Effect of temperature on optical spectra and relaxation dynamics of Sm3+ in Gd3Ga5O12 single crystals

Effect of temperature on optical spectra and relaxation dynamics of Sm3+ in Gd3Ga5O12 single crystals

Journal of Alloys and Compounds 582 (2014) 208–212 Contents lists available at ScienceDirect Journal of Alloys and Compounds journal homepage: www.e...

775KB Sizes 0 Downloads 21 Views

Journal of Alloys and Compounds 582 (2014) 208–212

Contents lists available at ScienceDirect

Journal of Alloys and Compounds journal homepage: www.elsevier.com/locate/jalcom

Effect of temperature on optical spectra and relaxation dynamics of Sm3+ in Gd3Ga5O12 single crystals Radosław Lisiecki a,⇑, Witold Ryba-Romanowski a, Piotr Solarz a, Adam Strze˛p a, Marek Berkowski b a b

Institute of Low Temperature and Structure Research, Polish Academy of Sciences, Okolna 2, 50-422 Wrocław, Poland Institute of Physics, Polish Academy of Sciences, Al. Lotnikow 32/46, 02-668 Warsaw, Poland

a r t i c l e

i n f o

Article history: Received 15 April 2013 Received in revised form 10 July 2013 Accepted 22 July 2013 Available online 14 August 2013 Keywords: GGG crystals Optical spectroscopy Samarium-doped materials

a b s t r a c t Single crystals of Gd3Ga5O12 (GGG) doped with 2 at.% and 10 at.% of Sm3+ were fabricated by the Czochralski method. Optical absorption and emission spectra as well as luminescence decay curves for these crystals were recorded at different temperatures ranging from 5 K to 300 K. The energies of the crystal field sublevels of selected multiplets were determined based on optical spectra recorded at T = 5 K. It has been found that widths of spectral lines related to transitions between individual crystal field levels of multiplets involved increase by factors of 5–7 when the sample temperature grows from 5 K to 300 K. In contrast to this the positions of spectral lines appear to be independent of temperature. Temperature dependence of the 4G5/2 luminescence lifetime was determined for both concentrations of samarium. Luminescence decay curves follow a single-exponential time dependence when Sm3+ concentration amounts to 2 at.% but a non-exponential decay kinetic and efficient quenching process were observed in highly-doped sample. The Inokuti–Hirayama classical energy transfer model was utilized successfully to account for time dependence of an experimental decay curves recorded for GGG:10 at.% Sm crystal. Ó 2013 Elsevier B.V. All rights reserved.

1. Introduction Single crystal Gd3Ga5O12 (GGG) was recognized as promising magneto-optics material. The advantageous spectroscopic and thermo-mechanical qualities of GGG crystals doped with erbium and neodymium allowed considering their systems as potential laser media [1–3]. The Czochralski method commonly used to produce the GGG single crystals makes it possible to fabricate high optical quality crystals provided the inconvenience related to evaporation of gallium oxide is overcame [4]. In contrast to well known Y3Al5O12 (YAG) host the crystal structure of GGG can accommodate appreciable amounts of luminescent rare earth ions thus offering an opportunity to design new luminescent systems, in particular phosphors and laser active media emitting in the visible. It seems that trivalent samarium is of great worth for this purpose. In the past the interest in Sm3+-doped luminescent materials was marginal owing to the insufficient absorption properties for optical pumping with classical lamps. This shortcoming can be overcame at near future owing to rapid development of blue emitting laser diodes with quite high power [5]. The promising luminescence features of single crystals containing Sm3+ admixture were documented [6,7]. In the past the spectroscopic properties of Sm3+ have been examined in

⇑ Corresponding author. Tel.: +48 (71) 343 50 21; fax: +48 (71) 344 10 29. E-mail address: [email protected] (R. Lisiecki). 0925-8388/$ - see front matter Ó 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jallcom.2013.07.148

yttrium-based garnets, in particular in Y3Al5O12 (YAG) and Y3Ga5O12 (YGG) single crystals. Visible luminescence of samarium in the yttrium gallium garnets has been discussed by Keller and Pettit [8]. Furthermore, the crystal field energy levels of the 6H and 6F multiplets of samarium in the YAG and YGG crystals have been found [9]. The preliminary crystal field calculations for Sm3+ transitions in the YAG crystal were also reported by Gruber et al. [10]. Subsequently, the analysis of the energy levels of Sm3+ in the Y3Al5O12 (YAG) crystal was extended to include 4G5/2 multiplet [11]. Eventually, absorption and emission spectra of Sm3Ga5O12 crystal were documented in [12]. The large energy gap between the metastable level 4G5/2 and a next lower-lying state results in low contribution of nonradiative multiphonon relaxation and thereby in efficient yellow–red luminescence. In addition, it has been shown that this emission can be enhanced in numerous inorganic materials applying the europium–samarium energy transfer [13–15]. The preliminary spectroscopic investigations of Sm3+ doped GGG and (Ca, Mg, Zr) GGG crystals grown by Czochralski technique have been reported [16,17]. In particular, absorption and emission spectra were measured at room temperature and analyzed in the framework of the Judd–Ofelt theory. High absorption cross-section at 405 nm and high quantum efficiency of the luminescence from metastable level were estimated. The emission cross-section of 3.18  1021 cm2 at 613 nm related to the transition 4G5/2 ? 6H7/2 was found for GGG:Sm crystal. Considerably lower value of emission cross-section

R. Lisiecki et al. / Journal of Alloys and Compounds 582 (2014) 208–212

amounting to 4.28  1022 cm2 was estimated for (Ca, Mg, Zr) GGG crystal. Intention of the present work is to get new information relevant to luminescence phenomena in Gd3Ga5O12 crystals doped with Sm3+, encompassing their spectral features and excited state relaxation dynamics. Taking advantage of low temperature spectroscopy the factors governing luminescence spectra, among others crystal field splitting of multiplets, intensity distribution and linewidths of transitions between individual crystal field levels were determined. 2. Experimental Single crystals of (Gd1xSmx)3Ga5O12 with x = 0.02 and 0.1 were grown by the Czochralski method. The real concentrations of Sm3+ are 8.6  1019 ions/cm3 for x = 0.02 and 4.3  1020 ions/cm3 for x = 0.1. The polished samples in the form of rectangular plates several millimeters thick were prepared for optical measurements. The absorption spectra were measured with a Varian Model 5E UV–VIS– NIR spectrophotometer. High-resolution emission spectra were recorded with a experimental system consisting of a Dongwoo Optron DM 158i excitation monochromator and DM711 emission monochromator having 750 mm focal length. An ozone-free Xenon lamp DL 80-Xe was utilized as an excitation source. Luminescence decay curves were excited with the Continuum Surelite I optical parametric oscillator (OPO) pumped by a third harmonic of Nd:YAG laser and a Tektronix Model TDS 3052 digital oscilloscope was used to record the luminescence decay curves. For low-temperature measurements, crystals were assembled in a continuous flow liquid helium cryostat Oxford Instruments equipped with a temperature controller.

3. Results and discussion Gd3Ga5O12 (GGG) crystal belongs to cubic garnet system with Ia–3d space group and is characterized by the following cell parameters a = 12.38 Å, Z = 8 and d = 7.09 g cm3. Spectroscopic properties of GGG single crystal activated with luminescent ions are determined by its structural features. Fig. 1 presents the polyhedrons and coordination spheres of gadolinium and gallium ions in the GGG. In this structure Gd3+ ions are located in dodecahedral sites whereas the Ga3+ ions reside in octahedral and tetragonal sites. Incorporated samarium ions substitute gadolinium ions and owing to very small difference in radii of the two ions the crystal structure is not distorted even when the doping level is high. All optical spectra in this system are related to intra-configuration transitions within the 4f5 configuration of Sm3+. Fig. 2 presents survey absorption spectrum of the GGG:2 at.% Sm crystal measured at T = 5 K. At such a low temperature spectral lines are related to transitions from the lowest energy Stark component of 6H5/2 ground state to higher-lying multiplets of Sm3+. It can be seen in Fig. 2 that

Fig. 1. Structure of Gd3Ga5O12 crystal.

209

there are two groups of spectral lines separated by relatively large energy gap. Referring to energy level diagram of Sm3+ the absorption lines in infrared region are assigned to transitions terminating on low energy multiplets of 6H and 6F sextet terms whereas absorption lines in the visible and UV region are due to transitions to higher energy multiplets of 4F, 4G, 4H, 4I, 4K, 4L, 4M quartets terms. The energy gap between the two groups of multiplets corresponds to energy distance between the highest energy crystal field component of the 6F11/2 state and lowest energy Stark sublevel of 4 G5/2 multiplet. In GGG:Sm system it is equal to 6908 cm1. Assignment of well separated absorption lines of Sm3+ in infrared region is quite straightforward but it is problematic for transitions within visible and UV region where energy distances between multiplets are frequently lower than crystal field splitting of individual multiplets. Nevertheless, careful examination of the absorption spectrum measured at T = 5 K made it possible to determine energies of crystal field levels of excited multiplets located between 4500 and 25,000 cm1. Evaluated energy values, numbers of crystal field levels and overall crystal field splitting of selected multiplets of Sm3+ in GGG crystal are presented in Table 1. Figs. 3 and 4 show emission spectra recorded at T = 5 K for 2 at.% Sm3+:GGG crystal. Spectral lines in Fig. 3 are related to transitions from metastable 4G5/2 level to the ground state 6H5/2 and the first excited state 6H5/2. Since at 5 K solely the lowest energy crystal field component of the initial multiplet is populated the spectra should correspond to transitions ending on three crystal field components of the ground state 6H5/2 and four crystal field components of the 6H7/2 first excited state. In fact, for the 4G5/2 ? 6H5/2 transition, three emission lines appear at 17,432, 17,525 and 17,596 cm1 while four emission lines corresponding to 4G5/ 6 2 ? H7/2 transition are located at 16,604, 16,442, 16,296 and 16,275 cm1. Spectral widths (FWHM) of these lines range from 3 to 7 cm1. Fig. 4 shows T = 5 K emission spectra related to transitions from luminescent 4G5/2 level of Sm3+ to next higher energy 6 H9/2 and 6H11/2 manifolds. Five emission lines with the most intense one at 15,380 cm1 are associated with the 4G5/2 ? 6H9/2 transition whereas four lines attributed to the 4G5/2 ? 6H11/2 transition emerge within 14,000–14,450 cm1 spectral range. Their FWHM values are in the range of 3–11 cm1. Energies of crystal field sublevels and overall splitting of terminal 6HJ manifolds were determined and included in Table 1. To obtain these results assessment of energy differences between the lowest energy absorption line of 4G5/2 level peaking at 17,596 cm1 and emission lines related to the 4G5/2 ? 6HJ transitions have been found. Crystal field splitting of ground state 6H5/2 of Sm3+ in the Gd3Ga5O12 crystal was found to be 162 cm1 while roughly two times higher value of 328 cm1 was estimated for the first excited state 6H7/2 of Sm3+. Knowledge of number of experimentally determined crystal field sublevels of multiplets is important to assess number of local positions of Sm3+ ions in the Gd3Ga5O12 host lattice. In agreement with Kramers theorem the crystal field levels of Sm3+ manifolds are doubly degenerate and their number is expected to be J + 1/2. As it can be seen in Table 1, numbers of crystal field components derived from low temperature spectra do not exceed those theoretically predicted except for the 6F7/2 level for which the number of its experimentally estimated Stark components is higher then J + 1/2. These findings imply that although majority of Sm3+ ions are accommodated in a well-defined local position in GGG lattice the occurrence of other minority sites cannot be excluded. Indeed, four non-equivalent Nd3+ sites in the GGG have been found by Guyot et al. when employing site selective technique [18]. It has been then concluded that merely the so called c site contributes to emission spectrum. Subsequent study on powder sample of GGG:Nd corroborated this finding. [19]. The nature of minority sites of rare earth ions in GGG host is not clear. It was assumed from the study of the rare-earth doped yttrium gallium garnet host

210

R. Lisiecki et al. / Journal of Alloys and Compounds 582 (2014) 208–212

16 T=5K

GGG:2 at.% Sm

GGG:2 at.% Sm

T=5K

14

-1

Absorbtion coefficient [cm ]

F5/2

-1

Absorbtion coefficient [cm ]

8 6

12

10

8

6

F7/2

6 6

F9/2

4

6

4

2 2

6

H13/2

6

6

F3/2, H15/2

6

F11/2

0 5000

6000

7000

8000

9000 10000 11000

0

20000

25000

30000

35000

-1

-1

Wavenumber [cm ]

Wavenumber [cm ]

Fig. 2. Absorption spectra of GGG:2 at.% Sm3+ crystal recorded at T = 5 K.

Table 1 The Stark components of Sm3+ selected multiplets in the Gd3Ga5O12. SLJ

6

H5/2 H7/2 6 H9/2 6 H11/2 6 H13/2 6 F1/2 6 F3/2, 6H15/2 6 F5/2 6 F7/2 6 F9/2 6 F11/2 4 G5/2 4 F3/2 4 G7/2 4 I9/2 4 M15/2 4 I11/2 4 I13/2 4 F5/2 4 I15/2 6 P5/2 4 L13/2 4 F7/2 6 P3/2 6

Stark sublevels Exp.

Theor.

3 4 5 4 7 1 10 3 7 5 6 2 2 4 5 8 6 7 3 8 3 7 4 2

3 4 5 6 7 1 10 3 4 5 6 3 2 4 5 8 6 7 3 8 3 7 4 2

[8] that luminescent ions may marginally reside into gallium sites. The gallium ions are characterized by sixfold and fourfold oxygen coordination numbers respectively. Knowing the structural characteristic of lanthanide ions in oxide crystals, the Sm3+ ions would be aimed to occupy local position with higher oxygen coordination number and as the consequence the additional Stark components of Sm3+ multiplets may be observed. The energy values of crystal field sublevels of Sm3+ multiplets have been estimated in the past for the Y3Al5O12:Sm3+ crystal [11,20], which is isostructural compound to the Gd3Ga5O12:Sm3+. Considering the 6H5/2 ground state of Sm3+, it can be seen that its crystal field splitting of 247 cm1 in the YAG:Sm3+ [11] is markedly higher than 162 cm1 in the GGG:Sm3+. Also, the crystal field

Energy (cm1)

DE

0, 71, 162 992, 1155, 1300, 1320 2216, 2336, 2379, 2505, 2520 3140, 3175, 3540, 3573 4874, 4902, 4917, 4994, 5130, 5163, 5255 6479 6616, 6630, 6647, 6679, 6711, 6742, 6782, 6820, 6839, 6864 7187, 7273, 7291 8040, 8059, 8071, 8076, 8113, 8164, 8202 9186, 9213, 9230, 9273, 9310 105,40, 10,553, 10,569, 10,627, 10,656, 10,688 17,596, 17,868 18,774, 18,787 19,818, 19,936, 19,978, 19,996 20,202, 20,269, 20,310, 20,362, 20,388 20,654, 20,621, 20,672, 20,756, 20,815, 20,833, 20,997, 21,050 21,362, 21,448, 21,513, 21,609, 21,653, 21,674 21,980, 21,990, 22,017, 22,141, 22,230, 22,254, 22,275, 22,477, 22,645, 22,745 23,068, 23,130, 23,709, 23,768, 23,809, 23,836, 23,887, 23,938 24,357, 24,393, 24,429 24,546, 24,646, 24,663, 24,689, 24,745, 24,754, 24,794 24,848, 24,894, 24,931, 25,016 25,176, 25,205

162 328 304 433 381 – 248 104 164 124 148 269 13 178 186 396 312 295 268 870 72 248 168 29

splitting for higher energy multiplets of the 6H term of Sm3+ in the yttrium aluminum garnet crystal is larger as compared to respective splitting values for the samarium-doped gadolinium gallium garnet crystal. This trend is not valid for higher energy levels, however. For example, two experimentally observed crystal field levels of the 4G5/2 multiplet have energies of 17,597 cm1 and 17,784 cm1 in the YAG:Sm3+ whereas 17,596 cm1and 17,868 cm1 were found for the GGG:Sm3+. It is worth noticing here that numbers of experimentally determined crystal field levels for the 2H11/2 and 4G5/2 multiplets are lower than predicted. As a consequence their actual overall crystal field splitting is not known. Fundamental difficulty in determining the 4G5/2 splitting stems from an extremely weak intensity of absorption related to

211

R. Lisiecki et al. / Journal of Alloys and Compounds 582 (2014) 208–212

6

4

G5/2

4

G5/2

6

H 7/2

H 7/2

4 4

16000

16200

16600 17300

16400

G 5/2

6

H 5/2

17400

17500

17600

17700

-1

Wavenumber [cm ]

6

G5/2

H 5/2

T=300K

Luminescence intensity [a.u.]

Luminescence intensity [a.u.]

2%Sm:GGG

GGG:2at.% Sm

T=5K

T=200K

Fig. 3. Emission spectra of GGG:2 at.% Sm crystal recorded at T = 5 K and related to 4 G5/2 ? 6H7/2 and 4G5/2 ? 6H5/2 transitions of Sm3+.

T=50K

4

6

G5/2 H11/2

4

G5/2

6

T=5K

H9/2 16000

16400

16800

17400

17600

-1

14400

14600 -1

Fig. 5. The emission spectra of 2% Sm:GGG related to 4G5/2 ? 6H5/2 and 4G5/2 ? 6H7/2 transitions of Sm3+ and recorded at T = 5–300 K.

Wavenumber [cm ]

15000

15100

15200

15300

40

15400

the 6H5/2 ? 4G5/2 transition. Contribution of the electric-dipole mechanism is very small because related values of matrix elements of unit tensor operators that govern this transition are essentially equal to zero. Selection rules for magnetic dipole mechanism are fulfilled in this case and significant contribution to the absorption intensity come from a magnetic dipole transition. The knowledge of the crystal field splitting for the 2H11/2 multiplet is not crucial but the 4G5/2 multiplets acts as metastable level and the population distribution over its crystal field levels may be relevant to interpret luminescence phenomena at room temperature. To verify this supposition the effect of temperature on luminescence spectra was examined. Fig. 5 compares spectra related to the 4G5/2 ? 6H7/2 and 4G5/2 ? 6H5/2 transitions recorded at several temperatures from 5 K to 300 K. It can be seen that the number of spectral lines contributing to spectra does not increase with growing temperature up to 200 K. In T = 300 K spectra very weak additional lines can be discerned at 16,692 cm1, 16,874 cm1 and 17,694 cm1. Their spectral positions are consistent with transitions from the crystal field component of the 4G5/2 located at 17,868 cm1. This finding implies that the missing crystal field component of the 4 G5/2 multiplet may be located higher than 17,868 cm1 or, if not, the intensity of transitions originating in this components are vanishingly small. Spectra shown in Fig. 5 reveal that widths of spectral lines increase markedly with growing temperature. In Fig. 6 widths (FWHM) of lines related to transitions from the lowest

40 4

G 5/2 - 6H5/2

4

-1

-1

FWHM [cm ]

Wavenumber [cm ] Fig. 4. The emission spectra of GGG:2 at.% Sm crystal recorded at T = 5 K and related to 4G5/2 ? 6H9/2 and 4G5/2 ? 6H11/2 transitions of Sm3+.

2%Sm:GGG

2%Sm:GGG

30

line 0-0 line 0-1

-1

13900 14000 14100

Wavenumber [cm ]

FWHM [cm ]

Luminescence intensity [a.u.]

T=5K

20

line 0-0 line 0-1

30

20

10

10

0

6

G5/2 - H 7/2

0 0

100

T [K]

200

300

0

100 200 T [K]

300

Fig. 6. Temperature dependence of linewidth of (0–0) and (0–1) spectral lines attributed to 4G5/2 ? 6H5/2 and 4G5/2 ? 6H7/2 transitions.

crystal field component of the 4G5/2 multiplet to the lowest (0–0 line) and second (0–1 line) crystal field components of terminal multiplets are plotted versus sample temperature. Luminescence decay curves for the 4G5/2 luminescence of samarium in 2% and 10%-doped samples have been recorded and analyzed. Luminescence decay curve of the 4G5/2 level recorded with the sample containing 2% of samarium is a single exponential with a time constant of 2.2 ms. This value is fully consistent with the experimental lifetime of 4G5/2 level amounting to 2.1 ms measured recently for Sm0.03Gd2.97Ga5O12 [16]. Single exponential decay curve imply that luminescent Sm3+ ions are located in well-defined sites. For the 10%-doped sample decay curves of 4 G5/2 luminescence become non-exponential as it is shown in Fig. 7. The mean value of the 4G5/2 lifetime was estimated to be

R. Lisiecki et al. / Journal of Alloys and Compounds 582 (2014) 208–212

e

7

e

6

e

5

e

4

e

3

e

2

e

1

e

0

2.0 1.5 1.0

e

-2

0

100

200

300

T [K] tmean = 628 µs 4

-1

10 at.%Sm

0.5 0.0

e

acceptor when the rate of energy transfer to the acceptor is equal to the rate of intrinsic decay of the donor. According to formula (2) our fitting procedure gave R0 = 10.3 Å for a = 3.5. The dipole–dipole coupling parameter Cda estimated 52 from the relation C da ¼ R60 s1 m6 s1. 0 amounts to 4.7  10 4 The lifetime values of the G5/2 luminescence for 2% Sm and 10% Sm-doped samples have been measured as a function of temperature as it is shown in inset to Fig. 7. These findings imply that the temperature does not affect significantly the 4G5/2 lifetime in Sm:Gd3Ga5O12 irrespective of Sm3+ concentration.

2 at.%Sm

2.5

Lifetime [ms]

Luminescence intensity [a.u.]

212

G 5/2

4. Conclusion GGG:10 at.% Sm

0

1000

2000

3000

4000

Time [µs] Fig. 7. Decay curve of the 4G5/2 luminescence recorded for 10 at.% Sm:GGG sample. Inset shows the lifetimes of 4G5/2 level measured for 2 at.% Sm:GGG and 10 at.% Sm:GGG crystals plotted versus sample temperature.

0.63 ms. It appears that for higher concentration of Sm3+ ions, a quenching effect is quite effective in the Gd3Ga5O12 crystal. For a comparison, a single exponential decay curve of samarium luminescence was detected for the 0.1% Sm3+ doped YAG crystal and 4 G5/2 lifetime was found to be of 1.96 ms [20]. It is worth noticing, that already for 0.5% Sm3+ doped sample the luminescence decay was significantly faster and non-exponential. Therefore, efficient cross relaxation process was observed in the yttrium aluminum garnet crystal doped with Sm3+ ions. The study on quenching mechanism of Sm3+ luminescence in Y3Al5O12:Sm3+ has been continued considering materials fabricated by sol–gel combustion method [21]. It has been then concluded that the significant cross-relaxation process in this system is caused by the interaction of electric quadrupoles between Sm3+ ions. Various approaches have been proposed to account for an experimental time dependence of a luminescence decay. Selfquenching of luminescence have been considered involving two processes namely, the migration of excitation energy over a donor ions and the excitation energy transfer to acceptors. A diffusion model is applied to account for experimental data when the energy migration takes place with a high probability [22]. Otherwise, nonexponential decay curves may by reasonably analyzed neglecting migration process [23]. The model predicted by the Inokuti–Hirayama has been used to fit the non-exponential decay curve of 4G5/2 luminescence recorded for 10 at.% Sm:GGG. The time dependence of the luminescence intensity is expressed by following formula:

IðtÞ ¼ I0 exp

" t

s0

 3=s # t

a

s0

ð1Þ

Based on absorption and emission spectra measured at T = 5 K, energies of crystal field sublevels were determine for low energy multiplets of Sm3+. Low temperature spectroscopic investigations of Sm:Gd3Ga5O12 crystals indicate that samarium ions occupy preferentially one type of sites in the GGG crystal. At low temperature luminescence spectra consist of well separated lines having linewidth values between 3 cm1 and 11 cm1. With growing temperature widths of lines start to increase at about 100 K and attain values between 30 cm1 and 40 cm1 at 300 K. Contribution of lines related to transitions from higher energy crystal field components of luminescent 4G5/2 multiplet to luminescence spectrum at 300 K is found to be marginal. Relaxation dynamic of 4G5/2 metastable level is independent of the temperature. With increasing Sm3+ concentration the luminescence becomes effectively quenched and luminescence decay curves become non-exponential. In the framework of Inokuti–Hirayama model the critical distance R0 and the interaction parameter Cda have been evaluated. The results showed that activator–activator interaction play a significant role in the relaxation of the luminescent level of Sm3+. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16]

where I(t) is the luminescence intensity after pulse excitation, s0 denotes intrinsic lifetime of donors in the absence of acceptors, s = 6 for a dipole–dipole interaction, and the a is given by the equation:

  4 3 a ¼ pC 1  N0 R30 3 s

ð2Þ

where C is the gamma function, N0 is the concentration of acceptors and R0 is the critical radius defined as a distance between donor and

[17] [18] [19] [20] [21] [22] [23]

H. Lundt, H. Weinder, Opt. Commun. 82 (1991) 484. Q. Wang, S. Zhang, S. Wu, S. Dong, J. Lumin. 40–41 (1988) 181. A. Kaminskii, Phys. Stat. Sol. A 87 (1985) 11. C.D. Brandle, A.J. Valentino, J. Cryst. Growth 12 (1972) 3. J.W. Raring, E.M. Hall, M.C. Schmidt, C. Poblenz, B. Li, N. Pfister, D.F. Feezell, R. Craig, J.S. Speck, S.P. Denbaars, S.J. Nakamura, Proc. SPIE 7602 (2010) 760218. G. Dominiak-Dzik, J. Alloys Comp. 391 (2005) 26. P. Solarz, M. Sobczyk, Opt. Mater. 34 (2012) 1826–1832. S.P. Keller, G.D. Pettit, Phys. Rev. 121 (1961) 1693. P. Grunberg, S. Hufner, E. Orlich, J. Schmitt, J. Appl. Phys. 40 (1969) 1501. J.B. Gruber, M.E. Hills, M.P. Nadler, M.R. Kokta, C.A. Morrison, Chem. Phys. 113 (1987) 175. S.B. Stevens, C.A. Morrison, M.D. Seltzer, M.E. Hills, J.B. Gruber, J. Appl. Phys. 70 (1991) 948. A.I. Bilyi, A.E. Nosenko, V.M. Lukyantes, V.I. Litvinenko, J. Appl. Spectrosc. 41 (1984) 763. J.P. Huang, Q.X. Li, D.H. Chen, Mater. Sci. Eng. B, Solid-State Mater. Adv. Technol. 172 (2010) 108. Q. Wei, D.H. Chen, Opt. Laser Technol. 41 (2009) 783. T. Aitasalo, P. Deren, J. Holsa, H. Jungner, M. Lastusaari, J. Niittykoski, W. Strek, Radiat. Meas. 38 (2004) 515. Wenpeng Liu, Qingli Zhang, Dunlu Sun, Jianqiao Luo, Changjiang Gu, Haihe Jiang, Shaotan Yin, J. Cryst. Growth 331 (2011) 83. Wenlong Zhoua, Qingli Zhanga, Jin Xiaoa, Jian qiao Luoa, Wenpeng Liua, Haihe Jianga, Shaotang Yina, J. Alloys Comp. 491 (2010) 618. Y. Guyot, L.E. Bausa, J. Garcia-Sole, I. Vergara, A. Montelli, R. Moncorge, J. Appl. Phys. 72 (1992) 5876. L. Lipin´ska, W. Ryba-Romanowski, A. Rzepka, S. Ganschow, R. Lisiecki, R. Diduszko, A. Zaja˛czkowska, Cryst. Res. Technol. 44 (2009) 477. M. Malinowski, R. Wolski, Z. Fukacz, T. Łukaszewicz, Z. Łuczyn´ski, J. Appl. Spectrosc. 62 (1995) 840. L. Mu, H. Yang, W. He, L. Wang, Q. Zhang, Xiyou Jinshu Cailiao Yo Gongcheng/ Rare Met. Mater. Eng. 38 (2009) 2032. M. Yohota, O. Tanimoto, J. Phys. Soc. Jpn. 22 (1967) 779. M. Inokuti, F. Hirayama, J. Chem. Phys. 43 (1965) 1978.