1. Introduction

Recently, great efforts have been made to search for one-phase and full-color phosphors because there is neither energy loss through reabsorption nor deterioration of color rendering index (CRI) due to different photochemistry in mixed phosphors. A common strategy to obtain the full color phosphors is based on the energy transfer from Eu²⁺ to Mn²⁺, such as La_0.827Al_11.9O_19.09: Eu²⁺, Mn²⁺ [1], Ca₂MgSi₂O₇: Eu²⁺, Mn²⁺ [2] and Ba₃MgSi₂O₈: Eu²⁺, Mn²⁺ [3]. However, it is well known that the emission of manganese ions depends on the valence state and distribution of Mn in host lattices. For example, green to yellow emission originates from the tetrahedrally coordinated Mn²⁺ (weak crystal-field), whereas red emission is from the octahedrally coordinated Mn²⁺ or Mn⁴⁺ (strong crystal-field) [4,5]. In most cases, the Mn-activated phosphors emit either pure green or red light. There have been very few reports concerning multiple color emission from Mn singly doped phosphors, which are of great commercial values considering the fact that manganese is a non-rare element.

The key lies in finding of a phosphor with the coexistence of tetrahedral and octahedral cationic sites, which are available from ZnGa₂O₄ spinel for example [6]. Recently, Cr³⁺-activated ZnGa₂O₄ spinel has shown persistent luminescence, which may be used in in-vivo targeted bio-imaging [7]. Since 2012, zinc gallogermanate has emerged to be a new kind of oxide phosphor with some properties superior to ZnGa₂O₄ spinel, for example, a super-long deep-red persistent luminescence has been found in Zn₃Ga₂Ge₂O₁₀:Cr³⁺ [8]. Allix presented a detailed study of the crystal structure of such kind of phosphor. It is found that the zinc gallogermanate is a solid solution of ZnGa₂O₄ spinel and Zn₂GeO₄ willemite [9]. Its crystal lattice also possesses both the tetrahedral Zn²⁺ and octahedral Ga³⁺ sites, thus meets the prerequisite for the multiple color emission.

In the present work, we show that the valence state of Mn and thus the luminescence can be tuned through the composition modification. The origin of the tunable luminescence is discussed from the viewpoint of solid state defect reaction. The criterion of achieving the multiple color emission is proposed.

2. Experimental

Phosphors were prepared by solid state reaction using ZnO, Ga₂O₃, GeO₂, and MnCO₃ as raw materials. Starting materials were weighted according to the formula of Zn_1+xGa_2-2xGe_xO₄: 0.5mol%Mn, x = 0 ~1 and finely mixed in an agate. All the homogeneous powders were then pre-fired at 900 °C for 2 hours in air or with carbon to get a reducing atmosphere. After that, the pre-fired powders were sintered at 1200 °C for 6 hours and then cooled downed to room temperature.

a) Powder X-ray diffraction (XRD) patterns of samples were recorded using a Bruker AXS D8 Advance diffractometer (Voltage 50 kV, current 40 mA, Cu Ka) with a step width of 0.02°; b) The composition of the phosphors was measured by X-ray florescence (XRF) (ARL9800XXP + , Angular accuracy: 0.001°, Thermo ARL, Ecublens, Switzerland); c) Photoluminescence (PL) and photoluminescence excitation spectra (PLE) were measured by a Fluorolog-3-P UV-vis-NIR fluorescence spectrophotometer (JobinYvon, Longjumeau, French) with a Xenon lamp as the excitation source; d) The diffuse reflection spectrum was recorded using a Shimadzu UV-3100 UV-vis-NIR spectrometer and the high white powder BaSO₄with high reflectivity was utilized as a standard; e) Electron paramagnetic resonance (EPR) spectra were recorded on an EXM-8/2.7 EPR spectrometer (BrukerBioSpin GmbH), operating at the X-band frequency (9.457 GHz); f) The decay curves were recorded by FLSP920 (Edinburgh Instruments).

3. Result and discussion

The compositions of the samples are presented in Table 1.It is found that Ga₂O₃ and GeO₂ get evaporated slightly during the preparation course. Previous EPR study of the non-doped zinc gallogermanate indicates that the host matrix contains the gallium and germanium vacancies, which act as the emitting centers responsible for the self-luminescence of the non-doped phosphors [10,11].

Table 1. Nominal and measured compositions of the samples

View Table | View all tables in this article

XRD patterns of the samples are shown in Fig. 1.In accordance with our previous work, pure ZnGa₂O₄ or Zn₂GeO₄ phase is obtained when x equals 0 or 1; only the spinel ZnGa₂O₄ phase is seen when x< 0.5; samples with x ≥ 0.5 contain the Zn₂GeO₄ secondary phase which gradually gains the dominance. The Mn²⁺doped samples exhibit no impurity phases. The compositional evolution of the XRD patterns accord with Allix’s work on the same materials as those studied in this work [9]. A more detailed study on the structure of zinc gallogermanate phosphors can also be referred to our recent publication [11].

 

Fig. 1 X-ray diffraction patterns of the samples Zn_1+xGa_2-2xGe_xO₄:Mn, x = 0 ~1.

Download Full Size | PDF

To estimate the absorption induced by Mn doping, the difference reflection spectra which are obtained by subtracting the spectra of the Mn²⁺-doped samples from those of the non-doped ones are shown in Fig. 2.The doped samples exhibit two absorption bands: one is in a range of from 275 nm to 375 nm; the other is from 400 nm to 550 nm. The peak position of the former shifts to longer wavelength with the substitution until x = 0.4, from which point it starts to shift to shorter wavelength.

 

Fig. 2 Difference reflection spectra obtained by subtracting the spectra of the Mn²⁺-doped samples from those of the non-doped ones.

Download Full Size | PDF

The PLE and PL spectra (Fig. 3) of the Mn-activated ZnGa₂O₄ spinel (x = 0) and Zn₂GeO₄ (x = 1) willemite are alike to the previous work [12], that is, the former and latter phosphors emit bright green emissions at 503 and 532 nm when excited by 290 and 332 nm UV lights, respectively. It is interesting to note that a pure deep red emission at ~680 nm is observed when x = 0.05 ~0.25. The PLE spectra show two excitation bands, that is, a band in a range of from 250 nm to 375 nm, and a band from 375 nm to 500 nm when monitoring the red emission. Samples with x = 0.4 ~0.75 show a mixture of the 532 nm green and 680 nm red emissions. The green emission undoubtedly stems from the ⁴T₁→⁶A₁ electronic transition of Mn²⁺ occupying the Zn²⁺ sites. The origin of the red emission however is rather subtle. After a careful scrutinization of the emission spectra, it is found that the Mn-activated ZnGa₂O₄ (x = 0) also shows a very week red emission at ~680 nm. Similar red emission has been observed from the Mn-activated ZnGa₂O₄ by Yu [13] and Hsu [14] et al, and they both suggested that a certain amount of Mn²⁺ are oxidized to Mn⁴⁺ and then substitute for the octahedrally coordinated Ga³⁺ leading to the deep red emission. The formation of Mn⁴⁺ significantly quenches the green emission of Mn²⁺ because the absorption of Mn⁴⁺ overlaps with the emission of Mn²⁺. The formation of the octahedrally coordinated Mn⁴⁺ proceeds according to the following mechanism: Ge⁴⁺ substitutes for Ga³⁺ in ZnGa₂O₄ because the charge and size different between Ga³⁺-Ge⁴⁺ is smaller than Ga³⁺-Zn²⁺ [R (Zn²⁺) = 0.74 Å, R (Ga³⁺) = 0.62 Å and R (Ge⁴⁺) = 0.53 Å] [15]. However, to balance the extra positive charge of Ge⁴⁺, a pair of Ge⁴⁺ and Mn²⁺ will simultaneously replace two Ga³⁺, which is highly possible because the radius size of Mn²⁺ (0.66 Å) is also close to that of Ga³⁺. When prepared in the air at high temperature (e.g. in the present case), Mn²⁺ tends to oxidize to Mn⁴⁺ such that Mn⁴⁺ occupies the octahedral Ga³⁺ sites and emits red light [12]. This is similar to the process occurring in CaAl₁₂O₁₉:Mn, 2Al³⁺ → Ge⁴⁺ + Mn⁴⁺ + O^2- [16].

 

Fig. 3 Excitation (a) and emission (b) spectra of the samples Zn_1+xGa_2-2xGe_xO₄:Mn, x = 0 ~1. The excitation wavelengths are 290 nm for the x = 0 sample and 350 nm for the rest of samples.

Download Full Size | PDF

The red emission band is attributed to the electronic transition between ²E and ⁴A₂ of Mn⁴⁺ accompanied with vibronic transitions. Here, one may argue that the octahedrally coordinated Mn²⁺ may also emit red light, nevertheless, as reported by Kim et al, the octahedrally coordinated Mn²⁺ in ZnGa₂O₄ phosphor emits at 513 nm [17]. Therefore the possibility of Mn²⁺ emitting red is not considered in the present study.

We would like to mention that the deep red emission observed from the studied phosphors is far away from the eye sensitive region. Thus, similar to Cr³⁺ [7], the deep red emission of Mn⁴⁺ seems to be more suitable as a source for in-vivo bio-imaging.

To verify the red emission is indeed originated from Mn⁴⁺, the sample showing the strongest red emission (x = 0.1) was also prepared in a reducing atmosphere filled with CO gas. The reduced sample emits only the 503 nm green light similar to the Mn-activated ZnGa₂O₄ (x = 0) (result is not shown here). More supporting evidence can be gleaned from the EPR measurement as shown in Fig. 4.The reduced sample shows the typical hyperfine sextet arising from the Mn²⁺, whereas the EPR signal of the sample prepared in the air (oxidized) weakens dramatically indicative of the decreased concentration of Mn²⁺ [18]. Same changes also occur when comparing the EPR spectra of the Mn-activated ZnGa₂O₄ (x = 0) and the Ge⁴⁺ substituted sample (x = 0.1) with the latter shows much weak EPR signal of Mn²⁺ (Fig. 4(b)).

 

Fig. 4 EPR spectra (a) of the Mn-activated (x = 0.1) samples prepared in the air (red line, enlarged by 10 times) and CO atmosphere (black line). EPR spectra (b) of the Mn-activated samples with x = 0 (black line) and x = 0.1 (red line, enlarged by 10 times).

Download Full Size | PDF

The compositional variation of the integrated intensities of the green and red emission bands is shown in Fig. 5.With only a little amount of Ga³⁺ being substituted by Ge⁴⁺ in ZnGa₂O₄, the green emission is completely quenched, while a moderately intense deep red emission band appears (x = 0.05). The deep red emission increases until x = 0.1 and then its intensity starts to decline as more Ga³⁺ are substituted by Ge⁴⁺. On the other hand, the intensity of the 530 nm green emission grows. The results clearly show that the substitution of Ge⁴⁺ for Ga³⁺ encourages the formation of Mn⁴⁺, leading to the enhancement of the deep red emission. It may also suggest that the coexistence of Ge⁴⁺ and Ga³⁺ is necessary for inducing the red emission. As mentioned above, the high-temperature-open-air synthesis route favors the formation of Mn⁴⁺, while the substitution of Ge⁴⁺ for Ga³⁺ helps to distribute/separate the Mn⁴⁺ ions. The incorporation of Ge⁴⁺ was also used by Y. Li et al as an effective strategy to improve the deep red emission of Mn⁴⁺ [19]. However, based on the solid state nuclear magnetic resonance (NMR) studies of the chemical environment of Ga³⁺ in the same phosphors (x = 0 ~0.5), Allix et al found that as more Ga³⁺ are replaced by Ge⁴⁺, the less number of octahedrally coordinated Ga³⁺ sites are available for being occupied by Mn⁴⁺ [9], consequently, the deep red emission fades away with further substitution. The rhombohedral crystal structure of Zn₂GeO₄ has only tetrahedral sites, thus one can only see the green emission of Mn²⁺. To our best knowledge, the red emission of Mn⁴⁺ was observed only from thin film Zn₂GeO₄: Mn phosphor grown by pulsed laser deposition. A considerable structural change must occur to adapt to the change in the valence of Mn which is possible only for films with structural flexibility [12].

 

Fig. 5 Integrated intensities of the emissions at 532 and 680 nm as a function of Ge⁴⁺content (X). The insets are the digital photos of the samples (x = 0.1 and x = 1) being excited by UV lights in the dark.

Download Full Size | PDF

The luminescence decay curves of the green and red emissions of all the samples are measured. Here, the decay curves of the sample with x = 0.5 is shown in Fig. 6 as an example. All the decay time can be expressed by: I = A₁exp (-t/τ₁) + A₂exp (-t/τ₂) where I is the luminescence intensity; A₁ and A₂ are constants; t is the time; and τ₁ and τ₂ are the rapid and slow parts of lifetime. The average lifetime (τ) is calculated by the following equation:τ = (A₁τ₁ + A₂τ₂)/(A₁ + A₂) and is presented in Table 2.The decay time of the green emission is in the millisecond time regime typical for Mn²⁺. Whereas, the red emission has a much shorter lifetime in the microsecond time regime, which is again reminiscent of the green and red emissions originating from different valence state of Mn ions.

 

Fig. 6 Luminescence decay curves of the green (green curve) and red (red curve) emissions from the sample with x = 0.5 taken as an example. The insets indicate the different coordination environments of Mn²⁺ (four-fold coordinated) and Mn⁴⁺ (six-fold coordinated). The gray and violet balls represent the oxygen and Mn²⁺/Mn⁴⁺ ions, respectively.

Download Full Size | PDF

Table 2. Decay times of the samples monitored at 532 nm and 680 nm emissions.

View Table | View all tables in this article

4. Conclusion

The color of the Mn-activated zinc gallogermanate phosphors can be tuned by adjusting the green emission from Mn²⁺ and red emission from Mn⁴⁺ through controlling the concentration of Ge. The high-temperature-open-air synthesis route and the coexistence of Ge⁴⁺ and Ga³⁺ are the criteria for inducing the deep red emission of Mn⁴⁺. The Ge⁴⁺ substitution for Ga³⁺ helps to distribute/separate the Mn⁴⁺ ions and thus enhance the deep red emission. However, the more Ga³⁺ are replaced by Ge⁴⁺, the less number of octahedrally coordinated Ga³⁺ sites are available for being occupied by Mn⁴⁺, as a result, the intensity of the deep red emission falls. The present phosphors can be used as one-phase full-color phosphors and for in-vivo bio-imaging.