Photoluminescence and afterglow of deep red emitting SrSc₂O₄:Eu²⁺

Abstract

This work deals with the photoluminescence (PL) properties of SrSc₂O₄:Eu²⁺ and SrSc₂O₄:Eu²⁺,Dy³⁺. For this purpose a series of powder samples of SrSc₂O₄ with various concentrations of Eu²⁺ and Dy³⁺ was prepared. The samples were synthesised via a high temperature solid state route. Phase purity was investigated by conducting X-ray powder diffractometry. The PL properties of all prepared samples were elucidated by recording PL and PL excitation spectra. Furthermore, the temperature behaviour of the PL of SrSc₂O₄:Eu²⁺ was investigated from 100 to 500 K. Diffuse reflectance spectra were recorded to investigate the optical properties of Eu²⁺ doped SrSc₂O₄. Additionally, persistent luminescence of SrSc₂O₄:Eu²⁺ and SrSc₂O₄:Eu²⁺,Dy³⁺ was investigated. To this end, PL lifetime measurements were conducted. SrSc₂O₄:Eu²⁺ shows an emission band in the deep red spectral range. Moreover, it turned out that the red emission of SrSc₂O₄:Eu²⁺ shows persistent luminescence too.

---

1. Introduction

Today, materials which show red photoluminescence (PL) or persistent luminescence are of great interest for various applications. Their uses as phosphors in light emitting diodes or as fluorescence probes in bio-imaging are only two potential fields of application for these materials. For this reason, many research groups are investigating different host compounds and various dopants to find new red emitting materials.^1–3 Often, Eu²⁺ is taken as a dopant to obtain red emission.^3,4 Since the emission of Eu²⁺ is due to interconfigurational d–f-transitions, the energy of the emission strongly depends on the host material. Therefore, the emission band of Eu²⁺ can be shifted from the ultraviolet (UV) to deep red range of the electromagnetic spectrum by changing the chemical environment.⁵

For instance, in a material providing a surrounding with little covalent character like BaAlF₅, the Eu²⁺ emission occurs in the UV range at about 361 nm.⁶ With increasing covalent interaction the emission band of Eu²⁺ can be located all over the visible spectrum. In CaAl₂O₄:Eu²⁺ the emission band of Eu²⁺ is located in the blue spectral range at about 436 nm.⁷ Green emission of Eu²⁺ can be observed in oxides and oxy-nitrides like SrAl₂O₄:Eu²⁺ (λ_em,max ≈ 512 nm) and β-SiAlON:Eu²⁺ (λ_em,max ≈ 535 nm). Red 5d–4f-emission of Eu²⁺ is commonly observed in nitrides and sulphides. For example, Sr₂Si₅N₈:Eu²⁺ possesses an emission band at about 620 nm. Eu²⁺ doped sulfides such as CaS:Eu²⁺ or SrS:Eu²⁺ show emission at about 650 and 615 nm, respectively.

So far, not many red emitting luminescent materials on the basis of Eu²⁺-doped oxides are known and even less showing persistent luminescence in the deep red spectral range.¹ Most of the red or infrared emitting persistent phosphors are sulphide based compounds doped with lanthanide ions.⁸ Especially in the red region, this group of compounds provides a variety of different emission wavelengths. Unfortunately, such compounds usually are extremely sensitive against moisture and thus chemically unstable. Therefore, further efforts like encapsulation and surface modification are necessary to protect these compounds from degradation. A chemically more stable group of compounds are oxides. Usually, oxides are impervious against moisture. However, most oxides which show red persistent luminescence are doped with Cr³⁺.

SrSc₂O₄ was firstly described and its structure solved by Carter and Feigelson in 1964.⁹ Later Gaume et al. reported on the spectroscopic properties of Yb³⁺-doped SrSc₂O₄.¹⁰ To the best of our knowledge there is no report concerning the PL properties of Eu²⁺ doped SrSc₂O₄. Therefore, a series of powder samples of SrSc₂O₄ with various contents of Eu²⁺ and Dy³⁺ was prepared. SrSc₂O₄ crystallizes in the orthorhombic crystal system with space group Pnam. In SrSc₂O₄ the Sr²⁺ ions are surrounded by eight O²⁻ ions. The ionic radius of 8-fold coordinated Sr²⁺ is 1.26 Å. 8-Fold coordinated Eu²⁺and Dy³⁺exhibit ionic radii of 1.25 Å and 1.027 Å.¹¹ Therefore, it is assumed that the Eu²⁺and Dy³⁺ ions tend to occupy the Sr²⁺ sites.

2. Experimental

The SrSc₂O₄:Eu²⁺ and SrSc₂O₄:Eu²⁺,Dy³⁺samples as investigated in this work were synthesized by high temperature solid state reaction. Therefore, high purity educts SrCO₃ (Aldrich, 99.9%), Sc₂O₃ (Treibacher Industrie AG, 99.99%), Eu₂O₃ (Treibacher Industrie AG, 99.99%), and Dy₂O₃ (Treibacher Industrie AG, 99.99%) were weighted in stoichiometric amounts and were thoroughly blended in acetone in an agate mortar. After drying at ambient temperatures, the obtained powder blends were calcined at 1400 °C for 12 h in Mo boats in a reducing hydrogen atmosphere (Westfalen, 99.999%). After calcination, red sinter bodies were obtained which were ground to a fine μ-powder.

Phase purity of the synthesized samples was investigated using X-ray powder diffractometry (XRD). XRD patterns were collected on a Rigaku MiniFlex II diffractometer working in Bragg–Brentano geometry using Cu K_α radiation. Step width and integration time were set to 0.02° and 1 s, respectively.

Particle size and morphology were investigated using scanning electron microscopy (SEM). Therefore, the scanning electron microscope Zeiss EVO MA10 equipped with a LaB₆-cathode was used. Pressure in the sample chamber was 5 × 10⁵ Pa and acceleration voltage was 8 kV.

PL as well as PL excitation (PLE) spectra were recorded on an Edinburgh Instruments FSL900 spectrometer equipped with a Xe arc lamp (450 W) and a cooled (−20 °C) single-photon counting photomultiplier (Hamamatsu R2658P). Obtained PL spectra were corrected by applying a correction file obtained from a tungsten incandescent lamp certified by the National Physics Laboratory U.K.

Persistent luminescence decay (PLD) curves were recorded after exciting the samples for 15 min with an excitation wavelength of λ_ex = 525 nm. Absolute luminance and radiance was measured using the optical power meter 1830-C of Newport Corporation equipped with a Si-detector.

Temperature dependent PL measurements from 100 to 500 K were performed using the Oxford Instruments cryostat MicrostatN2. Liquid nitrogen was used as a cooling agent. Temperature stabilization time was 60 s and tolerance was set to ±3 K.

Diffuse reflectance (DR) spectra were recorded on an Edinburgh Instruments FS900 spectrometer equipped with a Xe arc lamp (450 W), a cooled (−20 °C) single-photon counting photomultiplier (Hamamatsu R928) as well as a Teflon-coated integration sphere. BaSO₄ (99.998%, Sigma-Aldrich) was used as a reflectance standard.

3. Results and discussion

The collected XRD patterns as well as the ICDD reference card of SrSc₂O₄ are depicted in Fig. 1 for comparison purposes. The diffractograms prove the formation of the orthorhombic SrSc₂O₄ phase without any impurity phases. In addition, with increasing Eu²⁺ doping concentration no significant shift of the reflexes is observed. The diffractograms confirm the used synthesis route for the preparation of SrSc₂O₄:Eu²⁺.

Fig. 1 XRD patterns of Sr_1−xEu_xSc₂O₄ with various doping concentrations and ICDD reference card of SrSc₂O₄.

Morphology of the surface as well as particle size of the prepared samples were investigated by using SEM. Fig. 2 shows the obtained image of SrSc₂O₄. The prepared powders of SrSc₂O₄ are in micrometer range. Further, the particles are heavily cracked by crushing them in the agate mortar.

Fig. 2 SEM photograph of SrSc₂O₄.

Reflectance behaviour of the doped and undoped SrSc₂O₄ powder samples was investigated by recording DR spectra. BaSO₄ was used as a reflectance standard. The obtained spectra are illustrated in Fig. 3. Undoped SrSc₂O₄ shows about 70% reflectance compared to BaSO₄, which is in line with the white to greyish body colour. This greying is ascribed to Mo impurities caused by the Mo foil used during the high temperature annealing. The Eu²⁺ doped SrSc₂O₄ samples show strong absorption bands at about 320 and 475 nm. These bands are assigned to transitions from the ⁸S ([Xe]4f⁷) ground state to the 5d multiplet. With increasing Eu²⁺ concentration this absorption increases and broadens towards the long wavelength range. Therefore, the body colour of SrSc₂O₄:Eu²⁺ is shifted from light pink to red violet.

Fig. 3 DR spectra of Sr_1−xEu_xSc₂O₄ with various doping concentrations.

PL as well as PLE spectra of Sr_0.98Eu_0.02Sc₂O₄ are depicted in Fig. 4. The PLE spectrum was recorded monitoring the Eu²⁺ emission at λ_em = 720 nm. The obtained PLE spectrum shows a broad band which consists of various unresolved bands. This band originates from the transition of the ⁸S ground state to the excited states ⁶P and ⁸P ([Xe]4f⁶5d¹).¹² PL of SrSc₂O₄:Eu²⁺was measured upon excitation of λ_ex = 525 nm. The obtained spectrum shows a broad band in the deep red range of the visible spectrum. This emission band ranges from about 595 to 840 nm with a maximum at about 687 nm and a full width half maximum of about 103 nm (2098 cm⁻¹). Hence, the PL maximum of SrSc₂O₄:Eu²⁺ is located in the first bio-imaging window.

Fig. 4 Room temperature PL and PLE spectra of Sr_0.98Eu_0.02Sc₂O₄.

The PLE spectra obtained from Sr_1−xEu_xSc₂O₄ with x = 0.02, 0.03, and 0.05 are depicted in Fig. 5. The PLE maximum shifts towards longer wavelengths with increasing Eu²⁺ concentration. In addition to that, the PLE spectra broaden with increasing Eu²⁺ concentration. The shift of the excitation maximum is mainly ascribed to a change in covalency. This is verified by calculating the centroid wavenumber of the excitation bands. The results are summarized in Table 1. With increasing Eu²⁺ concentration the calculated energy of the centroid wavenumbers decrease. This observation reflects an increase in the covalency of the bonds of SrSc₂O₄:Eu²⁺ and is attributed to the higher alkalinity of Eu²⁺ compared to Sr²⁺. To investigate the broadening of the excitation band in more detail, the crystal field splitting of the d-orbitals of Eu²⁺ was calculated (see Table 1). With increasing Eu²⁺ concentration crystal field splitting increases, too. Since Eu²⁺ is slightly smaller than Sr²⁺, the incorporation of Eu²⁺ into SrSc₂O₄ leads to a smaller Sr site and thus to higher crystal field splitting of the d-orbitals.

Fig. 5 Room temperature PLE spectra of Sr_1−xEu_xSc₂O₄ with various Eu²⁺ concentrations.

Table 1 Excitation bands, crystal field splitting, and centroid shift of Sr_1−xEu_xSc₂O₄ with various Eu²⁺ concentrations

Sample (x)	λ ex range (nm)	Crystal field splitting (cm^−1)	Centroid wavenumber (cm^−1)
0.02	238⋯697	27 670	26 596
0.03	237⋯689	27 680	25 381
0.05	235⋯685	27 954	23 810

---

PL spectra of Sr_1−xEu_xSc₂O₄ with x = 0.001, 0.005, 0.010, 0.020, 0.030, and 0.050 are illustrated in Fig. 6. The spectra were recorded using an excitation wavelength of λ_ex = 525 nm. No Eu³⁺ emission was observed. Therefore, it is assumed that the concentration of Eu²⁺ corresponds to the amount which was weighed out. With increasing Eu²⁺ concentration the PL maximum shifts from 688 to 702 nm. In addition to that, with increasing doping concentration the PL intensity of Sr_1−xEu_xSc₂O₄ increases up to an Eu²⁺ content of x = 0.005. At higher concentrations the PL intensity decreases due to concentration quenching (inset Fig. 6).

Fig. 6 PL spectra of Sr_1−xEu_xSc₂O₄ with various Eu²⁺ concentrations. Inset: integrated PL intensity in dependence of the Eu²⁺ concentration.

To investigate the mechanism of the energy transfer responsible for concentration quenching, van Uitert's approach was applied:¹³

here, I is the intensity, k as well as A are constants, and x is the concentration of Eu²⁺. θ = 3 corresponds to exchange interaction whereas θ = 6, 8, and 10 corresponds to dipole–dipole, dipole–quadrupole, and quadrupole–quadrupole interaction, respectively. Fig. 7 depicts the experimental data and the fitting curves. The best fitting is obtained for θ = 6 indicating that the mechanism of concentration quenching occurs via dipole–dipole interaction. This result is quite reasonable since the transitions of excitation and emission of Eu²⁺ are quantum mechanically allowed.

Fig. 7 Dependence of I/I₀ on the Eu²⁺ concentration.

Furthermore, the critical distance R_c between the Eu²⁺ ions was calculated. R_c is defined as the distance, at which the probability of energy transfer equals the probability of radiative transition. To estimate the critical distance R_c between Eu²⁺ ions in SrSc₂O₄:Eu²⁺ Dexter's formula for dipole–dipole interaction can be used:¹⁴

In this equation Q_A is the absorption cross section of Eu²⁺ and E is the energy of maximum spectral overlap. ∫F_S(E)F_A(E)dE represents the spectral overlap of the normalized PL (F_S(E)) and PLE (F_A(E)) spectrum of Eu²⁺. Q_A is 4.8 × 10⁻¹⁶f_d. Here, f_d is the oscillator strength of the electronic transitions of Eu²⁺ and was taken as 0.02. The spectral overlap integral was calculated to be 0.03 eV⁻¹. Plugging these values into eqn (2)R_c is calculated to be 20.3 Å.

In order to examine the temperature behaviour of the PL of SrSc₂O₄:Eu²⁺ PL spectra were recorded from 100 to 500 K. The resulting spectra are illustrated in Fig. 8. With increasing temperature the PL intensity gradually decreases. Moreover, the emission maximum shifts from 692 to 680 nm. Due to extending equilibrium distance between Eu²⁺ and its surrounding oxygen ligands with increasing temperature, crystal field splitting of the d-orbitals of Eu²⁺decreases. As a consequence, the energetic distance between the d and f-orbitals increases and the emission maximum is shifted towards higher energy. The inset of Fig. 8 depicts the integrated PL intensity of the investigated sample. Fitting these data points with a Fermi–Dirac distribution yields the activation energy E_A for thermal quenching.

Fig. 8 PL spectra of Sr_0.995Eu_0.005Sc₂O₄ from 100 to 500 K. Inset: integrated PL intensity in dependence of the temperature.

In eqn (3)I(T) is the PL intensity at a certain temperature and I₀ is the PL intensity at zero kelvin. B is the frequency factor for thermal quenching, k = 8.617 × 10⁻⁵ eV K⁻¹ is the Boltzmann constant, and T is the temperature. Parameters I₀ and B were derived from the fitting function and are about 0.97 and 2945, respectively. From this E_A was calculated to be 0.25 eV. By applying E_A to the following formula, T_1/2 can be calculated:

T_1/2 is the temperature where PL intensity of a luminescent centre decreases to one-half of its maximal intensity. For the herein investigated Sr_0.995Eu_0.005Sc₂O₄ sample T_1/2 was calculated to be 357 K.

To investigate the persistent luminescence properties of Eu²⁺ doped SrSc₂O₄, luminescent lifetime measurements were conducted. For that purpose, the investigated samples were charged for 15 min with an excitation wavelength of λ_ex = 525 nm. Fig. 9 shows the obtained PLD curves of Sr_1−xEu_xSc₂O₄ with x = 0.001, 0.005, and 0.010. The longest afterglow duration was found for Sr_0.999Eu_0.001Sc₂O₄. The corresponding PLD curve can be best fitted with the following equation:

here, I is the PL intensity, A₁, A₂, A₃, and A₄ are fitting parameters, and t is the time. τ₁, τ₂, τ₃, and τ₄ are the partial luminescence lifetimes of the exponential components. The average luminescence lifetime can be obtained using eqn (6):

Fig. 9 PLD curves of Sr_1−xEu_xSc₂O₄ with various doping concentrations.

Partial lifetimes and their corresponding emission fractions as well as the calculated average lifetime τ are summarized in Table 2. With increasing Eu²⁺ content the PL lifetime of SrSc₂O₄:Eu²⁺ decreases and the PLD curve becomes tri-exponential.

Table 2 Persistent luminescence lifetimes τ as well as partial lifetimes τ₁, τ₂, τ₃, and τ₄ and the emission fractions frac₁, frac₂, frac₃, and frac₄ of Eu²⁺ in Sr_1−xEu_xSc₂O₄ and Sr_1−x−yEu_xDy_ySc₂O₄

Sample (x, y)	τ 1 (s)	frac1 (%)	τ 2 (s)	frac2 (%)	τ 3 (s)	frac3 (%)	τ 4 (s)	frac4 (%)	τ (s)
0.001, 0.000	21.7	4.8	88.3	27.4	293.6	47.3	1007.4	20.6	371
0.005, 0.000	10.4	2.3	57.9	18.7	193.6	79.0			164
0.010, 0.000	9.6	6.2	56.1	59.7	283.8	34.1			131
0.001, 0.001	16.2	3	93.7	32.0	271.1	65.0			207
0.001, 0.005	9.2	2.0	84.8	32.2	257.0	65.9			197

---

It is well known that additional incorporation of Dy³⁺ into Eu²⁺ doped materials can lead to long lasting persistent phosphors.¹⁵ In order to investigate the influence of Dy³⁺ co-doping on the afterglow duration, additional luminescence lifetime measurements were performed on Sr_0.998Eu_0.001Dy_0.001Sc₂O₄ and Sr_0.994Eu_0.001Dy_0.005Sc₂O₄. Fig. 10 illustrates the PLD curves of co-doped SrSc₂O₄:Eu²⁺,Dy³⁺. Analogous to Eu²⁺, with increasing Dy³⁺ concentration the afterglow duration decreases. Another possibility to enhance the afterglow duration of SrSc₂O₄:Eu²⁺ could be co-doping with Cr³⁺.¹⁶ Further experiments should be performed to investigate the influence of Cr³⁺ co-doping on the persistent luminescence properties of SrSc₂O₄:Eu²⁺.

Fig. 10 PLD curves of Sr_0.999−yEu_0.001Dy_ySc₂O₄ with different concentrations of Dy³⁺.

Fig. 11a depicts the luminance in dependence of the time. For comparison the solid line visualises the luminance value 0.32 mcd × m⁻² which is the minimum value commonly used for safety signage.¹⁷ The dashed line illustrates the limit of the sensitivity of the dark-adapted eye. For the Sr_0.999Eu_0.001Sc₂O₄ sample 0.32 mcd × m⁻² is reached after about 6 min. Absolute radiance is plotted in Fig. 11b. The absolute radiance values obtained after charging the sample for 15 min are considerably smaller compared to other deep red emitting persistent phosphors.¹⁸

Fig. 11 Persistent luminescence curves of Sr_0.999Eu_0.001Sc₂O₄: luminance (a) and radiance (b) of the Eu²⁺ emission.

4. Conclusions

SrSc₂O₄:Eu²⁺ powder samples were prepared and its PL properties were investigated, also time and temperature dependent. SrSc₂O₄:Eu²⁺ shows a broad emission band in the deep red region of the spectrum with a maximum at about 687 nm. This deep red lying emission band is quite rare for oxidic host materials doped with Eu²⁺.¹ Further, no Eu³⁺ emission was observed. The highest PL intensity was found for a Eu²⁺ concentration of x = 0.005. At a higher Eu²⁺ concentration the PL intensity decreases due to concentration quenching caused by energy transfer between Eu²⁺ ions, likely resulting in energy migration to defect states. It turned out that in SrSc₂O₄:Eu²⁺ energy transfer between Eu²⁺ ions is of a resonance type and occurs via dipole–dipole interaction. The critical distance R_c was calculated to be about 20.3 Å. Temperature dependent PL measurements revealed a T_1/2 value of 357 K. Additionally, it was found that SrSc₂O₄:Eu²⁺ shows deep red persistent luminescence after charging it with an excitation wavelength of λ_ex = 525 nm. The longest afterglow duration was found for the sample Sr_0.999Eu_0.001Sc₂O₄. This sample reaches the value of 0.32 mcd × m⁻² after about 6 min what is considerably shorter than for comparable materials.¹⁸ No increase in afterglow duration could be observed if SrSc₂O₄:Eu²⁺ was additionally doped by Dy³⁺. For this reason, it would be worth the effort to try to increase the afterglow duration by doping with additional ions like Cr³⁺.

Acknowledgements

The authors are grateful to Merck KGaA Darmstadt, Germany for generous financial support.

References

1. D. Deng, H. Yu, Y. Li, Y. Hua, G. Jia, S. Zhao, H. Wang, L. Huang, Y. Li, C. Li and S. Xu, J. Mater. Chem. C, 2013, 1, 3194.
2. (a) K. Song, J. Zhang, Y. Liu, C. Zhang, J. Jiang, H. Jiang and H.-B. Qin, J. Phys. Chem. C, 2015, 119, 24558–24563; (b) R. Yu, N. Xue, T. Wang, Z. Zhao, J. Wang, Z. Hei, M. Li, H. Mi Noh and J. Hyun Jeong, Ceram. Int., 2015, 41, 6030–6036.
3. P. Pust, V. Weiler, C. Hecht, A. Tücks, A. S. Wochnik, A.-K. Henß, D. Wiechert, C. Scheu, P. J. Schmidt and W. Schnick, Nat. Mater., 2014, 13, 891–896.
4. (a) R.-J. Xie, N. Hirosaki, T. Suehiro, F.-F. Xu and M. Mitomo, Chem. Mater., 2006, 18, 5578–5583; (b) S. Tezuka, Y. Sato, T. Komukai, Y. Takatsuka, H. Kato and M. Kakihana, Appl. Phys. Express, 2013, 6, 72101.
5. (a) S. H. M. Poort, H. M. Reijnhoudt, H. O. T. van der Kuip and G. Blasse, J. Alloys Compd., 1996, 241, 75–81; (b) S. H. M. Poort and G. Blasse, J. Lumin., 1997, 72–74, 247–249.
6. W. Zhang, R. Hua, T. Liu, J. Zhao, L. Na and B. Chen, Mater. Res. Bull., 2014, 60, 247–251.
7. (a) C. Chang, J. Xu, L. Jiang, D. Mao and W. Ying, Mater. Chem. Phys., 2006, 98, 509–513; (b) B. Zhang, C. Zhao and D. Chen, Luminescence, 2010, 25, 25–29.
8. (a) J. Zhang, Y.-L. Liu and S.-q. Man, J. Lumin., 2006, 117, 141–146; (b) X. Duane, S. Huang, F. You and K. Kang, J. Rare Earths, 2009, 27, 43–46; (c) S. K. Singh, RSC Adv., 2014, 4, 58674–58698.
9. J. R. Carter and R. S. Feigelson, J. Am. Ceram. Soc., 1964, 47, 141–144.
10. R. Gaume, B. Viana, J. Derouet and D. Vivien, Opt. Mater., 2003, 22, 107–115.
11. R. D. Shannon, Acta Crystallogr., Sect. A: Cryst. Phys., Diffr., Theor. Gen. Crystallogr., 1976, 32, 751–767.
12. (a) G. Blasse, Phys. Status Solidi B, 1973, 55, K131–K134; (b) G. Blasse and B. C. Grabmaier, Luminescent Materials, Springer-Verlag, Berlin/Heidelberg, 1st edn, 1994.
13. L. G. van Uitert, J. Electrochem. Soc., 1967, 114, 1048.
14. (a) D. L. Dexter, J. Chem. Phys., 1953, 21, 836; (b) G. Blasse, Philips Res. Rep., 1969, 24, 131–144.
15. (a) I. P. Sahu, D. P. Bisen, N. Brahme and M. Ganjir, Luminescence, 2015, 30, 1318–1325; (b) B. K. Gupta, A. Kumar, P. Kumar, J. Dwivedi, G. N. Pandey and G. Kedawat, J. Appl. Phys., 2015, 117, 243104; (c) P. Wang, X. Xu, D. Zhou, X. Yu and J. Qiu, Inorg. Chem., 2015, 54, 1690–1697.
16. (a) H. Ryu and K. S. Bartwal, J. Alloys Compd., 2008, 464, 317–321; (b) J. Ueda, K. Kuroishi and S. Tanabe, Appl. Phys. Express, 2014, 7, 62201.
17. K. van den Eeckhout, D. Poelman and P. Smet, Materials, 2013, 6, 2789–2818.
18. (a) J. Xu, S. Tanabe, A. D. Sontakke and J. Ueda, Appl. Phys. Lett., 2015, 107, 81903; (b) Y. Zhuang, J. Ueda and S. Tanabe, Appl. Phys. Express, 2013, 6, 52602.

---