Theoretical design and experimental realization of Fe³⁺-doped dual-band near-infrared garnet phosphors

Abstract

Cr³⁺-free near-infrared (NIR) phosphors based on Fe³⁺ have garnered extensive attention due to their environmentally friendly and tunable optical properties. However, the reported luminescence predominantly originates from the Fe³⁺ ion in tetrahedral coordination, with wavelengths in the range of 670–830 nm. Phosphors with luminescence from octahedrally coordinated Fe³⁺, which are expected to shift to longer wavelengths over 900 nm, are limited due to the challenges such as quenching mechanisms. Garnets with the formula A₃B₂C₃O₁₂ are excellent hosts for phosphors due to their rigid structures and tunable luminescence properties. Theoretical analysis, supported by first-principles calculations, indicates that Fe³⁺ can occupy both tetrahedral (Fe(T)) and octahedral (Fe(O)) sites, potentially producing dual-band emission in garnet crystals with large octahedral host ions, such as Sr₃ (Sc/Lu/Y)₂Ge₃O₁₂ crystals. This has guided us in the experimental realization of dual-band NIR luminescence, peaking at 720–730 nm (Fe(T)) and 980–990 nm (Fe(O)), in these materials. Consistent with our optical transition analysis, the luminescence intensities of Fe(T) and Fe(O) show different temperature dependencies. Fe(T) exhibits weaker temperature dependence, while Fe(O) experiences severe temperature quenching via the ²T₂ intermediate energy level. The dual-band NIR phosphors Sr₃ (Sc/Lu/Y)₂Ge₃O₁₂:Fe³⁺ show potential applications in luminescence intensity ratio-based and luminescence decay time-based thermometers, with a significant maximum relative sensitivity of 1.24% K⁻¹ at 155 K. The materials designed here provide a foundation for related application explorations, and the strategy developed can be applied to the exploration and development of Fe³⁺-activated advanced optical materials.

---

1 Introduction

Near-infrared (NIR) phosphors have wide applications in plant cultivation, night-vision imaging, food analysis, and biomedicine.¹ NIR-emitting luminescent materials with high efficiency and good thermal properties are in increasing demand. Initially developed through trial-and-error experimental attempts, and guided by high-throughput theoretical calculations, the exploration of NIR luminescent materials based on rare-earth and Cr³⁺ ions has exhibited rapid developments.^2,3 The former activator, based on the f–f transition, shows a narrow bandwidth,⁴ while the latter suffers from a potential risk of oxidation of Cr³⁺ to Cr⁶⁺ and competition with tetrahedrally coordinated site occupation.⁵ As an environmentally benign and biologically compatible ion, Fe³⁺ has attracted increased attention serving as a promising activator candidate for efficient near-infrared luminescence.⁶ However, the development of Fe³⁺ NIR phosphors shows many difficulties and challenges.^7–9

The unclear luminescence mechanisms and uncertainties in site occupation are key obstacles in the development of Fe³⁺-based NIR-emitting phosphors. However, significant experimental and theoretical advancements have been achieved only recently. The site occupation and optical transitions of iron ions in solids from the literature have been elaborately rationalized by theoretical calculations.¹⁰ Typically, tetrahedrally coordinated Fe³⁺ in oxide insulators exhibit near-infrared emission in the range of 670–830 nm, which has been extensively reported in several hosts, such as MgAl₂O₄,^11,12 MgGa₂O₄,^13,14 LiGa₅O₈,¹⁵ Mg₂SnO₄,¹⁶ KAl₁₁O₁₇,¹⁷ NaAl₅O₈,¹⁸ and Li₂ZnAO₄ (A = Si, and Ge).¹⁹ Octahedrally coordinated Fe³⁺ phosphors are expected to exhibit significantly longer wavelength emissions due to stronger ligand field strength, but very limited cases have been conclusively reported. Specifically, octahedral Fe³⁺ luminescence was explored in A₂BB′O₆ (A = Sr²⁺, Ca²⁺; B, B′ = In³⁺, Lu³⁺, Sb⁵⁺, Sn⁴⁺) double perovskites, which exhibit tunable emission wavelengths and high internal quantum efficiency.^7,20 The Fe³⁺-activated Sr₂ScSbO₆ NIR phosphor exhibits near-unity internal quantum efficiency and excellent thermal stability, demonstrating its potential applications in anti-counterfeiting, information encryption, and night vision.²¹ The NIR luminescence from Fe³⁺ at octahedrally coordinated sites in Sr₉Ga(PO₄)₇ and NaScSi₂O₆ hosts shows potential applications in night vision and optical thermometry.^22,23 Furthermore, the limited luminescence from octahedrally coordinated Fe³⁺ prompts the exploration of quenching mechanisms of Fe³⁺ in solids. The presence or absence of luminescence in iron-doped crystals has recently been uncovered,⁹ but it requires further validation and exploration through experimental techniques.

Garnets are excellent hosts for phosphors and laser crystals because their rigid crystal structures suppress nonradiative relaxations due to electron–phonon couplings.^24,25 Garnets are cubic crystals belonging to the Ia3̄d (no. 230) space group. In the formula A₃B₂C₃O₁₂, A, B, C and O occupy the 24c, 16a, 24d and 96h Wyckoff positions, respectively, forming [AO₈] dodecahedra, [BO₆] octahedra, and [CO₄] tetrahedra. Each octahedron is connected to six [AO₈] dodecahedra by sharing edges, while each tetrahedron is connected to four [BO₆] octahedra by sharing corners, providing three types of cation sites for dopant ions. Garnet phosphors are unique in their tunable luminescence properties, which can be adjusted by varying the A, B, and C cations, or forming solid solutions, allowing for the tuning and optimization of optical properties. For example, the Ce³⁺ activator in garnets occupies the dodecahedral A lattice site, with its emission tunable from green to orange-red by varying the ligand field strength and nephelauxetic effect.²⁶ The Cr³⁺ activator in garnets occupies the octahedral B site, and its luminescence can be tuned from sharp ²E → ⁴A₂ emission for a relatively strong ligand field case to broad ⁴T₂ → ⁴A₂ emission under the situation of a weaker ligand field.^27,28 Additionally, garnets have tetrahedral sites to accommodate Cr⁴⁺ ions with ³T₂ → ³A₂ emissions.⁵ In principle, Fe³⁺ ions can occupy tetrahedral and octahedral lattice sites in solids, as indicated by EPR studies and first-principles calculations.⁹ However, only luminescence from a single Fe³⁺ site in garnets has been reported to date, such as in Ca₄ZrGe₃O₁₂, Y₃Al₅O₁₂, Y₃Ga₅O₁₂, and Lu₃Al₅O₁₂ garnets, with emission band peaks falling in the range of 750–830 nm.^29–32 This luminescence has been reliably identified as originating from tetrahedral Fe³⁺, while octahedral Fe³⁺ in these garnets relaxes nonradiatively via²T₂, which is lower than the otherwise luminescence level ⁴T₁ due to the excessively strong ligand fields in those hosts.⁹

Dual-modal or multi-modal phosphors have gained increasing popularity in the field of anti-counterfeiting techniques,^33,34 optical thermometry,^35,36 multi-channel fluorescence imaging,³⁷ photodynamic therapy,³⁸etc.

Guided by a phenomenological model predicting the potential of observing Fe³⁺ luminescence at an octahedral site from the luminescence properties of Cr³⁺ at the same site,⁹ and further validated through first-principles calculations, we experimentally realized a series of efficient dual-band near-infrared phosphors from both tetrahedral and octahedral Fe³⁺ in iron ion-doped garnets Sr₃Sc₂Ge₃O₁₂ (SSGG), Sr₃Lu₂Ge₃O₁₂ (SLGG) and Sr₃Y₂Ge₃O₁₂ (SYGG). The potential applications of the Fe³⁺-activated dual-band near-infrared garnet phosphor have also been discussed.

2 Results and discussion

2.1 Theoretical design

The Tanabe–Sugano diagram for Fe³⁺ with a d⁵ electronic configuration is shown in Fig. 1(a). As demonstrated in previous works by some of us,^9,10 Fe³⁺ at tetrahedral sites, denoted as Fe(T), typically exhibits deep red to NIR emission, unless suppressed by factors such as structural relaxation in soft crystals, strong ionization, or unidentified quenching centers. Fe³⁺ at octahedral sites (Fe(O)) can exhibit either emissive or quenched behavior depending on the ligand field strength. Fig. 1(b) and (c) display the coordinate configuration diagrams of two typical Fe(O) activators, with the relatively weak and strong ligand field strength. The critical ligand field strength was estimated to be about 1.50–1.70 eV,⁹ above which Fe³⁺ becomes nonemissive due to nonradiative relaxation to the ground state via the ²T₂ intermediate energy level. To predict the potential presence of Fe(O) luminescence, a phenomenological model was proposed, which correlates the optical behavior of Fe³⁺ with that of Cr³⁺ at the same octahedral sites.⁹ Specifically, for an octahedrally coordinated lattice site where Cr³⁺ exhibits broad ⁴T₂ → ⁴A₂ emissions at a sufficiently large wavelength, the weaker-than-critical ligand field leads to emissive NIR luminescence of Fe³⁺ occupying the same site, as the relatively high ²T₂ energy level of Fe³⁺ does not provide an efficient quenching channel. In the literature studies, Cr³⁺ doped SYGG and SSGG crystals were reported to show broad ⁴T₂ → ⁴A₂ emissions.^39,40 Thus, the three garnets, SSGG, SLGG, and SYGG, are considered as hosts for potential Fe³⁺-based dual-band NIR (i.e. Fe(T) and Fe(O), as shown in Fig. 1(d)) phosphors in our work.

Fig. 1 (a) The Tanabe–Sugano diagram of Fe³⁺ (3d⁵), with the ranges of the ligand field Δ to Racah parameter B ratio, Δ/B, for tetrahedral Fe(T) and octahedral Fe(O) schematically marked. (b and c) The coordinate configuration diagrams of the typical octahedral Fe(O) activators with the relatively weak and strong ligand field (LF) strengths, leading to radiative (emission) and nonradiative relaxation (no emission) to the ground state, respectively. (d) The crystal structure of the garnets and the potential sites iron ions may enter. (e) The defect formation energy of both intrinsic and extrinsic (related to iron dopants) defects as a function of the Fermi energy in SSGG. The vertical dashed (dotted) lines are the equilibrium Fermi energies E_F⁰ with (without) extrinsic defects.

We used SSGG as a prototype host to discuss the native defects, site occupation, and valence states of the iron dopants obtained from first-principles calculations. For convenience, the Kröger–Vink notation D^q_H is employed to denote a defect with net charge q (′, ×, ˙ being −1, 0, and +1, respectively) formed by replacing a host atom H with a donor atom D.⁴¹ In particular, D = V represents a vacancy, and H = i represents an interstitial defect. Antisite defects dominate among intrinsic defects due to their relatively lower formation energy and consequently higher concentration, as shown in Fig. 1(e). The defect formation energies in Fig. 1(e) are plotted under the chemical potential conditions of coexistence with Sc₂O₃ and Sr₂GeO₄ under a moderate oxygen atmosphere (Δμ_O = −1.5 eV). For the pristine crystal without an iron dopant, the equilibrium Fermi energy E_F⁰, governed by charge neutrality, is estimated at the intersection of the formation energy lines of the leading positive native defects with the leading negative native defect . In the case of iron doping, there are three potential cation sites for substitution. Iron substitution at the Sr²⁺ site can occur as Fe²⁺ and Fe³⁺ defects. However, both have high formation energies and are insignificant in concentration, so they are not shown in the figure. The preference at the tetrahedral Ge⁴⁺ site and the octahedral Sc³⁺ site varies along with the chemical potential condition. Iron substitution at the tetrahedral site mainly occurs as Fe⁴⁺ and Fe³⁺ ions, while substitution at the octahedral site mainly occurs as Fe³⁺ and Fe²⁺ ions, with the valence states depending on the Fermi energy. Under the conditions shown in Fig. 1(e) with iron dopants, neutral octahedral , neutral tetrahedral , and negatively charged are all important. and , with opposite charge states, compensate each other to maintain charge neutrality. Tuning the chemical potential condition can influence the relative defect formation energies, but the neutral octahedral remains dominant, followed by the defects. Besides, in the series of garnets, SSGG, SLGG, and SYGG, iron substitutions at both tetrahedral and octahedral sites as Fe³⁺ ions are always dominant among native and other extrinsic defects (see more details in Note S1 of the ESI†).

The ligand field strength (Δ) of Fe³⁺ in a series of garnets is calculated and listed in Table 1. The ligand field strength of the tetrahedral Fe³⁺ (Δ_Fe(T)) is slightly greater than 4/9 times that of the octahedral Fe³⁺, as anticipated by the shorter Fe–O bond lengths in the tetrahedral sites relative to those in the octahedral sites. In the iron-doped Y₃Al₅O₁₂ and Y₃Ga₅O₁₂ garnet hosts, the calculated Δ_Fe(O) values are 1.77 and 1.69 eV, about or surpassing the critical ligand field strength (1.5–1.7 eV)⁹ where Fe(O) emission quenches. This leads to the mono-band (tetrahedral Fe³⁺) luminescence. Here in the Sr₃ (Sc/Lu/Y)₂ Ge₃O₁₂ host, the calculated ligand field strength of octahedral Fe³⁺ is lower than the critical ligand field strength, indicating the potential for dual-band (tetrahedral and octahedral Fe³⁺) luminescence. Subsequently, the detailed energy levels have been calculated.

Table 1 The ligand field strength of tetrahedral and octahedral Fe³⁺ (Fe(T) and Fe(O)) in a series of garnet hosts (in units of eV)

Garnet	Fe(T)	Fe(O)
Sr3Sc2Ge3O12	0.736	1.432
Sr3Lu2Ge3O12	0.741	1.392
Sr3Y2Ge3O12	0.746	1.379
Y3Al5O12	0.848	1.772
Y3Ga5O12	0.822	1.685

---

In the SYGG host, the ⁴T₁ ↔ ⁶A₁ excitation and emission energies of octahedral Fe³⁺ are predicted at 1.44 eV and 1.22 eV, respectively. The ²T₂ energy is calculated as 1.56 eV by considering the mixing of different Slater determinants due to the coulombic interaction. This energy is substantially higher than the stable ⁴T₁ level (estimated at 1.30 eV), allowing emission from this octahedral Fe³⁺ ion. For tetrahedral Fe³⁺, the ⁴T₁ ↔ ⁶A₁ excitation and emission energies are correspondingly 1.79 eV and 1.70 eV. Therefore, iron-doped SYGG is also anticipated to function as a dual-band NIR phosphor.

The luminescence trends of tetrahedral and octahedral Fe³⁺ in the series of garnets, SSGG, SLGG, and SYGG, are discussed as follows. Detailed calculations show that the ⁴T₁ → ⁶A₁ emission energies of octahedral Fe³⁺ are 1.223 eV, 1.220 eV, and 1.192 eV in the garnet hosts of B = Y, Lu, and Sc, respectively. The shorter Fe–O bond length and larger ligand field strength lead to lower emission energy, assuming the similarity in nephelauxetic effects and crystal rigidity. The average Fe–O bond lengths are 2.077 Å, 2.070 Å, and 2.056 Å after Fe replaces B = Y, Lu, and Sc in the garnets, respectively. For the tetrahedral Fe³⁺ in the three garnets, the calculated ⁴T₁ → ⁶A₁ emission energies are approximately 1.70 eV, with slight variations from 1.696 eV for B = Y to 1.704 eV for B = Sc. The slightly blue-shifted emission energies can be attributed to the degree of angular distortion of tetrahedral Fe³⁺, as revealed in our previous work.¹⁰ The standard deviation of bond angles σ[A(O–Fe³⁺–O)] was introduced to evaluate the angular distortion.¹⁰ The σ[A(O–Fe³⁺–O)] values are 6.71°, 7.62°, and 8.38° in the garnet hosts of B = Sc, Lu, and Y, respectively (see more details in Table S1 of the ESI†). The smaller angular distortion of tetrahedral Fe³⁺ in SSGG leads to the slight blue shift of the tetrahedral Fe³⁺ luminescence compared to that in SYGG.

2.2 Experimental realization, characterization, and analysis

Crystal structure and X-ray diffraction (XRD) analysis. The iron-doped SSGG, SLGG, and SYGG samples were characterized in detail by XRD, as shown in Fig. 2(a), and are well-indexed to the standard card (PDF #29-1313, PDF #85-2410), indicating the successful synthesis of single-phased phosphors with 0.5%Fe. The shift of Bragg reflection peaks toward lower angles along Sc³⁺, Lu³⁺, and Y³⁺ in the garnet crystals suggests lattice expansion, which is attributed to the increase in the ionic radius. It is noted that the synthesized SSGG samples doped with x%Fe (x = 0.1, 0.5, 1, 3, 5, 7) show negligible shifts in Bragg reflection peaks as x varies (refer to Fig. S4 of ESI†). The Rietveld analyses of XRD spectra in Fig. 2(b–d) further verify the phase purity of the as-prepared samples. Fig. 2(e) displays the X-ray photoelectron spectroscopy (XPS) full survey with Sr, Ge, Sc, O, and Fe elements in the sample. As shown in Fig. 2(f), two peaks centered around 709.4 eV and 722.8 eV in the high-resolution XPS spectrum can be attributed to the 2p_3/2 and 2p_1/2 of Fe³⁺, respectively.²⁰ The SEM and EDS images of SSGG:Fe³⁺ in Fig. 2(h–i) show the particle size of the phosphor, with Sr, Sc, Ge, O, and Fe elements uniformly distributed throughout the material.

Fig. 2 (a–d) XRD patterns and Rietveld refinement of SSGG, SLGG and SYGG doped with 0.5%Fe. (e) XPS full survey, (f and g) high-resolution XPS of Fe³⁺, Sr²⁺, Sc³⁺, Ge⁴⁺ and O²⁻, (h) SEM images, and (i) SEM/EDS mapping images of SSGG:Fe³⁺.

Dual-band near-infrared PL and PLE spectra. According to our theoretical studies, the iron dopant in the three garnets can occupy both tetrahedral and octahedral lattice sites, potentially producing dual-band NIR emission. Experimental measurements of photoluminescence (PL), PL excitation (PLE) spectra, and lifetime decay curves in SSGG, SLGG, and SYGG doped with 0.5%Fe at room temperature are plotted in Fig. 3. Successfully, there are two emission bands in all the three hosts: the first emission peaks at around 730 nm with a relatively small full width at half maximum (FWHM) of 68 nm (0.15 eV), and the second emission extends from 800 to 1300 nm, peaking at around 980 nm with a FWHM of 158 nm (0.20 eV). The former is attributed to the ⁴T₁ → ⁶A₁ emission of tetrahedrally coordinated Fe³⁺ (labeled as Fe(T)), and the latter is from octahedrally coordinated Fe³⁺ (labeled as Fe(O)). The measured emission energies are consistent with our optical transition calculations. The luminescence behaviour of tetrahedral Fe³⁺ is similar to that in other garnets,^30,31,42 and the emission energies in the host Sr₃(Sc/Lu/Y)₂ Ge₃O₁₂ show a blue shift relative to that in the YAG host due to the smaller Fe(T) ligand field strength, as shown in Table 1. However, the luminescence from octahedral Fe³⁺ ions in those garnets, such as Y₃Al₅O₁₂ and Y₃Ga₅O₁₂, was not observed, which can be attributed to the larger Fe(O) ligand-field strength causing the downshift of the ²T₂ level, providing a pathway for non-radiative relaxation to the ground state¹⁰ (refer to Table 1 and Fig. 1(a) for the dependence of energy levels on ligand field strength). Across the SYGG, SLGG, and SSGG series, the first PL emission peak (Fe(T)) shows a slight blue-shift from 730 nm to 720 nm, tentatively attributed to the change in the degree of structural distortion. The second emission peak (Fe(O)) shifts slightly from 980 nm to 990 nm, probably due to variations in the ionic radius and ligand field strength. These trends have been discussed in detail in the Theoretical design section (section 2.1).

Fig. 3 (a and b) Room-temperature PL spectra of SSGG, SLGG and SYGG doped with 0.5%Fe upon excitation at 274 and 320 nm. (c and d) PLE spectra and (e and f) room-temperature decay profiles of Fe(T) and Fe(O) emissions of SSGG, SLGG, and SYGG doped with 0.5%Fe, with decay lifetimes τ determined by single-exponential decay fitting.

The PL intensities of both bands vary with the concentration of Fe³⁺, initially increasing and then decreasing with an increase in the doping amount in SSGG:x%Fe (x = 0.1, 0.5, 1, 3, 5, 7) (Fig. S5†). The optimal doping concentration is 1% for Fe(T) emission but about 3% for the Fe(O) emission, which is similar to those of Fe³⁺ in other hosts, such as 1–2% in (Sr/Ca)₂InSbO₆,⁷ 1.5% in NaAl₅O₈,¹⁸ and 2% in NaScSi₂O₆.²²

As shown in Fig. 3(c), the PLE spectra of Fe(T) in SSGG, SLGG and SYGG doped with Fe³⁺ contain three bands: a strong band at 274 nm, attributed to the Fe_tet³⁺ → Fe_tet⁴⁺ + e_CBM charge transfer (CT) transition, and two weak bands at about 410 and 450 nm, associated with d → d transitions. Similar excitation features were also reported in the Ca₃Ga₂Sn₃O₁₂ host, i.e. a strong band due to CT transition and two weak d → d transitions³¹ of Fe³⁺. Our first-principles calculation gives the Δ, B, and Δ/B values of tetrahedral Fe³⁺ as 0.74 eV, 0.067 eV, and 11.0, respectively. Based on the Tanabe–Sugano diagram, it can be inferred that the 410 and 450 nm excitation of tetrahedral Fe³⁺ is due to the transitions from ⁶A₁ to ⁴E(⁴D) and ⁴T₂(⁴D) excited states, respectively. While monitoring the emission from octahedral Fe³⁺ (Fe(O)), the PLE spectra show only one band peaking at about 302 nm in SSGG, and at 320 nm in SLGG and SYGG hosts (Fig. 3(d)). This band is attributed to the Fe_oct³⁺ → Fe_oct²⁺ + h_VBM CT transition. As the d → d transition of Fe³⁺ is both spin- and parity-forbidden to a first-order approximation, the excitation of octahedral Fe³⁺ due to d → d transitions is generally too weak to be observable.

For the samples SSGG, SLGG and SYGG doped with 0.5%Fe, the decay lifetimes of Fe(T) and Fe(O) emissions are in the ranges of 8.11–9.90 ms (Fig. 3(e)) and 1.33–1.93 ms (Fig. 3(f)) at room temperature, respectively. As the doping concentration increases, the emission decay lifetime of Fe(T) in SSGG:x%Fe decreases from 9.44 ms (x = 0.1) to 6.16 ms (x = 7), while the decay lifetime of Fe(O) decreases much smoothly from 1.35 ms (x = 0.1) to 1.16 ms (x = 7) (refer to Fig. S5(c and d)† for the concentration-dependent decay curves). This is similar to those of Fe³⁺ activators at tetrahedral or octahedral sites in other systems. For instance, the octahedral Fe³⁺ emission decay lifetimes in NaScSi₂O₆, Sr₉Ga(PO₄)₇, and CaSnO₃ hosts are 0.93 ms, 0.60 ms, and 2.76 ms, respectively.^7,22,23 The luminescence decay lifetime from tetrahedral Fe³⁺ in Li₂ZnSiO₄:xFe³⁺ decreases from 4.08 ms (x = 0.002) to 2.82 ms (x = 0.06) as the Fe³⁺ concentration increases.¹⁹ Similarly, the decay lifetime of tetrahedral Fe³⁺ activators in Ca₃Ga₂Sn₃O₁₂:xFe³⁺ decreases from 6.04 ms (x = 0.002) to 5.26 ms (x = 0.04) with increasing Fe³⁺ doping concentration.³¹ However, generally speaking, the radiative decay lifetime of activators at octahedral sites should be longer than that at tetrahedral sites, which inspires us to further explore the luminescence behaviors at low temperatures. Furthermore, there is also a stronger dependence of the Fe(T) intensity on the doping concentration than that of Fe(O), showing the lower quenching concentration for Fe(T) than that for Fe(O) (refer to Fig. S5(a and b)† for the concentration-dependent emission spectra).

Temperature-dependent luminescence. The optical properties of these two NIR luminescences are investigated as functions of temperature. The temperature-dependent PL spectra and corresponding contour maps of SSGG:1%Fe, upon excitation at 274 nm, 302 nm and 410 nm, are displayed in Fig. 4(d–i). The PL spectra of Fe(T) show weaker temperature dependence compared to that of Fe(O). Under 274 nm excitation, a clear zero-phonon line at approximately 702 nm, along with a structured broad band, is observed at low temperatures until over 130 K, where the whole band becomes gradually structureless. Under 302 nm excitation, the zero-phonon line of Fe(T) is not resolvable even at a low temperature of 5 K, presumably due to significant homogeneous broadening, while under 410 nm excitation, in addition to the Fe(T) emission, another NIR luminescence peak at 1070 nm is observed, which exhibits a zero-phonon line at about 1040 nm at low temperature below 80 K, as shown in Fig. 4(j).

Fig. 4 (a) Normalized PLE and PL spectra of SSGG:1%Fe at 5 K; (b and c) temperature-dependent PLE spectra of SSGG:1%Fe phosphors monitored at 1070 nm and 720 nm, respectively; (d–i) temperature-dependent PL spectra and related contour maps of SSGG:1%Fe phosphors upon excitation at 274 nm, 302 nm and 410 nm, respectively; (j) detailed PL spectra upon excitation at 410 nm at various low temperatures to show the change in band shapes; (k and l) comparison of the integrated emission intensity and the luminescence decay lifetime as a function of temperature over the range of 5–455 K between Fe(T) (λ_emi = 720 nm) and Fe(O) (λ_emi = 990 nm) centers.

To study the phenomena in more detail, the temperature-dependent PLE spectra, monitored at 1070 nm and 720 nm, were recorded from 5 K to 455 K in 25 K intervals, as shown in Fig. 4(b and c), respectively. For comparison, the normalized PLE and PL spectra at low temperature (5 K) are also provided in Fig. 4(a). As shown, aside from the charge-transfer-related excitation centering at 302 nm, other broad excitation bands centering at 360 and 426 nm appear when monitoring the 1070 nm luminescence. At a low temperature, such as 5 K, the luminescence intensity of 1070 nm is stronger than that of 720 nm upon excitation at 410 nm. This NIR luminescence peak at 1070 nm closely resembles the peak at 990 nm, except for a redshift of about 80 nm and the appearance of a zero-phonon line at low temperatures. It is reasonable to attribute this luminescence to perturbed Fe(O) centers. The observed d → d transitions in the excitation spectra and the resolved ZPL lines in emission spectra measured at low-temperature (below 80 K) are attributed to breaking of the parity-selection rule, which is forbidden for a strict Fe(O) but is broken due to perturbation. The broad excitation bands centering at 360 and 426 nm correspond to the transitions to ⁴E (⁴D) and ⁴T₂ (⁴D). Here, we consider the distorted Fe(O) center accompanied by a intrinsic defect, sharing two oxygen-ion ligands with Fe(O). This kind of perturbation leads to a red-shift in energies predicted at the same level (Table S2 of ESI†). Additionally, contributions of trace Cr³⁺ and Cr⁴⁺ impurities to the aforementioned NIR luminescence can be ruled out. In the case of deliberate Cr doping, only a broad-band luminescence peak at 848 nm, attributable to Cr³⁺ at octahedral sites with a weak ligand field,⁴³ was observed, while the longer-wavelength IR luminescence potentially attributable to Cr⁴ was not detected (see Fig. S7 of the ESI†). It has been reported that Cr⁴⁺ at a tetrahedral Ge⁴⁺ site exhibits a broad ³T₂ → ³A₂ emission peak at around or even beyond 1200 nm due to the weaker ligand field, with short decay time from a few microseconds to no more than a few hundred microseconds due to its spin-allowed nature.⁴⁴ The trace Ni²⁺ impurity is also ruled out, as the possible NIR luminescence of Ni²⁺ at octahedral Sc³⁺ sites in garnets should be around or even beyond 1550 nm according to the ligand-field strength based on Tanabe–Sugano theory.^45,46

Fig. 4(k) plots the integrated total emission intensity of Fe(T) (λ_emi = 720 nm) and Fe(O) (λ_emi = 990 nm) as a function of temperature. The stronger temperature dependence of the Fe(O) emission is due to relaxation to the ground state via the ²T₂ intermediate energy level, which is populated via thermal activation. Fig. 4(l) illustrates the NIR emission decay lifetimes as a function of temperature. As the temperature increases from 5 K to 455 K, the luminescence decay lifetime of Fe(T) (λ_emi = 720 nm) only decreases slightly from 9.23 ms to 7.50 ms, while that of Fe(O) (λ_emi = 990 nm) decreases sharply from 13.50 ms to 0.48 ms. The luminescence decay lifetime of octahedral Fe³⁺ luminescence (Fe(O)) is longer than that of the tetrahedral Fe³⁺ luminescence (Fe(T)) at low temperatures, as anticipated from the additional parity-forbidden rule for the former. When the ligand field is strong enough, the ²T₂ level shifts downwards to approach or even become lower than the ⁴T₁ level of Fe(O), playing an important role in nonradiative relaxation. This leads to a sharp decrease in the luminescence decay time of Fe(O) as temperature increases, or even quenching of the luminescence altogether.

2.3 Potential applications and future outlook

The SSGG:Fe³⁺ samples show a dramatic temperature-dependent dual-band NIR luminescence, prompting us to explore their potential as LIR thermometers. The temperature-dependent PL spectra, shown in Fig. 4, cover a temperature range of 5–455 K. The LIR (I_{720 nm}/I_{990 nm}) as a function of temperature is plotted in Fig. 5(a) upon 274 and 302 nm excitation separately, and the relative temperature sensitivity defined as S_r = |dln(LIR)/dT—for the two LIRs—is plotted in Fig. 5(b). The maximum sensitivities for the two LIRS reach 0.91% K⁻¹ at 230 K under 274 nm excitation and 0.90% K⁻¹ at 330 K under 302 nm excitation. By switching between the two excitation, high sensitivity across a broad temperature range can be achieved with SSGG:Fe³⁺, demonstrating its potential as a dual-modal LIR thermometer.

Fig. 5 (a) Luminescence intensity ratio LIR = I_{720 nm}/I_{990 nm}versus temperature T upon 274 nm (red) and 302 nm (blue) excitation in SSGG:1%Fe³⁺. (b) Relative temperature sensitivity S_r of the LIR thermometer. (c) Luminescence decay lifetime of the 990 nm emission as a function of temperature T. (d) Temperature-dependent S_a and S_r values of the lifetime-based thermometer.

The temperature quenching behaviors of the luminescence of SSGG:Fe³⁺ can be investigated for thermometry based on luminescence decay lifetime. Fig. 5(c) illustrates the luminescence decay lifetime of 990 nm emission as a function of temperature in a broad range of 5 K–455 K, and Fig. 5(d) displays the temperature dependence of S_a = |dτ/dT| and S_r = |dln(τ)/dT|. Typical lifetime-based thermometers are referenced in Table S3.† The high maximum relative sensitivity of 1.24% K⁻¹ at 155 K achieved in this study aligns with those values reported using other materials in the literature.

As a photoluminescent material with excellent tunable broadband near-infrared emission, SSGG:Fe³⁺ can be further explored towards its potential applications in night vision imaging, non-destructive detection of organic molecules, and anti-counterfeiting (Fig. S8†). The internal quantum efficiency and external quantum efficiency of an optimal sample SSGG:3%Fe³⁺ is 14.1% and 9.6%, respectively, when excited with a 274 nm light (Fig. S9†).

The theoretical predictions and experimental realizations in this study are expected to provide insights for the design and optimization of near-infrared materials based on garnet matrices and iron activator ions for subsequent applications.

3 Conclusions

In summary, this study successfully demonstrates the potential of Fe³⁺-doped SSGG, SLGG, and SYGG garnets as dual-band near-infrared (NIR) phosphors. Theoretical analysis and first-principles calculations indicate that Fe³⁺ ions can occupy both tetrahedral and octahedral sites. The energy-level calculations show that both centers can emit in the NIR range via⁴T₁ → ⁶A₁ transitions, while the ²T₂ state serves as an effective quenching path for the octahedral Fe³⁺ at elevated temperatures. Experimental validation confirms dual-band NIR emissions at about 730 nm and 980 nm in the three iron-doped garnets. The optical properties of these phosphors, studied over a wide temperature range, reveal significantly different temperature-dependent behavior of the tetrahedral and octahedral Fe³⁺ centers, consistent with our theoretical analysis of the quenching mechanism. These findings highlight the potential applications of these garnet phosphors in luminescence intensity ratio thermometers and luminescence decay time thermometry, with a high maximum relative sensitivity of 1.24% K⁻¹ at 155 K. The materials designed here provide a foundation for related application explorations, and the strategy developed can be applied to the exploration and development of Fe³⁺-activated advanced optical materials.

Data availability

The authors confirm that the data supporting the findings of this study are available within the article and its ESI.†

Conflicts of interest

The authors declare no conflict of interest.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (Grant No. 12474242), the Science and Technology Research Program of Chongqing Municipal Education Commission (Grant No. KJQN202300652), the Postgraduate Scientific Research and Innovation Project of Chongqing (Grant No. CYS23438), and the Innovation Program for Quantum Science and Technology (Grant No. 2021ZD0302200). The numerical calculations were partially performed on the supercomputing system at the Supercomputing Center of the University of Science and Technology of China.

References

1. J. Qiao, G. Zhou, Y. Zhou, Q. Zhang and Z. Xia, Divalent Europium-Doped Near-Infrared-Emitting Phosphor for Light-Emitting Diodes, Nat. Commun., 2019, 10, 5267.
2. H. Li, Y. Wang and L. Liao, Near-Infrared Luminescent Materials Incorporating Rare Earth/Transition Metal Ions: From Materials to Applications, Adv. Mater., 2024, 36, 2403076.
3. Y. Ma, Z. Wang, Y. Wang, Z. Liu, Y. Wang and R. Lv, Coumarin derivative dye sensitized NaYGdF₄:Yb,Er nanoparticles with enhanced NIR II luminescence for bio-vascular imaging, J. Rare Earths, 2023, 41, 1843.
4. J. J. Joos, D. Van der Heggen, L. I. Martin, L. Amidani, P. F. Smet, Z. Barandiarán and L. Seijo, Broadband infrared leds based on europium-to-terbium charge transfer luminescence, Nat. Commun., 2020, 11, 3647.
5. F. Zhao, Z. Song and Q. Liu, Advances in chromium-activated phosphors for near-infrared light sources, Laser Photonics Rev., 2022, 16, 2200380.
6. F. Zhao, Z. Song and Q. Liu, Advances in Fe³⁺-activated luminescent materials for near-infrared light sources, Prog. Solid State Chem., 2024, 74, 100456.
7. D. Liu, G. Li, P. Dang, Q. Zhang, Y. Wei, L. Qiu, M. S. Molokeev, H. Lian, M. Shang and J. Lin, Highly efficient Fe³⁺-doped A₂BB′O₆ (A = Sr²⁺, Ca²⁺; B, B^′ = In³⁺, Sb⁵⁺, Sn⁴⁺) broadband near-infrared-emitting phosphors for spectroscopic analysis, Light:Sci. Appl., 2022, 11, 112.
8. L. Yan, G. Zhu, S. Ma, S. Li, Z. Li, X. Luo and B. Dong, Emerging Fe³⁺ Doped Broad NIR-Emitting Phosphor Ca_2.5Hf_2.5 (Ga, Al)₃ O₁₂: Fe³⁺ for LWUV Pumped NIR LED, Laser Photonics Rev., 2024, 18, 2301200.
9. Q. Chen, Q. Quan and C.-K. Duan, Uncovering the Presence or Absence of Photoluminescence from Iron Ions in Crystals, Phys. Rev. Mater., 2024, 8, 095201.
10. Q. Chen, C. Ji and C.-K. Duan, Luminescence of iron ions in crystals: Site occupancy, valence states, and excited-state properties, Phys. Rev. B, 2024, 109, 165124.
11. K. Kniec, W. Piotrowski, K. Ledwa, L. D. Carlos and L. Marciniak, Spectral and Thermometric Properties Altering through Crystal Field Strength Modification and Host Material Composition in Luminescence Thermometers Based on Fe³⁺ Doped AB₂O₄ Type Nanocrystals (A = Mg, Ca; B = Al, Ga), J. Mater. Chem. C, 2021, 9, 517.
12. Z. Zhou, H. Jiang, J. Wei, Z. Fei, B. Yin, J. Qiu, Z. Yang and G. Dong, Achieving Spectrally Tunable and Thermally Stable Near-Infrared Emission in Fe³⁺-Activated Spinel Phosphors via the Cation Site Modulation Strategy, J. Mater. Chem. C, 2024, 12, 13268.
13. Z. Zhou, S. Zhang, Y. Le, H. Ming, Y. Li, S. Ye, M. Peng, J. Qiu and G. Dong, Defect Enrichment in Near Inverse Spinel Configuration to Enhance the Persistent Luminescence of Fe³⁺, Adv. Opt. Mater., 2021, 10, 2101669.
14. Y. Ye, Y. Ding, H. Yang, Q. Mao, L. Pei, M. Liu and J. Zhong, Boosting Near-Infrared Emission in Spinel-Type Phosphor via Oxygen Vacancy Engineering for Versatile Application, Adv. Funct. Mater., 2024, 2405048.
15. X. Lu, Y. Wang, J. Yang, P. Townsend and D. Hreniak, LiGa₅O₈: Fe³⁺: A Novel and Super Long Near-Infrared Persistent Material, Ceram. Int., 2024, 50, 35359–35367.
16. M. Li, Y. Jin, L. Yuan, B. Wang, H. Wu, Y. Hu and F. Wang, Near-Infrared Long Afterglow in Fe³⁺-Activated Mg₂SnO₄ for Self-Sustainable Night Vision, ACS Appl. Mater. Interfaces, 2023, 15, 13186.
17. G. Liu, S. Zhang and Z. Xia, Multi-Sites Energy Transfer in Fe³⁺-Doped KAl₁₁O₁₇ Phosphor Toward Zero Thermal Quenching Near-Infrared Luminescence, Opt. Lett., 2023, 48, 1296.
18. K. Cheng, W. Huang, X. Liu, X. Gong and C. Deng, Cr³⁺-Free Broadband Near-Infrared Phosphors NaAl₅O₈:Fe³⁺, J. Alloys Compd., 2023, 964, 171240.
19. Y. Yang, L. Shen, J. Zhang, S. Zhao, Q. Pang, X. Zhang, P. Chen and L. Zhou, Tetracoordinate Fe³⁺ Activated Li₂ZnAO₄ (A = Si, Ge) Near-Infrared Luminescent Phosphors, Inorg. Chem., 2023, 62, 12862.
20. J. Su, R. Pang, T. Tan, S. Wang, X. Chen, S. Zhang and H. Zhang, A novel near-infrared emitting Sr₂LuSbO₆:Fe³⁺ phosphor with persistent luminescence performance, Adv. Opt. Mater., 2024, 12, 2303187.
21. M. Yeerlan, M. Zhang and P. Dai, Near-unity quantum efficiency wavelength-tunable NIR phosphors (Sr_2(yCa_y)Sc_1(xSbO₆: xFe³⁺ with excellent thermal stability, J. Am. Ceram. Soc., 2024, 107, 5588.
22. X. Zhang, D. Chen, X. Chen, C. Zhou, P. Chen, Q. Pang and L. Zhou, Broadband near-infrared luminescence from Fe³⁺-activated NaScSi₂O₆ phosphors for luminescence thermometry and night-vision applications, Dalton Trans., 2022, 51, 14243.
23. F. Zhao, Y. Shao, Z. Song and Q. Liu, Structural confinement toward suppressing concentration and thermal quenching for improving near-infrared luminescence of Fe³⁺, Inorg. Chem. Front., 2023, 10, 6701.
24. S. Hariyani, M. Sójka, A. Setlur and J. Brgoch, A guide to comprehensive phosphor discovery for solid-state lighting, Nat. Rev. Mater., 2023, 8, 759.
25. P. Bolek, T. van Swieten, J. Zeler, A. Meijerink and E. Zych, Luminescence Thermometry of Pr³⁺-Doped Sr₃Y₂Ge₃O₁₂ and Sr₃Sc₂Ge₃O₁₂ Submicron Garnets Spanning the 13–1025 K Range and New Insight to Their Spectroscopy, Chem. Mater., 2024, 36, 8894.
26. Z. Xia and A. Meijerink, Ce³⁺-Doped garnet phosphors: composition modification, luminescence properties and applications, Chem. Soc. Rev., 2017, 46, 275.
27. D. Wang, L. Yan, G. Zhu, S. Ma, N. Zhou, S. Li, Z. Li, Y. Cong, X. Bai, X. Han and B. Dong, Achieving Flat Ultra-Broadband NIR Emission in Cr³⁺ Doped Garnet-Type Phosphors Enabled by Structural Regulation Toward Multi-Functional Spectroscopy Applications, Laser Photonics Rev., 2024, 2401256.
28. J. Xu, J. Ueda and S. Tanabe, Toward tunable and bright deep-red persistent luminescence of Cr³⁺ in garnets, J. Am. Ceram. Soc., 2017, 100, 4033.
29. G. Blasse, J. de Blank and D. Ijdo, The Luminescence of the Garnet Ca₄ZrGe₃O₁₂, Mater. Res. Bull., 1995, 30, 845–850.
30. K. Kniec, K. Ledwa, K. Maciejewska and L. Marciniak, Intentional modification of the optical spectral response and relative sensitivity of luminescent thermometers based on Fe³⁺, Cr³⁺, Nd³⁺ co-doped garnet nanocrystals by crystal field strength optimization, Mater. Chem. Front., 2020, 4, 1697.
31. Y. Wang, Q. Bu, D. Li, S. Yang, L. Li, G. Xiang, S. Jiang, Y. Chang, C. Jing, X. Zhou, L.-I. Bulyk and A. Suchocki, UV- and X-ray-activated broadband Nir garnet-type Ca₃Ga₂Sn₃O₁₂:Fe³⁺ phosphors with efficient persistent luminescence, Dyes Pigm., 2024, 225, 112091.
32. R. Zhao, X. Guo, J. Zhang and R. Cui, Near-infrared luminescent properties and applications of Fe³⁺-doped YAG phosphors, J. Lumin., 2024, 267, 120352.
33. Y. Tang, Y. Cai, K. Dou, J. Chang, W. Li, S. Wang, M. Sun, B. Huang, X. Liu, J. Qiu, L. Zhou, M. Wu and J.-C. Zhang, Dynamic multicolor emissions of multimodal phosphors by Mn²⁺ trace doping in self-activated CaGa₄O₇, Nat. Commun., 2024, 15, 3209.
34. S. Li, P. Zhang, Z. Wang, Y. Yang, X. Miao and W. Liu, Multicolor and multimodal luminescence in an Er³⁺ single-doped double perovskite for advanced anti-counterfeiting and encryption, Inorg. Chem. Front., 2024, 11, 4721.
35. J. Zhong, D. Chen, Y. Peng, Y. Lu, X. Chen, X. Li and Z. Ji, A Review on Nanostructured Glass Ceramics for Promising Application in Optical Thermometry, J. Alloys Compd., 2018, 763, 34.
36. D. Chen, Z. Wan, Y. Zhou, X. Zhou, Y. Yu, J. Zhong, M. Ding and Z. Ji, Dual-phase glass ceramic: Structure, dual-modal luminescence, and temperature sensing behaviors, ACS Appl. Mater. Interfaces, 2015, 7, 19484.
37. Y. Fan, P. Wang, Y. Lu, R. Wang, L. Zhou, X. Zheng, X. Li, J. A. Piper and F. Zhang, Lifetime-Engineered NIR-II Nanoparticles Unlock Multiplexed In Vivo Imaging, Nat. Nanotechnol., 2018, 18, 941–946.
38. K. Liu, X. Liu, Q. Zeng, Y. Zhang, L. Tu, T. Liu, X. Kong, Y. Wang, F. Cao, S. A. G. Lambrechts, M. C. G. Aalders and H. Zhang, Covalently Assembled NIR Nanoplatform for Simultaneous Fluorescence Imaging and Photodynamic Therapy of Cancer Cells, ACS Nano, 2012, 6, 4054.
39. L. Li, H. Yang, Y. Wang, X. Zhou, F. Ling, H. Gao, Z. Yang, X. Tang and Y. Hua, Cr³⁺-Yb³⁺ energy transfer drives ultra-wideband thermally-stable NIR-emitting Sr₃Sc₂Ge₃O₁₂ garnet phosphors for pc-LED application, Ceram. Int., 2024, 50, 32662.
40. X. Sun, J. Li, K. Feng, R. Zheng and H. Yuan, Luminescence characteristics of Bi³⁺, Cr³⁺ and Bi³⁺/Cr³⁺ activated Sr₃Y₂Ge₃O₁₂ phosphors, J. Lumin., 2022, 248, 118984.
41. U. Guth, Kröger-Vinks Notation of Point Defects, in Encyclopedia of Applied Electrochemistry, Springer New York, 2014, pp. 1159–1160.
42. X. Zhou, X. Luo, B. Wu, S. Jiang, L. Li, X. Luo and Y. Pang, The broad emission at 785 nm in YAG:Ce³⁺,Cr³⁺ phosphor, Spectrochim. Acta, Part A, 2018, 190, 76.
43. P. Dang, Y. Wei, D. Liu, G. Li and J. Lin, Recent advances in chromium-doped near-infrared luminescent materials: Fundamentals, optimization strategies, and applications, Adv. Opt. Mater., 2023, 11, 2201739.
44. X. Feng, L. Lin, R. Duan, J. Qiu and S. Zhou, Transition metal ion activated near-infrared luminescent materials, Prog. Mater. Sci., 2022, 129, 100973.
45. E. Zannoni, E. Cavalli, A. Toncelli, M. Tonelli and M. Bettinelli, Optical spectroscopy of Ca₃Sc₂Ge₃O₁₂:Ni²⁺, J. Phys. Chem. Solids, 1999, 60, 449.
46. J. Koetke, G. Huber and K. Petermann, Spectroscopy of Ni²⁺-doped garnets and perovskites for solid state lasers, J. Lumin., 1991, 48–49, 564.

---

Footnotes

† Electronic supplementary information (ESI) available: Computational settings and discussions on the defect formation energies in the series garnets; experimental details; XRD patterns; PL spectra and luminescence decay curves of Cr-doped and various Fe-doped samples; anti-counterfeiting experiments; quantum efficiency measurements. See DOI: https://doi.org/10.1039/d4qi02523g

‡ These authors contributed equally to this work, from experimental and theoretical aspects, respectively.

---