Rare earth co-doping nitride layers for visible light

Materials Chemistry and Physics 134 (2012) 716e720

Contents lists available at SciVerse ScienceDirect

Materials Chemistry and Physics journal homepage: www.elsevier.com/locate/matchemphys

Rare earth co-doping nitride layers for visible light J. Rodrigues a, *, S.M.C. Miranda b, N.F. Santos a, A.J. Neves a, E. Alves b, K. Lorenz b, T. Monteiro a a b

Departamento de Física & I3N, Universidade de Aveiro, Campus Universitário de Santiago, 3810-193 Aveiro, Portugal Instituto Tecnológico e Nuclear, Instituto Superior Técnico, Universidade Técnica de Lisboa, Estrada Nacional 10, 2686-953 Sacavém, Portugal

a r t i c l e i n f o

a b s t r a c t

Article history: Received 28 September 2011 Received in revised form 22 February 2012 Accepted 23 March 2012

AlN and GaN films were co-implanted with europium and praseodymium ions followed by thermal annealing treatments for the lattice recovery and ions activation. Their structural and optical properties were studied by Rutherford Backscattering Spectrometry and optical spectroscopy techniques. The evolution of the ions photoluminescence intensity with temperature was analyzed for both AlN:Eu,Pr and GaN:Eu,Pr hosts, and compared with the ones of individually doped layers (GaN:Eu, GaN:Pr, AlN:Eu and AlN:Pr). Sharp emission lines due to intra-4fn shell transitions can be observed even at room temperature for the Eu3þ and Pr3þ for both hosts. Ó 2012 Elsevier B.V. All rights reserved.

Keywords: Nitrides Rare earth Ion implantation Photoluminescence spectroscopy

1. Introduction Nitride systems such as GaN, InN and AlN layers are known to be important technological materials for lighting applications [1e4]. During the last decades, the increased market demands for high efficient visible light sources, have promoted an intense research on these materials promoting bandgap engineering using the ternary alloys and tuning the electron-hole recombination at different wavelengths [2e4]. Another approach to tune visible light emission is based on the doping with efficient activators in the nitride hosts which lead to the desired light emission. Among these activators, rare earth ions (RE) assume an important role in nitrides, due to their narrow atomic-like transitions and the span of their emission lines from ultraviolet to infrared, depending on the ion configuration [5e8]. In the last years, several nitride systems have been doped with RE ions either during growth or by ex-situ methods using ion implantation and subsequent thermal annealing treatments [8]. Europium ions are among the most studied rare earths incorporated in the nitrides either by in-situ or ex-situ doping processes. Due to their intra-4f 6 electronic configuration, the visible emission from Eu3þ ions arises from transitions between the crystal-field split energy levels of the 5D0,1 and 7FJ manifolds, which occur in the orange-red spectral region [5e14]. Recently, an efficient Eu3þ related light emitting diode was reported for europium-

* Corresponding author. Tel.: þ351 234 370 299; fax: þ351 234 378 197. E-mail address: [email protected] (J. Rodrigues). 0254-0584/$ e see front matter Ó 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.matchemphys.2012.03.056

doped GaN host [5]. Despite the fact that several works have been published concerning Eu3þ luminescence in nitrides [5e14], much less attention has been paid to the co-doping effects with this ion and other lanthanide ions in order to improve the potentialities of the RE emitters in the nitride systems. In this work, we have co-doped AlN and GaN layers with europium and praseodymium ions, both important activators in the red spectral region in several wide bandgap hosts [5e20]. The nitride layers were implanted with both ions followed by thermal annealing treatments in order to recover the damage and promote the RE ion activation. It was found that, the co-doping process leads to simultaneous Eu3þ and Pr3þ emission, widening the contribution of the red spectral range light provided by both ions. The evolution of the ions photoluminescence intensity with temperature was analyzed for both AlN:Eu,Pr and GaN:Eu,Pr hosts, and compared with the ones of individually doped layers (GaN:Eu, GaN:Pr, AlN:Eu and AlN:Pr). 2. Experimental details Commercial GaN/Al2O3 and AlN/Al2O3 (Technologies and Devices International Inc.) templates grown by hydride vapor phase epitaxy were used in the current experiments. The samples were implanted with 150 keV praseodymium and subsequently with 150 keV europium ions both with a fluence of 8  1014 cm2. Postimplant annealing to remove implantation damage and optically activate the RE ions was performed at 1000  C during 20 min in flowing N2 and placing a piece of GaN/sapphire or AlN/sapphire

J. Rodrigues et al. / Materials Chemistry and Physics 134 (2012) 716e720

face to face with the samples as a proximity cap to protect the surface during the high temperature treatment. Additional GaN and AlN layers doped separately with Eu3þ and Pr3þ ions and annealed with similar conditions were studied previously [11,17] and are used for comparison purposes. Rutherford Backscattering Spectrometry/Channelling (RBS/C) studies were performed with a 1 mm diameter collimated beam of 2 MeV Heþ ions. The backscattered particles were detected at 165 with respect to the incoming beam direction using a pin-diode detector. Steady state photoluminescence (PL) was generated using the 325 nm light from a cw HeeCd laser and an excitation power density less than 0.6 W.m2. The samples were mounted in a cold finger of a closed-cycle helium cryostat and the sample temperature was controlled in a range from 14 K to room temperature (RT). The luminescence was measured using a dispersive system SPEX 1704 monochromator (1 m, 1200 mm1) fitted with a cooled Hamamatsu R928 photomultiplier tube. For the PL excitation (PLE) spectra a modular double grating excitation spectrofluorimeter with a TRIAX 320 emission monochromator (Fluorolog-3, Horiba Scientific) coupled to an R928 Hamamatsu photomultiplier was used. The excitation source was a 450 W Xe arc lamp. The emission spectra were corrected for detection and optical spectral response of the spectrofluorimeter and the excitation spectra were corrected for the spectral distribution of the lamp intensity using a photodiode reference detector.

Fig. 2. Normalized 14 K PL spectra of Eu and Pr doped nitride layers obtained with 325 nm excitation. Full and open circles stand for separated GaN layers doped with Eu3þ (GaN:Eu) and Pr3þ(GaN:Pr), respectively, full and open triangles for separated AlN layers doped with Eu3þ (AlN:Eu) and Pr3þ (AlN:Pr) respectively. Red and blue lines correspond to the spectra of GaN and AlN samples co-doped with both ions (GaN:Eu,Pr; AlN:Eu,Pr), respectively. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

a

Counts

2000 1500 1000 500 0 100

F2

3

F2

14 K

RT R

200

300

400 550

600

650

615

700

620

625

650

655

660

Wavelength (nm)

Channel

RE concentration (at-%)

Pr : P0

7

PL intensity (arb. units)

random Pr+Eu aligned Pr+Eu random Pr aligned Pr

2500

b

3+ 3

3+ 5

Eu : D0

3000

b

0.20

Pr Pr+Eu

0.15

0.10

0.05

Normalized integrated intensity

a

717

1.4 5

D0

3

P0

1.2

7

F2

3

F2

1.0

0.8

0.6

0

50

100

150

200

250

300

0.00 0

50

100

150

200

250

300

Depth (nm) Fig. 1. (a) RBS/C random and aligned spectra of the AlN:Pr and AlN:Eu,Pr layers. (b) RE profiles extracted from the spectra in Fig. 1a.

Temperature (K) Fig. 3. (a) Temperature dependent PL spectra of the Eu3þ and Pr3þ highest intensity transitions for the co-doped GaN layer. The spectra were obtained with above bandgap excitation. (b) Evolution of the integrated intensity of the 5D0 / 7F2 (Eu3þ) and the 3 P0 / 3F2 (Pr3þ) lines with the temperature.

718

J. Rodrigues et al. / Materials Chemistry and Physics 134 (2012) 716e720

3. Results and discussion Fig. 1(a) shows RBS/C random and aligned spectra for AlN, after implantation of Pr only and after co-implantation with Pr and Eu. The RE profiles extracted from the random RBS spectra are shown in Fig.1(b). In the case of GaN the maximum RE concentration is found at w20e30 nm depth while the same implantation conditions lead to a wider profile in AlN with a range of w50 nm. The maximum concentrations in the co-implanted samples are 0.18 at% and 0.1 at% for GaN and AlN, respectively. In both cases the RE profiles do not change after annealing. Note that it is not possible to distinguish between Eu and Pr due to their similar mass. The RBS/C spectra in Fig. 1 reveal the expected increase of backscattering yield from Al within the implanted region due to implantation defects. However, the effect is much stronger for GaN than for AlN (not shown). As a measure for the crystalline quality and level of implantation damage we calculate the minimum yield (yield in the aligned spectrum divided by the yield in the random spectrum) for Ga and Al within a window comprising the entire implanted region. The minimum yield cmin for Al within the implanted area increases from 2% in the as-grown to 5% for the AlN:Pr sample and to 7% for the co-implanted sample. The Ga minimum yield for the as-grown sample is similar to that of as-grown and around 2%. However, after Pr-implantation cmin increases to 18% and reaches 39% for the co-implanted sample. Annealing recovers the lattice damage only partly resulting in minimum yields of 6% and 30% for AlN and GaN, respectively.

It is well established that RE ions in nitride layers are essentially incorporated in substitutional sites or in sites slightly displaced from the substitutional ones [8e12,17]. The fraction of the RE in these near-substitutional sites, fs, can be estimated by

fs ¼




1  cRE min

.
 1  cAl min

(1)

Directly after implantation, values of fs of 87% and 86% were obtained for the AlN:Pr and AlN:Eu,Pr layers, respectively. For GaN the substitutional fraction is estimated to be 89% for simple Primplantation but only 69% for Pr/Eu co-implantation. In both materials the substitutional fraction decreases slightly after annealing possibly due to the formation of RE-defect clusters. Fig. 2 shows the normalized 14 K PL spectra of the europium and praseodymium co-implanted and annealed GaN and AlN layers. The intensity was normalized to the emission maxima of the rare earth ions and the PL was obtained by above (for the GaN layers) and below (for the AlN layers) bandgap excitation using the 325 nm line of a HeeCd laser. For comparison purposes we have added on the Figure the corresponding spectra of the GaN:Eu, GaN:Pr, AlN:Eu and AlN:Pr doped systems. After implantation and annealing the co-doped GaN:Eu,Pr and AlN:Eu,Pr samples exhibit simultaneously the 5D0 / 7F2 and the 3P0 / 3F2 transitions corresponding to the highest intensity transitions of the Eu3þ and Pr3þ ions in the nitride layers, respectively [9,11]. For the Eu3þ luminescence in the GaN:Eu,Pr layer, the 5D0 / 7F2 transition is resolved in three main

Fig. 4. (a) Temperature dependent PL spectra of the Eu3þ and Pr3þ highest intensity transitions for the co-doped AlN layer. The spectra were obtained with below bandgap excitation. (b) Evolution of the integrated intensity of the 5D0 / 7F2 (Eu3þ), the 3P0 / 3F2 (Pr3þ) and 3P1 / 3H5 (Pr3þ) lines with the temperature. (c) RT PLE spectra for the AlN:Eu,Pr sample monitored at the 5D0 / 7F2 and the 3P1 / 3H5 transitions of the Eu3þ and Pr þ ions. (d) Enlarged 14 K PL spectra for co-doped AlN and GaN layers obtained with the 325 nm line of a HeeCd laser (blue and red solid lines, respectively) and 265 nm excitation (black solid line). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

J. Rodrigues et al. / Materials Chemistry and Physics 134 (2012) 716e720

lines with strongest intensity for the 621.6 and 622.5 nm lines which have been recently assigned as one of the dominant Eu3þ centers (Eu1) in GaN [8,10]. The observed results in the co-doped sample indicate that as for the GaN:Eu layers, the coimplantation followed by the annealing procedure at 1000  C, favors the formation of the Eu1 centre. For the praseodymium emission in the GaN:Eu,Pr layer, the 3P0 / 3F2 transition occurs at the same energy as the one detected in GaN:Pr layers with the main lines w650 nm and w653 nm [9,11]. Therefore, despite the stronger lattice damage and reduced substitutional fraction in the co-doped sample the peak position and spectral shape of the intraionic luminescence are comparable to samples implanted only with one ion species and lower total fluence, indicating similar site symmetry of the optical active Eu and Pr centres in co-implanted and individually implanted GaN and AlN samples. The temperature dependence of the PL intensity for the 5 D0 / 7F2 (Eu3þ) and the 3P0 / 3F2 (Pr3þ) lines for the co-doped nitride layers is depicted in Figs. 3 and 4. For the GaN layers doped with Eu and Pr ions separately, several reports indicate that the intraionic luminescence decreases gradually with increasing temperature [8,9,11]. As shown in Fig. 3, for the GaN layer co-doped with Eu and Pr ions the integrated intensity of the 5D0 / 7F2 and 3 P0 / 3F2 transitions follow the same trend, both increasing up to w100 K and for higher temperatures a decrease of the luminescence is observed. Such behavior in the 14 Ke100 K temperature range could be explained by a thermal population of both RE ions by thermal detrapping of shallow trap levels. This interpretation is in good agreement with the elevated level of residual implantation defects after annealing in this sample and the broad defect bands detected by PL (Fig. 4). For higher temperatures, the enhancement of the nonradiative recombination processes leads to a decrease of the intraionic PL intensity. Fig. 4(a) and (b) shows the temperature dependent PL spectra and the behavior of the integrated intensity for the RE transitions in the AlN:Eu,Pr layer. In contrast with the GaN:Eu,Pr layer, for which the luminescence is obtained upon above GaN bandgap excitation, in this case the luminescence was excited by photons with below bandgap energy as shown in the RT PLE spectra of Fig. 4c). Eu3þ luminescence is excited via the previously reported excitation bands (X2) and (X1) with maxima at w265 nm and 345 nm, respectively [10,11,21e24]. By monitoring the 3P1 / 3H5 transition of Pr3þ ion in the co-doped sample a main broad excitation band peaked near the same peak position of the X2 band and a small intensity shoulder at w320 nm can be recognized in the spectrum. These peak positions are close to the energy values earlier stated for Pr3þ in AlN layers [23,24]. Additionally, an excitation band in the shortest wavelength range is also known to populate a DAP transition involving Al vacancies in AlN layers, which is responsible for a broad luminescence band at w450 nm [25]. With 325 nm below bandgap excitation (the used conditions for the PL measurements) we fed simultaneously the excited levels of both ions in the codoped sample. Despite the fact that the Pr3þ ions could be excited at shorter wavelengths their emission intensity is mainly hidden by the strong AlN defect band. With 265 nm excitation besides the violet/blue AlN defect band, only the europium luminescence could be distinguished (Fig. 4d). While for the GaN:Eu,Pr layer the Eu3þ and Pr3þ peak intensities at low temperature are similar, for the AlN:Eu,Pr layer the Pr3þ emission has a lower intensity than that of Eu3þ as expected from the PLE data (Fig. 4c). Additionally, while for the GaN:Eu,Pr layer the temperature dependence of the integrated intensity for the main emission lines of both RE ions follow the same trend, for the case of the AlN:Eu,Pr layer a distinctly different behavior is observed for both ions. In particular, the decrease in intensity of the Eu3þ 5D0 / 7F2 transition is described by two distinct nonradiative paths in a fair agreement with previous

719

results for the AlN:Eu layers [11]. On the other hand, the decrease of the intensity for Pr3þ luminescence, coming from both the 3P0 and 3 P1 multiplets, reveals that the luminescence quenching can be accounted for a single nonradiative extinction mechanism. 4. Conclusions GaN and AlN were co-implanted with Eu and Pr ions. Postimplant annealing to remove implantation damage and optically activate the RE ions was performed. RBS/C results show that the minimum yield cmin for GaN within the implanted area is higher than the one observed for AlN. Annealing recovers the lattice damage only partly. In both materials the substitutional fraction decreases slightly after annealing. After implantation and annealing the co-doped samples exhibit the characteristic transitions in the red spectral region corresponding to the highest intensity transitions of the RE ions in the nitride layers. The emission properties of both samples are comparable to the ones implanted only with one ion and lower total fluence. Temperature dependent PL measurements show that the two samples exhibit different behavior of the integrated intensity for the RE transitions. In the case of the GaN sample the integrated intensity of the 5D0 / 7F2 and 3P0 / 3F2 transitions follow the same trend, both increasing up to w100 K and decreasing for higher temperatures. In the AlN sample two distinct nonradiative paths describe the decrease of the Eu3þ transition and only one nonradiative extinction mechanism is associated with the Pr3þ ions. From our experimental results no indications for an energy transfer between Eu and Pr ions have been found. Acknowledgments Funding by FCT Portugal (Ciência 2007, PTDC/CTM/100756/2008 and PEst-C/CTM/LA0025/2011) is gratefully acknowledged. J. Rodrigues thanks FCT for her PhD grant, SFRH/BD/76300/2011. Acknowledgement is due to R.A.S. Ferreira and V.T. Freitas for their help in the PLE experiments. References [1] F.A. Ponce, D.P. Bour, Nature 386 (1997) 351e359. [2] S. Nakamura, S.J. Pearton, G. Fasol, The Blue Laser Diode, Spring-Verlag, New York, 2000. [3] T. Steiner, Semiconductor Nanostructures for Optoelectronic Applications, Artech House, Inc, 2004. [4] H. Morkoç, Handbook of Nitride Semiconductors and Devices, Wiley-VCH, 2008. [5] A. Nishikawa, T. Kawasaki, N. Furukawa, Y. Terai, Y. Fujiwara, Appl. Phys. Exp. 2 (2009) 71004(1)e71004(3). [6] J. Heikenfeld, M. Garter, D.S. Lee, R. Birkhahn, A.J. Steckl, Appl. Phys. Lett. 75 (1999) 1189e1191. [7] Y.Q. Wang, A.J. Steckl, Appl. Phys. Lett. 82 (2003) 502e504. [8] K.P. O’Donnell, Rare Earth Doped III-Nitrides for Optoelectronic and Spintronic Applications, Springer, 2010, (and references therein). [9] T. Monteiro, C. Boemare, M.J. Soares, R.A. Sá Ferreira, L.D. Carlos, K. Lorenz, R. Vianden, E. Alves, Physica B 308e310 (2001) 22e25. [10] K. Wang, K.P. O’Donnell, B. Hourahine, R.W. Martin, I.M. Watson, K. Lorenz, E. Alves, Phys. Rev. B 80 (2009) 125206(1)e125206(2). [11] K. Lorenz, E. Alves, F. Gloux, P. Ruterana, M. Peres, A.J. Neves, T. Monteiro, J. Appl. Phys. 107 (2010) 023525(1)e023525(4). [12] I.S. Roqan, K.P. O’Donnell, R.W. Martin, P.R. Edwards, S.F. Song, A. Vantomme, K. Lorenz, E. Alves, M. Bo ckowski, Phys. Rev. B 81 (2010) 85209(1)e85209(5). [13] N. Woodward, J. Poplawsky, B. Mitchell, A. Nishikawa, Y. Fujiwara, V. Dierolf, Appl. Phys. Lett. 98 (2011) 011102(1)e011102(3). [14] W.M. Jadwisienczak, H.J. Lozykowski, I. Berishev, A. Bensaoula, I.G. Brown, J. Appl. Phys. 89 (2001) 4384e4390. [15] R. Birkhahn, M. Garter, A.J. Steckl, Appl. Phys. Lett. 74 (1999) 2161e2163. [16] H.J. Lozykowski, W.M. Jadwisienczak, I. Brown, J. Appl. Phys. 88 (2000) 210e222. [17] M. Peres, S. Magalhães, N. Franco, M.J. Soares, A.J. Neves, E. Alves, K. Lorenz, T. Monteiro, Microelectr. J. 40 (2009) 377e380.

720

J. Rodrigues et al. / Materials Chemistry and Physics 134 (2012) 716e720

[18] E. Brown, U. Hömmerich, T. Yamada, H. Yamane, J.M. Zavada, J. Alloys Compounds 488 (2009) 628e631. [19] M.R.N. Soares, C. Nico, M. Peres, N. Ferreira, A.J.S. Fernandes, T. Monteiro, F.M. Costa, J. Appl. Phys. 109 (2011) 013516(1)e013516(5). [20] M.R.N. Soares, C. Nico, J. Rodrigues, M. Peres, M.J. Soares, A.J.S. Fernandes, F.M. Costa, T. Monteiro, Opt. Mater. 34 (2011) 27e29. [21] S. Magalhães, M. Peres, V. Fellmann, B. Daudin, A.J. Neves, E. Alves, T. Monteiro, K. Lorenz, J. Appl. Phys. 108 (2010) 084306(1)e084306(6).

[22] M. Peres, S. Magalhães, J. Rodrigues, M.J. Soares, V. Fellmann, A.J. Neves, E. Alves, B. Daudin, K. Lorenz, T. Monteiro, Opt. Mat 33 (2011) 1045e 1049. [23] H.J. Lozykowski, W.M. Jadwisienczak, A. Bensaoula, O. Monteiro, Microelectr. J. 36 (2005) 453e455. [24] H.J. Lozykowski, W.M. Jadwisienczak, Phys. Stat. Sol. (B) 244 (2007) 2109e2126. [25] T. Schultz, PhD Dissertation, Tech. Univ. Berlin (2010).