MOVPE growth and characterization of AlxGa1-xN

MOVPE growth and characterization of AlxGa1-xN

Materials Science and Engineering B50 (1997) 219 – 222 MOVPE growth and characterization of Alx Ga1-x N S. Ruffenach-Clur a, O. Briot a,*, J.L. Rouvi...

394KB Sizes 2 Downloads 40 Views

Materials Science and Engineering B50 (1997) 219 – 222

MOVPE growth and characterization of Alx Ga1-x N S. Ruffenach-Clur a, O. Briot a,*, J.L. Rouvie`re b, B. Gil a, R.L. Aulombard a a

b

GES, CC074, UM II, Place E. Bataillon, 34095 Montpellier Cedex 5, France CEA-Grenoble, DRFMC/SP2M, 17 rue des Martyrs, 38054 Grenoble Cedex 9, France

Abstract AlGaN is an important material in the design of nitride devices. However, little is known concerning its growth with high Al contents. We have studied the growth of Alx Ga1-x N epilayers on c-face sapphire by low pressure MOVPE (76 Torr), using triethylgallium, trimethylaluminum and ammonia as precursors. The solid versus gas phase composition relationship was determined experimentally and was fitted using a kinetic model. Then the structural properties of the layers (x =0–1) were studied, using X-ray diffraction, scanning electron microscopy and transmission electron microscopy. We demonstrate that at high Al content, the buffer layer defects are replicated into the AlGaN layer. © 1997 Elsevier Science S.A. Keywords: AlGaN; MOVPE growth; Epilayers

which is the temperature usually used in our equipment to grow high quality GaN.

1. Introduction Gallium nitride and gallium nitride alloys are probably, to date, the most promising material among the wide band gap semiconductors. AlGaN has been successfully employed in many devices like lasers [1,2] and transistors [3], and because of its direct band gap AlGaN is needed for UV devices like photodetectors [4]. Therefore, it is very important to understand the growth mechanisms involved in AlGaN growth, to improve the epilayers quality by optimising the growth parameters. In this work, we grew Alx Ga1-x N epilayers in the whole range of composition from x = 0 –1 and we modeled the Al incorporation in the solid phase. The structural analysis of these layers demonstrate an evolution in the layer quality which decreases with increasing Al content. This degradation of the crystalline quality becomes very strong for x ] 0.5, which should explain why there are so few data available in the literature for AlGaN containing more than 50% Al. Our investigations lead us, through TEM analysis, to demonstrate that the texture of the buffer layer results at high Al contents in a ‘two directional’ growth along the c-axes and the (101( 1) directions in the epilayer. All the epilayers were grown at the same temperature, * Corresponding author. Tel.: +33 467143604; 467144240; e-mail: [email protected]

fax:

+ 33

0921-5107/97/$17.00 © 1997 Elsevier Science S.A. All rights reserved. PII S 0 9 2 1 - 5 1 0 7 ( 9 7 ) 0 0 1 6 6 - 9

2. Experimental The AlGaN layers were grown by low pressure MOCVD using triethylgallium, trimethylaluminum and ammonia. In order to explore all the range of composition for Alx Ga1-x N [4], we have grown samples from GaN up to AlN varying the Al composition by 10% in the gas phase from one sample to another. The sapphire substrate was cleaned in a 1 H3PO4:3 H2SO4 hot solution for 10 min. After being introduced in the reactor, the substrate was nitridated for 10 min under 300 sccm ammonia at 76 Torr. We deposited a ˚ AlN buffer layer at 800°C and recrystallised it at 500 A the epilayer’s growth temperature (980°C). For all compositions of AlGaN, all growth parameters were kept constant except of course the molar flow rates of TEGa and TMAl. However, the total group III molar flow rate was kept constant as well as the V/III molar ratio (10 000). The Al composition was determined by energy dispersive analysis of X-ray (EDAX), making the proper corrections for atomic number, absorption and fluorescence. These results were in good agreement with the results found using X-ray diffraction, and assuming that the c-lattice parameter of the alloy satisfies Veg-

S. Ruffenach-Clur et al. / Materials Science and Engineering B50 (1997) 219–222

220

ard’s law. Cross sections for TEM were prepared using the standard procedures: mechanical polishing and Argon ion milling. TEM observations were realized on a JEOL 4000 EX electron microscope.

3. Results and discussion First, we studied the incorporation of aluminum in the solid phase during the LP-MOVPE growth. This point may be controversial because it is strongly dependent of the amount of premature reactions in the reactor [5]. Indeed, it was demonstrated that, for example, for growth at atmospheric pressure, premature reactions between TMAl and ammmonia could strongly occur, resulting in a depletion of aluminum in the gas phase. Still, it is possible to grow AlGaN at atmospheric pressure by taking special care to reduce this phenomenon [6]. Fortunately, in this case, at 76 Torr and due to the geometry of our reactor inlet, adduct formation is negligible and it was possible to correlate the Al composition in the gas phase and in the solid AlGaN using a simple model. Making the assumptions that the growth mechanism is limited by mass-transport of group III elements through the gas phase [7] and that the solid composition will be determined by the relative Al and Ga fluxes reaching the interface by diffusion, we can write [8]:

 

AlGaN(0001)[21( 1( 0]//AlN(011( 1)[21( 1( 0] AlGaN(0001)[21( 1( 0]//AlN(01( 11)[1( 21( 0] AlGaN(0001)[21( 1( 0]//AlN(101( 1)[1( 21( 0] AlGaN(0001)[21( 1( 0]//AlN(1( 011)[1( 21( 0] AlGaN(0001)[21( 1( 0]//AlN(11( 01)[1( 1( 20] AlGaN(0001)[21( 1( 0]//AlN(1( 101)[1( 1( 20]



1 1 1+k −1 xg where x is the Al solid phase composition, xg is the Al gas phase composition and k is the ratio between the gallium precursor diffusion coefficient DGa through the mixture of ammonia and hydrogen (used as carrier gas) and the aluminum one DAl. To calculate these last two values, we used the Wilke and Lee model for the diffusion coefficients [9] and corrected it to take into account the thermal diffusion due to the temperature gradient in the reactor [10]. We found for k the value 0.83, which fits the experimental data (Fig. 1). The details of this model will be reported elsewhere. To analyse the quality of our AlGaN epilayers, we have performed X-ray diffraction measurements. We observe a strong change in the spectra for the layers containing 50% and more aluminum. In fact, the results reported in Fig. 2 show that for x ] 0.5 a second orientation appears, the (101( 1) reflexion. This indicates that the layer is not fully monocrystalline but contains two well defined orientations. Scanning electron microscopy displays the surface morphology change which appears for x ] 0.5 (Fig. 3). At low Al content, the surface morphology appears smooth and becomes acicular for high Al content. The reasons for these modifications will be given below, from the TEM results. x=

Four Alx Ga1-x N samples with different Al compositions (x=0.16, 0.45, 0.55 and 0.85) have been studied using transmission electron microscopy. The two last samples greatly differ from the first two. The samples with a low Al concentration look like traditional GaN layers. They are monocrystalline with a Ga polarity [11–13] and contain a high density of dislocations (about 1010 cm − 2) and many nanopipes [14]. The samples with an Al content higher than 0.55 contain two kinds of textures: the planes parallel to the initial (0001) sapphire surface plane can be the (0001) planes or the pyramidal {101( 1} planes. These two textures clearly appear in the diffraction pattern of Fig. 4(b). The angular width of these textures is about 94°. Due to the symmetry of the wurtzite material, the {101( 1} texture has six variants. These six variants can be characterized by the orientation of the c and a=[21( 1( 0] axis with respect to the texture of the (0001) AlN buffer layer which is defined by: AlN(0001)[21( 1( 0]// Al2O3(0001)[101( 0]. The six different orientations of the grains are:

Fig. 1. AlGaN solid phase versus gas phase composition. The squares represent the experimental data fitted by our model (full line).

S. Ruffenach-Clur et al. / Materials Science and Engineering B50 (1997) 219–222

221

Fig. 2. X-ray diffraction measurements for Alx Ga1-x N. A second orientation appears for x] 0.5.

Two of these variants can be clearly seen on a HREM picture (Fig. 4(a)) taken along the a = [21( 1( 0] direction. The four other variants cannot be resolved on this HREM image, but are present due to the symmetry of the crystal. It appears that at high Al concentration, the {101( 1} is favored. Such a texture has already been reported in GaN layers grown on (0001) sapphire [15]. But in this case, the {101( 1} textured grains were very small and appeared only in the buffer layer. In our case, HREM images reveal that the {101( 1} textures start growing at once on the AlN buffer layer (not on the sapphire surface) and that they propagate throughout the layer. These grains are responsible for the rough surfaces observed by SEM (Fig. 3): {101( 1} grains are higher than (0001) grains and forms grains elongated along the {21( 1( 0} directions. For x=

Fig. 3. SEM images of Alx Ga1-x N alloys for three different significative compositions. Each picture is 2 mm wide.

Fig. 4. Transmission electron microscopy of an Al0.85Ga0.15N sample: (a) high resolution image; (b) electron diffraction; and (c) low resolution image showing the grains shapes.

0.85, the average width of these grains is 75 nm; their length cannot be determined on the TEM cross-section. Since the buffer texture is the same in all the samples, we have to correlate the fact that the {101( 1} textures appear in the epilayers with the increasing aluminum content. All these experiments lead us to assume that the evolution of the surface morphologies is clearly correlated to the Al content, due to the difference of the surface mobility of the reactants. Indeed, all the layers were grown at 980°C, which is the optimised growth temperature for GaN, but not for AlN. In order to verify our hypothesis, we have grown AlN layers at different growth temperatures, and we have observed a clear improvement of the epilayer quality when increasing the growth temperature up to 1080°C [16]. Therefore, it seems reasonable to think that at 980°C, the surface mobility of Ga adatoms is much higher than for Al adatoms. These results indicate that at low Al contents, the epilayer quality seems to be controlled by the surface diffusion of Ga adatoms which leads to smooth surfaces. At higher Al content the lower surface mobility of Al adatoms induces nucleation both on the c-axis and on the (101( 1) oriented grain. Taking these

S. Ruffenach-Clur et al. / Materials Science and Engineering B50 (1997) 219–222

222

results into account, it seems that in order to grow high quality AlGaN in the whole range of composition, it is necessary to find for each composition the optimised growth temperature which is certainly increasing with increasing Al content.

4. Conclusion We have successfully grown AlGaN in the whole range of composition and modeled the incorporation of Al from the gas phase to the solid phase using diffusion consideration. Structural analysis of the layers indicate an evolution of the crystallinity which is clearly correlated to the Al content. Indeed, SEM, X-ray diffraction and TEM display a sharp transition for x ]0.5. At low Al contents, Ga adatoms with high surface mobility dominate the growth mechanisms, leading to smooth surface morphologies. At high Al contents, the low mobility of Al is harming the epilayer quality: it results in a ‘two directional growth’ induced by the buffer layer. Therefore, to achieve high quality AlGaN epilayers it is necessary to raise the growth temperature according to the Al content of the epilayer.

Acknowledgements This work has been supported by the DRET/DGA and by THOMSON-CSF-LCR.

.

References [1] S. Nakamura and G. Fasol, The Blue Laser Diode, SpringerVerlag, New York, 1997. [2] S. Nakamura, M. Senoh, S. Nagahama, N. Iwasa, T. Yamada, T. Matsushita, Y. Sugimoto, H. Kiyotu, Appl. Phys. Lett. 70 (7) (1997) 868. [3] M.A. Khan, Q. Chen, C.J. Sung, J.W. Yang, M. Blasingame, M.S. Shur, H. Park, Appl. Phys. Lett. 68 (4) (1996) 514. [4] D. Walker, X. Zhang, A. Saxler, P. Kung, J. Xu, M. Razeghi, Appl. Phys. Lett. 70 (8) (1997) 949. [5] C.H. Chen, H. Liu, D. Steigerwald, W. Imler, C.P. Kuo, M.G. Craford, M. Ludowise, S. Lester, J. Amano, J. Electron. Mater. 25 (6) (1996) 1004. [6] Y. Koide, H. Itoh, N. Sawaki, I. Akasaki, M. Hashimoto, J. Electrochem. Soc. 133 (9) (1986) 1956. [7] M.A. Khan, R. Skogman, R. Shulze, M. Gershenzon, App. Phys. Lett. 43 (5) (1983) 492. [8] G.B. Stringfellow, Organometallic Vapor Phase Epitaxy: Theory and Practice, Academic Press, San Diego, 1989. [9] C.R. Wilke, C.Y. Lee, Ind. Eng. Chem. 47 (1955) 1253. [10] W.L. Holstein, J. Electrochem. Soc. 135 (7) (1988) 1788. [11] J.L. Rouvie`re, M. Arlery, R. Niebuhr, K.H. Bachem, O. Briot, MRS Internet J. Nitride Semicond. Res. 1 (1996) 33. [12] B. Daudin, J.L. Rouvie`re, M. Arlery, Appl. Phys. Lett. 69 (17) (1996) 2480. [13] F.A. Ponce, D.P. Bour, W.T. Young, M. Saunders, J.W. Steeds, Appl. Phys. Lett. 69 (3) (1996) 337. [14] W. Qian, M. Skowronski, K. Doverspike, L.B. Rowland, D.K. Gaskill, J. Cryst. Growth 151 (1995) 396. [15] S. Christiansen, M. Albrecht, W. Dorsch, H.P. Strunk, C. Zanotti-Fregonara, G. Salviati, A. Pelzmann, M. Mayer, M. Kamp, K.J. Ebeling, MRS Internet J. Nitride Semicond. Res. 1 (1996) 19. [16] S. Clur, O. Briot, J.L. Rouvie`re, A. Andenet, Y.-M. Le Vaillant, B. Gil, R.L. Aulombard, J.F. Demangeot, J. Frandon, M. Renucci, MRS Symp. Proc. 468, San Francisco, 1997, in press.