Luminescence and structural analysis of Ce³⁺ and Er³⁺ doped and Ce³⁺–Er³⁺ codoped Ca₃Sc₂Si₃O₁₂ garnets: influence of the doping concentration in the energy transfer processes

Abstract

We present research on Ca₃Sc₂Si₃O₁₂ garnets doped with Ce³⁺ and Er³⁺ ions that were synthesized by the freeze-drying precursor method. The structural characterization was performed by X-ray diffraction (XRD) and X-ray photoelectron spectroscopy (XPS). Scanning electron microscopy (SEM) images of the calcined material were studied. Nanocrystals of Ca₃Sc₂Si₃O₁₂ were obtained by calcining the precursor at 1200 °C for 4 h with a mean size of about 100 nm. The luminescence features of Ce³⁺ and Er³⁺ doped samples indicated that these ions are effectively incorporated into the crystalline phase. In addition, codoping with Ce³⁺–Er³⁺ ions results in energy transfer processes from Ce³⁺ to Er³⁺ ions which led to the enhancement of near-infrared luminescence at 1.5 μm. The efficiency of this non-radiative energy transfer process has been studied by means of excited state dynamics as a function of the doping concentration. From these results we have found that near-infrared Er³⁺ luminescence efficiency under VIS Ce³⁺ excitation is at maximal for 1Ce³⁺–1Er³⁺ codoped Ca₃Sc₂Si₃O₁₂ garnets (in mol%). Luminescence quenching at higher Er³⁺ ion concentrations is observed.

---

1. Introduction

Rare earth (RE) doped luminescent materials have attracted much attention due to their potential applications in optoelectronic technology. In particular, Er³⁺ doped materials could be helpful in germanium based photovoltaic systems to make better use of the solar spectrum. They could enhance photon access in the near infrared (NIR) region where germanium cells show the highest conversion efficiency. The down-conversion involved processes are based on an energy transfer mechanism between RE pairs, starting with sensitizers that have strong absorption in the UV-blue region.^1,2

One of the most promising candidates for use as a sensitizer is the Ce³⁺ ion. Absorption of Ce³⁺ originates from the allowed electric-dipole transition from the 4f ground state to the 5d excited one, resulting in a high absorption cross section in the UV-blue spectral region.³ This absorption can be tuned by selection of the host material due to the strong dependence of the Ce³⁺ ion level diagram on the crystalline field, which also has appreciable shifts on the luminescence band.^4,5 On the other hand, with respect to the typical option of Yb³⁺ ions as the luminescent center after energy transfer from the Ce³⁺ 5d level,⁶ Er³⁺ ions are an excellent option for energy conveyance from UV to NIR photons at 1550 nm, because they could involve the obtainment of three NIR photons after absorption of UV-blue photons,⁷ with a potential effective quantum efficiency close to 300%. The use of Ce³⁺ ions as sensitizers enhances the absorption in the UV-blue region, where Er³⁺ ions only present relatively weak and narrow absorption peaks.

The combination of Ce³⁺ and Er³⁺ ions with a suitable host would optimize the conversion of UV-VIS radiation into NIR. In this sense, nano-garnets with the general formula A₃B₂C₃O₁₂ have been widely studied after doping with RE ions due to their potential applications in optoelectronics. In particular, a silicate garnet with the composition Ca₃Sc₂Si₃O₁₂ (CSSO) has been synthesized by different methods such as solid-state reaction, microwave or sol–gel. Moreover their luminescence properties have been studied after doping with several lanthanides (Ce³⁺, Tb³⁺, Pr³⁺, Eu³⁺ or Nd³⁺).^8–15 But, as far as we know, they have not been studied after codoping with Ce³⁺ and Er³⁺.

In this work, Ce³⁺ and Er³⁺ single and codoped Ca₃Sc₂Si₃O₁₂ garnets have been synthesized using a freeze-drying precursor route. Structural characterization by XRD and SEM techniques was performed to confirm the crystal structure of the CSSO garnet and to observe the microstructure of the sintered samples. Cerium ions could be present in the samples as Ce⁴⁺ ions, which are not luminescent. By this reason XPS was used for estimating the relative concentration of the two oxidation states of Ce (III and IV) on the surface of the samples. Finally, photoluminescence properties of single doped and codoped garnets and excite state dynamics in codoped samples have been studied in order to analyze the efficiency of the energy transfer mechanisms.

2. Experimental

2.1. Synthesis

Undoped, Ce³⁺–Er³⁺ single- and co-doped Ca₃Sc₂Si₃O₁₂ polycrystalline powders were synthesized by the freeze-drying precursor method,^16–18 using CaCO₃ (99%, Merck), Sc₂O₃ (99.9%, Aldrich), tetraethyl orthosilicate (C₈H₂₀O₄Si) (98%, Aldrich), Ce(NO₃)₃·6H₂O (99.99%, Aldrich) and Er(NO₃)₃·5H₂O (99.9%, Aldrich) as reagents. Table 1 presents the different dopant concentrations that have been considered in this work. Hydrated metal nitrates were previously studied by thermogravimetric analysis to determine the cation content. Stoichiometric quantities of the reagents were individually dissolved in distilled water and then mixed together. Afterwards, ethylenediaminetetraacetic acid (EDTA) (99.7% Aldrich) was added as the complexing agent, in a 1 : 1 ligand to metal molar ratio, to prevent precipitation.

Table 1 Sample codes and dopant content, in mol%, in the Ca₃Sc₂Si₃O₁ (CSSO) garnets studied in this work

Sample	Ce^3+	Er^3+
CSSO	0	0
CSSO–1Ce	1	0
CSSO–4Ce	4	0
CSSO–0.5Er	0	0.5
CSSO–1Er	0	1
CSSO–1Ce0.5Er	1	0.5
CSSO–1Ce1Er	1	1
CSSO–1Ce2Er	1	2
CSSO–1Ce4Er	1	4

---

The pH of the solution, which was initially acidic, was adjusted to around 7–8 by adding aqueous ammonia. Drops of the as-obtained solution were subsequently flash-frozen by pouring them into liquid nitrogen. In this way, each frozen drop retains the homogeneous cation distribution of the original solution. Then, the sample was freeze-dried in a HetoLyolab 3000 (Heto-Holten A/S) freeze-dryer apparatus for 3 days. The obtained highly hygroscopic amorphous precursor was immediately pre-calcined at 400 °C for 2 hours to prevent rehydration and to eliminate organic matter. Finally, the dry powders were pressed at 10 Tm in a hydrostatic press in order to obtain pellets which were calcined in alumina crucibles at 1200 °C for 4 hours.

2.2. Experimental setup

X-ray analyses (XRD) were carried out by using a Philips diffractometer equipped with an X’Celerator detector and a Cu anode (Cu Kα₁, α₂) in the 20–80 2θ diffraction range. The microstructure of the samples was studied by scanning electron microscopy (SEM) (Jeol LTD, mod. JSM-6300). All samples were covered with a thin film of silver to avoid charging problems and to obtain better image definition. The high-resolution X-ray photoelectron spectroscopy (XPS) spectrum was collected on an ESCALAB 250 spectrometer, using a monochromatized Al K_α X-ray radiation (hν = 1486.6 eV) with a spot size of 650 μm. The spectrometer energy calibration was performed using the Au 4f_7/2 and Cu 2p_3/2 photoelectron lines. The chemical shifts in the spectra were corrected with reference to the binding energy of the C 1s core level spectrum, corresponding to the carbon contamination layer at 284.6 eV. This peak arises from a minor quantity of contaminants and is generally accepted to be independent of the chemical state of the sample under investigation. The spectrum was collected in constant analyzer energy (CAE) mode, with pass energy of 20 eV and with an energy resolution of about 0.1 eV. For all the measurements, the pressure in the ultra-high vacuum analysis chamber was lower than 8 × 10⁻⁹ mbar, avoiding ejected photoelectron interaction with gas molecules. Laser Raman spectra were recorded on a Jobin Yvon/Labram HR spectrometer using 632.8 nm excitation from a He–Ne laser. Double beam Shimadzu UV-2450 Scan UV-visible spectrophotometer was used to record the reflectance spectra over a wavelength range of 200–800 nm. Emission and excitation spectra were recorded by using a 300 W xenon lamp with a double-grating monochromator (SPEX 1680). These spectra were collected by using a spectrometer (ANDOR Shamrock) equipped with an UV-VIS-NIR CCD camera (Newton CCD) and a NIR InGaAs or a single-grating monochromator (SPEX 1681) equipped with a photomultiplier tube (Hamamatsu R928). The spectral response of the set-up was taken into account. The luminescence decays in the nanosecond scale were measured by using a LifeSpec II spectrometer from Edinburgh Instruments. The samples were excited by 70 ps pulses with a 1 MHz repetition rate produced by a semiconductor laser operating at 405 nm. A multi-channel plate photomultiplier tube was used for the detection. The system was operated in time correlated single photon counting mode and had an instrumental response function (IRF) with a half maximum width of about 70 ps. The millisecond kinetics of luminescence were recorded using an optical parametrical oscillator (OPO) pumped by the 3rd harmonic of a YAG:Nd laser as the excitation source. The signal was detected with a digital storage oscilloscope controlled by a personal computer. The samples were excited by 6 ns pulses with a 20 Hz repetition rate. The time resolution of this setup was 20 ns.

3. Results and discussion

The freeze-drying synthesis technique shows some advantages with reference to other precursor methods like co-precipitation or sol–gel method. The co-precipitation methods require that the cation solubility in the reaction mixture be very low to achieve nominal composition, while in the sol–gel method residual carbon from the complexing agent or from the gelation agent can remain, which do not decompose until reaching high temperatures. The amorphous product obtained using the freeze-drying method only contains nitrates which decompose at low temperatures. This method has been successfully used in previous works, obtaining pure phases and nanostructured materials at relatively low temperatures.^16–18 For all of these reasons, we have used the freeze-drying synthesis method for studying the influence of doping concentration in the energy transfer processes of CSSO doped with Ce³⁺ and Er³⁺ ions.

3.1. Structural characterization

3.1.1. XRD phase characterization and crystal size. XRD patterns of Ce³⁺–Er³⁺ doped garnets previously heat treated at 1200 °C for 4 hours show a major phase, Ca₃Sc₂Si₃O₁₂, with a secondary crystalline phase, like Ca₃SiO₅, Sc₂O₃ and SiO₂, see Fig. 1. These phases were also observed in our previous work about CSSO garnets codoped with Ce³⁺ and Tb³⁺ ions.¹⁹ Moreover, mean sizes of the nanocrystals have been obtained using the Scherrer equation,²⁰ from the full width at half maximum (FWHM) and the XRD peak position:where d is the crystal size, λ is the X-ray wavelength, θ is the diffraction angle, β is the full-width at half maximum (FWHM) of the diffraction peak and k is a constant depending on the particle shape (currently k = 0.9). The obtained nanocrystal mean sizes, shown in Table 2, are similar to those previously obtained by us in Ce³⁺–Tb³⁺ codoped CSSO garnets.¹⁹

Fig. 1 XRD patterns of undoped, Ce³⁺ and Ce³⁺–Er³⁺ doped Ca₃Sc₂Si₃O₁₂ heat treated garnets; the main reflection planes are indicated. Standard peaks of Ca₃Sc₂Si₃O₁₂ (JCPDS file 72-1969) have been included for comparison purposes.

Table 2 Mean sizes of Ce³⁺–xEr³⁺ (x = 0, 1 or 4) co-doped CSSO nanocrystals, in mol%, after thermal treatment at 1200 °C for 4 hours

CSSO–1Ce	110 nm
CSSO–1Ce1Er	138 nm
CSSO–1Ce4Er	88 nm

---

Regarding the crystal structure of the CSSO garnets, it corresponds to the cubic system with Ia3d̄ space group (#230).⁶ The Ca²⁺ ions are situated in the center of eight-coordinated dodecahedra (24c), while Sc³⁺ ions are located in the center of six coordinated octahedra (16a) and Si⁴⁺ ions are in the center of four-coordinated tetrahedra (24d). It is known that in the Ca₃Sc₂Si₃O₁₂ garnet structure one dodecahedron is associated to four other dodecahedra and four octahedral; two tetrahedra by shared edges and the other four octahedral by corner sharing.²¹ According to the literature, Ca²⁺ ions (ionic radius 114 pm) are substituted for Ce³⁺ and Er³⁺ ions in the dodecahedral position and do not substitute Sc³⁺ ions because of the difference of the respective ionic radius, 75 pm for Sc³⁺ and up to 100 pm for Ce³⁺ (115 pm) and Er³⁺ (103 pm).^22–25

3.1.2. Morphology. The morphology of the prepared doped powders calcined at 1200 °C for 4 hours has been studied by SEM. Fig. 2 illustrates the irregularity in shape and size of the sintered material.

Fig. 2 SEM micrographs of 1Ce³⁺–xEr³⁺ (x = 0, 1, 2, 4), in mol%, co-doped Ca₃Sc₂Si₃O₁₂ garnets as a function of Er³⁺ ion concentration.

The mean size of the aggregated grains, around 40 μm, is bigger than those obtained by other conventional synthesis processes like sol–gel combustion or emulsion-evaporation methods, in which higher heat treatment temperatures are necessary to obtain aggregated grains around 1–10 μm.^26,27 On the other hand, it can be noted that the morphology of the samples seems to be affected by the Er³⁺ dopant concentration. When the Er³⁺ concentration increases from 0 to 1 mol%, an increment of the sinterization level could be observed. Further increases in the Er³⁺ concentration, from 2 to 4 mol%, does not result in a significant change of the sintering.

3.1.3. Relative concentration of Ce oxidation states (XPS). The garnet samples doped with a high content of Ce (4 mol%), were also analyzed by X-ray photoelectron spectroscopy (XPS) to detect possible chemical changes in the surface of the samples. In particular, we used this technique in order to estimate the relative concentration of the two oxidation states of Ce (III and IV).

Fig. 3 shows the high resolution XPS spectra of Ce 3d in the binding energy (BE) range of 870–930 eV. The peaks between 870 and 895 eV belong to the Ce 3d_5/2 levels while peaks between 895 and 913 eV correspond to the Ce 3d_3/2 levels. The subscripts of U and V correspond to the oxidation state of Ce and the superscripts are an arbitrary label in order to differentiate the peaks of the same oxidation state.

Fig. 3 High resolution XPS spectra of the Ce 3d region (solid line) in a CSSO–4Ce garnet. Deconvolution of the peaks is indicated in dotted lines. The red dashed line represents the best fit obtained taking into account all deconvolutions.

The satellite peak at around 917 eV, which is not observed in the initial state of trivalent Ce, is a characteristic fingerprint of the presence of cerium(IV), indicating that there is a mixture of Ce(III) and Ce(IV) ions in the sample.²⁸

As it can been observed, the shape of the Ce 3d spectral line is complex due to the overlapping of several doublet lines resulting from different final states of cerium after the emission of photoelectrons. Thus, it is well accepted that the presence of Ce⁴⁺ ions leads to the appearance of three doublets and the Ce³⁺ ions produce two doublets in the XPS spectrum, with a total sum of ten lines. For the quantitative calculations of the Ce³⁺ and the Ce⁴⁺ concentrations, the high resolution XPS spectra of Ce 3d were deconvoluted in ten lines (see Fig. 3), according to the oxides of Ce(III)₂O₃ and Ce(IV)O₂.²⁹ In this work, U and V notations are used for the assignment of Ce 3d corresponding to 3d_3/2 and 3d_5/2, respectively.

In the Ce³⁺ species, V₃₊¹–U₃₊¹ are the shake down peaks and V₃₊²–U₃₊² are the main peaks. In Ce⁴⁺, two shake down peaks V₄₊¹–U₄₊¹ and V₄₊²–U₄₊² were assigned, and the main peaks V₄₊³–U₄₊³ were the characteristic peaks of Ce⁴⁺.

The concentration of Ce³⁺ and Ce⁴⁺ can be determined from the ratio of the integrated peak areas according to the following equations:³⁰

where the notations Ce(III) and Ce(IV) represent the corresponding sums of the integrated peak areas related to the Ce XPS spectrum. By applying this equation, we obtained rates of about 40 and 60% for Ce³⁺ and Ce⁴⁺, respectively.

It could be said that XPS is a surface technique which explores materials until a depth of about 10 nm. However, when this technique is used, as in our case, with crystal sizes of about 100 nm, the analysed volume corresponds to about half of the total volume of the nanocrystals. Therefore, the XPS results indicate that a large fraction of the cerium ions is present in our sample as Ce³⁺ and then participate in the luminescence processes.

3.1.4. Laser Raman spectra. Raman experiments of undoped and Ce³⁺ doped CSSO garnets were carried out in order to complete the structural characterization. Fig. 4 shows the Raman spectra of these garnets excited at 632.8 nm. These spectra were measured at different areas of the garnets in order to confirm the homogeneity of the samples. The spectra mainly present the usual features of Ca₃Sc₂Si₃O₁₂ garnets, together with a peak of the Sc₂O₃ phase³⁰ at 420 cm⁻¹.

Fig. 4 Raman spectra of CSSO and CSSO–Ce obtained under 632.8 nm laser excitation.

According to previous studies, Ca₃Sc₂Si₃O₁₂ garnets have 25 Raman active vibration modes expected on the basis of the analysis of the cubic space group Ia3d̄ of the silicate garnets.^21,31–35 In particular, 13 of them are observed in the region of 200–950 cm⁻¹ as can be observed in Fig. 4, together with the peak at 420 cm⁻¹ corresponding to the mode Fg of Sc₂O₃ phase.³⁰ According to these studies, the band assignments can be summarized as follows.

The high frequency modes (800–950 cm⁻¹) are related to symmetric and asymmetric internal stretching vibrations of rigid SiO₄ tetrahedra. The modes lying from 450 to 700 cm⁻¹ are assigned to bending motions of these tetrahedra. Finally, the remaining lattice modes (200–415 cm⁻¹) involve rotations and translations of the SiO₄ groups, octahedrally coordinated trivalent cations and dodecahedrally coordinated divalent cations. On the other hand, in a previous work we demonstrated that the intensity reduction of the peak corresponding to the Sc₂O₃ phase when doping with RE³⁺ ions is in agreement with the XRD results. Moreover, this Sc₂O₃ phase almost vanishes by RE³⁺ doping.¹⁹

3.2. Spectroscopy analysis

3.2.1. Absorption spectra. Absorption spectra of Ce³⁺ and Er³⁺ doped and Ce³⁺–Er³⁺ codoped CSSO garnets are shown in Fig. 5. A strong absorption peak is observed at 265 nm together with another peak at 440 nm, which correspond to 4f → 5d transitions of Ce³⁺ ions. On the other hand, weaker Er³⁺ absorption peaks are also observed at 520 and 647 nm, corresponding to intraconfigurational 4f–4f transitions in the visible range to the ²H_11/2 and ⁴F_9/2 levels of Er³⁺ ions, respectively. As result, it is found that Ce³⁺ ions are more efficiently excited than Er³⁺ ions in the UV-blue spectral range.

Fig. 5 Absorption spectra of Ce³⁺ and Er³⁺ doped and Ce³⁺–Er³⁺ codoped CSSO garnets. Spectra are multiplied by a factor of 20 in the range of 500–700 nm.

3.2.2. Luminescence.
3.2.2.1. Ce³⁺ and Er³⁺ single doped garnets. The photoluminescence spectra of Ce³⁺ and Er³⁺ 1 mol% single-doped CSSO garnets are shown in Fig. 6. The excitation and emission spectra corresponding to 1 mol% Ce³⁺ doped garnets are similar to those obtained by authors in a previous work.¹⁹ Moreover, these spectra are also similar to those obtained by Cheng et al.³³ for Ce³⁺ ions in a closed matrix. The emission spectrum shows a broad emission band with two components, at 505 and 550 nm. The component at 505 nm corresponds to the interconfigurational transition of the Ce³⁺ ions from the lowest energy 5d level to the ground level, i.e. ⁵D_3/2 → ²F_5/2, shown in Fig. 6. The component located at 550 nm corresponds to the transition from the same emitting level to the first excited 4f level, i.e. ⁵D_3/2 → ²F_7/2. Moreover, the excitation spectrum of these emissions presents two peaks, at 311 and 450 nm. The peak at 450 nm is two orders of magnitude more intense.

Fig. 6 Normalized excitation and emission spectra of Ce³⁺ (black) and Er³⁺ (red) 1 mol% single-doped CSSO garnets with Ce³⁺ and Er³⁺, respectively.

On the other hand, in the sample single-doped with Er³⁺ under excitation at 378 nm, that corresponds to the ⁴I_15/2 → ⁴G_11/2 transition of Er³⁺ ions, emissions at 545, 560, 660 and 850 nm are observed, see Fig. 6. These emissions correspond to transitions from the ²H_11/2, ⁴S_3/2, ⁴F_9/2 and ⁴I_9/2 levels to the ground state ⁴I_15/2, respectively. These peaks, with well resolved Stark components, are typical of ions in crystalline environments and they are indicative of the effective incorporation of Er³⁺ ions into the CSSO garnets.

3.2.2.2. Ce³⁺–Er³⁺ codoped garnets. The incorporation of Er³⁺ ions into Ce³⁺–Er³⁺ codoped CSSO garnets results in a decrease of the Ce³⁺ emission intensity obtained under direct excitation of these ions, see Fig. 7. An excitation wavelength of 425 nm was selected in order to excite Ce³⁺ ions without simultaneous excitation of Er³⁺ ions. However, VIS and NIR luminescence of Er³⁺ ions at 540, 850 and 1550 nm is also observed, with the 1550 nm emission being much more intense than the others, as is shown in Fig. 7. It is noted that the highest emission intensity at 1550 nm is obtained for the Er³⁺ intermediate concentration of 1.0 mol%.

Fig. 7 (A) VIS-NIR emission spectra of Ce³⁺–Er³⁺ codoped CSSO garnets under excitation at 425 nm. Inset shows a magnified VIS spectrum of a 1Ce³⁺–4Er³⁺ codoped CSSO garnet. (B) Energy level diagrams of Ce³⁺ and Er³⁺ ions with main excitation and emission transitions, indicated by solid arrows, and energy transfer mechanisms, indicated by dashed arrows.

These Er³⁺ emissions are due to an energy transfer from Ce³⁺ to Er³⁺ ions. These energy transfer processes are caused by the overlap of the broad Ce³⁺ emission at 505–550 nm, see Fig. 6, with the Er³⁺ absorption, mainly at 520 nm (⁴I_15/2 → ²H_11/2), see Fig. 5.

On the other hand, the decrease of the Ce³⁺ emission with the increase of the Er³⁺ concentration is not followed by an equivalent enhancement of the NIR emission of these ions, as observed in Fig. 7. For this reason, the excited state dynamics after pulsed excitation were studied in order to analyse the mechanisms involved in the luminescence of these materials.

3.2.3. Excited state dynamics. The decays of the Ce³⁺ emission in Ce³⁺–xEr³⁺ codoped CSSO garnets (x = 0, 0.5, 1, 2 and 4 mol%), obtained after excitation and detection at 405 and 505 nm, respectively, are shown in Fig. 8. These decays present non-exponential behaviour related to the energy transfer from Ce³⁺ to Er³⁺ ions after Ce³⁺ excitation. These results can be analysed by using the Inokuti–Hirayama expression for dipole–dipole interaction between donors and acceptors:³⁶where τ is the intrinsic lifetime of the donors and Q is a parameter related to the energy transfer probability, which can be calculated from the fits. Good fits are obtained, as shown in Fig. 8, with values for Q increasing from 0.48 to 2.92 when the Er³⁺ concentration increases from 0.5 to 4 mol%.

Fig. 8 Decays of Ce³⁺ doped and Ce³⁺–Er³⁺ codoped CSSO garnets with excitation at 405 nm and detection at 505 nm. The red lines represent the best fits of the experimental results to eqn (3).

On the other hand, the energy transfer efficiency from Ce³⁺ to Er³⁺ ions, η_ET, can be calculated by using eqn (4),

where τ_Ce-0Er and τ_Ce-xEr are the effective lifetime of Ce³⁺ emission in single and codoped samples, respectively, calculated from the decays shown in Fig. 8. The obtained results are shown in Table 3, for the different garnets. As can be seen, the lifetime of the Ce³⁺ ion gradually decreases with the Er³⁺ concentration while the energy transfer efficiency increases up to 0.82 for the highest Er³⁺ doping level of 4 mol%. The increase of this energy transfer efficiency from Ce³⁺ ions with the Er³⁺ concentration is roughly in agreement with the decrease of the Ce³⁺ emission, shown in Fig. 7, when the Er³⁺ concentration increases. A similar effect was observed by G. Zhang et al.³⁷ in the case of Ce³⁺–Pr³⁺ codoped CaLaGa₃S₆O garnets. However, the NIR Er³⁺ emission initially increases with the concentration of these ions but it decreases for concentrations over 1 mol%, see Fig. 7.

Table 3 Effective lifetime of Ce³⁺ emission, τ_Ce-xEr. Energy transfer efficiency η_ET, obtained by means of eqn (3). Effective lifetime of the Er³⁺ NIR emission, τ_xEr. Quantum efficiency of the Er³⁺ NIR emission η_Er-emis. Calculated and experimental values for the emission ratio I_Er/I_Ce

Samples	τCe-xEr (ns)	ηET	τxEr (ms)	ηEr-emis
					Calc.	Exp.
CSSO–1Ce	56.0	—	—	—
CSSO–1Ce0.5Er	44.5	0.21	6.0	1.0	0.27	0.10
CSSO–1Ce1Er	33.6	0.40	5.9	0.98	0.65	0.32
CSSO–1Ce2Er	19.8	0.65	1.6	0.27	0.50	0.28
CSSO–1Ce4Er	10.0	0.82	0.9	0.15	0.68	0.45

---

Having such hints, we measured the decay of the NIR Er³⁺ emission at 1550 nm after pulsed excitation at 405 nm, corresponding to direct excitation of Er³⁺ ions. These decays, shown in Fig. 9, present a single exponential behaviour for the lowest Er³⁺ concentrations of 0.5 and 1 mol%, with a lifetime of 6.0 ms. Meanwhile, for concentrations above 1 mol% the emission decays are quite faster, with effective lifetimes presented in Table 3. This behaviour would be due to quenching of the emission by energy migration until luminescence is trapped.

Fig. 9 Decays of Ce³⁺–Er³⁺ codoped CSSO garnets excited at 405 nm and detected at 1550 nm, corresponding to the ⁴I_13/2 → ⁴I_15/2 transition.

From these effective lifetimes, we obtained the quantum efficiency of the NIR Er³⁺ emission by using eqn (5):

where τ_xEr and τ_0.5Er are the effective lifetime for the different Er³⁺ concentrations. The results obtained for the quantum efficiency of this emission, given in Table 3, present large values, close to 1, for an Er³⁺ concentration up to 1 mol% with an important decrease for larger concentrations.

Now we can calculate the expected ratio of Er³⁺ to Ce³⁺ integrated emissions, taking into account the energy transfer efficiency from the Ce³⁺ ions and the quantum efficiency of the NIR Er³⁺ emission, by using eqn (6).

The obtained results are shown in Table 3 and would be compared with the experimental values obtained directly from the integrated photon fluxes of Er³⁺ and Ce³⁺ emissions, from spectra in Fig. 7.

The experimental emission ratios obtained experimentally from spectra in Fig. 7 are lower than the values calculated by using eqn (5). This indicates that in Ce³⁺–Er³⁺ co-doped garnets the energy transfer from Ce³⁺ to Er³⁺ is not the only mechanism involved in the reduction of Ce³⁺ emission. A similar result was obtained in garnets codoped with Ce³⁺ and Tb³⁺ ions.¹⁹ But in this case we obtain a relatively high efficiency in the transfer from Ce³⁺ ions to acceptor ions. For the optimum composition of 1 mol% of Er³⁺, following Fig. 6, about half of the non-radiative decays of Ce³⁺ ions take to the NIR emission of Er³⁺ ions (from the comparison of experimental and calculated ratios I_Er/I_Ce in Table 3).

4. Conclusions

Ca₃Sc₂Si₃O₁₂ garnets doped with Ce³⁺ and Er³⁺ ions have been successfully synthesized by a freeze-drying precursor method. After heat treatment at 1200 °C, Ca₃Sc₂Si₃O₁₂ nanocrystals, with a mean size of about 100 nm have been obtained. From XPS measurements we found that a large fraction of the cerium ions are present as Ce³⁺ ions in these garnets. Moreover, luminescence features indicated that Ce³⁺ and Er³⁺ dopant ions are effectively incorporated into the garnets. In addition, codoping with Ce³⁺–Er³⁺ ions results in energy transfer processes from Ce³⁺ to Er³⁺ ions, allowing for efficient obtainment of NIR emission at 1.5 μm under blue excitation. The efficiency of this non-radiative energy transfer process has been analyzed by means of excited state dynamics as a function of the doping concentration. From these results we found that NIR luminescence of the garnets is maximal for CSSO–C1E. For higher Er³⁺ concentrations, the energy transfer efficiency increases, up to 82% for 4 mol% of Er³⁺, but also an appreciable increase of quenching effects is observed for Er³⁺ concentrations over 1 mol%.

Acknowledgements

The authors would like to thank Ministerio de Economía y Competitividad (MAT2013-42407-R cofunded by FEDER funds) and Fundación CajaCanarias (Proyecto HIDROSOLAR, ENER02) for financial support. J. J. Velázquez also thanks Ministerio de Economía y Competitividad for Posdoctoral grant (FPDI-2013-16895). Fernando Rivera-López thanks Cajasiete for financial support related with Servicio de Sistema Multitécnicas de Análisis de Superficies (SEGAI-ULL). We also thank the group MATMOL, from University of La Laguna (ULL), for the possibility of using the UV-Visible absorbance spectrophotometer. This research was partially performed in the framework of the projects CNR-PAS (2014–2016) and COST ACTION MP1401. R. R. Gonçalves and M. Ferrari acknowledge Brazilian Scientific Mobility Program “Ciências sem Fronteiras”.

References

1. B. M. van der Ende, L. Aarts and A. Meijerink, Phys. Chem. Chem. Phys., 2009, 11, 11081,  10.1039/b913877c.
2. J. Zhou, Y. Teng, X. Liu, S. Ye, Z. Ma and J. Qiu, Phys. Chem. Chem. Phys., 2010, 12, 13759,  10.1039/c0cp00204f.
3. F. Piccinelli, A. Speghini, G. Mariotto, L. Bovo and M. Bettinelli, J. Rare Earths, 2009, 27, 555,  DOI:10.1016/s1002-0721(08)60287-2.
4. D. Q. Chen, Y. S. Wang, Y. L. Yu, P. Huamg and F. Y. Weng, J. Appl. Phys., 2008, 104, 116105,  DOI:10.1063/1.3040005.
5. P. Dorenbos, J. Lumin., 2000, 91, 106,  DOI:10.1016/s0022-2313(00)00197-6.
6. J. Ueda and S. Tanabe, J. Appl. Phys., 2009, 106, 043101,  DOI:10.1063/1.3194310.
7. X. Chen, J. Wu, X. Xu, Y. Zhang, N. Sawanobori, C. Zhang, Q. Pan and G. J. Salamo, Opt. Lett., 2009, 34, 887,  DOI:10.1364/ol.34.000887.
8. Y. Suzuki, M. Kakihana, Y. Shimomura and N. Kijima, J. Jpn. Soc. Powder Powder Metall., 2007, 54, 44,  DOI:10.2497/jjspm.54.44.
9. Y. Shimomura, T. Honma, M. Shigeiwa, T. Akai, K. Okamoto and N. Kijima, J. Electrochem. Soc., 2007, 154, 35,  DOI:10.1149/1.2388856.
10. K. V. Ivanovskikh, A. Meijerink, F. Piccinelli, A. Speghini, E. I. Zinin, C. Ronda and M. Bettinelli, J. Lumin., 2010, 130, 893,  DOI:10.1016/j.jlumin.2009.12.031.
11. M. Bettinelli, A. Speghini, F. Piccinelli, A. N. C. Neto and O. L. Malta, J. Lumin., 2011, 131, 1026,  DOI:10.1016/j.jlumin.2011.01.016.
12. C. Feng, Y. Meng, X. Liu, J. Wang, L. Yang and X. Jinshu, Chin. J. Rare Met., 2013, 87 DOI:10.3969/j.issn.0258-7076.2013.01.016.
13. Y. Chen, J. Feng, Y. Pan, J. Li, S. Zeng, M. Liang, Z. Liu, N. Li and Y. Su, J. Lumin., 2014, 148, 153,  DOI:10.1016/j.jlumin.2013.12.015.
14. J. Qiao, J. Zhang, X. Zhang, Z. Hao, Y. Liu and Y. Luo, Dalton Trans., 2014, 43, 4164,  10.1039/c3dt52902a.
15. W. Wan, Q. Xiao, C. Yang and K. Meng, J. Chin. Ceram. Soc., 2010, 38(10), 1862–1866.
16. M. O. Mazan, J. Marrero-Jerez, A. Soldati, P. Núñez and S. Larrondo, Int. J. Hydrogen Energy, 2015, 40, 3981,  DOI:10.1016/j.ijhydene.2015.01.006.
17. M. Amsif, D. Marrero-López, A. Magrasó, J. Peña-Martínez, J. C. Ruiz-Morales and P. Núñez, J. Eur. Ceram. Soc., 2009, 29, 131,  DOI:10.1016/j.jeurceramsoc.2008.06.001.
18. J. Marrero-Jerez, S. Larrondo, E. Rodríguez-Castellón and P. Núñez, Ceram. Int., 2014, 40, 6807,  DOI:10.1016/j.ceramint.2013.11.143.
19. J. J. Velázquez, R. Fernández-González, J. Marrero-Jerez, V. D. Rodríguez, A. Lukowiak, A. Chiappini, A. Chiasera, M. Ferrari and P. Núñez, Opt. Mater., 2015, 46, 109,  DOI:10.1016/j.optmat.2015.03.057.
20. B. D. Cullity, Elements of X-ray Diffraction, Addison-Wesley, Massachusetts, 1978.
21. W. Yun-Fang, C. Ya-Han, N. Yung-Tang and C. In-Gann, J. Am. Ceram. Soc., 2013, 96, 234,  DOI:10.1111/jace.12034.
22. Y. Liu, X. Zhang, Z. Hao, Y. Luo, X. Wang and J. Zhang, J. Mater. Chem., 2011, 21, 16379.
23. M. S. Kishore, N. P. Kumar, R. G. Chandran and A. A. Setlur, Electrochem. Solid-State Lett., 2010, 13(6), J77–J80.
24. R. G. Burns, Ionic radii of transition metals and related cations, in Mineralogical Applications of Crystal Field Theory. Cambridge Topics in Mineral Physics and Chemistry. (No 5), Cambridge University Press, Cambridge, 2nd edn, 1993, pp. 464–465,  DOI:10.1017/cbo9780511524899.016.
25. V. Zepf, Rare Earth Elements. A New Approach to the Nexus of Supply, Demand and Use: Exemplified along the Use of Neodymium in Permanent Magnets, Springer Theses, Augsburg University, 2013, ISBN 978-3-642-35458-8.
26. Y. Liu, J. Hao, W. Zhuang and Y. Hu, J. Phys. D: Appl. Phys., 2009, 42, 245102,  DOI:10.1088/0022-3727/42/24/245102.
27. N. Enomoto, T. Sakai, M. Inada, Y. Tanaka and J. Hojo, J. Ceram. Soc. Jpn., 2010, 118, 1067,  DOI:10.2109/jcersj2.118.1067.
28. E. Bêche, P. Charvin, D. Perarnau, S. Abanades and G. Flamant, Surf. Interface Anal., 2008, 40, 264,  DOI:10.1002/sia.2686.
29. S. Gavarini, M. J. Guittet, P. Trocellier, M. Gauthier-Soyer, F. Carrot and G. Matzen, J. Nucl. Mater., 2003, 322, 111,  DOI:10.1016/s0022-3115(03)00306-4.
30. J. Zhang, X. Ju, Z. Y. Wu, T. Liu, T. D. Hu, Y. N. Xie and Z. L. Zhang, Chem. Mater., 2001, 13, 4192,  DOI:10.1021/cm010235p.
31. N. D. Todorov, M. V. Abrashev, V. Marinova, M. Kadiyski, L. Dimowa and E. Faulques, Phys. Rev. B: Condens. Matter Mater. Phys., 2013, 87, 104301.
32. A. Ubaldini and M. M. Carnasciali, J. Alloys Compd., 2008, 454, 374.
33. Y. Cheng, K. Cheah and M. Gong, J. Lumin., 2011, 131, 1589.
34. A. M. Hofmeister and A. Chopelas, Phys. Chem. Miner., 1991, 17, 503.
35. S. P. Feofilov, D. V. Arsentyev, A. B. Kulinkin and R. I. Zakharchenya, J. Lumin., 2011, 131, 438.
36. M. Inokuti and F. Hirayama, J. Chem. Phys., 1965, 43, 1978,  DOI:10.1063/1.1697063.
37. G. Zhang, C. Liu, J. Wang, X. Kuang and Q. Su, J. Mater. Chem., 2012, 22, 2226,  10.1039/c1jm14942c.

---