Blue LED-excitable ultra broadband near-infrared luminescence based on KCdF3:Cr3<bold>+</bold>/Ni2+ nanocrystals embedded in fluorosilicate glass

Long Chen; Junhuan Lai; Yong Xu; Yinsheng Xu; Xueyun Liu

doi:10.3788/COL202523.081601

1. Introduction

Near-infrared (NIR) spectroscopy, covering the NIR-I (700–1000 nm) and NIR-II (1000–1700 nm) windows, features its strong penetrating ability into tissues and thus holds a wide range of applications in areas such as food safety, night vision, medical imaging, plant growth, and information identification^[1–13]. As a widely used broadband NIR light source, tungsten halogen lamps have limitations such as low energy efficiency and large size, and it is difficult to achieve portable application requirements^[14,15]. Although traditional NIR light-emitting diodes (LEDs) with compact structures present long lifetime and high efficiency, the narrowband emission (850–940 nm) limits their application in biomedical and food analysis due to the spectral overlap with organic molecules in food and biological tissues^[16,17]. In this regard, developing an NIR light source with broadband emission and high efficiency becomes urgently required. Recently, commercial high-power blue LED chips combined with broadband NIR-emitting phosphors, namely phosphor-converted NIR-LEDs, were found to be a promising approach for obtaining broadband-emitting NIR sources because of their small size, low cost, long lifetime, convenience, and high efficiency, and thus are called next-generation NIR light sources^[6,18–22].

As a crucial component of such NIR light sources, broadband NIR luminescent materials depend greatly on the matrix and doping species. Among various activators, chromium ion ( ${Cr}^{3 +}$ ) with a unique $3 d^{3}$ electronic configuration, is commonly used as a promising NIR emission center upon blue light excitation, which can achieve deep red to NIR tunable broadband emission (600–1400 nm) in a weak crystal field owing to highly sensitive optical characteristics of ${Cr}^{3 +}$ ions in the local environment^[23–27]. Abundant research on ${Cr}^{3 +}$ -activated broadband NIR materials utilized for NIR light sources has been widely investigated during the past decade^{[3,4,28–31]}. Unfortunately, the lack of effective NIR-II emission of ${Cr}^{3 +}$ -doped materials severely hinders NIR spectroscopy applications. Nickel ion ( ${Ni}^{2 +}$ ), another common NIR activator, typically produces a broadband NIR emission spanning the 1000–2000 nm range within the NIR-II region^[32,33]. However, due to poor absorption in the blue light region, its luminescent efficiency is extremely low. A donor-acceptor strategy is proposed utilizing efficient energy transfer (ET) from ${Cr}^{3 +}$ to ${Ni}^{2 +}$ to enhance the blue light excitation efficiency of ${Ni}^{2 +}$ -based NIR-II emission^[3,6,32,34]. Considering coordination selection, a six-coordinated crystal field environment must be provided for these two activators to exhibit photoactivity. Hence, blue LED-excitable ${Cr}^{3 +} / {Ni}^{2 +}$ co-doped phosphors with octahedrally coordinated structures have been developed for realizing attractive ultra-wide NIR luminescence containing both NIR-I and NIR-II regions, such as ${LaMgGa}_{11} O_{19} : {Cr}^{3 +} / {Ni}^{2 +}$ ^[6], ${Zn}_{1 + y} {Sn}_{y} {Ga}_{2 - 2 y} O_{4} : {Cr}^{3 +} / {Ni}^{2 +}$ ^[35], ${ZnGa}_{2} O_{4} : {Cr}^{3 +} / {Ni}^{2 +}$ ^[21], ${Mg}_{2} {SnO}_{4} : {Cr}^{3 +} / {Ni}^{2 +}$ ^[36], ${LiMgPO}_{4} : {Cr}^{3 +} / {Ni}^{2 +}$ ^[20], ${MgGa}_{2} {(O, F)}_{4} : {Cr}^{3 +} / {Ni}^{2 +}$ ^[34], and ${RbCdF}_{3} : {Cr}^{3 +} / {Ni}^{2 +}$ ^[37].

Compared to phosphors used for NIR-LEDs, glass ceramics achieved via controlling the crystallization of nanocrystals in an amorphous glass matrix, exhibit the advantages of combining glass and crystal, such as a simple manufacturing process, higher transparency, better chemical and thermal stability, ordered crystal field environment for luminescent ions, and epoxy resin free in assembly process^[27,38,39]. Consequently, research on ${Cr}^{3 +} / {Ni}^{2 +}$ co-doped glass ceramics that were primarily crystallized by oxide nanocrystals, such as ${LiGa}_{5} O_{8}$ ^[40], $β - {Ga}_{2} O_{3}$ ^[41], ${ZnAlO}_{4}$ ^[42], ${MgSiO}_{4}$ ^[43], or $MgO$ ^[44] nanocrystals, has been implemented earlier for broadening the NIR emission band. Nevertheless, the higher phonon energy of the oxide nanocrystals often leads to an increase in the probability of nonradiative relaxation of the ions, thereby impacting their luminescent performance. In contrast, embedding fluoride nanocrystals into the glass matrix provides a low phonon energy environment for the incorporated ions, which contributes to the enhancement of the luminescent efficiency of fluoride-based materials^[38,45]. In addition, the crystal field strength in fluoride nanocrystals is generally weaker than that in oxide nanocrystals, which facilitates a red shift of the transition metal ion emission peaks toward longer wavelengths in the fluoride crystal environment^[38,46,47]. Accordingly, it is highly rewarding to develop fluoride-based glass ceramics co-activated by ${Cr}^{3 +}$ and ${Ni}^{2 +}$ for obtaining enhanced and ultra-wide NIR luminescence covering the NIR-I and NIR-II regions. According to our information, there has been very little research on ${Cr}^{3 +} / {Ni}^{2 +}$ co-doped NIR luminescent fluoride-based glass ceramics.

Herein, we report blue LED-excitable NIR broadband luminescence based on ${Cr}^{3 +} / {Ni}^{2 +}$ co-doped transparent fluorosilicate glass ceramics containing ${KCdF}_{3}$ nanocrystals. A simple ${SiO}_{2} - KF - {CdF}_{2}$ ternary fluorosilicate glass system^[48] with octahedral coordination and a lower phonon energy environment due to the precipitated fluoride nanocrystals, is intentionally designed to accommodate ${Cr}^{3 +}$ and ${Ni}^{2 +}$ ions. With the site-occupancy of these two types of doping ions in the cation ${Cd}^{2 +}$ ions, an effective NIR emission from 700 to 1800 nm upon blue light excitation is observed for the glass ceramics. The structural properties and optical performance of the glass ceramic samples under different heat-treated temperatures and doping contents were investigated using transmission electron microscopy (TEM), X-ray diffraction (XRD), and fluorescence spectroscopy. The energy transfer mechanism of ${Cr}^{3 +}$ to ${Ni}^{2 +}$ is analyzed. Finally, an NIR LED device by integrating the glass ceramic sample with a blue LED chip is fabricated, and its applications in covert information recognition and night vision illumination are demonstrated.

2. Experiment

The glass samples with a composition of ${55 SiO}_{2} - 20 K F - {25 Cd F}_{2} - x {Cr}_{2} O_{3} - y NiO$ (in mole fraction, $x = 0$ , 0.03, 0.05, 0.07, and 0.09; $y = 0$ , 0.03, 0.05, 0.1, 0.2, and 0.3) were synthesized using the melt quenching method. ${SiO}_{2}$ , KF, ${CdF}_{2}$ (AR, Macklin Reagents), ${Cr}_{2} O_{3}$ , and NiO (99.99%, Aladdin Reagents) have been selected as ingredients. Approximately 20 g of the components were weighed according to the given stoichiometric ratios, and the mixture was thoroughly loaded into an alumina crucible and melted at 1450°C for 10 min. Then the melt liquid was cast onto a steel plate to form the precursor glass (marked as $x {Cr}^{3 +} / y {Ni}^{2 +} - PGs$ ). The PGs were sliced into pieces and subjected to heat treatment at 550°C, 560°C, 570°C, and 580°C in a Nabertherm N60 precision-controlled muffle furnace to obtain glass ceramics (marked as $x {Cr}^{3 +} / y {Ni}^{2 +} - GCs$ , respectively). Finally, each sample was optically polished for further characterization.

XRD patterns were measured by a Bruker D2 X-ray diffractometer ( $λ = 1.5406 Å$ ). Optical transmittance spectra were recorded with a PerkinElmer Lambda 950 spectrophotometer. TEM, high-resolution (HRTEM) images, and element mapping were conducted on an FEI Talos F200X. A fluorescence spectrometer (Edinburgh FLS980) was utilized to measure the emission/excitation spectra and fluorescence lifetime, with a microsecond flashlamp (µF900) and a continuous 450 W Xe lamp as excitation sources, respectively. An Ocean Optics spectrometer was employed to obtain the electroluminescence spectra of the glass ceramics-converted NIR LEDs. Photographs for the applications of the NIR LEDs device were captured using an NIR camera (DinoCapture 2.0). All measurements were performed at room temperature unless otherwise specified.

3. Results and Discussion

Figure 1(a) displays the crystal structure of ${KCdF}_{3}$ , which shows an orthorhombic perovskite structure with space group Pnma (No. 62). In such a structure, six $F^{-}$ ions coordinate with each ${Cd}^{2 +}$ ion to form [ ${CdF}_{6}$ ] octahedra, while $K^{+}$ ions occupy the cavities formed by eight corner-sharing [ ${CdF}_{6}$ ] octahedra^[49]. The doping ${Ni}^{2 +}$ ( $r = 0.72 Å$ ) and ${Cr}^{3 +}$ ( $r = 0.69 Å$ ) tend to occupy the octahedrally coordinated ${Cd}^{2 +}$ ( $r = 0.95 Å$ ) site due to their smaller ionic radii than that of ${Cd}^{2 +}$ and thus locate in a suitable NIR-activated environment. Figure 1(b) presents the XRD results of the PG and the GCs at different heat treatment temperatures. The PG sample exhibits a typical amorphous structure. After heat treating the precursor glass at 550°C, 560°C, and 570°C, respectively, the corresponding diffraction peaks gradually sharpen and align closely with the standard pattern of the ${KCdF}_{3}$ nanocrystals (PDF#22-0822), indicating the precipitation and gradual growth of the ${KCdF}_{3}$ nanocrystals within the glass matrix during heat treatment. Further raising the temperature to 580°C, the intensity of the diffraction peaks related to the ${KCdF}_{3}$ nanocrystals no longer increases, while a distinct hetero peak originating from the $K_{6} {Si}_{3} O_{9}$ phase appears at a $2 θ$ angle of 18° upon closer observation. The presence of the $K_{6} {Si}_{3} O_{9}$ phase leads to enhanced light scattering, thereby reducing the transparency of the glass ceramics^[38]. Therefore, the optimum heat treatment temperature for the glass ceramics was chosen at 570°C, and all subsequent samples were heat-treated at this temperature accordingly. Figures 1(c) and 1(d) display the XRD patterns of the GC-570 samples doped with varying concentrations of ${Cr}^{3 +}$ [varying from 0.03% to 0.09% (mole fraction) when the content of ${Ni}^{2 +}$ is fixed at 0.2% (mole fraction)] and ${Ni}^{2 +}$ [varying from 0.03% to 0.2% (mole fraction) when the content of ${Cr}^{3 +}$ is fixed 0.07% (mole fraction)], respectively. A blank GC-570 sample was also prepared, and its XRD pattern is presented in Figs. 1(c) and 1(d) for comparison. The results indicate that the intensity of the diffraction peaks remains unchanged with increasing concentrations of ${Cr}^{3 +}$ and/or ${Ni}^{2 +}$ , demonstrating that the addition of ${Cr}^{3 +}$ and ${Ni}^{2 +}$ into the glass has no effect on the precipitated crystalline phase. However, it is noted that the position of the diffraction peaks gradually shifts towards a larger $2 θ$ angle with increasing concentration of ${Cr}^{3 +}$ and (or) ${Ni}^{2 +}$ in the samples, which implies that the shrinkage of the crystal lattice occurs because of the partial substitution of the larger ${Cd}^{2 +}$ ions ( $r = 0.95 Å$ ) by ${Cr}^{3 +}$ ions ( $r = 0.69 Å$ ) and ${Ni}^{2 +}$ ions ( $r = 0.72 Å$ )^[27].

Figure 1.(a) Cell structure of the KCdF₃ crystal; (b) XRD patterns of 0.07Cr³⁺/0.2Ni²⁺-PG and GC samples at different heat-treatment temperatures, and the inset shows the photograph of the glass samples in daylight; (c), (d) XRD patterns of xCr³⁺/0.2Ni²⁺-GCs (x = 0, 0.03, 0.05, 0.07, 0.09) and 0.07Cr³⁺/yNi²⁺-GCs (y = 0, 0.03, 0.05, 0.1, 0.2, 0.3), heat-treated at 570°C for 2 h, respectively.

Download full size

View all figures

To get a deeper insight into the crystallized glass, TEM characterization of a representative sample $0.07 {Cr}^{3 +} / 0.2 {Ni}^{2 +} - GC 570$ was performed, as depicted in Figs. 2(a)–2(h). The bright-field TEM image shown in Fig. 2(a) reveals that spherical nanocrystals with an average diameter of around 7 nm are randomly distributed within the glass matrix. The distinct lattice fringes in the HRTEM image shown in Fig. 2(b) confirm the crystalline nature of the precipitated nanoparticles. The corresponding lattice spacing ( $d$ value) is calculated to be about 0.302 nm, which can be assigned to the (121) crystal plane of ${KCdF}_{3}$ ( $d_{(121)} = 0.304 nm$ ). To further identify the formed nanocrystals, the dark-field HAADF-TEM and two-dimensional element mapping images of K, Cd, F, Cr, and Ni were measured, as given in Figs. 2(c)–2(h). It shows that the distribution of the K, Cd, and F elements coincides very well with that of the nanocrystals. According to the TEM and XRD results, it is confirmed that the ${KCdF}_{3}$ crystalline phase was formed in the glass matrix and that a portion of the ${Cr}^{3 +}$ and ${Ni}^{2 +}$ ions can be embedded in the crystallized ${KCdF}_{3}$ nanocrystals.

Figure 2.(a) Bright-field TEM image of the 0.07Cr³⁺/0.2Ni²⁺-GC570 sample, and the inset shows the size distribution plot of the KCdF₃ nanocrystals; (b) HRTEM image; (c)–(h) Dark-field HADDF-TEM images and element mapping of K, Cd, F, Cr, and Ni, respectively.

Download full size

View all figures

The optical transmission spectra of the blank sample and ${Cr}^{3 +} / {Ni}^{2 +} - GC 570$ samples were examined over the spectral range of 300 to 2000 nm, as illustrated in Fig. 3(a). Both samples exhibit high transmittance ( $> 80 %$ ) owing to the small size of the precipitated ${KCdF}_{3}$ nanocrystals. The blank glass ceramic appears colorless as a result of its lack of absorption in the visible region, while the co-doped GC presents a deep yellow appearance [see the insets of Figs. 1(b) and 3(a)]. The obvious absorption bands centered at $\sim 825 nm$ and 1420 nm for the co-doped GC originate from the ${{}^{3}A}_{2} ({}^{3}F)$ to ${{}^{3}T}_{1} ({}^{3}F)$ and ${{}^{3}T}_{2} ({}^{3}F)$ transitions of ${Ni}^{2 +}$ , respectively^[38]. A relatively weak absorption peak at approximately 620 nm, attributed to the ${{}^{4}A}_{2} ({}^{4}F) \to {{}^{4}T}_{2} ({}^{4}F)$ transition of ${Cr}^{3 +}$ , is also detected in the absorption spectrum of the co-doped GC^[50]. Notably, in comparison to the blank sample, the co-doped GC samples exhibit a significant red shift in the absorption cutoff edge, shifting from 400 to 475 nm. This red shift is likely ascribed to the absorption of ${Cr}^{3 +}$ with octahedral coordination in the blue region, corresponding to the ${Cr}^{3 +} {{}^{4}A}_{2} ({}^{4}F) \to {{}^{4}T}_{1} ({}^{4}F)$ transition^[25].

Figure 3.(a) Transmission spectra of the blank- and 0.07Cr³⁺/0.2Ni²⁺-GC570 samples; (b) PLE (monitored at 784 nm) and PL (excited by 450 nm) spectra of the 0.07Cr³⁺-GC570 sample; (c) PL spectra of the 0.2Ni²⁺-GC570 and 0.07Cr³⁺/0.2Ni²⁺-GC570 samples excited by 450 nm; (d) PEL spectra of the 0.2Ni²⁺-GC570 and 0.07Cr³⁺/0.2Ni²⁺-GC570 samples monitored at 1584 nm.

Download full size

View all figures

Figures 3(b)–3(d) display the photoluminescence excitation (PLE) and photoluminescence (PL) spectra of the ${Cr}^{3 +}$ and ${Ni}^{2 +}$ single-doped samples as well as the ${Cr}^{3 +} / {Ni}^{2 +}$ co-doped GC samples. As seen from Fig. 3(b), the characteristic excitation broadband divided into two bands with central wavelengths of approximately 450 and 625 nm (corresponding to the transitions of octahedrally coordinated ${Cr}^{3 +}$ ions from the ${{}^{4}A}_{2}$ to ${{}^{4}T}_{2} ({}^{4}F)$ and ${{}^{4}T}_{2} ({}^{4}F)$ states, respectively^[5]) are observed in the PEL spectrum (monitored at 784 nm) of the ${Cr}^{3 +}$ single-doped sample ( $0.07 {Cr}^{3 +} - GC 570$ ). When excited by 450 nm light, the $0.07 {Cr}^{3 +} - GC 570$ sample exhibits an NIR-emitting broadband ranging from 700 to 1100 nm, which is ascribed to the characteristic ${{}^{4}T}_{2} ({}^{4}F) \to {{}^{4}A}_{2} ({}^{4}F)$ transition of ${Cr}^{3 +}$ in an octahedral coordination environment. Noticeably, in addition to the aforementioned broadband from ${Cr}^{3 +}$ , a new emission band extending from 1100 to 1800 nm, centered at approximately 1584 nm, emerges in the PL spectrum of the ${Cr}^{3 +} / {Ni}^{2 +}$ co-doped GC sample ( $0.07 {Cr}^{3 +} / 0.2 {Ni}^{2 +} - GC 570$ ) under 450 nm excitation [Fig. 3(c)]. This new NIR luminescence actually arises from the ${{}^{3}T}_{2} ({}^{3}F) \to {{}^{3}A}_{2} ({}^{3}F)$ transition of octahedrally coordinated ${Ni}^{2 +}$ . Although the ${Ni}^{2 +}$ -doped GC sample ( $0.2 {Ni}^{2 +} - GC 570$ ) exhibits a similar emission spectrum in the 1100–1800 nm range under the same excitation conditions, the ${Ni}^{2 +}$ -related NIR emission intensity of the $0.07 {Cr}^{3 +} / 0.2 {Ni}^{2 +} - GC 570$ sample is approximately 11 times stronger than that of the $0.2 {Ni}^{2 +} - GC 570$ sample, as shown in Fig. 3(c). Furthermore, two distinct broad excitation bands were observed in the $0.07 {Cr}^{3 +} / 0.2 {Ni}^{2 +} - GC 570$ sample when monitoring the emission of ${Ni}^{2 +}$ at 1584 nm. The positions and shapes of these excitation bands resemble those observed in the $0.07 {Cr}^{3 +} - GC 570$ sample [Fig. 3(d)]. In contrast, the PLE spectrum of the $0.2 {Ni}^{2 +} - GC 570$ sample is comprised of two very weak excitation bands centered at $\sim 400$ and 825 nm (as seen in the enlarged PLE spectrum), corresponding to the electron transitions of ${{}^{3}A}_{2} ({}^{3}F)$ to ${{}^{3}T}_{1} ({}^{3}P)$ and ${{}^{3}T}_{1} ({}^{3}F)$ , respectively^[38]. These results indicate the presence of efficient ET from ${Cr}^{3 +}$ to ${Ni}^{2 +}$ within the co-doped glass ceramics.

The concentration-dependent photoemission spectra of the ${Cr}^{3 +} / {Ni}^{2 +}$ co-doped samples were systematically investigated to gain a deeper understanding of the ET behavior. Figure 4(a) displays the PL spectra of the $x {Cr}^{3 +} / 0.2 {Ni}^{2 +} - GC 570$ samples upon excitation by 450 nm ( $x = 0.03$ , 0.05, 0.07, and 0.09). It is clearly seen that a weak NIR emission band peaking at approximately 918 nm originating from ${Cr}^{3 +}$ is observed, and almost no ${Ni}^{2 +}$ -related emission signal is observed in the ${Cr}^{3 +} / {Ni}^{2 +}$ co-doped PG sample. However, after crystallization, dual NIR-emitting broad bands related to ${Cr}^{3 +}$ (700–1100 nm) and ${Ni}^{2 +}$ (1100–1800 nm) can be detected in all the ${Cr}^{3 +} / {Ni}^{2 +} - GC 570$ samples. In particular, the wavelength of the ${Cr}^{3 +}$ -related emission shows a pronounced blue shift from 918 to 784 nm, accompanied by a substantial enhancement in emission intensity of ${Cr}^{3 +}$ compared to the PG sample. This phenomenon further demonstrates that in the ${Cr}^{3 +} / {Ni}^{2 +} - GC 570$ samples, ${Cr}^{3 +}$ and ${Ni}^{2 +}$ were embedded into the ${KCdF}_{3}$ nanocrystals, occupying an ordered crystal field environment with octahedral coordination and low phonon energy, and hence resulting in an enhanced ultra-wide NIR-emitting broadband co-activated by ${Cr}^{3 +}$ and ${Ni}^{2 +}$ . In addition, it is notable that with the increase of the ${Cr}_{2} O_{3}$ content, both ${Cr}^{3 +}$ and ${Ni}^{2 +}$ -related emissions are increased continuously (reaching a maximum at $x = 0.07$ ), and the enhancement of the latter is obviously greater than that of the former. This result further confirms that the effective emission of ${Ni}^{2 +}$ essentially originates from the ET process from ${Cr}^{3 +}$ to ${Ni}^{2 +}$ . However, when the ${Cr}^{3 +}$ content is subsequently increased, the emission intensities of both ${Cr}^{3 +}$ and ${Ni}^{2 +}$ decrease due to the occurrence of a typical concentration quenching effect between the activators^[15]. The PL spectra of the $0.07 {Cr}^{3 +} / y {Ni}^{2 +} - GC 570$ samples ( $y = 0$ , 0.03, 0.05, 0.1, 0.2, and 0.3) also provide evidence for the presence of the ET process. Figure 4(b) shows that with the successive addition of ${Ni}^{2 +}$ , the emission intensity of the ${Cr}^{3 +}$ declines rapidly, while the emission intensity of the ${Ni}^{2 +}$ rises initially until reaching $y = 0.2$ . Beyond this concentration, a further increase in ${Ni}^{2 +}$ concentration leads to a decline in its emission intensity. All the above phenomena directly corroborate the effective ET from ${Cr}^{3 +}$ to ${Ni}^{2 +}$ in the ${KCdF}_{3}$ -based glass ceramics.

Figure 4.PL spectra of the (a) xCr³⁺/0.2Ni²⁺-GC570 (x = 0.03, 0.05, 0.07, and 0.09) and (b) 0.07Cr³⁺/yNi²⁺-GC570 (y = 0, 0.03, 0.05, 0.1, 0.2, and 0.3) samples under 450 nm excitation.

Download full size

View all figures

The luminescence dynamics of ${Cr}^{3 +}$ were examined to give more convincing evidence of the local environment change and ET behavior in the crystallized glass. Figures 5(a) and 5(b) show the luminescence decay curves monitored at 784 nm under 450 nm excitation for the $0.07 {Cr}^{3 +} / y {Ni}^{2 +} - GC 570$ ( $y = 0$ , 0.03, 0.05, 0.1, 0.2, and 0.3) and 0.07 ${Cr}^{3 +} / 0.2 {Ni}^{2 +} - PG$ samples. A double-exponential decay equation accurately describes all decay curves as follows^[20,34]: $I (t) = A_{1} \exp (- t / τ_{1}) + A_{2} \exp (- t / τ_{2}),$ (1)where $I (t)$ represents the fluorescence intensity at time $t$ . The parameters $A_{1}$ and $A_{2}$ are the fitting constants, with $τ_{1}$ and $τ_{2}$ corresponding to the short and long decay time components, respectively. The decay lifetimes of ${Cr}^{3 +}$ are determined using the following equation^[7,8]: $τ = (A_{1} τ_{1}^{2} + A_{2} τ_{2}^{2}) / (A_{1} τ_{1} + A_{2} τ_{2}) .$ (2)

Figure 5.Decay curves of the (a) 0.07Cr³⁺/0.2Ni²⁺-PG and GC570 as well as (b) 0.07Cr³⁺/yNi²⁺-GC570 samples under 450 nm excitation when monitoring the emission at 784 nm.

Download full size

View all figures

As summarized in Table 1, the decay lifetime of the ${Cr}^{3 +}$ in the crystallized glass is approximately 181 µs, which is apparently longer than that in the corresponding PG sample (114 µs). This strongly demonstrates the incorporation of ${Cr}^{3 +}$ into an environment with lower phonon energy, namely, fluoride ${KCdF}_{3}$ nanocrystals, after crystallization, reducing the nonradiative relaxation of the ${Cr}^{3 +}$ , consequently. Moreover, it can be seen that the value of the decay lifetime of ${Cr}^{3 +}$ in the crystallized glass is gradually decreased from 364 to 173 µs with the content of ${Ni}^{2 +}$ increasing from 0 to 0.3% (mole fraction), which further proves the ET process from ${Cr}^{3 +}$ to ${Ni}^{2 +}$ . On the basis of this, the ET efficiency ( $η$ ) from ${Cr}^{3 +}$ to ${Ni}^{2 +}$ is estimated using the equation below^[35], $η = 1 - τ_{s} / τ_{s 0},$ (3)where $τ_{s}$ and $τ_{s 0}$ stand for the decay lifetime of ${Cr}^{3 +}$ with and without ${Ni}^{2 +}$ , respectively. Using Eq. (3), the values of $η$ are calculated to be 38.5%, 41.5%, 45.8%, and 50.2% for the doping concentration of ${Ni}^{2 +}$ ( $y = 0.03$ , 0.05, 0.1, and 0.2), respectively. The energy transfer process from ${Cr}^{3 +}$ to ${Ni}^{2 +}$ becomes more effective as the concentration of ${Ni}^{2 +}$ increases.

Table 1. Decay Lifetimes of Cr³⁺ [τ(Cr³⁺)] in the Cr³⁺/Ni²⁺ Co-Doped Samples with Different Ni²⁺ Contents and Energy Transfer Efficiencies (η) from Cr³⁺ to Ni²⁺

View table

View all Tables

Table 1. Decay Lifetimes of Cr³⁺ [τ(Cr³⁺)] in the Cr³⁺/Ni²⁺ Co-Doped Samples with Different Ni²⁺ Contents and Energy Transfer Efficiencies (η) from Cr³⁺ to Ni²⁺

Sample	τ(Cr³⁺) (μs)	η
0.07Cr³⁺/0.2Ni²⁺-PG	114	—
0.07Cr³⁺-GC570	364	—
0.07Cr³⁺/0.03Ni²⁺-GC570	224	38.5%
0.07Cr³⁺/0.05Ni²⁺-GC570	213	41.5%
0.07Cr³⁺/0.1Ni²⁺-GC570	197	45.8%
0.07Cr³⁺/0.2Ni²⁺-GC570	181	50.2%
0.07Cr³⁺/0.3Ni²⁺-GC570	173	—

The energy level diagram of the ${Cr}^{3 +}$ and ${Ni}^{2 +}$ ions in the current glass ceramic system and the possible ET mechanism between them are illustrated in Fig. 6. Upon excitation by 450 nm, the electrons in the ground state ( ${{}^{4}A}_{2}$ ) of ${Cr}^{3 +}$ are first excited to the ${{}^{4}T}_{1} ({}^{4}F)$ excited state. The excited electrons preferentially undergo non-radiative relaxation to the ${{}^{4}T}_{2} ({}^{4}F)$ state, where a portion of them transition back to the ground state ( ${{}^{4}A}_{2}$ ) via a radiative process, emitting energy as near-infrared light that peaks at 784 nm. Other parts of the excitation energy of ${Cr}^{3 +}$ can also be transferred to the neighboring ${Ni}^{2 +}$ through a nonradiative ET process by which electrons are preferentially excited from the ground state ${{}^{3}A}_{2} ({}^{3}F)$ to the excited state ${{}^{3}T}_{1} ({}^{3}F)$ of ${Ni}^{2 +}$ due to the energetic proximity of the ${{}^{4}T}_{1} ({}^{4}F)$ excited state of ${Cr}^{3 +}$ to the ${{}^{3}T}_{1} ({}^{3}F)$ level of ${Ni}^{2 +}$ . After that, the excited electrons non-radiatively relax to the ${{}^{3}T}_{2} ({}^{3}F)$ energy level and then radiatively relax to the ground state ${{}^{3}A}_{2} ({}^{3}F)$ , giving the enhanced NIR emission centered at 1584 nm from ${Ni}^{2 +}$ .

Figure 6.Schematic diagram for the energy transfer from Cr³⁺ to Ni²⁺ in the glass ceramics.

Download full size

View all figures

The ultra-broadband NIR-emitting ${KCdF}_{3} : {Cr}^{3 +} / {Ni}^{2 +}$ -based glass ceramics were integrated with a 460 nm LED chip to fabricate NIR LED devices aimed at assessing their potential application prospects. The insets in Fig. 7(a) show the pictures of the as-fabricated NIR LED device taken in daylight (i) and its working state captured with an infrared camera in the dark (ii). The electroluminescence (EL) spectra of the NIR LED device, operated at different driving currents, exhibit an ultra-broadband NIR emission spanning 700 to 1800 nm [Fig. 7(a)]. Figures 7(b)–7(d) sequentially illustrate the utilization of the integrated NIR LED light source in encryption, night vision, and information identification. The left photographs were captured with a normal camera in daylight, and the right photographs were obtained by an NIR camera in dark conditions. As shown in Fig. 7(b), using the NIR LED as the light source, the real information “1379” written in black can be clearly read due to the strong NIR light absorption of black carbon, while the interference message with colorful texts is eliminated^[20]. Furthermore, taking advantage of the powerful penetration ability of the NIR light, the NIR LED light source can be utilized to provide a clear image through different materials. For example, the message “ ${KCdF}_{3} : {Cr}^{3 +} / {Ni}^{2 +}$ ” covered by dark sunglasses is nearly invisible to the naked eye under daylight conditions, whereas it can be successfully detected when illuminated with the NIR LED device [Fig. 7(c)]. Figure 7(d) displays the images of an apple under the illumination of natural light and NIR LED light. In a dark field, the image captured by the NIR camera clearly manifests the outline of an apple compared to a normal camera. These results prove that the NIR LED light source developed in this study has potential applications in the fields of encryption, information identification, and night vision.

Figure 7.(a) Driven current-dependent electroluminescence spectra of the as-fabricated NIR LED; (b)–(d) Applications of the as-fabricated NIR LED light source, and photographs taken with a normal camera in daylight (left) and a NIR camera under illumination of the prepared NIR LED light (right), respectively. A 1000 nm long-pass filter was used when capturing images in (c) and (d).

Download full size

View all figures

4. Conclusion

In summary, transparent fluorosilicate glass ceramics embedded with perovskite ${KCdF}_{3}$ nanocrystals were fabricated using a melt-quenching method followed by heat treatment. The characterizations of XRD, TEM, photoemission spectra, and fluorescence lifetimes confirm that ${Ni}^{2 +}$ and ${Cr}^{3 +}$ ions can enter into the ${KCdF}_{3}$ nanocrystals with low phonon energy and occupy the octahedral site [ ${CdF}_{6}$ ] after crystallization. By controlling the ET from ${Cr}^{3 +}$ to ${Ni}^{2 +}$ in the glass ceramics, a broadband NIR emission ranging from 700 to 1800 nm with 450 nm excitation was achieved with a total bandwidth of 519 nm (225 nm + 294 nm). The ET efficiency from ${Cr}^{3 +}$ to ${Ni}^{2 +}$ is calculated to be as high as 50.2% by recording the fluorescence lifetime. The as-fabricated NIR LED device based on the combination of our optimal glass ceramic with a 460 nm blue light chip demonstrates its multi-functional applications in concealed information recognition, non-destructive detection, and night vision. Our results point out a direction for the design of fluoride nanocrystal-based transparent optical materials co-activated by ${Cr}^{3 +}$ and ${Ni}^{2 +}$ .

Category: Optical Materials

Received: Dec. 30, 2024

Accepted: Apr. 1, 2025

Published Online: Jul. 4, 2025

The Author Email: Xueyun Liu (liuxueyun@nbu.edu.cn)

DOI:10.3788/COL202523.081601

CSTR:32184.14.COL202523.081601

Table 1. Decay Lifetimes of Cr3+ [τ(Cr3+)] in the Cr3+/Ni2+ Co-Doped Samples with Different Ni2+ Contents and Energy Transfer Efficiencies (η) from Cr3+ to Ni2+

Table 1. Decay Lifetimes of Cr3+ [τ(Cr3+)] in the Cr3+/Ni2+ Co-Doped Samples with Different Ni2+ Contents and Energy Transfer Efficiencies (η) from Cr3+ to Ni2+

Table 1. Decay Lifetimes of Cr³⁺ [τ(Cr³⁺)] in the Cr³⁺/Ni²⁺ Co-Doped Samples with Different Ni²⁺ Contents and Energy Transfer Efficiencies (η) from Cr³⁺ to Ni²⁺

Table 1. Decay Lifetimes of Cr³⁺ [τ(Cr³⁺)] in the Cr³⁺/Ni²⁺ Co-Doped Samples with Different Ni²⁺ Contents and Energy Transfer Efficiencies (η) from Cr³⁺ to Ni²⁺