L. Chen). 0022-0248/$ - see front matter r 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.jcrysgro.2006.03.016

lateral overgrowth [4] and buffer layers [5], it is more ideal to use high-quality GaN crystals as substrates for homoepitaxial growth. Unfortunately, the growth of large-size GaN single crystals with high quality remains a challenge for researchers by now. Hydride vapor phase epitaxy (HVPE) [6,7] has been extensively studied to fabricate large-sized GaN plates recently, but, the hetero-epitaxial growth nature of the technology makes it difficult to improve the quality of the products for high-performance LDs. Through high nitrogen pressure growth of GaN crystals over 1 cm in sizes have been achieved [8,9], the method cannot be widely applied due to its critical growth conditions (pressure over 1.0 GPa and temperature over 1500 1C). Using NaN3 as flux and metallic Ga as raw material, Yamane et al. [10] reported transparent GaN single crystals grown at 750 1C under 5 MPa nitrogen pressure. In addition, it has been demonstrated that metallic Na or mixtures of Na and Ca can serve as flux for GaN crystals growth in a similar approach [11,12]. A pressure threshold at a level of several MPa is needed in the

ARTICLE IN PRESS J.K. Jian et al. / Journal of Crystal Growth 291 (2006) 72–76

Na flux methods. It was found that the pressure threshold was decreased and the quality of GaN crystals was improved by the addition of metallic Ca to Na flux [11,12]. But, only metallic Ca failed to grow GaN single crystals as flux [11]. Growth of GaN single crystals at nitrogen pressures as low as 1–2 atm was reported by a Li3N flux method at about 780 1C [13–15]. The work suggests that flux method is perhaps promising for low-cost and large-scale growth of GaN crystals. Up to now, no other fluxes are known to be used in growing GaN single crystals to our knowledge except Na [10], Li [13–15] and Na–Ca [11,12]. New flux is highly desired for growth of high-quality large GaN single crystals. Alkaline-earth Ca is found to have a higher ability to solubilize nitrogen than Li according to Ca–N [16] and Li–N [17] phase diagrams, and AlN crystals could be obtained from Ca3N2–AlN melt under 1500–1600 1C [18]. It is reasonable to speculate that Ca3N2 may be a good flux for GaN crystal growth. In the present work, we report our experimental results on transparent GaN single crystals grown by a Ca3N2 flux method, which demonstrates that Ca3N2 can function as a good flux for GaN single crystal growth under moderate nitrogen pressure (about 0.2 MPa). Transparent prism-like GaN single crystals with 1.5 mm in length and 100 mm in width are grown at 900 1C under a nitrogen pressure of about 0.2 MPa. Optical and scanning electron microscopy (SEM) observations reveal that those crystals prefer to grow along the [0 0 0 1] direction. Raman spectra are employed to characterize the quality of the crystals. 2. Experimental procedure The flux Ca3N2 is synthesized by the direct reaction between metal Ca plates (Alfa Aesar, 99.5%, metals basis) and high purity N2 (99.999%) in a quartz tube at about 900 1C (X-ray diffraction analysis indicates that there is a little impurity of CaO or Ca(OH)2 in the as-prepared Ca3N2). The starting materials for GaN crystals growth are metal Ga (99.999%) and as-prepared Ca3N2, which were put in a tungsten crucible (50 mm inner diameter, 60 mm depth) in a proper proportion. The growth apparatus used here was same with that of Li3N flux research reported before [13]. After evacuated to a vacuum of about 5  103 Pa, the system was filled with N2 gas and the pressure was fixed at about 0.2 MPa. The tungsten crucible filled with starting materials was heated to 900 1C by an induction-heater, kept at the temperature for about 12 h, and then was slowly cooled at a rate of about 3 1C/day. The pressure inside the system changed a little (raised to 0.21 MPa) during the process because the heated tungsten crucible is properly insulated in the system by adiabatic materials, as depicted in Ref. [13]. The whole growth process lasted about 160 h. Lastly, the sample in the crucible was cooled to room temperature by switching off the power. The product was washed by HCl solution and distilled water, and some colorless transparent crystals were obtained. The as-obtained samples were characterized

73

by X-ray powder diffraction (XRD) using a MACM18XHF diffractometer with Cu Ka radiation. The morphologies of the products are observed by an optical microscope and an SEM (FEI, XL-30). The compositions of samples are addressed by energy-dispersive X-ray (EDX) spectroscope equipped on the SEM. Raman scattering spectra of products are collected at room temperature by a multichannel modular triple Raman system (JY-64000) using 532 nm line of a solid-state laser as excitation source. The resolution of the Raman spectrometer is lower than 1 cm1 and the diameter of incident light spot about 1 mm. 3. Results and discussion Fig. 1 presents XRD pattern of the milled products prepared at a molar ratio of Ca3N2: Ga ¼ 1:8. All reflections in the pattern correspond to those of hexagonal GaN with lattice constants a ¼ 3.189 A˚ and c ¼ 5.185 A˚ listed in the standard data of XRD [19]. It is worth noting that (1 0 0) reflection is obviously preferred compared to other reflections of the pattern although the sample has been milled before XRD examination. Optical microscope images of the products are shown in Fig. 2a. It can be seen that most of those crystals with hexagonal prismatic shape are colorless and transparent. The GaN crystals have length from tens of microns to 1.5 mm and diameter of several ten microns. Fig. 2b displays some GaN crystals with sizes of about 1 mm. It has been reported that GaN single crystals prepared by Li flux method have platelet-like shapes [13–15]. But, the GaN crystals grown here have prism-like shapes, no platelet crystals are found. The change of crystal morphology reveals that GaN crystals have different growth habit using Ca3N2 flux. SEM images in Fig. 3 further display the morphological features of the products. Most prismatic GaN crystals have hexagonal pyramidal caps. Some prismatic crystals have smooth side faces and some crystals have growth stripes on their side faces. Considering

Fig. 1. X-ray diffraction pattern of the milled products, Cu Ka radiation.

ARTICLE IN PRESS 74

J.K. Jian et al. / Journal of Crystal Growth 291 (2006) 72–76

Fig. 3. SEM images of the GaN crystals revealing their detailed morphological features. Fig. 2. Optical images of GaN crystals. (a) Overall observation of the products showing the crystals are transparent and (b) three prismatic GaN crystals with lengths about 1 mm.

hexagonal structure of GaN, it can be deduced that the length orientation of those crystals is the [0 0 01] direction. The side faces of those hexagonal prisms are 1 01¯ 0 and the side faces of hexagonal pyramidal caps are 1 0 1¯ 1 . Such morphological feature coincides with the (1 0 0) reflection-preferred XRD pattern of the sample shown in Fig. 1. Raman scattering examinations are employed to further characterize the products. The configurations of our Raman examinations are arranged by the incident light perpendicular to the axis of GaN prism and the polarization of light perpendicular (configuration 1) or parallel (configuration 2) to the c-axis of GaN prism. Fig. 4 represents the Raman spectra of GaN crystal taken under two configurations. Five phonon modes are observed at 144, 532, 560, 568 and 740 cm1 in the spectrum a, which are in good agreement with the reported E2(low), A1(TO), E1(TO), E2(high) and E1(LO) phonons of GaN [20]. Under configuration 2, the spectrum b greatly changes. There are

only three modes corresponding to A1(TO), E2(high) and E1(LO) apparently observed. E1(TO) mode can be carefully detected from the shoulder of E2(high). In addition, the intensity of all modes become very weak and A1(TO) mode becomes stronger than that of E2(high). Considering the Raman selection rules of hexagonal GaN [21], the forbidden modes, such as E1(LO) in configuration 1, E2(high) and E1(LO) in configuration 2, are observed because that the polarization of incident light is not exactly perpendicular or parallel to the c-axis of GaN prismatic crystal. The results of the Raman examinations reveal that the axis of the GaN prism grown here has a [0 0 0 1] orientation, consistent with that of XRD and SEM investigations. Fig. 5 schematically illustrates the typical morphologies of GaN crystals grown by Li3N and Ca3N2 flux, respectively. GaN platelets enclosed by {0 0 0 1} and  1 0 1¯ 0 planes are usually obtained by Li3N flux [13–15], as shown in Fig. 5a. The dominant morphology of GaN single crystals here is prism with side  synthesized faces of 1 0 1¯ 0 and 1 0 1¯ 1 , as depicted in Fig. 5b. For GaN crystals grown by alkali metal fluxes (Li or Na), their

ARTICLE IN PRESS J.K. Jian et al. / Journal of Crystal Growth 291 (2006) 72–76

Fig. 4. Raman spectra of one GaN prismatic crystals collected under different configurations: (a) incident light and polarization are both perpendicular to axis of prism and (b) polarization parallel to its axis.

75

using NaN3 flux and thought that those prismatic crystals grew from a near equilibrium condition. Here, there are some unreacted Ga, intermetallic compound of CaGa4 and little Ca3N2 found in the crucible after experiment, implying the GaN crystals grow in the Ga-rich environment. Another reason for the prismatic morphology of the GaN crystals is probably due to the higher binding energy of Ca to nitrogen. A more recent report of Morishita et al. [24] showed that the addition of Li or Ca to Ga–Na melt resulted in the increase of nitrogen solubility in system and attributed the effect to the higher binding energies of Li and Ca to N. Based on Ca–N, Li–N phase diagrams [16,17] and related experiments [11,13,24], it can be deduced that the nitrogen in Ga–Ca system should be more tightly bound and become more stable than in Ga–Li or Ga–Na melt under the same conditions. The transportation of nitrogen in the system is slower and the supersaturation of GaN will decrease, which perhaps accounts for the prismlike morphology of the GaN crystals obtained here. Step edges can be observed on the surfaces of many GaN crystals grown here, as shown in Fig. 3, suggesting that the growth of those crystals is probably controlled by an edgenucleated layer growth mechanism [23]. The detailed growth process of the GaN crystals is unclear at present and more experiments are needed to get the optimum growth conditions for larger crystals. 4. Conclusions In summary, Ca3N2, a new flux was exploited to grow transparent GaN single crystals under moderate condition (900 1C, 0.2 MPa of N2 pressure). Hexagonal GaN crystals with prismatic morphology were synthesized and the length of crystals was up to 1.5 mm. SEM and Raman investigations revealed that the GaN crystals grew along [0 0 0 1] direction. The results presented here provided a new alternative flux for GaN single crystal growth. Acknowledgments This work was financially supported by the National Natural Science Foundation of China through a Project (59925206) and the Ministry of Science and Technology of PRC through a Project (2002AA311210). References

Fig. 5. Schematic illustrations of morphologies of GaN crystals grown by Li3N (a) and Ca3N2 (b) flux, respectively.

morphologies are thought to be related to the Ga concentration in system. In Li3N case, our previous results have demonstrated that GaN platelet crystals grow from the Ga-rich melt of Li–Ga–N [13,14,22]. On the contrary, Yamane et al. [23] reported that prismatic morphology was preferred for GaN crystals grown in high Ga concentration

[1] S. Nakamura, M. Senoh, S. Nagahama, N. Iwasa, T. Yamada, T. Matsushita, H. Kiyoku, Y. Sugimoto, Jpn. J. Appl. Phys. 35 (1996) L74. [2] S. Nakamura, Mater. Res. Soc. Sympos. Proc. 482 (1998) 1145. [3] S. Nakamura, Science 281 (1998) 956. [4] A. Sakai, H. Sunakawa, A. Usui, Appl. Phys. Lett. 71 (1997) 2259. [5] H. Amano, N. Sawaki, I. Akasaki, Y. Toyoda, Appl. Phys. Lett. 48 (1986) 353. [6] S.T. Kim, Y.J. Lee, D.C. Moon, C.H. Hong, T.K. Yoo, J. Crystal Growth 194 (1998) 37. [7] H. Morkoc, Mater. Sci. Eng. 5–6 (2001) 135. [8] S. Porowski, J. Crystal Growth 189/190 (1998) 153.

ARTICLE IN PRESS 76

J.K. Jian et al. / Journal of Crystal Growth 291 (2006) 72–76

[9] I. Grzegory, J. Phys.: Condens. Matter 13 (2001) 6875. [10] H. Yamane, M. Shimada, S.J. Clarke, F.J. Disalvo, Chem. Mater. 9 (1997) 413. [11] F. Kawamura, M. Morishita, T. Iwahashi, M. Yoshimura, Y. Mori, T. Sasaki, Jpn. J. Appl. Phys. 41 (2002) L1440. [12] F. Kawamura, T. Iwahashi, M. Morishita, K. Omae, M. Yoshimura, Y. Mori, T. Sasaki, Jpn. J. Appl. Phys. 42 (2003) L729. [13] Y.T. Song, W.J. Wang, W.X. Yuan, X. Wu, X.L. Chen, J. Crystal Growth 247 (2003) 275. [14] Y.T. Song, X.L. Chen, W.J. Wang, W.X. Yuan, X. Wu, J. Crystal Growth 260 (2004) 327. [15] X.L. Chen, Sci. Technol. Adv. Mater. 6 (2005) 766. [16] T.B. Massalski, Binary Alloy Phase Diagrams, second ed., vol.1, pp. 928.

[17] T.B. Massalski, Binary Alloy Phase Diagrams, second ed., vol.3. pp. 2448. [18] C.O. Dugger, Mater. Res. Bull. 9 (1974) 331. [19] Joint Committee on Powder Diffraction Standards, JCPDS Card: 500792. [20] J.W. Orton, C.T. Foxon, Rep. Progr. Phys. 61 (1998) 1. [21] T. Kozawa, T. Kachi, H. Kano, Y. Taga, M. Hashimoto, N. Koide, K. Manabe, J. Appl. Phys. 75 (1994) 1098. [22] W.J. Wang, Y.T. Song, W.X. Yuan, Y.G. Cao, X. Wu, X.L. Chen, Appl. Phys. A 78 (2004) 29. [23] H. Yamane, M. Shimada, T. Sekiguchi, F.J. DiSalvo, J. Crystal Growth 186 (1998) 8. [24] M. Morishita, F. Kawamura, M. Kawahara, M. Yoshimura, Y. Mori, T. Sasaki, J. Crystal Growth 284 (2005) 91.