Spectroscopic Analysis of Eu3+ Implanted and Annealed GaN Layers

Page 1 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Spectroscopic Analysis of Eu3+ Implanted and Annealed GaN Layers and Nanowires

J. Rodrigues1, M. F. Leitão1, J. F. C. Carreira1, N. Ben Sedrine1, N. F. Santos1, M. Felizardo2, T. Auzelle3, B. Daudin3, E. Alves2, A. J. Neves1, M. R. Correia1, F. M. Costa1, K. Lorenz2, T. Monteiro1* *[email protected], phone: +351 234370824

1

Departamento de Física & I3N, Universidade de Aveiro, Campus Universitário de

Santiago, 3810-193 Aveiro, Portugal 2

IPFN, Instituto Superior Técnico, Universidade de Lisboa, Campus Tecnológico e

Nuclear, EN10, 2695-066 Bobadela LRS, Portugal 3

Univ. Grenoble Alpes, INAC-SX, SP2M, F-38000 Grenoble, France

1 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 37

Abstract In this work, a detailed spectroscopic analysis of Eu3+ implanted and annealed GaN nanowires (NWs) and layers is presented by using temperature dependent steady-state photoluminescence, room temperature photoluminescence excitation and time resolved photoluminescence. Independently of the used implantation angle and ion fluence, all the studied post-implant annealed samples evidence red

5

D0→7FJ luminescence

transitions of the Eu3+ (4f 6) ions. One dominant Eu3+ optical centre was found for both GaN NWs and layers, together with the presence of different overlapping Eu3+ minority optical centers. The thermal stability of the intra-4f 6 lines was found to be higher for the NWs where the red emission is observed with the naked eye even at room temperature. Besides the lanthanide emission, the photoluminescence spectra of the NWs and layers exhibit a broad yellow luminescence band (YL) differing slightly in the spectral shape and peak position in the different samples. While the YL in the layers is commonly ascribed to a free to bound (e-A) or donor acceptor pair (DAP) transitions, the recombination kinetics of the YL in the NWs supports a model for a surface mediated recombination process.

Keywords: NWs, Eu3+, Ion implantation, YL, PL/PLE, TRPL


Page 3 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1. Introduction Rare-earth (RE)-doped semiconductors have been widely investigated as a class of materials with promising properties, especially in the optoelectronic area, taking advantage of the remarkable atomic-like intra-4f RE ions

n

transitions of the trivalent charged

1–3

. Gallium nitride (GaN) is a wide band gap semiconductor with a direct

bandgap energy of 3.4 eV at room temperature 4. GaN and its ternary alloys are wellestablished materials in solid-state optics/lighting/optoelectronic applications such as light emitting diodes (LEDs) and laser diodes (LDs) 5–8. Among the main advantages as a host for RE ions, GaN exhibits a high transparency which makes it suitable for the incorporation of the ladder-like RE3+ electronic energy levels promoting the ion optical activation with reduced luminescence thermal quenching

2,9,10

. Furthermore, the 3+

charge state is easily achieved since RE are preferentially on isovalent Ga-site

11

. The

large values for both thermal conductivity and breakdown fields provide further benefits for high temperature and high power electronic devices 10. Trivalent europium (Eu3+) ions are known to be efficient red activators in several wide bandgap hosts. Europium ions are amongst the most studied RE incorporated in nitrides either by in-situ or ex-situ doping processes. Due to their intra-4f

6

electronic

configuration, the visible emission from Eu3+ ions in the orange-red spectral region arises from transitions between the crystal-field split energy levels of the 5D0 and 7FJ manifolds

12–14

. In 2009, Nishikawa and co-workers reported the fabrication of an

efficient, low voltage Eu3+ related light emitting diode obtained by doping a GaN host 12

.

The use of nanowires (NWs) as potential nanoscale building blocks for future optoelectronic devices offers some advantages over traditional layered structures, mostly due to their ability to be grown in different, cheaper and highly mismatched



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 37

substrates, with absence of extended defects and efficient release of strain by surface relaxation

15,16

. The possibility to grow GaN NWs on silicon substrate enables the

compatibility with the nowadays existing microelectronic technology. Higher light extraction efficiency is also expected from these low dimensional structures when compared with the layers due to their non-planar geometry. This geometry is favorable for light extraction due to the decreasing in the fraction of photons experiencing total internal reflection (due to the large refractive index of GaN). In addition, the extraction efficiency strongly depends also on the density of the nanowire array as well as on the refractive index of the substrate

17

. A number of other wide bandgap semiconductor

NWs has been doped with RE ions, as is the case of ZnO 25 or LaPO4 26, revealing their potential to give rise to efficient multicolor emissions that may find applications in optoelectronic and display devices 25. To the best of our knowledge the optical properties of RE doped GaN NWs were not extensively studied by the scientific community, deserving an in-depth analysis due to their potential application in optoelectronics. Doping NWs is not an easy task due to the low diffusion coefficients of the RE ions, inhibiting ex-situ doping by diffusion 18. The probable segregation of the dopant elements to the surface of the NWs constitutes a challenge when in-situ doping is used, as was reported also for other materials

19–21

.

Considering the ex-situ methodology, one approach to dope these nanostructures is by ion implantation

18,22

. This method is a key technology in the semiconductor industry

and allows the introduction of dopants in a controlled way and without the limitations of the solubility of the dopant in the matrix 23. Previous works from our group 9,24 revealed that it is possible to achieve optical activation of the RE ions in GaN NWs by ion implantation and post-implantation thermal annealing, and showed that the emission of


Page 5 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


the doped NWs, in terms of intensity and thermal stability, is comparable with those obtained for GaN layers processed in similar conditions 9,24. In the present work different GaN structures (layers and NWs) were doped with Eu3+ by ion implantation. The samples were implanted with tilted angles of 20o and 30o relatively to the c-axis, in order to evaluate the influence of the implantation angle on the samples optical properties. Post-implantation thermal annealing was performed in the same conditions for all the samples in order to recover the lattice damage and promote the optical activation of Eu3+. The spectroscopic features of the doped structures were studied by means of Raman spectroscopy, photoluminescence (PL), PL excitation (PLE) and time resolved photoluminescence (TRPL). All the implanted and annealed samples exhibit the intra-4f 6 emission from the Eu3+ ion along with a broad unstructured yellow luminescence (YL). This work exploits the possibility of the incorporation of Eu ions in GaN NWs as an approach to contribute to the development of efficient red emitting devices at the nanoscale with comparable emission and higher thermal stability than the one found for the more commonly used layers.

2. Experimental details GaN NWs were grown by molecular beam epitaxy (MBE) on (111) Si substrates, as reported elsewhere

27

. A thin (2-3 nm) aluminium nitride (AlN) buffer layer was

deposited prior to the NWs growth which was performed in nitrogen-rich conditions. Table 1 summarizes a detailed description of the studied samples. Two sets of NWs samples (hereafter labeled NWA and NWB) were grown under similar conditions at ∼ 860 oC during ~ 13 h. Implantation of the samples was carried out using 300 keV Eu3+ ions at angles of 20° (NWA) and 30° (NWB) from the surface normal, at room temperature. The tilted geometry minimizes the number of ions reaching the Si-



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 37

substrate. The samples were rotated during the implantation to avoid bending of the NWs due to implantation-induced strain. NWA samples were implanted with fluences of 5×1013 Eu/cm2 (NWA1) and 3×1015 Eu/cm2 (NWA2), whereas the NWB samples were implanted only with the highest fluence of 3×1015 Eu/cm2. Commercial GaN layers (Lumilog), ~ 3•m thick, implanted simultaneously with the NWs were used as reference samples (GaN RefA and RefB). Monte Carlo simulations using the SRIM code

28

were performed to estimate the ion distribution in the samples.

For layered GaN samples these simulations yield values for ion range and straggle of 52.0 nm and 18.6 nm for 20º incidence angle and 48.1 nm and 18.0 nm for 30º incidence angle, respectively. In the case of the NWs, a part of the ions will be implanted into the side facets of the NWs with SRIM estimated values for range/straggle (calculated from the side facet) of 16.5 nm/10.9 nm and 21.8 nm/13.1 nm, respectively, for 20º and 30º incidence angle. Although the non-planar geometry of the NWs will influence the final distribution of ions inside the NWs

18

, these

simulations show that most of the ions will come to rest well inside the NW crystal. This assumption is also supported by recent molecular dynamics simulation for Er implantation into GaN NWs using similar parameters as those used in our experiment 29

. Post-implantation rapid thermal annealing (RTA) was performed at 1000 ºC during

30 s in flowing N2, using an ANNEALSYS halogen lamp furnace in order to recover the lattice damage caused by the implantation process and promote the optical activation of the ions.

Sample

Implantation Fluence (ion/cm2) Tilt angle (o)

Type of structure

NWA1

5x1013

20

Nanowires

NWA2

3x1015

20

Nanowires

NWB

3x1015

30

Nanowires


Page 7 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


RefA1

5x1013

20

Layer

RefA2

3x1015

20

Layer

RefB

3x1015

30

Layer

Table 1 – Implantation parameters of the studied samples. All the samples were implanted at RT with an energy of 300 keV and subject to the same thermal annealing (30 s at 1000 ºC in flowing N2).

Scanning electron microscopy (SEM) measurements for the as-grown samples were acquired using a ZEISS Ultra 55 microscope while a TESCAN Vega3 SBH SEM microscope was used for the implanted and annealed samples. The Raman spectra were performed at room temperature (RT) on a Horiba Jobin Yvon HR800 spectrometer, under the incidence of the 442 nm He-Cd (Kimmon IK Series) laser line and focusing with an objective of ×100 magnification (NA= 0.9; f= 0.1 cm). The backscattering geometry x(y,·)̅ was used to minimize the signal from Si substrate. Steady state photoluminescence (PL) was generated using the 325 nm light from a cw He-Cd laser with an excitation power density less than 0.6 W.cm-2. 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 RT. The luminescence was measured using a dispersive system SPEX 1704 monochromator (1m, 1200 gr.mm-1) fitted with a cooled Hamamatsu R928 photomultiplier tube. The RT PLE was assessed in a Fluorolog-3 Horiba Scientific modular equipment with a double additive grating scanning monochromator (2×180 mm, 1200 gr.mm-1) in the excitation and a triple grating iHR550 spectrometer in the emission (550 mm, 1200 gr.mm-1). A 450 W Xe lamp was used as excitation source. The PLE was assessed by monitoring the energy of the maxima of the PL emission lines and the excitation was 7 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 37

scanned to higher energies. The measurements were performed using front face acquisition geometry, and the presented spectra were corrected to the optical components, PMT detector and to the Xe lamp spectral responses. RT TRPL spectra were acquired with the same Fluorolog-3 system using a pulsed Xe lamp coupled to a monochromator.

3. Results and Discussion 3.1 Morphological and structural analysis Figure 1 shows SEM images of the (a) side and (b) top views of the as-grown GaN NWs (NWB). The NWs are vertically aligned exhibiting lengths between 1500 and 2300 nm and a mean diameter of 140 nm. The density of the NWs was estimated to be higher than 8.4×109 NWs/cm2. No 2D GaN layer between the buffer layer and the NWs was found for the present samples. Some coalescence between the wires is also observed due to the high density of NWs. Figure 1 (c) depicts top view SEM images of the implanted and annealed NWs: NWA1, NWA2 and NWB. As can be inferred by the SEM images of the nanostructures, the implantation and thermal annealing procedures does not seem to affect significantly the NWs morphology.

(a)

(b)


Page 9 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(c)

5×1013 Eu/cm2; 20o NWA1

3×1015 Eu/cm2; 20o NWA2

3×1015 Eu/cm2; 30o NWB

Figure 1 – (a) Side view and (b) top view SEM images of the as-grown GaN NWs (NWB before implantation). (c) SEM micrographs of the NWs NWA1, NWA2 and NWB. All the samples were annealed in the same conditions as mentioned in the main text.

The structural characterization was evaluated by Raman spectroscopy as shown in Figure 2. Raman studies revealed that the E2H phonon mode of the as-grown NWs is shifted by 2 cm-1 to lower energies relatively to the as-grown GaN layer reference sample, indicating that the NWs are under tensile strain [Figure 2 (a)]. After implantation, Raman spectroscopy measurements allowed the observation of the phonon density of states (DOS) activation induced by the damage in the crystalline lattice

30,31

[Figure 2 (b)]. After the annealing, the Raman spectrum evolved to a

spectrum dominated by first-order processes, where the GaN phonon modes of the Brillouin center (E1(TO); A1(TO) and E2), predicted by the selection rules for x(y,·)̅ scattering geometry, are expected. The measured phonon frequencies coincide with the ones obtained for the as-grown NWs, indicating that the state of strain is the same for as-grown and annealed samples and the recovery of the crystalline lattice was achieved.



GaN-Layer

GaN-Layer

-1 @ 533 cm E1(TO)

(a) A (TO) 1

-1

@ 560 cm

x(y,.)x

-1

Si-Subs @ 520 cm

H GaN-Layer

Raman Intensity ( arb.uni.)

E2

-1

@ 569 cm

λexc@ 442 nm

NW ( as-grown) GaN-Layer (as-grown)

GaN-Layer

E1(LO)

500

550

700

-1

@ 744 cm

750

800

-1

Raman shift (cm ) -1

E1(TO) @ 558cm

(b)

15

-2

x(y,.)x

NWA2-3x10 (cm ); 20º: 15 -2 NWA2-3x10 (cm ) + RTA; 20º: DOS[Davidov et al (1998)]

1.5

Raman Intensity (arb.uni.)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 37

λexc@ 442 nm

a)

Sigle et al (1997)

1.0

0.5 A1 (acoustic overtone) L

-1

E2 @ 142 cm 0.0

a)

H

(acoustic overtone)

a)

QLO

-1

GaN

A1(TO)@ 532 cm

150

300

450

-1

E2 @ 567 cm

A1+E2

600

-1

@ 738 cm

750

-1

Raman shift (cm )

Figure 2 – Raman spectra of the (a) as-grow GaN NWs and layer samples and for (b) NWA2 sample as-implanted and after annealing. The expected DOS is represented for comparison.


Page 11 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


3.2 Low temperature PL

3.2.1 UV-Vis-NIR spectral range Figure 3 displays the 14 K PL spectra of the NWs and GaN layers implanted and annealed under the same conditions. The extended PL spectra over the ultraviolet (UV), visible and near infrared (NIR) spectral range [Figure 3 (a)] were obtained with 325 nm photon excitation, corresponding to an energy above the GaN bandgap. Figure 3 (b) depicts high resolution spectra of the Eu3+ ion-related emission lines for all the samples. Independently of the implantation angle and fluence, the samples (NWs and layers) exhibit sharp red lines corresponding to the 5D0→7FJ (1-4) transitions [Figure 3 (b)], due to the intra-4f 6 transitions of the Eu3+ ion. In addition to the intraionic luminescence, other optically active centers, such as, donor-bound excitons (D0X), donor acceptor pairs (DAP) and broad visible bands can be also identified in the present samples [Figure 3 (a)]. The GaN layers show well defined D0X and DAP transitions, whereas for the NWs only the NWA1 sample, implanted with the lower fluence, evidences a broad near band edge (NBE) recombination and DAP transitions. The intensity decrease of the NBE emission associated with a strong DAP emission constitutes an indication of a high dopant concentration introduced unintentionally during the growth process, as corroborated by the spectrum of the as-grown NWs sample [Figure 3 (a)]. An increase of the broadening of the NBE is further promoted by the implantation and annealing processes. With the exception of the NWA1, and as a general trend, all the implanted and annealed NWs exhibit higher intensity ratio of the Eu3+ emission/NBE when compared with the GaN layers submitted to the same treatments. Notwithstanding, it is important to note that a higher implanted volume fraction is found in the NWs as compared with the layers, so a contribution from the unimplanted deep layers is also present.



Energy (eV) 3.2

2.8

2.4

15

o

3x10 , 20 5x1013, 20o 3x1015, 30o as-grown

λexc=325 nm 14 K DAP 5

D0→ 7F2

0

Energy (eV)

(b) 2.1

2

NWs

DX

Layers

*

Normalized intensity

3.6

350 400 450 500 550 600 650 700 750

2.05

(c)

2

λexc=325 nm 14 K

5

1.95

1.9

7

D 0 → F2 5

D 0 → 7 F4

5

D0→ 7F3

5

D 0 → 7 F1

5x1013, 20o 3x1015, 20o 3x1015, 30o

NWs Layers

590

5x1013, 20o 3x1015, 20o 3x1015, 30o

600

610

Wavelength (nm)

620

630

640

Wavelength (nm)

650

Energy (eV) 2.6

2.4 15

2.2

5

NWs

2

1.8

14 K

o

3x10 , 20 5x1013, 20o 3x1015, 30o as-grown


(a) Normalized intensity

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 37

YL D1→ 7F1

Layers GL+YL GL

450

500

550

600

650

700

Wavelength (nm) Figure 3 – (a), (b) and (c) Low temperature PL spectra of the selected NWs and layers implanted with Eu3+ ions, acquired with above bandgap excitation (He-Cd, 325 nm). Figure 3 (b) depict high resolution spectra of the ion-related emission lines while Figure (c) displays the PL spectra of the visible spectral region (green/yellow luminescence spectral region). The asterisk denotes the second order of the D0X emission.

Concerning the deep defect recombination, a special attention should be paid to the YL identified both in the implanted and annealed NWs and layers [Figure 3 (c)]. It is well established nowadays that undoped and doped GaN layers typically exhibit a broad YL 12 ACS Paragon Plus Environment

660

670

Page 13 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


described by a DAP (involving a shallow donor) or free to bound (e-A) transitions 32–38. Currently, it is commonly believed that the YL is associated with the presence of native defects namely gallium vacancies, VGa, and their VGaON complexes in the 2-/- charge states

32–38

. On the other hand, besides the YL, annealed GaN layers and Ga-face

terminated bulk samples frequently exhibit a broad green luminescence (GL) that can be ascribed to the -/0 states of the same defect

37,38

. In the same spectral region different

types of defects can be found, for instance Reshchikov and co-workers

39

reported a

study where they propose that a GL band peaked at 2.35 eV (∼ 528 nm), denominated GL2, was associated with the presence of nitrogen vacancies (VN). In the present study, some differences were found for the layers concerning the spectral shape and peak position of the broad bands in the visible region. While for the RefB sample the emission is dominated by the presence of the GL, with the YL showing a negligible intensity when compared with the remaining emissions, for the as-grown, RefA1 and RefA2 samples both GL and YL appear to be present with comparable intensities. These observations suggest the presence of different defects or the same defects in different charge states. The GL observed in RefB exhibits the a peak position and a spectral shape similar to the one measured by Reshchikov

39

. The different relative

intensity of the GL and YL is likely to be due to inhomogeneity in the defects distribution in the layers. In the case of the NWs, the YL is also identified for the implanted and annealed samples, excepting the case of the NWA2 sample, where the YL intensity is negligible when compared with the ion-related emission. It is important to note that, contrary to what was verified in the layers, this band is not present in the as-grown samples, suggesting an induced/enhanced YL recombination in GaN NWs by ion implantation and post-growth annealing, as was previously reported by our group in similar samples



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 37

40

. The presence of this YL in the implanted and annealed samples is more clearly

observed at RT due to a higher YL/ion emission intensity ratio [not shown here, but present in the RT PL/TRPL spectra (see section 3.4)]. Despite the fact that the emission occurs in the same spectral region of the YL in the layers, its spectral shape and peak position are slightly different. A similar luminescence band was also identified in NWs doped with different ions 24,40 where the emission intensity was found to be sensitive to the sample environmental atmosphere, as well as on the surface modification processes. Being so, while the YL in the layers follows the common behavior reported for the GaN host 32–36, the YL in the NWs was ascribed to a surface mediated recombination process, as suggested by previous studies 40. Furthermore, recent theoretical calculations indicate that the most stable location for the native point defects in GaN NWs is in the surface region41 and the analysis of the spatial distribution of surface defects by using panchromatic CL clearly reveals a surface layer with strong YL in thick NW42 likely to be associated with the predisposition of the defects/impurities to segregate at the surface 19

. In the case of the NWs evaluated in this work this effect will be further discussed

based on the RT recombination kinetics (section 3.4).

3.2.2 intra-4f 6 Eu3+ luminescence Enlarged spectra of the 14 K intra-shell Eu3+ luminescence for implanted and annealed NWs and GaN layers are depicted in Figure 4 (a). As currently identified in nitrides, the Eu3+ emission is dominated by the forced dipole electric transition between the 5D0 and 7

F2 multiplets

10,43–45

. The peak position and line assignments of the Eu3+ transitions

identified in the studied samples are compiled in Table 2.


Page 15 of 37


(a)

Energy (eV) 2.01

2

λexc=325 nm 14 K

1.99

1.98

1.97

5

D0→ 7F2

NWs

5x1013, 20o 3x1015, 20o 3x1015, 30o

Layers

5x1013, 20o 3x1015, 20o 3x1015, 30o

616

618

620

622

624

626

628

630

Wavelength (nm)

(b) Normalized intensity

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Energy (eV) 2.01

2

1.99

5

1.98

1.97

I0

D0→ 7F2

I0/10

NWs Layer 616

618

620

622

624

626

628

630

Wavelength (nm)

Figure 4 – (a) High-resolution spectra of the 5D0→7F2 transition at 14 K. (b) PL spectra recorded at 14 K under above bandgap excitation (325 nm) with two different excitation densities for the (a) layer and (NWs) implanted with with 3x1015 Eu/cm2 and an angle of 30o.

For the hypersensitive 5D0→7F2 transition, the energy's peak are different, depending on are observed whether in NWs or layers. As a general trend the crystal field splitted Stark



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 37

lines deviate ~1 meV to lower energies in the GaN films when compared with those in NWs. This shift could be attributed to different strain states on the GaN layers and NWs, as confirmed by the Raman measurements. A similar effect was obtained for Pr3+ doped GaN layers and NWs 9. In addition to the shift, the relative intensities between the emission lines from the 5D0 multiplet to the unfolded sublevels of the 7F2 multiplet are dissimilar for NWs and GaN layers. These findings point out that strain induced different transition probabilities between the 5D0 level and the unfolded 7F2 Stark levels, as also noticed for europium doped GaN layers with different thicknesses 46. GaN NWs Transitions

Fluence (ions.cm-2)

5 × 10 5

D0→7F1

Angle (o ) 20

20

3 × 10

30

5 × 10

5

20

20

7

D0→ F2 3 × 10

30

5 × 10 5

D0→7F3

20 20

3 × 10

30 5 × 10 5

20

D0→7F4 3 × 10

20

30

GaN Layers

Peak positions (± 0.1 nm) 598.8 600.4 601.1 602.2 598.7 601.1 602.2 598.8 601.1 602.3 618.8 620.8 621.7 622.6 618.7 620.7 621.6 622.6 618.8 619.4 620.9 621.7 622.7 632.8 634.1 632.8 634.1 632.8 634.2 660.5 662.1 664.0 665.2 656.5 660.5 662.1 664.0 665.1 656.5

600.5 601.2 602.2 601.0 598.7 601.1 621.0 622.0 622.9 620.9 621.9 622.8 621.0 621.9 622.8

634.6 634.8 632.8 634.3 662.3 664.3 665.4

664.0 665.4 662.2


Page 17 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


662.1 664.1 665.2

664.1 665.3

Table 2 – Assignments and peak position of the Eu3+ intra-shell lines for the GaN NWs and layers implanted in the same conditions.

In the case of the NWs, and regardless the used implantation angle, the most intense PL emission comes from the line peaked at ∼ 622.6 nm, while for the layers the line at

∼ 621.9 nm is the dominant one. Additionally, differences in the short wavelength region of this transition are also identified: one easily notes that for the layers a welldefined line is identified near 621 nm whereas only a shoulder slightly shifted to higher energies is observed in the NWs. Besides the main intense three lines, additional lines at

∼ 619 nm are identified in all the NWs. Such transitions are absent in the studied GaN layers. It is well established that when the ions are placed in a local crystal field environment with low symmetry, a state with J=2, such as the 7F2 multiplet, suffers a maximum fivefold splitting corresponding to the 2J+1 Stark levels of the degenerate state

47–49

.

When Eu3+ ions are introduced in the hexagonal GaN host often replace the Ga3+ ions occupying substitutional (or near- substitutional) sites with a C3v local symmetry 10,43,48. In such case only three closely spaced lines are expected to be observed for the 5

D0→7F2 transition. Particularly, under this symmetry the degeneracy of a state with J=2

cannot be fully lifted 43,46,48,49 being however unfolded in a nondegenerate A level and a two-fold degenerate E level. The fact that more than three splittings are observed for the implanted and annealed NWs implies that either the Eu3+ ions could be located in lower site symmetry than the trigonal one or that there are more than one symmetry site/environment resulting in different europium related optically active centers. In GaN 17 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 37

host, the presence of Eu3+ in more than one optically active site was already reported in the literature 1,10,50–52. In the work of Dierolf et al. 52 at least nine different incorporation sites of Eu ions were identified in doped GaN layers produced by different methods. In other works

43,50,51,53

involving GaN layers doped by ion implantation and annealed at

high temperatures two major europium luminescence centers, commonly designated by Eu1 and Eu2, emitting in the same spectral region have been found. Their spectral overlap can be lifted using different photon excitation wavelengths, since these centers 1,43,52

differ mainly in their preferential excitation pathways

. In order to clarify the

different assignments found in the literature, Table 3 depicts a summary of the different reported Eu3+ optical centers in GaN as well as the results found in the present work. It was reported that the Eu1 center generally appears at relatively low annealing temperatures and low nitrogen pressures (such as those used in this work) while Eu2 is favored by high annealing temperatures and nitrogen pressures

43,50,53

. The spectral

shape and peak positions of the Eu3+ emission lines identified in the GaN layers studied in this work are in very good agreement with previously reported Eu1 center

1,51–53

.A

careful analysis of the spectra reveals also the presence of the minority centre labelled MS8 (Eu2I), whose dominant line is at 621 nm. Another line at 621.9 nm belonging to MS8 centre can be overlapped with Eu1 centre transition. The MS8 center is commonly found in implanted samples as indicated in Table 3

1,52

. The spectra recorded for the

NWs evidence additional features in the spectral region of the 5D0→7F2 transition, with additional lines in the higher wavelength region (Figure 4). Similar to what was observed in the layers, the dominant emission corresponds to the Eu1 center, while the spectral position of the additional lines suggest the presence of more than one Eu3+ optically active minority centers, MS4 and MS8. The broad line that appears close to

∼ 619 nm can be attributed to the MS4 center, while the shoulder at 621 nm is in the


Page 19 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


same position as the dominant line in the MS8 center (see Table 3). These differences in the spectral signatures of the centers indicate the presence of different local environments for the emitting ion, which could include Eu3+ ions in substitutional Ga sites but being perturbed by different nearby defects

52

. This last case is particularly

important in the case of the NWs since ions in the same lattice site can experience different strain states depending on the distance with respect to the NWs walls, which constitutes a specific feature of NWs, as was studied in detail by Brandt et al. 21,54. Decreasing the power excitation [Figure 4 (b)], by one order of magnitude, revealed no changes in the peak position and spectral shape for the Eu3+ emission in the NWs and layers. As a general tendency it is possible to assume that the used implantation angle, fluence and annealing temperature result in a dominant Eu1 center in the GaN for both layers and NWs, overlapped with different minority centers.

Reference

Present work

Sites/Peak position (nm)

NWs (3×1015, 30o)

621.7 622.7

Layers (3×1015, 30o)

621.9 622.8 621.6 622.6 634.3

Dierolf et al. [43] ▲

Main site

Bodiou et al. [1] ◊

618.8 619.4

620.9 621.0

624.1

621.0 623.3 630.1

620.4 620.7 633.9

618.9 619.6 632.7

621.1 623.1 631.5

630.0

617.2 623.7 628.9

621.0 621.9

MS 1

MS 2

MS 3

MS 4

MS 5

MS 6

MS 7

MS 8

622.4 (m) 622.6 (I)

620.2

620.8

Eu1(m/I)

Eu2m

Eu2I

622.5

620.6

618.8

615.1

Site I

Site II

Site III

Site IV

Peng et al. [10] +

O’Donnell et al. [36,41,44] *

620.8 621.7 622.5

618.7 619.3

Eu1

Eu2

617.2

620.8 221.9

Eu2



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 37

▲ MS 2, 3, 4, 5 and 8 can be excited non-resonantly with visible light; Main site and MS 1 excited with sub-band gap UV light; Main site, MS 3, 4 and 5 excited with above band gap energy; (MS 2 and 8 preferentially found in ion-implanted samples, MS 3 preferentially found in in-situ doped samples) ◊ Eu2(m/I) dominant under low photon flux, Eu1(m/I) dominant under high photon flux and excited with both above and below band gap excitation (m stands for in-situ doped samples and I for ion-implanted samples) + Sites I and IV excited both above and below band gap; Site II excited only above band gap; Site III better observed with below band gap energy * Eu1 excited both above and below band gap; Eu2 excited only above band gap

Table 3 – Peak positions for the 5D1→7F2 transition lines as well as the Eu sites’ assignments found in the literature.

3.3 Temperature dependency of the intraionic Eu3+ luminescence Figure 5 shows the temperature dependent PL for NWB [Fig. 5 (a)] and RefB [Fig. 5 (b)] and the temperature dependence of the 5D0→7F2 integrated intensity for all the NW samples [Fig. 5 (c)] and layers [Fig. 5 (d)]. In all cases, the overall hypersensitive luminescence intensity decreases gradually with increasing temperature due to competitive thermally activated nonradiative processes. The inset in Fig. 5 (a) reveals slight differences in the temperature behavior for the dominant lines in the 5D0→7F2 transition of the NWs sample, corroborating the assumption that more than one Eu3+ center are in fact contributing for the luminescent emission. Considering the internal quantum efficiency of the overall luminescence estimated by the ratio

55 ,

the

Eu3+ emission intensity for the NWs exhibits a slightly higher thermal stability when compared with the GaN layers. At RT ∼ 50% of the low temperature intensity is still observed for all the implanted and annealed NWs.


Page 21 of 37

Energy (eV) 2.04

2

Intensity (arb. units)

15

1.96

o

NWs 3x10 , 30 5 D0→ 7F2 14 K RT

D0→ 7F1

5

590

600

5

610

620

1.92

2.08

2.04

2 15

618

7

D0→ F

630

Energy (eV)

(b)

1.88

620

622

624

Wavelength (nm) 5 7 0 3

D → F4

640

650

660

5

14 K

1.92

1.88

D0→ 7F2

RT 5

5

D0→ 7F1

590

670

1.96 o

Layer 3x10 , 30


2.08


(a)

600

610

D0→ 7F4

5

D0→ 7F3

620

630

640

650

660

670

Wavelength (nm)

Wavelength (nm)

(c)

(d)

1.0

NWs 0.8 0.6 0.4

13

o

5x10 , 20 3x1015, 20o 3x1015, 30o

0.2 0.0

0

50

100

150

200

250

300

Normalized integrated area

Normalized integrated area

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1.2

5x1013, 20o 3x1015, 20o 3x1015, 30o

1.0 0.8 0.6 0.4 0.2

Layers

0.0

0

50

Temperature (K)

100

150

200

250

Temperature (K)

Figure 5 – Temperature dependent PL of the lines of the Eu3+ ion for the (a) NWs and (b) layer samples implanted with 3x1015 Eu/cm2 and an angle of 30o. The spectra were obtained upon 325 nm excitation. (c) and (d) Integrated intensity of the 5D0 →7F2 transition as a function of temperature for all the NWs and layers, respectively.

3.4 RT luminescence studies

3.4.1 PLE and excitation energy dependent PL The evaluation of the preferential excitation paths for the intraionic luminescence in both kinds of studied samples was performed at RT via PL excitation measurements. Figure 6 (a) and (b) displays the PLE spectra for the NWs and layers, respectively, when monitored at the 5D0→7F2 transition. As the emission is spectrally overlapped with 21 ACS Paragon Plus Environment

300


the aforementioned YL [Figure 3(c)], the PLE also measures the excitation pathways for the YL recombination. Similar excitation spectral features are observed for all the samples. For the NWs [Figure 6 (a)] the preferential population of the ions’ energy levels can be made via a broad subgap excitation band with onset absorption at

∼ 425 nm. The band extends to shorter wavelengths overlapping with an unclear GaN band edge absorption. A subgap non-resonant excitation both in the visible and in the UV for the Eu3+ emission was previously reported by Dierolf et al. 52 and K. Wang et al. 45

. 3

2.8

2.6

(b)

2.4

350

F0,1→5D3

400

o

5x10 , 20 3x1015, 20o 3x1015, 30o

PLE@ 621 nm NWs

7

7

7

F0→5L6

13

450

500


F0,1→5L10

3.2

550

Energy (eV) 3.2 15

NWs 3x10 , 30

2.8

2.4

400

450

500

550

5

7

D0→ F2

600

650

2.4

5x1013, 20o 3x1015, 20o 3x1015, 30o

3x1015, 20o

350 400 450 500 Wavelength (nm)

400

450

Wavelength (nm)

500

550

Energy (eV) 3.6

3.2

o

PLE@ 621 nm PLE@ 612 nm PLE@ 662 nm PL@ 325 nm PL@ 350 nm PL@ 366 nm PL@ 390 nm PL@ 407 nm

350

(d)

2

2.8

Layers PLE@ 621 nm

2.8

2.4

PLE@ 580 nm PLE@ 621 nm PL@ 325 nm PL@ 350 nm PL@ 366 nm PL@ 390 nm PL@ 407 nm


3.6

3.2

350

Wavelength (nm)

(c)

Energy (eV) 3.6


Energy (eV) 3.6 3.4


(a)


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 37

350

400

450

500

2 15

Layer 3x10 , 30

5

D0→ 7F2

550

600

Wavelength (nm)

Wavelength (nm)

Figure 6 – RT PLE spectra for the GaN sample monitored at the 5D0→7F2 transition of the Eu3+ ions: (a) NWs and (b) layers implanted with Eu. RT PL/PLE spectra (c) for the 22 ACS Paragon Plus Environment

o

650

Page 23 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(c) NWs and (d) layers implanted with 3x1015 Eu/cm2 and 30o. The inset in Figure 6 (b) is a magnified view of the PLE spectrum of one of the layers, showing the broad excitation band as well as the lines/shoulders with the arrows indicating the possible position of the Eu3+ ion’s energetic levels.

In the case of the GaN layers [Figure 6 (b)], and despite the presence of a similar subgap excitation band with low intensity, the preferential population mechanism occurs via the well-defined GaN band edge absorption. This behavior indicates, in both cases, an efficient energy transfer from the GaN host to the Eu3+ defects as reported in the literature

10

. In addition, several lines/shoulders overlapped with the broad excitation

band, likely to be associated with the higher energetic levels of the Eu3+ ions, can be identified, especially in the case of the NWs. For instance, the maximum at

∼

350 nm

could be associated with the 7F0, 1→5L10 transition, while the shoulders at ∼ 390 nm and

∼ 407 nm may correspond to the 7F0→5L6 and 7F0,1→5D3 transitions, respectively. In order to explore this hypothesis, excitation energy dependent PL was performed and the resulting spectra for the samples with the highest RT PL intensity are depicted in Figure 6 (c) and (d). It is interesting to note that, in the case of the NWs, exciting the samples with an energy equal or above the GaN bandgap promotes, besides the intraionic emission, the observation of the YL. However, the latter becomes almost negligible by pumping the NWs with below bandgap energy. Nevertheless, subgap excitation with wavelength photons of 390 nm and 407 nm clearly promote the population of the Eu3+ emitting levels. Unfortunately, due to the low intensity of the RT emission under Xe lamp excitation (low excitation density) and therefore necessary large width of the slits used in the signal detection, we are unable to resolve the spectral shape of the intraionic recombination. For the case of the GaN layers a strong contribution of the broad



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 37

emission band was identified for both above and below bandgap excitations masking the resonant excitation paths via the Eu3+ lines. It was observed that with above bandgap excitation by the He-Cd laser (325 nm), the luminescence is dominated by the YL in all the layers. When excited at the bandgap energy (wavelength 366 nm) the strong YL completely dominates the PL spectrum, hindering the observation of the ion-related emission. For the other excitation wavelengths (325, 350, 390 and 407 nm), besides the broad band, it was also possible to identify the ion 5D0→7F2 transition. When excited with below bandgap excitation, the broad band exhibits a slightly shift towards higher energies (∼ 10 nm), corroborating the assumption that different luminescence centres emitting in the green/yellow spectral region are present in the GaN layers, as mentioned previously.

3.4.2 Time resolved photoluminescence Time resolved spectroscopy constitutes a powerful tool to explore the dynamics of the identified optical centers. As such, RT TRPL spectra for the highest fluence implanted NWs and GaN layers are shown in Figure 7 (a) to (d) for different time delays for above [(a) and (b)] and below [(c) and (d)] bandgap excitation. The samples were excited with 325 nm and 390 nm photon excitation and the spectra were obtained with different delay times after the pulse lamp for a fixed time window of 0.5 ms. In both samples, the intraionic emission and the YL are observed. However, as indicated by the PLE spectra with below bandgap excitation, the intensity of the YL emission is much weaker than the one observed when excited with 325 nm. In all the cases, the spectra clearly evidence that the Eu3+ emission vanishes for delays higher than 1 ms, meaning that a shorter lifetime (hundreds of µs) is expected for the 5D0→7F2 transition for both NWs and layers, corresponding to the 1/e decrease of its intensity. The decay is similar in 24 ACS Paragon Plus Environment

Page 25 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


NWs and GaN layers and is in the same order of magnitude of the values reported in the literature for layers

43,44,50

. The most noticed difference evidenced from the TRPL

spectra is related with the kinetics of the YL in NWs and GaN layers. Here, taking into account the results for above bandgap excitation, despite the identification of the luminescence in similar spectral range for NWs and layers (see also Figure 3(c) corresponding to steady state PL at 14 K), the TRPL spectra clearly show that the YL in NWs and layers present different recombination kinetics. As aforementioned, the YL in GaN layers have been ascribed to e-A and/or DAP recombination involving VGa complexes

32–36

. Such recombination models imply, in the case of DAP transitions, a

nonexponential decay dependent on the distance separating shallow donors and deep acceptors in the host. This means that short (long) lived transitions are expected for closest (apart) pairs due to the high (low) recombination probability provided by the overlap of the carriers’ wave functions. On the other hand, e-A transitions typically exhibit a high transition probability, and therefore a short lifetime. The recombination of an e-A transition involving the same acceptor of a DAP is commonly deviated to shorter wavelengths due to the ionization of the shallow donor involved. Figure 7 (e) shows a comparison between the TRPL spectra of GaN NWs and layers for the time delay of 0.05 ms and fixed time window of 0.5 ms obtained with above bandgap excitation. As already observed (see Figure 3 (c)), the spectra clearly evidence that the YL in the NWs have a slightly different spectral shape and peak position when compared with those measured for the layers. On the other hand, the emission in the NWs is shifted to longer wavelengths excluding the hypothesis that the measured fast decay could be related with the e-A transition from the same defect identified in the layers. In fact, while the YL in the GaN layers follows the tendency of a DAP recombination as found in the literature

56

, the YL in the NWs evidence a very short



lifetime, faster than the intraionic emission, in the range of tens/hundreds of microseconds or lower. As such, the kinetics data are consistent with a different recombination model for the YL in the NWs, supporting previous reports which assign

D0→ 7F2

λexc=325 nm

(b)

5

Layer 3x1015, 30o

D0→ 7F2

λexc=325 nm

0.05 0.1 0.2 0.5 0.7

Ti m

1

e

e

de la y

de la y

(m

(m

s)

s)

0.05 0.1 0.2 0.5 0.7

1 5

5 550

600

650

450

500

NWs 3x1015, 30o


(c)

550

600

650

Wavelength (nm)

Wavelength (nm)

5

D0→ 7F2

λexc=390 nm

(d)

Layer 3x1015, 30o

5

D0→ 7F2

λexc=390 nm

0.05 0.1 0.2 0.5 0.7

(m s)

(m s)

0.05 0.1 0.2 0.5 0.7

de la y

de la y

1

e

1

5 450

500

550

600

Wavelength (nm)

650


500

5 450

500

550

600

650

Wavelength (nm)


Ti m

450

Ti m

5

NWs 3x1015, 30o


(a)


the emission to surface states mediated recombination 24,40.

Ti m e

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 37

Page 27 of 37

Energy (eV)

(e) Normalized intensity

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


2

2.2

2.4

2.6

2.8

3

5

D0→ 7F2

NWs Layer

3x1015, 30o Delay 0.05 ms λexc=325 nm RT

400

450

500

550

600

Wavelength (nm)

650

Figure 7 – RT TRPL measurements for the (a) and (c) NWs and (b) and (d) layer implanted with 3x1015 Eu/cm2 and 30o. The measurements were obtained under (a) and (b) 325 nm (above bandgap excitation) excitation and (c) and (d) 390 nm (below bandgap excitation) of a flash lamp. (e) Normalized RT PL spectra of the NWs and layer implanted with the same conditions, obtained with a time delay of 0.05 ms and above bandgap excitation (325 nm).

4. Conclusions Europium implanted and annealed GaN NWs and layers were studied by Raman, PL, PLE and TRPL. Independently of the used implantation angle and fluence, the GaN NWs and layers annealed at 1000 oC in N2 exhibit Eu3+ emission dominated by the 5

D0→7F2 transition. NWs showed higher intensity of the ion emitting lines than the

layers implanted in the same conditions. A slightly higher thermal stability of the intraionic emission (~ 50%) was found for the implanted and annealed NWs. The intraionic emission was discussed in the basis of a dominant Eu3+ center in the GaN samples, with a contribution from minority center that differ in the layers and NWs, as suggested by the identified number of Stark lines. With above bandgap excitation, a



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 37

lifetime in the range of hundreds of microseconds was found for the Eu3+ 5D0→7F2 transition for both layers and NWs. Besides the intraionic emission the implanted and annealed NWs and layers exhibit YL bands with distinct characteristics. The YL in GaN layers follows the typical behavior of the commonly assigned e-A and DAP recombination models in GaN samples. Contrarily, the YL identified in the NWs with a faster decay than the intraionic emission place in evidence a distinct recombination model supporting a surface state mediated recombination process, likely present in these NWs with high surface/volume ratio.

Acknowledgements This work is funded by FEDER funds through the COMPETE 2020 Programme and National Funds through FCT - Portuguese Foundation for Science and Technology under the project UID/CTM/50025/2013. The

authors

also

NAN/2156/2012,

acknowledge

financial

PTDC/FIS-NAN/0973/2012

support and

from

FCT:

PTDC/CTM-

RECI/FIS-NAN/0183/2012

(FCOMP-01-0124-FEDER-027494). J. Rodrigues and N. F. Santos thank to FCT for their PhD grants, SFRH/BD/76300/2011 and SFRH/BD/90017/2012. K. Lorenz acknowledges funding by the program FCT Investigador.

References

(1)

Bodiou, L.; Braud, A.; Doualan, J.-L.; Moncorgé, R.; Park, J. H.; Munasinghe, C.; Steckl, A. J.; Lorenz, K.; Alves, E.; Daudin, B. Optically Active Centers in Eu Implanted, Eu in Situ Doped GaN, and Eu Doped GaN Quantum Dots. J. Appl. Phys. 2009, 105, 043104. 28 ACS Paragon Plus Environment

Page 29 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(2)

Fleischman, Z.; Tafon, P. S.; Dierolf, V.; Munasinghe, C.; Steckl, A. J. Identification of Defect-Trap-Related Europium Sites in Gallium Nitride. Phys. Status Solidi (c) 2007, 4, 834–837.

(3)

Kenyon, A. J. Recent Developments in Rare-Earth Doped Materials for Optoelectronics. Prog. Quant. Electron. 2002, 26, 225–284.

(4)

Suski, T.; Perlin, P.; Teisseyre, H.; Leszczyński, M.; Grzegory, I.; Jun, J.; Boćkowski, M.; Porowski, S.; Moustakas, T. D. Mechanism of Yellow Luminescence in GaN. Appl. Phys. Lett. 1995, 67, 2188–2190.

(5)

Ponce, F. A.; Bour, D. P. Nitride-Based Semiconductors for Blue and Green Light-Emitting Devices. Nature 1997, 386, 351–359.

(6)

Ozgur, U.; Huiyong Liu; Xing Li; Xianfeng Ni; Morkoç, H. GaN-Based LightEmitting Diodes: Efficiency at High Injection Levels. P. IEEE 2010, 98, 1180– 1196.

(7)

Fasol, G. Room-Temperature Blue Gallium Nitride Laser Diode. Science 1996, 272, 1751-1752.

(8)

Nakamura, S.; Pearton, S. J.; Fasol, G. The Blue Laser Diode: The Complete Story; Springer Verlag, 2000.

(9)

Rodrigues, J.; Ben Sedrine, N.; Felizardo, M.; Soares, M. J.; Alves, E.; Neves, A. J.; Fellmann, V.; Tourbot, G.; Auzelle, T.; Daudin, B.; et al. GaN:Pr3+ Nanostructures for Red Solid State Light Emission. RSC Adv. 2014, 4, 62869– 62877.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(10)

Page 30 of 37

Peng, H.; Lee, C.-W.; Everitt, H. O.; Munasinghe, C.; Lee, D. S.; Steckl, A. J. Spectroscopic and Energy Transfer Studies of Eu3+ Centers in GaN. J. Appl. Phys. 2007, 102, 073520.

(11)

Alves, E.; Lorenz, K.; Vianden, R.; Boemare, C.; Soares, M. J.; Monteiro, T. Optical Doping of Nitrides by Ion Implantation. Mod. Phys. Lett. B 2001, 15, 1281–1287.

(12)

Nishikawa, A.; Kawasaki, T.; Furukawa, N.; Terai, Y.; Fujiwara, Y. RoomTemperature Red Emission from a P-Type/europium-Doped/n-Type Gallium Nitride Light-Emitting Diode under Current Injection. Appl. Phys. Express 2009, 2, 1004.

(13)

Jadwisienczak, W. M.; Lozykowski, H. J.; Berishev, I.; Bensaoula, A.; Brown, I. G. Visible Emission from AlN Doped with Eu and Tb Ions. J. Appl. Phys. 2001, 89, 4384-4390.

(14)

Rodrigues, J.; Miranda, S. M. C.; Santos, N. F.; Neves, A. J.; Alves, E.; Lorenz, K.; Monteiro, T. Rare Earth Co-Doping Nitride Layers for Visible Light. Mater. Chem. Phys. 2012, 134, 716–720.

(15)

Li, S.; Waag, A. GaN Based Nanorods for Solid State Lighting. J. Appl. Phys. 2012, 111, 071101.

(16)

Glas, F. Critical Dimensions for the Plastic Relaxation of Strained Axial Heterostructures in Free-Standing Nanowires. Phys. Rev. B 2006, 74, 121302(R).


Page 31 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(17)

Henneghien, A.-L.; Tourbot, G.; Daudin, B.; Lartigue, O.; Désières, Y.; Gérard, J.-M. Optical Anisotropy and Light Extraction Efficiency of MBE Grown GaN Nanowires Epilayers. Opt. Express 2011, 19, 527–539.

(18)

Ronning, C.; Borschel, C.; Geburt, S.; Niepelt, R. Ion Beam Doping of Semiconductor Nanowires. Mater. Sci. Eng. R Reports 2010, 70, 30–43.

(19) Dalpian, G.; Chelikowsky, J. Self-Purification in Semiconductor Nanocrystals. Phys. Rev. Lett. 2006, 96, 226802.

(20)

Fernández-Serra, M. V.; Adessi, C.; Blase, X. Surface Segregation and Backscattering in Doped Silicon Nanowires. Phys. Rev. Lett. 2006, 96, 166805.

(21)

Calarco, R.; Stoica, T.; Brandt, O.; Geelhaar, L. Surface-Induced Effects in GaN Nanowires. J. Mater. Res. 2011, 26, 2157–2168.

(22)

Lorenz, K.; Nogales, E.; Miranda, S. M. C.; Franco, N.; Méndez, B.; Alves, E.; Tourbot, G.; Daudin, B. Enhanced Red Emission from Praseodymium-Doped GaN Nanowires by Defect Engineering. Acta Mater. 2013, 61, 3278–3284.

(23)

O’Donnell, K. P. Rare-Earth Doped III-Nitrides for Optoelectronic and Spintronic Applications; Springer, 2009.

(24)

Rodrigues, J.; Miranda, S. M. C.; Peres, M.; Nogales, E.; Alves, L. C.; Alves, E.; Tourbot, G.; Daudin, B.; Méndez, B.; Lorenz, K.; et al. A Comparative Study of Photo-, Cathodo- and Ionoluminescence of GaN Nanowires Implanted with Rare Earth Ions. Nucl. Instrum. Meth. B 2013, 306, 201–206.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(25)

Page 32 of 37

Wang, D.; Xing, G.; Gao, M.; Yang, L.; Yang, J.; Wu, T. Defects-Mediated Energy Transfer in Red-Light-Emitting Eu-Doped ZnO Nanowire Arrays. J. Phys. Chem. C 2011, 115, 22729–22735.

(26)

Yu, L.; Song, H.; Lu, S.; Liu, Z.; Yang, L.; Kong, X. Luminescent Properties of LaPO4 :Eu Nanoparticles and Nanowires. J. Phys. Chem. B 2004, 108, 16697– 16702.

(27)

Landré, O.; Bougerol, C.; Renevier, H.; Daudin, B. Nucleation Mechanism of GaN

Nanowires Grown

on

(111) Si

by

Molecular Beam Epitaxy.

Nanotechnology 2009, 20, 415602.

(28)

Ziegler,

J.

F.

SRIM-2013

software

package,

available

online

at

http://www.srim.org.

(29)

Ullah, M. W.; Kuronen, A.; Stukowski, A.; Djurabekova, F.; Nordlund, K. Atomistic Simulation of Er Irradiation Induced Defects in GaN Nanowires. J. Appl. Phys. 2014, 116, 124313.

(30)

Davydov, V. Y.; Kitaev, Y. E.; Goncharuk, I. N.; Smirnov, A. N.; Graul, J.; Semchinova, O.; Uffmann, D.; Smirnov, M. B.; Mirgorodsky, A. P.; Evarestov, R. A. Phonon Dispersion and Raman Scattering in Hexagonal GaN and AlN. Phys. Rev. B 1998, 58, 12899–12907.

(31)

Siegle, H.; Kaczmarczyk, G.; Filippidis, L.; Litvinchuk, A. P.; Hoffmann, A.; Thomsen, C. Zone-Boundary Phonons in Hexagonal and Cubic GaN. Phys. Rev. B 1997, 55, 7000–7004.


Page 33 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(32)

Reshchikov, M. A.; Morkoç, H. Luminescence Properties of Defects in GaN. J. Appl. Phys. 2005, 97, 061301.

(33)

Ogino, T.; Aoki, M. Mechanism of Yellow Luminescence in GaN. Jpn. J. Appl. Phys. 1980, 19, 2395–2405.

(34)

Neugebauer, J.; Van de Walle, C. G. Gallium Vacancies and the Yellow Luminescence in GaN. Appl. Phys. Lett. 1996, 69, 503-505.

(35)

Seitz, R.; Gaspar, C.; Monteiro, T.; Pereira, E.; Leroux, M.; Beaumont, B.; Gibart, P. Temperature Behaviour of the Yellow Emission in GaN. MRS Internet J. N. S. R. 1997, 2, e36.

(36)

Reshchikov, M. A.; Morkoç, H.; Park, S. S.; Lee, K. Y. Transient Photoluminescence of Defect Transitions in Freestanding GaN. Appl. Phys. Lett. 2001, 78, 2882-2884.

(37)

Reshchikov, M. A.; Morkoç, H.; Park, S. S.; Lee, K. Y. Two Charge States of Dominant

Acceptor

in

Unintentionally

Doped

GaN:

Evidence

from

Photoluminescence Study. Appl. Phys. Lett. 2002, 81, 4970-4972.

(38)

Darakchieva, V.; Monteiro, T.; Rodrigues, J.; Son, M. T.; Xie, M.-Y.; Ben Sedrine, N.; Alves, E.; Usui, A.; Hemmingson, C.; Monemar, B. Electron Irradiation of Bulk HVPE GaN: Structural Properties and Photoluminescence. In Oral presentation; ICNS-9, 2011.

(39)

Reshchikov, M. A.; Demchenko, D. O.; McNamara, J. D.; Fernández-Garrido, S.; Calarco, R. Green Luminescence in Mg-Doped GaN. Phys. Rev. B 2014, 90, 035207. 33 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(40)

Page 34 of 37

Rodrigues, J.; Miranda, S. M. C.; Fernandes, A. J. S.; Nogales, E.; Alves, L. C.; Alves, E.; Tourbot, G.; Auzelle, T.; Daudin, B.; Méndez, B.; et al. Towards the Understanding of the Intentionally Induced Yellow Luminescence in GaN Nanowires. Phys. Status Solidi (c) 2013, 10, 667–672.

(41)

Wang, Z.; Li, J.; Gao, F.; Weber, W. J. Defects in Gallium Nitride Nanowires: First Principles Calculations. J. Appl. Phys. 2010, 108, 044305.

(42)

Li, Q.; Wang, G. T. Spatial Distribution of Defect Luminescence in GaN Nanowires. Nano Lett. 2010, 10, 1554–1558.

(43)

Bodiou, L.; Oussif, A.; Braud, A.; Doualan, J.-L.; Moncorgé, R.; Lorenz, K.; Alves, E. Effect of Annealing Temperature on Luminescence in Eu Implanted GaN. Opt. Mater. (Amst). 2006, 28, 780–784.

(44)

Lee, C.-W. Temperature Dependence of Energy Transfer Mechanisms in EuDoped GaN. J. Appl. Phys. 2004, 95, 7717-7724.

(45)

Wang, K.; Martin, R. W.; O’Donnell, K. P.; Katchkanov, V.; Nogales, E.; Lorenz, K.; Alves, E.; Ruffenach, S.; Briot, O. Selectively Excited Photoluminescence from Eu-Implanted GaN. Appl. Phys. Lett. 2005, 87, 112107.

(46)

Katchkanov, V.; O’Donnell, K. P.; Dalmasso, S.; Martin, R. W.; Braud, A.; Nakanishi, Y.; Wakahara, A.; Yoshida, A. Photoluminescence Studies of EuImplanted GaN Epilayers. Phys. Status Solidi (b) 2005, 242, 1491–1496.

(47)

Andreev, T.; Liem, N. Q.; Hori, Y.; Tanaka, M.; Oda, O.; Dang, D. L. S.; Daudin, B. Optical Transitions in Eu3+ Ions in GaN:Eu Grown by Molecular Beam Epitaxy. Phys. Rev. B 2006, 73, 195203. 34 ACS Paragon Plus Environment

Page 35 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(48)

Monteiro, T.; Boemare, C.; Soares, M. J.; Sá Ferreira, R. A.; Carlos, L. D.; Lorenz, K.; Vianden, R.; Alves, E. Photoluminescence and Lattice Location of Eu and Pr Implanted GaN Samples. Physica B 2001, 308-310, 22–25.

(49)

Wybourne, B. G. Spectroscopic Properties of Rare Earths; Interscience Publishers, Jonh Wiley & Sons Inc.: New York, London , Sydney, 1965.

(50)

Roqan, I. S.; O’Donnell, K. P.; Martin, R. W.; Edwards, P. R.; Song, S. F.; Vantomme, A.; Lorenz, K.; Alves, E.; Bockowski, M. Identification of the Prime Optical Center in GaN: Eu3+. Phys. Rev. B 2010, 81, 085209.

(51)

Wang, K.; O’Donnell, K. P.; Hourahine, B.; Martin, R. W.; Watson, I. M.; Lorenz, K.; Alves, E. Luminescence of Eu Ions in AlxGa1-xN across the Entire Alloy Composition Range. Phys. Rev. B 2009, 80, 125206.

(52)

Fleischman, Z.; Munasinghe, C.; Steckl, A. J.; Wakahara, A.; Zavada, J.; Dierolf, V. Excitation Pathways and Efficiency of Eu Ions in GaN by Site-Selective Spectroscopy. Appl. Phys. B 2009, 97, 607–618.

(53)

O’Donnell, K. P.; Roqan, I. S.; Wang, K.; Lorenz, K.; Alves, E.; Boćkowski, M. The Photoluminescence/excitation (PL/E) Spectroscopy of Eu-Implanted GaN. Opt. Mater. (Amst). 2011, 33, 1063–1065.

(54)

Brandt, O.; Pfüller, C.; Chèze, C.; Geelhaar, L.; Riechert, H. Sub-meV Linewidth of Excitonic Luminescence in Single GaN Nanowires: Direct Evidence for Surface Excitons. Phys. Rev. B 2010, 81, 045302.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(55)

Page 36 of 37

Reshchikov, M. A.; McNamara, J. D.; Behrends, A.; Mohajerani, M. S.; Bakin, A.;

Waag,

A.

Higher

than

90%

Internal

Quantum

Efficiency

of

Photoluminescence in GaN:Si,Zn. Phys. Status Solidi (c) 2013, 10, 507–510.

(56)

Hofmann, D. M.; Kovalev, D.; Steude, G.; Meyer, B. K.; Hoffmann, A.; Eckey, L.; Heitz, R.; Detchprom, T.; Amano, H.; Akasaki, I. Properties of the Yellow Luminescence in Undoped GaN Epitaxial Layers. Phys. Rev. B 1995, 52, 1670216706.


Page 37 of 37

Graphical abstract


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


5

D0→ 7F2

NWs Layer 616

618

620

622

624

626

628

630

Wavelength (nm)


Spectroscopic Analysis of Eu3+ Implanted and Annealed GaN Layers

Recommend Documents