Non-Rare-Earth Na3AlF6:Cr3+ Phosphors for Far-Red Light-Emitting Diodes.

Emerging phototherapy in a clinic and plant photomorphogenesis call for efficient red/far-red light resources to target and/or actuate the interaction of light and living organisms. Rare-earth-doped phosphors are generally promising candidates for efficient light-emitting diodes but still bear lower quantum yield for the far-red components, potential supply risks, and high-cost issues. Thus, the design and preparation of efficient non-rare-earth activated phosphors becomes extremely important and arouses great interest. Fabrication of Cr3+-doped Na3AlF6 phosphors significantly promotes the potential applications by efficiently converting blue excitation light of a commercial InGaN chip to far-red broadband emission in the 640-850 nm region. The action response of phototherapy (∼667-683 nm; ∼750-772 nm) and that of photomorphogenesis (∼700-760 nm) are well overlapped. Based on the temperature-dependent steady luminescence and time-resolved spectroscopies, energy transfer models are rationally established by means of the configurational coordinate diagram of Cr3+ ions. An optimal sample of Na3AlF6:60% Cr3+ phosphor generates a notable QY of 75 ± 5%. Additionally, an InGaN LED device encapsulated by using Na3AlF6:60% Cr3+ phosphor was fabricated. The current exploration will pave a promising way to engineer non-rare-earth activated optoelectronic devices for all kinds of photobiological applications.

Introduction

Plant cultivation in horticulture, as well as phototherapy to treat physical illness, has been significantly stimulated by the emergence of efficient red and far-red light-emitting diodes (LEDs).[1,2] For plant growth, photosynthesis requires red light radiation in the 600–690 nm range to overlap well with the respective absorption peaks of chlorophylls a and b at 662 and 642 nm.[2,3] Particularly, plant morphogenesis encompasses the light-mediated seed germination, seedling development, and photoperiodism, which is critical to healthy plant development. On the other hand, for phototherapy, there exist several active response bands located in the red and far-red light regions: (i) 613–623 nm, (ii) 667–683 nm, and (iii) 750–772 nm.[4,5] A growing body of evidence indicates photoacceptors in cells, like cytochrome c oxidase and the terminal enzyme of mitochondrial electron transport chain, exhibit response to low-intensity red light and even near-infrared light.[6] Exposure to 670 or 726 nm LEDs red light was reported to decrease the healing time in chronic ischemic ulcers in rats.[7,8] Therefore, achievement of novel red/far-red sources would significantly promote the development of plant cultivation and phototherapy.

Phosphor-converted light-emitting diodes (pc-LEDs) have been emerging as indispensable solid state light sources because of their unique merits such as energy savings, environment-friendliness, small footprint, long operational lifetime, and so on.[9] However, the deficiency of far-red light components limits LEDs applications in natural light sources, plant cultivation, phototherapy, and so on.[10] Typically, rare-earth (RE) ions like Eu3+, Eu2+, Pr3+, and Sm3+ are doped as activators into numerous hosts for red/far-red emissions under ultraviolet (UV) or blue excitation,[11] but the wide applications are limited by the following features: (i) Eu3+, Pr3+, and Sm3+ ions feature intra-4f parity-forbidden electronic transitions which have relatively weak absorption and therefore a low external quantum yield (QY);[12−14] (ii) Eu2+ has an allowed 4f7 → 4f65d1 transition with about 1000 times larger absorption than the intra-4f parity-forbidden transitions, but nearly all Eu2+-doped red/far-red phosphors exhibit a relatively low external QY ≤ 70%.[11,15,16] Besides, RE elements have additional issues due to potential supply risk and high cost, and hence non-RE activated phosphors become extremely important and currently arouse great interest.[17−19] Moreover, the recent commercial success of potassium fluorosilicate-based phosphors (K2SiF6:Mn4+) in LED applications has spurred new interest in the properties of non-oxide materials.[20,21] The manganese(IV) system has shown that peak emission wavelengths can be broadly modified when moving from an oxide system to a fluoride system.[19]

The Cr3+ ion has been well studied since the first generation of working lasers, i.e., the ruby (α-Al2O3:Cr3+) laser’s coherent far-red light at 694.3 nm.[22] Absorption bands of Cr3+ originate from interconfiguration spin-allowed transitions of 4A2 → 4T2 (4F), 4A2 → 4T1 (4F), and 4A2 → 4T1 (4P). Their oscillator strengths are typically on the order of 10–4,[23] which is at least hundreds of times more than that of the representative intra-4f transitions of RE ions, such as Eu3+: 7F0 → 5D0 about 1.3 × 10–8,[24] Pr3+: 3H4 → 1D2 about (0.3–1.5) × 10–6,[14] and Sm3+: 6H5/2 → 6P3/2,4F7/2 about 9.3 × 10–6.[25] Therefore, Cr3+-activated materials feature a significantly larger absorption cross-section, especially the 4A2 → 4T1(4F) blue absorption on the order of 10–18 cm2,[26] about 1000-fold greater than that of the RE ions on the order of 10–21 cm2.[27] Additionally, in the case of low crystal field surrounding Cr3+, the site emitting will occur from the 4T2 (4F) level to the 4A2 ground level, which is a spin-allowed transition.[28] Manipulation of the environment surrounding Cr3+ can tailor the emission peak of 4T2(4F) → 4A2 in the 650–1600 nm range.[29] Also, the 4T2(4F) → 4A2 transition of Cr3+ is typically about 2 orders of magnitude stronger than the 2E → 4A2 parity-forbidden transition of Cr3+ occurring in high crystal field.[30] Accordingly, these optical advantages of Cr3+ significantly enable the fabrication of efficient, advanced red/far-red emitting phosphors for blue-light-excited LEDs for photobiology.

In the current work, we synthesized a series of polycrystalline Na3AlF6:x% Cr3+ (x = 0–100) phosphors by means of a facile hydrothermal reaction. A broad 4T2(4F) → 4A2 emission band spanning 640–850 nm can be efficiently generated under excitation of UV/blue/orange light. Moreover, emission peak of 4T2(4F) → 4A2 was obtained with a continuous red-shifting from 710 to 745 nm as Cr3+ concentration increases to x ∼ 100. It is of great interest that the intensity of 4T2(4F) → 4A2 monotonically increase until higher Cr3+ concentration x ∼ 60 with a maximum QY of 75 ± 5%. Energy transfer dynamics of Cr3+ ions and the capabilities of photobiological application were systematically analyzed by means of time-resolved spectra and by an encapsulated blue light InGaN LED device, respectively.

Experimental Section

Hydrothermal Synthesis of Na3AlF6:x% Cr3+ Particles

Utilizing NH4F (≥98%), NaF (≥98%)], Al(NO3)3·9H2O (≥98%), and Cr(NO3)3·9H2O (99%) as raw materials (all purchased from Sigma-Aldrich), we synthesized a series of Na3AlF6:x% Cr3+ (x = 0.5, 1, 5, 10, 20, 30, 40, 50, 60, 70, 80, 90, 100) phosphors by a hydrothermal reaction. In all cases, the percentages given for Cr3+ dopant is in mol % with respect to the Al3+ site. First, the specified stoichiometric Al(NO3)3·9H2O and Cr(NO3)3·9H2O were weighed and then dissolved by 30 mL of deionized (DI) water (18.2 MΩ) under magnetic stirring in a glass beaker. Second, excess NaF and NH4F with >9 times molar ratio to metal ions (Al3+ and/or Cr3+) were measured and added into another 30 mL of DI water for dissolution. Third, the NaF and NH4F solution was transferred into the above Al(NO3)3 and Cr(NO3)3 solution for reaction, and then 10 mL of absolute ethanol was added additionally. At last, after vigorous stirring for about 30 min, the resultant white suspension was transferred into a 125 mL Teflon-lined stainless-steel autoclave (Parr Instrument Company), filled up to 70% of its capacity, sealed tightly, and heated at 240 °C in an electrical furnace for 15 h. After being slowly cooled to room temperature, the filtered precipitate was centrifuged several times with DI water to remove any possible ionic remnant and finally dried at 80 °C for 24 h in an ambient atmosphere.

Post-Heat Treatment of Na3AlF6:60% Cr3+ Powder

The as-obtained Na3AlF6: 60% Cr3+ phosphor was pressed into a quarter inch diameter pellet by using a hydraulic press and then placed in a quartz ampule which was sealed under vacuum for post-heat treatment. The ampule was inserted into a furnace and heated from room temperature to 700 °C for 2.5 h, dwelled at 700 °C for 36 h, and cooled to room temperature over 2.5 h. After heating, the ampule was opened in the air and the pellet was removed. The pellet was ground in an agate mortar and pestle for several minutes to produce a powder. The powder was stored with a desiccant after grinding.

Characterization

Phase identification of all the as-obtained phosphor samples were performed on a Siemens D500 X-ray powder diffractometer (XRD, Bruker AXS Inc., Madison, WI) using Cu Kα (λ = 1.5406 Å) radiation at 40 kV and 40 mA. XRD patterns were collected with a resolution of 0.018°/step and 2 s/step in a 2θ range of 10°–60°. Scanning electron microscopy (SEM) images of the respective powder samples were taken by means of the Carl Zeiss Sigma field emission SEM (Carl Zeiss, Carl Zeiss SMT Inc., Peabody, MA) using the secondary electron detector and operating at an accelerating voltage of 5.0 kV with a working distance of 8.8 mm. Energy-dispersive X-ray (EDX) spectroscopy area scans of the Na3AlF6: x% Cr3+ powder samples were further performed to determine the elemental composition using an accelerated voltage of 15 kV and a reduced working distance of 8.5 mm for an aperture of 60 μm.

Steady photoluminescence and excitation spectra were determined via single photon counting technique on a FSP920 spectrometer (Edinburgh Instruments, Livingston, UK) equipped with a 450 W xenon (Xe) lamp, TMS300 monochromators, and thermo-electronic cooled Hamamatsu R928P photomultiplier tube (PMT) detector. Time-resolved spectra were measured by time-correlated single photon counting technique on the FSP920 system with pulsed excitation sources of a microsecond μF900 Xe lamp. To eliminate signal noise from excitation sources, long-pass filters of 400, 490, 695, and 710 nm were appropriately applied in the front of detectors. Temperature-dependent fluorescence spectra of Na3AlF6: 60% Cr3+ were measured by using a Fluorolog-3 spectrophotometer (Horiba Jobin Yvon) equipped with a FL-1073 PMT detector and a SPEX DM302 P.C. Acq. module. The phosphors were heated on an external hot plate, and each temperature point was allowed to equilibrate as monitored by a thermocouple in a sand bath adjacent to the phosphors. Once the temperature was equilibrated, the emission spectrum of the hot sample was measured (λex = 450 nm) and compared to a room temperature sample of the phosphor to eliminate the impact of lamp intensity fluctuations. For comparison, all optical measurements were rigorously performed under identical conditions for each series of testing.

Fluorescence QY of phosphors was measured on a C9220-03 system (Hamamatsu, Bridgewater, NJ) with a 150 W Xe monochromatic lamp and an integrating sphere. A high-intensity 365 nm UV lamp (B-100A, UVP, LLC, CA) was used to shine phosphors for optical photographs. LEDs devices were fabricated by applying a 25 wt % phosphor slurry composed of the phosphor powder and Dow Corning OE-6550 silicone on top of a blue-emitting InGaN based LED mounted in a 2835 PLCC package (Power Opto Co., Taiwan), followed by thermal curing of the LEDs package on a hot plate at 125 °C.

Results and Discussion

As shown in Figure a and Figure S1, all the XRD profiles of as-obtained Na3AlF6:x% Cr3+ (x = 0.5, 1, 5, 10, 20, 40, 60, 80, 100) powders can be indexed as peaks corresponding to pure cryolites of Na3AlF6 with PDF card no. 01-082-0218 (also ICSD card no. 74202) and Na3CrF6 (PDF card no. 27-0675). When the Cr3+ concentration exceeds 60–70%, a secondary phase emerges with observed XRD patterns around 17°, 20.8°, and 29.7° (see the frame labels in Figure a and Figure S1). By completely going through XRD database, the impurities can be well indexed into the phase of (NH4)2NaAlF6 (PDF #97-024-9157; Figures S1 and S2a). Crystalline Na3AlF6 and Na3CrF6 both have structures with space group of P21/n and space group number of 14 (monoclinic Na3AlF6 for ICSD #74202: a = 5.454 Å, b = 5.616 Å, c = 7.822 Å, α = γ = 90°, and β = 90.118°; orthorhombic Na3CrF6: a = 5.460 Å, b = 5.680 Å, c = 7.880 Å, and α = β = γ = 90°). Powder XRD pattern Rietveld refinement for Na3AlF6:5% Cr3+ and Na3AlF6:80% Cr3+ was further performed to gain more structural information (Figures b and 1c). The main parameters of processing and refinement of the Na3AlF6:x% Cr3+ (x = 5, 80) samples are listed in Table S1. The refinement results in Figure b reveal that the as-obtained Na3AlF6:5% Cr3+ sample is all pure monoclinic Na3AlF6 (pink short vertical line in Figure b), and the weighted R-factor (Rwp) and profile R-factor (Rp) are determined to be 3.76% and 2.74%, respectively, indicating that the refined results are reliable. Whereas for Na3AlF6:80% Cr3+, the Rietveld refinements from XRD patterns of Na3CrF6 phase (pink short vertical line in Figure c) and (NH4)2NaAlF6 phase (neon blue vertical lines in Figure c) further confirm that the concentrated Cr3+ doping, such as more than 60–70%, does introduce a secondary phase of byproducts, (NH4)2NaAlF6 and/or (NH4)2NaCrF6. Through calculation, there generally exists more than 3% impurities in the Na3AlF6:80% Cr3+ phosphors, which will not influence the following studies on the optical properties of Na3AlF6:Cr3+ phosphors (see fluorescence comparison in Figure S2b and Figure ).

Figure 1

(a) XRD patterns of Na3AlF6:x% Cr3+ (x = 1, 5, 20, 60, 100) representatives and that of Na3AlF6 (PDF #01-082-0218) and Na3CrF6 (PDF #27-0675) standards. The green frames label the coexisting byproducts, (NH4)2NaCrF6, as the Cr3+ doping concentration exceeds 60–70% (see detailed discussion in Figures S1 and S2). (b, c) XRD Rietveld refinement results for the samples of 5% Cr3+ (b) and 80% Cr3+ (c) doped Na3AlF6 showing raw data (black crosses) and calculated (red solid line) XRD profiles, difference between the raw data and calculated patterns (blue solid line), and Bragg reflection positions (pink and neon blue short vertical lines). (d) SEM micrographs and (e) EDX spectra analysis of the as-prepared Na3AlF6:60% Cr3+ sample.

Figure 2

(a) Excitation (λem: 720 nm) and emission (λex: 420 and 580 nm) spectra of Na3AlF6:1% Cr3+ phosphors. The pink star labels the sharp interference light peaks typically from Xe lamp around 467 nm. (b) Time-resolved emission spectra of Na3AlF6:10% Cr3+ under pulsed light excitation of 420 nm. (c) Emission spectra and (d) decay curves of Na3AlF6:x% Cr3+ (x = 1, 5, 10, 20, 40, 60, 80, 100) phosphors. The inset of (c) shows the normalized emission peaks of Na3AlF6:x% Cr3+ phosphors.

Given that the effective ionic radius of Al3+ [r = 0.535 Å, coordination number (CN) = 6] is considerably close to that of Cr3+ (r = 0.615 Å, CN = 6)[31] and that no charge compensation needed during substitution, Na3AlF6 is a perfect host to continuously vary Cr3+ concentration from a regime where it acts as a dopant to a regime where it likely serves as a constituent. As a direct proof of the solid solution, increasing substitution of Al3+ (smaller radius) by Cr3+ (larger radius) from 1% to 100%, the XRD peaks like (002) profile of the as-prepared phosphors smoothly shift toward smaller angle relative to the peaks of pure Na3AlF6 and finally match well with that of pure Na3CrF6 (Figure a and Figure S1). This phenomenon verifies the incorporation of Cr into the lattice. Furthermore, the lattice parameters (a, b, c, and V) refined from the measured XRD patterns feature increasing behaviors with increasing x (Figure S3), which, well consistent with Vegard’s rule,[32] demonstrate the Na3AlF6:x% Cr3+ (x = 0.5–100) solid solutions are formed by us. Within the monoclinic ‘‘mixed-cation fluoride perovskite’’ structure of Na3AlF6 cryolite crystals, the corner-sharing octahedral network comprises alternating [AlF6]3– and [NaF6]5– octahedra with Na+ ions in interstitial sites (Figure S4).[33] The low symmetry of the [AlF6]3– octahedral site in this structure that would be further disordered by Cr3+ doping (Figure S4) is expected to benefit the spin-allowed transitions of Cr3+, especially the 4T2(4F) → 4A2 red/far-red emission in a low crystal field.[34]

From the SEM micrograph of Na3AlF6: 60% Cr3+ representative in Figure d, it can be seen that the prepared phosphors are well crystallized into polyhedron and have size distribution from 250 nm to 1.25 μm. To explore elemental composition as well as its distribution of Cr3+-doped Na3AlF6 phosphors, EDX mapping analysis was done for the Na3AlF6:60% Cr3+ sample (Figure e and Figure S5). The EDX elemental composition was determined by comparing relative peak intensities together with the corresponding sensitivity factors of each element and assuming their total intensities to be 100%. All the elements of Na+, Al3+, F–, and Cr3+ were clearly detected (Figure e) and homogeneously mapped in the as-obtained phosphors (Figure S5). EDX results reveal the mole percent of Cr3+ and that of Al3+ are further calculated to be about 66% and 34%, respectively, which are very close to the stoichiometric compositions of 60% Cr3+ and 40% Al3+ in the prepared sample.

To investigate the photoluminescence properties of Na3AlF6:Cr3+ phosphors, excitation and emission spectra were recorded for the Na3AlF6:1% Cr3+ representative, shown in Figure a. By monitoring the emission wavelength at 720 nm, we detected two excitation peaks (420 and 620 nm) due to the d–d inner transitions of Cr3+ from the 4A2 ground state to the 4T1(4F) and T2(4F) intermediate states,[35−37] respectively. By increasing Cr3+ concentration, the excitation peaks were observed to performs noticeable red-shift after normalizing all the excitation spectra at wavelength ∼420 nm (Figure S6a). Under excitation of 282, 420, and 580 nm, the Na3AlF6:1% Cr3+ phosphors yield one single broad peak at about 720 nm (Figure a and Figure S6b) due to the spin-allowed transition of 4T2(4F) → 4A2 in the case of Cr3+ located in the low crystal field.[28] The color rendering index of Na3AlF6:x% Cr3+ phosphors was calculated to be around (0.723, 0.276) and displayed by CIE chromaticity coordinates in Figure S7. Taking into account the crystal field strength and site symmetry of Na3AlF6 host, the underlying mechanisms of photoexcitation and energy transfer of Cr3+ activators were discussed and illustrated by a simplified Tanabe–Sugano diagram of 3d3 electron configuration in Figure S8. It is noted that the broadband emission of Na3AlF6:Cr3+ generally spans from 640 to 850 nm, has a full width at half-maximum of about 95 nm, and overlaps well with the “active” peaks at about 670, 760, and 825 nm of cells and DNA/RNA synthesis for phototherapy.[5,6]

To explore whether or not any different Cr3+ luminescent centers exist in Na3AlF6:x% Cr3+, time-resolved emission spectra with a fine time interval of 0.5 μs were recorded for the sample of Na3AlF6:10% Cr3+ under pulsed light excitation of 420 nm (Figure b). At initial delay time around 106 μs, no emission band appears, while a single broad band centered at about 720 nm emerges clearly at delay time around 107 μs due to the energy depopulation of excited Cr3+ ions. By prolonging delay time, the intensity of the whole Cr3+ characteristic band monotonically increases to a maximum at about 120 μs and then decreases. Notably, the band shapes and peak position of all the emissions remain unchanged at different delay time points. These results suggest that all the Cr3+ activators do have similar surrounding environments.

To optimize the Na3AlF6:x% Cr3+, a series of emission spectra were recorded under excitation of 420 nm, as shown in Figure c. The emission intensity of 4T2(4F) → 4A2 increases to a maximum until an elevated concentration of 60% Cr3+ and then decreases obviously because of the concentration quenching effect. Such an effect mainly results from the energy consumption via energy migration over Cr3+ sublattice sites and finally to the quenching sites, like impurities/defects particularly as the higher Cr3+ concentration in Na3AlF6 makes the Cr3+-to-Cr3+ interatomic distance much smaller.[38,39] In comparison with the emission intensity of Na3AlF6:1% Cr3+, that of Na3AlF6:60% Cr3+ exhibits about 5-fold enhancement. The QY value was measured to be 75 ± 5% under excitation of 420 nm, which is comparable to the QY value ∼68% reported by Torchia et al.[40] It is of great interest that the normalized emission peaks in the inset of Figure c show the distinct red-shift from 715 to 745 nm, which would benefit the color component manipulation according to the varied specific applications like in phototherapy and/or photomorphogenesis. The red-shift results from the perturbation effects of crystal field strength on the 4T2(4F) spin-allowed state of Cr3+ with increasing Cr3+ concentration in Na3AlF6. Generally, the crystal field strength (D) can be formulized as[41,42]where Z is the anion charge or valence, e is the electron charge, r is the radius of the d wave function, and R is the bond length. So the dependence of D on bond length can be generalized as D ∝ 1/R5 in a specific host. Doping Cr3+ ions (radius ∼0.615 Å) into Na3AlF6 by substituting Al3+ sites (radius ∼0.535 Å) will lead to a longer R of the Cr–F bond and therefore a weaker D crystal strength. Correspondingly, a smaller D leads to emission at lower energy, namely, the resultant red-shift at higher Cr3+ concentration. Experimentally, the obvious red-shift of absorption peaks of Na3AlF6:Cr3+ phosphors (Figure S6) well demonstrate such an effect of weaker Dq on lowering the energy of 4T2(4F) spin-allowed state of Cr3+. Besides, by greatly increasing Cr3+ concentration, the severe reabsorption occuring in the 650–710 nm range (see the overlapping range between absorption and emission in Figure a) would be another promotion to this red-shift for the photoluminescence of Na3AlF6:Cr3+ phosphors.

To give insight into energy-transfer process of Na3AlF6:x% Cr3+ (x = 1, 5, 10, 20, 40, 60, 80, 100) phosphors, the decay curve of 4T2(4F) → 4A2 was recorded versus Cr3+ concentration upon 420 nm pulsed light excitation. As shown in Figure d, the decay curve of dilute 1% Cr3+ doping behaves single exponential (similar behaviors in Figure S9 for the 0.5% Cr3+ and 2% Cr3+ doping), and that of concentrated x% Cr3+ doping (x ∼ 5–100) has a faster decline and becomes obviously nonexponential. These phenomena are typically caused by the introduction of additional energy decay paths as Cr3+-doping concentration increases, like cross-relaxation, energy migration, and the impurity/defect-induced nonradiative relaxation (NR) process, especially for the concentrated dopants.[9] In practice, the monoexponential decay curve can be well fitted to a first-order exponential function of I = A0 exp(−t/τ0), and the nonexponential one is well fitted to a second-order exponential function of I = A1 exp(−t/τ1) + A2 exp(−t/τ2), where I is the luminescence intensity; A0, A1, and A2 are constants of fitting parameters, respectively; t is time; τ0 is the lifetime for the monoexponential decay curve; and τ1 and τ1 are fast and slow lifetime of exponential components for the nonexponential decay curve, respectively. Using the formula τ = (A1τ12 + A2τ22)/(A1τ1 + A2τ2), we can obtain the average decay time (τ) of the second-order exponential decay for the Na3AlF6:x% Cr3+ samples. The trise, fitting function, R-squared (R2), fitting parameters, and τ are summarized as a function of Cr3+ concentration in Table .

Table 1

Fitting Functions, R-Squared (R2), Fitting Parameters, and the Calculated Decay Time of Decay Curve of 4T2(4F) → 4A2 Luminescence as a Function of Cr3+ Concentration in Na3AlF6 under Pulsed Light Excitation of 420 nm

Cr3+ (%)	fitting function	R2	component lifetime (μs)	fitting parameter	decay time (μs)
1	I = A0 exp(−t/τ0)	0.9992		A0 ∼ 1.4826	312.9
5	I = A1 exp(−t/τ1) + A2 exp(−t/τ2)	0.9993	τ1 = 116.7	A1 ∼ 0.8176	283.3
			τ2 = 329.6	A2 ∼ 1.0421
10	I = A1 exp(−t/τ1) + A2 exp(−t/τ2)	0.9993	τ1 = 136.1	A1 ∼ 1.1364	257.5
			τ2 = 333.1	A2 ∼ 0.7446
20	I = A1 exp(−t/τ1) + A2 exp(−t/τ2)	0.9992	τ1 = 127.2	A1 ∼ 1.4922	240.6
			τ2 = 349.2	A2 ∼ 0.5678
40	I = A1 exp(−t/τ1) + A2 exp(−t/τ2)	0.9992	τ1 = 79.5	A1 ∼ 2.0832	219.7
			τ2 = 320.8	A2 ∼ 0.7158
60	I = A1 exp(−t/τ1) + A2 exp(−t/τ2)	0.9991	τ1 = 87.7	A1 ∼ 2.1914	213.7
			τ2 = 342.8	A2 ∼ 0.5477
80	I = A1 exp(−t/τ1) + A2 exp(−t/τ2)	0.9992	τ1 = 65.7	A1 ∼ 2.7234	194.4
			τ2 = 304.7	A2 ∼ 0.6856
100	I = A1 exp(−t/τ1) + A2 exp(−t/τ2)	0.9985	τ1 = 30.1	A1 ∼ 19.6474	85.2
			τ2 = 267.9	A2 ∼ 0.6666

To further optimize the emission properties of Na3AlF6:x% Cr3+ phosphors, the sample of Na3AlF6:60% Cr3+ was heat-treated at 700 °C for 36 h in a quartz ampule under vacuum. Figure a shows that the particle size of heat-treated Na3AlF6:60% Cr3+ clearly becomes larger (up to 2–3 μm) than that before. The EDX result in Figure b validates the detection of all elements of Na+, Al3+, Cr3+, and F–. As expected, all elements were homogeneously distributed over the heat-treated Na3AlF6:60% Cr3+ phosphors (Figure S10). Transformation of wt % (at. %) Cr3+ and wt % (at. %) Al3+ recorded by EDX element mapping (inset of Figure b) to mole percent in cryolite formula results in the respective values of 57.9% and 42.1%, which are close to the stoichiometric composition of Na3AlF6:60% Cr3+. These results indicate that the heat treatment at a higher temperature ∼700 °C does not change the composition of phosphors. Figure c comparatively shows the emission spectra of Na3AlF6:60% Cr3+ phosphors before and after heat treatment. These two spectra have same band shape in the same region, but the heat treatment process increases emission intensity by about 11%. Also, the luminescence decay of heat-treated Na3AlF6:60% Cr3+ (average τ ∼ 263.7 μs) becomes obviously slower than that of sample without heat treatment (Table and Figure d). This phenomenon results from the decreased amount of additional energy quenching centers, such as typical lattice/surface defects and impurities introduced by higher 60% Cr3+ dopants. It is believed that a continuous increase of particle size will greatly increase luminescence intensity and efficiency of our synthesized Na3AlF6:Cr3+ phosphors.

Figure 3

(a) SEM micrograph and (b) EDX spectra analysis of the Na3AlF6:60% Cr3+ phosphors heat-treated at 700 °C for 36 h. (c) Comparative emission spectra and (d) decay curves of the Na3AlF6:60% Cr3+ before (green dotted curve) and after (red dotted curve) heat treatment.

For lighting phosphors, the working temperature would be much higher than room temperature. Although pc-LEDs are significantly more energy efficient light source than incandescent lighting, about 50–70% of the electricity supplied to an LED still becomes heat rather than light, and in high power packages in a more confined space, this can lead to an operating temperature of 150–200 °C during LEDs operation, which efficiently quenches photoluminescence via fast NR of excited electrons to the ground state of activators.[43,44] Thus, study of thermal stability on the far-red emission of Na3AlF6:Cr3+ phosphor is extremely important. Figure a shows the decay curves of the 4T2(4F) → 4A2 for Na3AlF6:60% Cr3+ sample, and obviously the decay becomes much faster as temperature increases. By calculation, a shortest average lifetime was obtained to be 126.7 μs at 470 K (∼200 °C) (blue sphere symbols in Figure b), which suggests that the nonradiative decay of the Cr3+:4T2(4F) excited state is thermally promoted at elevated temperature. Besides, Figure b exhibited the integrated emission intensity of 4T2(4F) → 4A2 as a function of temperature for Na3AlF6:60% Cr3+ under excitation of 420 nm (red star symbols in Figure b). The far-red emission of Cr3+ is clearly quenched with increasing temperature and decreases to about 60% at 470 K. The reason could be that a much higher 60% Cr3+ concentration in Na3AlF6 would feature severe energy migration among Cr3+ sublattice sites,[38,39] and energy is quenched on defects/impurities.

Figure 4

(a) Decay curves of 4T2(4F) → 4A2 of Na3AlF6:60% Cr3+ phosphors against temperature. (b) Thermal characteristics of 4T2(4F) → 4A2 of Na3AlF6:60% Cr3+: emission intensity marked by red stars and the corresponding lifetime by blue spheres. (c) Single configurational coordinate model for Cr3+ far-red emission under effect of temperature increment in Na3AlF6 low-field crystal. R denotes the configurational coordinate, the green arrow indicates the thermal activation process, and the pink curved arrow represents the NR process. (d) Optical images of Na3AlF6:60% Cr3+ phosphors shined by natural white light (left top) and UV light (left bottom) and that of a bare blue-emitting InGaN LED device (right top) and the InGaN LED device encapsulated with a silicone Na3AlF6: 60% Cr3+ phosphor slurry (right bottom).

A single configurational coordinate model of Cr3+ (simplified with only one 4T2(4F) excited state) in Figure c describes the effects of thermal quenching on the far-red emission of Cr3+ qualitatively. Once Cr3+ is excited from the 4A2 ground state to the 4T2(4F) state, energy will go through a fast NR process to populate the lowest energy point of the 4T2(4F). Basically, the energy in 4T2(4F) excited state radiatively decays to the 4A2 ground state by efficiently yielding far-red photons at room/lower temperature. However, in the case of an elevated temperature, the absorbed energy in 4T2(4F) excited state is feasibly promoted to the intersystem crossing (Q) of 4T2(4F) excited state and 4A2 ground state by thermal activation (T-activation) and then is quenched by fast nonradiative decay to the bottom of 4A2 ground state. Typically, the temperature-dependent fluorescence intensity can be expressed as[27,28]where I0 is initial fluorescence intensity recorded at ideal case with negligible thermal quenching, ΔEq is the thermal activation barrier to the level crossing Q, k is the Boltzmann constant, T is the absolute temperature, and the unit e–Δ therefore represents the probability of an excited ion being elevated over ΔEq barrier to the Q crossing. Accordingly, a higher temperature will thermally promote more excited ions to easily reach the level crossing Q, and meanwhile, the stronger NR process happens. On the assumption that the photon absorption and scattering of luminescent sample keep constant in the temperature region,[28] fluorescence intensity of the phosphors will definitely decrease with increasing temperature under the same excitation conditions. These theoretical analyses well reveal why our Na3AlF6:60% Cr3+ phosphor exhibits descending emission intensity by raising temperature from 300 to 470 K.

Figure d shows optical photographs of Na3AlF6:60% Cr3+ phosphors illuminated by natural white light (top panel) and UV lamp (bottom panel). The distinct observation of deep red color to the naked eye confirms that the Na3AlF6:60% Cr3+ radiates efficiently. The obtained deep red color nearly features the same color effects as Color Hex Color Codes of #960000.[45] Also, optical photographs of commercial blue-emitting InGaN LED packages encapsulated in silicone without and with inclusion of Na3AlF6:60% Cr3+ phosphor are shown in Figure d. The performance of four LEDs devices operated at constant current (drive voltage 2.72 V, drive current 14 mA) is shown in Figure S11. The first device is the bare blue LED; devices 1–3 are blue LEDs encapsulated with increasing amounts of Na3AlF6:60% Cr3+ phosphors, which efficiently convert a higher percentage of the blue light to far-red light.

Conclusions

Polycrystalline Cr3+-doped Na3AlF6 phosphors are synthesized by means of a facile hydrothermal reaction. Tunable broadband far-red fluorescence of Cr3+ was feasibly achieved under blue light excitation by increasing the substitution of Cr3+ ions into Al3+ sublattice sites. Na3AlF6:60% Cr3+ phosphors exhibit an appreciable QY ∼ 75% which is more efficient than some typical Eu2+-doped red/far-red phosphors, and the luminescence intensity and efficiency further increased by heat treatment. ET mechanisms, as well as energy decay dynamics of Cr3+, are rationally analyzed on the basis of photoemission, excitation, and time-resolved spectra. Non-RE activated red/far-red emitting phosphors in commercial LED further promote the design and fabrication of advanced pc-LEDs for photobiological applications.