Introduction

Near-infrared (NIR) light sources, operating within the 700–1000 nm spectrum, have garnered considerable interest across various fields, including biological imaging, non-destructive testing, plant growth illumination, and night vision technologies1,2,3,4. Recent research efforts have particularly focused on the development of NIR phosphors, with a special emphasis on those that can be efficiently activated by blue light-emitting diodes (LEDs). Once these blue LED-compatible and highly efficient NIR phosphors are realized, the well-established phosphor-converted white-LEDs (pc-WLEDs) technology can be harnessed to miniaturize NIR light sources. This could usher in a new era of intelligent, portable applications while addressing the inefficiencies, bulkiness, and short lifespans of conventional NIR light sources like incandescent bulbs and halogen lamps1,5,6,7.

Currently, the activator ions capable of emitting in the NIR region include rare earth ions (Pr3+, Tm3+, Yb3+, Nd3+, Dy3+, and Eu2+)6,8,9,10,11,12,13,14, transition metal ions (Ni2+, Mn2+, Fe3+, and Cr3+)1,5,6,7,15,16,17,18,19,20, and bismuth (Bi)21. Each of these activators offers distinct advantages, but also comes with inherent limitations, such as a narrow emission spectrum, restricted absorption, or low luminescence efficiency—all of which present significant challenges to overcome. Among these, Cr3+ ions remain a preferred option since they are capable of absorbing blue light (4A2g → 4T1g(4F)) and emitting broadband NIR light (4T2g(4F) → 4A2g(4F)) when positioned at octahedra sites with a weak crystal field strength. In recent years, considerable research endeavors have been directed towards the development of high-performance NIR phosphors centered on Cr3+, yielding continuous progress, especially in boosting internal and external quantum efficiencies (IQE/EQE). Notable examples1,22,23,24,25 include Ca3Sc2Si3O12: Cr3+ (λex = 460 nm, λem = 478 nm, IQE/EQE = 92.3%/25.5%), Y3Ga3MgSiO12:Cr3+ (λex = 438 nm, λem = 782 nm, IQE/EQE = 79.9%/33.7%), Gd3Ga4AlO12:Cr3+ (λex = 448 nm, λem = 730 nm, IQE/EQE = 85%/44%), and Mg4Ta2O9:Cr3+ (λex = 448 nm, λem = 730 nm, IQE/EQE = 87.5%/61.25%). Nevertheless, these NIR emissions arise from partially forbidden d–d transitions of Cr3+, characterized by a small absorption cross-section (ranging from 10−19 to 10−20 cm2)2,26,27, thereby posing a challenge in achieving exceptionally high EQE for NIR phosphors. In contrast, the rare earth Ce3+ ion benefits from spin-allowed 4f–5d transitions, facilitating the development of phosphors with exceptionally high quantum efficiencies (QEs). This is exemplified by the commercially successful yellow phosphor Y3Al5O12:Ce3+ (YAG:Ce3+), which boasts an EQE of up to 70%28.

The most widely utilized commercial pc-WLEDs was produced using an InGaN-based blue LEDs chip that was coated with the yellow phosphor YAG:Ce3+. Nevertheless, the pc-WLEDs possesses inherent limitations, primarily the absence of red emission components, which results in white light exhibiting high correlated color temperatures (CCT > 4500 K) and low color rendering indexes (Ra < 80)29. In recent years, extensive research efforts have been devoted to the design of innovative red phosphors that harness the Ce3+–Mn2+ energy transfer. The robust absorption of visible light exhibited by Ce3+ effectively addresses the challenges posed by the weak absorption capabilities of Mn2+ ions27. Furthermore, unlike the Ce3+–Cr3+ system30,31, the Ce3+–Mn2+ system is devoid of luminescence quenching traps, rendering it a highly appealing candidate for enhancing the red emission component in pc-WLEDs. If the existence of Mn2+-activated NIR emission is verified, the establishment of an energy transfer from Ce3+ to Mn2+, along with the precise manipulation of the local crystal field surrounding Mn2+, could pave the path towards the development of highly efficient NIR phosphors. However, a fundamental comprehension of the origins underlying NIR emission for Mn2+ remains an essential prerequisite for attaining this objective.

Transition metal ions, specifically Mn2+ and Cr3+, have played a pivotal role in the progress of luminescent materials. Notably, Mn2+ demonstrates an exceptional capacity to emit green, red, and potentially even NIR light, contingent upon its specific local crystal field environment. In phosphors32,33,34 including BaZnAl10O17:Mn2+ (λex = 450 nm, λem = 517 nm), Sr2MgAl22O36:Mn2+ (λex = 450 nm, λem = 518 nm), and Na2MgSiO4:Mn2+ (λex = 450 nm, λem = 520 nm), the presence of Mn2+ occupying tetrahedral sites contributes significantly to the green emission. Conversely, red emission is discernible when Mn2+ ions occupy octahedral sites, as exemplified in compounds35,36,37 like NaScSi2O6:Mn2+ (λex = 410 nm, λem = 654 nm), Ca18K3Sc(PO4)14:Mn2+ (λex = 405 nm, λem = 640 nm), and Na2Mg2Si6O15:Eu2+, Mn2+ (λex = 365 nm, λem = 630 nm). Nevertheless, the underlying mechanisms responsible for the NIR emission triggered by Mn2+ continue to be a subject of intense discussion. Earlier hypotheses ascribed the NIR emission to Mn2+ ions occupying cubically coordinated sites38,39, whereas contemporary perspectives favor the coupling effect arising from Mn2+–Mn2+ pairs16,40,41,42,43. However, Meijerink et al. have expressed skepticism towards the latter perspective44, contending that Mn2+–Mn2+ coupling does not significantly contribute to the red-shifted emissions. Instead, it is hypothesized that defects or the oxidation of Mn2+ to Mn3+ underlie the observed mechanisms.

Manganese (Mn) exhibits remarkable versatility in assuming a wide range of valence states45, ranging from +1 to +7, with Mn2+ ions demonstrating a particular proficiency in adapting to local structure characterized by four-, six-, or eight-fold coordination configurations. Notably, a significant number of Mn2+-activated NIR phosphors, including MnAl2O416, α-MnS41, and KMnF342 display a unique behavior where Mn2+ ions occupy neighboring sites. These distinguishing features pose a significant hurdle in deciphering the underlying mechanism governing the NIR emission exhibited by Mn2+ ions. Fortunately, Mn2+ exhibits both NIR and red emissions within the garnet structure, thereby offering an invaluable opportunity to delve deeper into the origins of NIR luminescence in Mn2+ ions. Within the garnet structure, we can not only scrutinize the valence states of manganese (Mn) but also rigorously explore the fundamental causes of its NIR and red emissions. Moreover, the establishment of a Mn2+–Mn2+ model within the garnet structure for first-principles calculations greatly aids in validating the ion pair coupling effects. Leveraging our previous achievements27, where the Ce3+–Mn2+ energy transfer within the garnet facilitated ultra-wideband emission, we have broadened the application spectrum of single-matrix pc-WLEDs. Our ongoing endeavor leverages the venerable garnet-based phosphor YAG:Ce3+ as a fundamental building block in establishing a Ce3+–Mn2+ energy transfer system, ultimately yielding an Mn2+-activated garnet phosphor that predominantly emits in the NIR region. The insights gained from our research not only enhance our understanding NIR emission of Mn2+ but also pave promising paths for future investigative endeavors aimed at developing NIR phosphors.

Results

Phase analysis

The phase purity of YAG:2%Ce3+, xMn2+, with x varying from 0% to 20%, has been extensively analyzed through meticulous XRD analysis and Rietveld refinement. Figure 1a and Table 1, along with Fig. S1 and Table S1 (Supporting Information), provide comprehensive details and understanding of the purity characteristics of these samples. Remarkably, the Rietveld refinement outcomes for YAG doped with 2%Ce3+ and 20%Mn2+ ions (Rp = 1.46%, Rwp = 1.93%, χ2 = 2.12) unequivocally validate the successful integration of Ce3+ and Mn2+ ions into the YAG crystal lattice. Nevertheless, when the Mn2+ content exceeds x = 20%, a minor impurity phase appears in the YAG structure, as clearly illustrated in Fig. S2. As a result, our meticulous investigation has narrowed down our focus to the specific composition of YAG:2%Ce3+, xMn2+, where x ranges from 0% to 20%, in order to accurately determine the origin of NIR emission. Furthermore, to facilitate the analysis of the luminescence phenomenon in YAG:2%Ce3+, 20%Mn2+, we also synthesized YAG:20%Mn2+ and Y2Mg0.92Mn0.08Al4SiO12 for QEs and coordination number analysis, as detailed in the following text. Their diffraction patterns are depicted in Fig. S3.

Fig. 1: Phase analysis of YAG:2%Ce3+, 20%Mn2+.
figure 1

a Rietveld refinements for XRD pattern; b the crystal structure and the coordination environments; c STEM–HAADF images; d HRTEM images; e, f CL spectra at points I, II and III and corresponding CL image; g, h SEM image, and EDS elemental mapping images

Full size image
Table. 1 The Rietveld refinement data of YAG:2%Ce3+, 20%Mn2+
Full size table

YAG exhibits a cubic Ia \(\bar{3}\) d symmetry, as shown schematically in Figs. 1b and S4, encompassing three distinct cation sites: the distorted dodecahedral YO8 with C2 point symmetry, the octahedral AlO6 with C3d point symmetry, and the tetrahedral AlO4 with S4 point symmetry. These cation sites collaborate to intricately forge the three-dimensional garnet framework through shared edges and vertices. Given the significant doping of Mn2+ at a concentration of 20%, high-resolution scanning transmission electron microscopy (STEM), in conjunction with high-angle annular dark-field (HAADF) imaging techniques, was utilized to meticulously investigate the atomic-scale architecture of YAG:2%Ce3+, 20%Mn2+. As clearly depicted in Fig. 1c, the vivid spots accurately pinpoint the precise locations of cation sites within a highly structured and well-defined garnet lattice, conspicuously free of any perceptible defects.

High-resolution transmission electron microscopy (HRTEM) images offer a revealing glimpse into micrometer-sized YAG:2%Ce3+, 20%Mn2+ phosphors, exhibiting irregular morphologies as depicted in Fig. S5. Notably, the observed lattice fringe spacing of 3.08 Å, as shown in Fig. 1d, likely corresponds to the (1 2 3) plane of the YAG structure. Figures 1e and S6 comprehensively present cathodoluminescence (CL) spectra, scanning electron microscopy (SEM) imagery, and energy-dispersive X-ray spectroscopy (EDS) elemental mappings of randomly sampled particles extracted from this phosphor. Despite variations in CL intensity, both brightly illuminated regions (I) and those with more muted luminescence (II, III) consistently exhibit a characteristic NIR emission that peaks at approximately 750 nm. The EDS mapping highlights the uniform distribution of Cerium (Ce), Manganese (Mn), silicon (Si), Yttrium (Y), Aluminum (Al), and Oxygen (O) throughout the particle, further corroborating the homogeneity of the doped system.

Collectively, these phase analyses confirm that Ce3+ and Mn2+ can be seamlessly and uniformly incorporated into the YAG crystal lattice without causing any structural degradation or generating impurities and defects.

Valence state, steady-state and transient photoluminescence characteristics

To unequivocally determine the valence state of Mn within YAG matrix, we employed a comprehensive approach that encompassed X-ray absorption near edge structure (XANES) analysis, X-ray photoelectron spectroscopy (XPS), and electron paramagnetic resonance (EPR) techniques on the YAG:2%Ce3+, 20%Mn2+ samples. Drawing parallels from our previous inquiries27, the valence state of Mn within the Lu2BaAl4SiO12 garnet structures exhibited a striking resemblance to the near-edge absorption pattern characteristic of MnO (Mn2+), as revealed through XANES analysis. Similarly, Fig. 2a demonstrates that the Mn absorption edge in YAG is intermediate between the absorption edges of Mn2O3 (Mn3+) and Mn foil, indicating the retention of a positive divalent state for Mn (e.g., Mn2+) in the YAG structure.

Fig. 2: Analysis of the valence state, steady-state and transient photoluminescence characteristics were conducted.
figure 2

a–c This includes XANES spectra, a high-resolution XPS spectrum for Mn-2p, and EPR spectra recorded at 77 K, all specific to Mn within the sample of YAG doped with 2%Ce3+ and 20%Mn2+; d Additionally, PLE and PL spectra were obtained for YAG doped with 2% Ce3+, YAG doped with 20%Mn2+, and YAG co-doped with both 2%Ce3+ and 20%Mn2+; e PL spectra were also recorded for YAG doped with 2%Ce3+ and varying concentrations of Mn2+ (ranging from x = 2% to 16%); f, g Furthermore, the fluorescence decay curves of Mn2+ in YAG doped with 2% Ce3+ and varying concentrations of Mn2+ (x = 2% to 20%) upon pulsed 450 nm excitation, monitored at 600 nm and 750 nm, are presented. Detailed fitting equations and parameters for these decay curves are outlined in Tables S2–S4

Full size image

The XPS spectra in Figs. 2b and S7 exhibit distinct peaks at 653.6 eV and 641.6 eV, respectively. These peaks can be attributed to the characteristic Mn 2p1/2 and Mn 2p3/2 transitions, which are associated with Mn2+ ions46. The EPR spectrum in Fig. 2c clearly exhibits a well-resolved sextet pattern, indicative of Mn2+, resulting from the transition Ms = |–1/2> to |+1/2>, which is superimposed with the hyperfine interaction between the electron spin and the 55Mn nucleus (I = 5/2). Furthermore, the isotropic g-value, determined to be approximately 2.02 based on the spin Hamiltonian (Eq. (S1) in the Supporting Information), suggests the presence of Mn2+ ions occupying octahedral symmetry sites within the YAG structure47. Moreover, as clearly illustrated in Fig. 2d, the photoluminescence excitation (PLE) spectrum of YAG:20Mn2+ prominently displays a characteristic d–d transition of Mn2+ ions, manifested by the NIR emission peak centered around ~750 nm. Remarkably, a narrow band located approximately at ~400 nm arises from the transition of 6A16(S) to [4A1(4G), 4E(4G)] states, which is the most prominent attribute of Mn2+ ions48. Drawing upon the converging evidence from XANES, XPS, EPR, and the spectral findings presented in Fig. 2d, we can unequivocally establish that the valence state of Mn ions in the YAG host lattice is divalent, specifically Mn2+.

Generally, the emission of Ce3+-activated oxide phosphors primarily occurs within the violet to blue spectral regions49,50,51. Notably, the Ce3+-doped YAG phosphor exhibits a distinct characteristic due to the unique positioning of Ce3+ within a dodecahedral YO8 site. In this configuration, the substantial overlap of electron clouds from the eight surrounding oxygen ions significantly enhances the Ce–O covalency, resulting in a pronounced centroid shift. Furthermore, the distorted YO8 configuration significantly enhances the splitting between the eg and t2g energy levels of Ce3+ in comparison to octahedral or cubic configurations, thereby facilitating the robust absorption of blue light and emission of yellow light in YAG:Ce3+51. Figure 2d illustrates comparative PLE and photoluminescence (PL) spectra for YAG doped with 2%Ce3+, 20%Mn2+, and a combination of 2%Ce3+ and 20%Mn2+. Specifically, YAG doped with 2%Ce3+ exhibits a prominent yellow emission peak centered precisely at 540 nm, arising from transitions involving the lowest 5d state of Ce3+ to its 2F5/2 and 2F7/2 energy levels. Notably, the significant overlap observed between the PL spectrum of Ce3+ and the PLE spectrum of Mn2+ suggests a potential energy transfer from Ce3+ to Mn2+ within the YAG lattice52. Upon excitation of Ce3+ ions at approximately 450 nm within the YAG:2%Ce3+, 20%Mn2+ phosphor, dual emissions are observed: one emanating from Ce3+ at approximately 540 nm and another originating from Mn2+ at approximately 600 nm and 750 nm. Furthermore, as illustrated in Figs. S8, S9, Table S2 and Eqs. (S2)–(S4) in the Supporting Information, the fluorescence lifetime of Ce3+ decreases gradually from 56 ns (at x = 0) to 8 ns (at x = 20%). These observation confirm the occurrence of energy transfer from Ce3+ to Mn2+ and suggests that the energy transfer efficiency (ηT) approaches 86% at x = 20% within the YAG matrix. Thanks to the assistance of Ce³⁺ ions, Mn²⁺ effectively captures blue light (~450 nm), achieving an absorption efficiency (AE) of up to 60%. As shown in Fig. 2e, at lower Mn2+ doping concentrations ranging from 2% to 8%, the primary Mn2+ emission in YAG is predominantly situated in the red region (~600 nm). Importantly, when the Mn2+ doping level exceeds 16%, the NIR emission (~750 nm) gradually becomes the dominant feature in the PL spectrum. To more clearly illustrate the impact of Mn2+ doping (x) on the energy distribution of emission spectra, we conducted peak fitting on the emission spectra of YAG:2%Ce3+, xMn2+ (where x ranges from 2% to 20%). Additionally, as demonstrated in Figs. 2d, e and S10, we calculated the proportions of the spectral integral intensities of Ce3+ (~540 nm), Mn2+ (~600 nm), and Mn2+ (~750 nm) within the overall emission spectrum.

Within the context of YAG:Ce3+, Mn2+, a prime opportunity emerges to delve deeper into the genesis of NIR emission emanating from Mn2+ ions. Initially, it is unlikely that the NIR emission originates from the coupling of Mn2+–Mn2+ ion pairs, as the generation of Mn2+–Mn2+ pair effects leading to the transition |6A1,4T1> → |6A1,6A1> typically necessitates the doping of higher concentrations of Mn2+ ions into the crystal lattice16,40,41,42,43. As shown in Fig. 2e, it is evident that the NIR emissions from Mn2+ ions are discernible even at a minimal doping concentration of 2%. Typically, during the preparation stage, raw materials enriched with Mn2+ undergo meticulous grinding procedures before being subjected to high-temperature calcination. This process results in Mn2+ being disseminated throughout the YAG lattice in a random yet uniform manner. Consequently, at lower concentrations of doping, the occurrence of Mn2+–Mn2+ ion pairing diminishes notably, suggesting that the observed NIR emission in Mn2+-doped YAG is unlikely to stem from the interaction between Mn2+–Mn2+ ion pairs. We can also categorically eliminate Mn3+ as the source of NIR emission by extending the PL spectrum of YAG:2%Ce3+, 20%Mn2+ across the 500 to 1200 nm range. The most distinguishing characteristic of Mn3+ emissions within the garnet structure is the intense luminescence arising from the transition 1T2 → 3T2, which is centered precisely between 1100 nm and 1200 nm53. It is noteworthy that, as clearly depicted in Fig. S11, both the spectra of YAG:2%Ce3+, 20%Mn2+ at room temperature (RT) and at 77 K exhibit a distinct absence of detectable luminescence within the crucial 1100 nm to 1200 nm region, thus firmly validating our initial conclusion.

The time-resolved spectroscopic analysis of YAG doped with Ce3+ and Mn2+ ions reveals a strikingly rapid decay rate for the red emission (~590 nm). As clearly demonstrated in Fig. S12, following a 10 ms delay, only the NIR emission centered at approximately 750 nm remains, indicating that these two emissions originate from distinct luminescent centers within the YAG crystal structure54,55. Remarkably, as shown in Fig. 2f, g, the NIR emission demonstrates a fluorescence lifetime of approximately ~10 ms, significantly surpassing the ~1 ms observed for the red emission (~590 nm). This suggests that the NIR emission cannot be attributed to Mn2+–Mn2+ coupling within the red emission center, as such coupling is expected to relax the spin selection rule and thus shorten the fluorescence lifetime (see Fig. S13 and Eq. (S5))56. More precisely, the extended NIR emission lifetime (~10 ms) indicates Mn2+ residing in high-symmetry positions within the garnet crystal structure. As schematically illustrated in Figs. 1b and S4, the 8-fold coordination gives rise to a distorted dodecahedral YO8 configuration, whereas the 6-fold coordination corresponds to an octahedral AlO6 site, and the 4-fold coordination presents a tetrahedral AlO4 arrangement. By leveraging previous research that employed synchrotron radiation to elucidate the Mn2+ coordination in garnet structures, we know that Mn2+ ions occupy both 8- and 6-coordinated sites27. Therefore, we hypothesize that the NIR emissions shown in Fig. 2d, e originate from these Mn2+ ions residing within 6-coordinate octahedral AlO6 sites. As illustrated in Fig. S14, we believe that the transitions 4T1(4G) → 6A1(6S) of isolated Mn2+ ions occupying octahedral sites not only result in the red emission but potentially also initiate an NIR emission, attributed to the intensified splitting of d–d energy levels due to by the intense local crystal field of AlO6 configurations characterized by shorter Al–O bonds. Regarding the origin of the red emission (~600 nm) exhibited by YAG:Ce3+, Mn2+, we postulate that it arises from Mn2+ ions residing within the distorted dodecahedral YO8 sites. Despite the elongation of the Y–O bonds, the distortion of the YO8 environment fosters a robust crystal field, thereby enabling Mn2+ to emit red light upon occupancy. This mechanism is analogous to that observed in Ce3+ occupies the YO8 site of YAG structure, resulting in yellow emission51. In addition, the factors pertaining to the crystal field strength are mirrored in Eq. (S6) of the Supporting Information.

Forecasting spectroscopic characteristics through first-principles calculations

First-principles calculations are recognized as a robust method for forecasting the energy level structure and spectroscopic characteristics of Mn2+ ions57. To validate our hypothesis, as illustrated in Figs. 3, S15 and Table S5, we constructed four distinct computational models based on the local structure of YAG. These models encompass: an individual Mn2+ ion occupying an isolated octahedral site (Fig. 3a), designated as \({\rm{Mn}}^{2+}_{\rm{oct}}\); a solitary Mn2+ ion residing within a dodecahedral coordination sphere (Fig. 3b), labeled as \({\rm{Mn}}^{2+}_{\rm{dod}}\); a scenario wherein octahedral sites are adjacent to dodecahedral sites, creating an interconnected environment (\({\rm{Mn}}^{2+}_{\rm{oct}}{-}{\rm{Mn}}^{2+}_{\rm{dod}}\), Fig. 3c); and finally, a configuration featuring dodecahedral sites alongside their geometrically equivalent neighboring sites (\({\rm{Mn}}^{2+}_{\rm{dod}}{-}{\rm{Mn}}^{2+}_{\rm{dod}}\), Fig. 3d). Our calculations reveal the emission energies of the 4T1 → 6A1 transition for Mn2+ in isolated octahedral (\({\rm{Mn}}^{2+}_{\rm{oct}}\)) and dodecahedral (\({\rm{Mn}}^{2+}_{\rm{dod}}\)) sites are 1.48 eV and 1.87 eV, respectively. Notably, first-principles analysis underscores that the NIR emission in YAG:Mn2+ originates from Mn2+ ions occupying octahedral AlO6 sites (\({\rm{Mn}}^{2+}_{\rm{oct}}\)), while the red emission emanates from Mn2+ occupying distorted dodecahedral YO8 sites (\({\rm{Mn}}^{2+}_{\rm{dod}}\)). However, contrary to the prevalent belief, our findings do not corroborate the notion that Mn2+–Mn2+ pairs contribute to NIR emission in the YAG structure. The electronic environments surrounding these pairs closely resemble those observed in isolated \({\rm{Mn}}^{2+}_{\rm{oct}}\) and \({\rm{Mn}}^{2+}_{\rm{dod}}\) configurations. Furthermore, the emission energies of Mn2+–Mn2+ pairs, estimated at 1.57 eV for \({\rm{Mn}}^{2+}_{\rm{oct}}\) and 1.91 eV for \({\rm{Mn}}^{2+}_{\rm{dod}}\) in pair system, closely align with those of their isolated counterparts, emphasizing their similarity. While previous studies on Mn2+–Mn2+ pairs has shown that these paired ions can modulate the absorption spectra, fluorescence lifetime, and other characteristics of Mn2+56,58. There remains a lack of conclusive direct evidence to support a significant red-shift in the emission spectrum caused by Mn2+–Mn2+ pairs in the YAG structure.

Fig. 3: First-principles calculations for YAG:2%Ce3+, 20%Mn2+.
figure 3

Configuration coordinate diagrams for the ground state 6A1 and excited states 4T1 of Mn2+-doped YAG structure, along with the corresponding partial charge density distribution of the highest occupied orbital and the lowest unoccupied orbital, are presented. In this context, four distinct computational models are derived from the local structure of YAG. a Isolated Mn2+ ion occupies octahedron site (\({\rm{Mn}}^{2+}_{\rm{oct}}\)); b Isolated Mn2+ ion occupies dodecahedron site (\({\rm{Mn}}^{2+}_{\rm{dod}}\)); c Mn2+ ions occupy the adjacent octahedron and dodecahedron sites (\({\rm{Mn}}^{2+}_{\rm{oct}}{-}{\rm{Mn}}^{2+}_{\rm{dod}}\)); d Mn2+ ions occupy the adjacent two dodecahedron sites (\({\rm{Mn}}^{2+}_{\rm{dod}}{-}{\rm{Mn}}^{2+}_{\rm{dod}}\))

Full size image

XAS characterizations

X-ray absorption spectroscopy (XAS) provides an exceptional method for unveiling the intricate structural details of Mn2+ within the garnet structure, including its valence state, coordination number, bond length, and other relevant parameters. Our group’s prior investigations have unequivocally determined that the average coordination number of Mn2+ in the garnet structure is seven, indicating its presence in both six- and eight-coordinated sites27. This leads us to envision the creation of Mn2+-activated garnet structures tailored to exclusively exhibit either red or NIR emission centers. Then, through meticulous XAS analysis of their coordination numbers, we can directly pinpoint the origin of NIR emission in Mn2+-activated garnet structures. Fortunately, as shown in Fig. 2e, Mn2+ exhibits a preferential propensity for forming red luminescent centers in garnet structure. In this pursuit, as depicted in Fig. S3, we have achieved a successful substitution of Y3+–Al3+ in YAG lattice with Mg2+–Si4+. Additionally, we further substituted Mg2+ with Mn2+, ultimately yielding a Mn2+-activated garnet phosphor (Y2Mg0.92Mn0.08Al4SiO12) that exclusively exhibits red emission. The extended X-ray absorption fine structure (EXAFS) spectra, accompanied by fitting data for both Y2Mg0.92Mn0.08Al4SiO12 and YAG:2%Ce3+, 20%Mn2+, are presented in Fig. 4a–f, with corresponding data are detailed in Tables 2, S6 and Figs. S16, S17. It is noteworthy that the fitting outcomes attribute the red luminescence of Y2Mg0.92Mn0.08Al4SiO12 to Mn2+ occupying an 8-coordinate site (\({\rm{Mn}}^{2+}_{\rm{dod}}\)). Consequently, the dual red and NIR emission centers observed in YAG:2%Ce3+, 20%Mn2+ can be traced to Mn2+ occupying 8- and 6-coordinate sites, respectively, as illustrated in Fig. 4g, h and Table 2. It is crucial to note that the majority of Mn2+-activated NIR phosphors, as exemplified by MgAl2O4:Mn2+, ZnAl2O4:Mn2+ and LaZnAl11O19:Mn2+ in Fig. S1816,40,43,44, share a commonality in their octahedral sites. This suggests that the NIR emission of Mn2+ is attributed to its occupancy within octahedral sites (\({\rm{Mn}}^{2+}_{\rm{oct}}\)), which are characterized by strong crystal field strength. This provides fundamental insights into the mechanism behind this intriguing phenomenon.

Fig. 4: XAS characterizations of Y2Mg0.92Mn0.08Al4SiO12 and YAG:2%Ce3+, 20%Mn2+.
figure 4

a, b FT-EXAFS spectra; c, d Experimental and calculated EXAFS spectra χ(k)k2; e, f Wavelet transform of the k3-weighted EXAFS data; g, h Photoluminescence and corresponding coordination environments

Full size image
Table 2 EXAFS fitting parameters at the Mn K-edge for Y2Mg0.92Mn0.08Al4SiO12 and YAG:2%Ce3+, 20%Mn2+ (Ѕ02=0.718)
Full size table

Luminescence properties of the fabricated NIR pc-LEDs

We have successfully crafted a NIR pc-LEDs device by seamlessly integrating a NIR phosphor, specifically YAG:2%Ce3+, 20%Mn2+, with an InGaN blue-LED chip emitting at approximately 450 nm, as depicted in Fig. 5a. The accompanying inset photograph showcases the fabricated pc-LEDs in both idle and activated states. Notably, the emission intensity of the device increases with the gradual rise in injected current. As documented in Fig. S19, precise measurements reveal that the NIR output power is 36.5 mW at 800 mA, while the photoelectric efficiency is 2.4% at 10 mA. Furthermore, a thermal imaging system was utilized to meticulously track the temperature fluctuations of the pc-LEDs device as the injection current increased. As shown in Fig. 5b, the device’s temperature steadily increased, rising from 45 °C at 50 mA to 77 °C at 800 mA. It is remarkable that the NIR emission of YAG:2%Ce3+, 20%Mn2+ demonstrates outstanding stability with negligible shifts even during temperature hikes, maintaining approximately 66% of its ambient temperature luminescence intensity even at 100 °C, as illustrated in Fig. S20. Moreover, the potential of this device in practical applications, such as night vision systems, detection technologies, and biological tissue imaging, has been preliminarily verified, as shown in Figs. 5c and S21. Under dim lighting conditions, the NIR camera, enhanced by our pc-LEDs, effortlessly captures intricate details of optical materials, identifies foreign bodies within fruits, and enables high-fidelity vascular imaging of the palm. These tasks are proven to be beyond the capabilities of a visible-light camera, even under bright illumination. Despite the residual presence of the red and far-red regions in the luminescence spectrum of YAG:Ce3+, Mn2+, Fig. S22 demonstrates a significant red shift of ~10 nm in the NIR emission of Lu3Al5O12:Ce3+, Mn2+ (LuAG:Ce3+, Mn2+) compared to YAG:Ce3+, Mn2+, which underscores the immense potential for spectral tuning and optimization of Mn2+-based materials. A precise and comprehensive understanding of the NIR emissions originating from Mn2+ ions holds the key to unlocking advanced NIR materials centered around the Mn2+ ion.

Fig. 5: Luminescence properties of NIR pc-LEDs using YAG:2%Ce3+, 20%Mn2+.
figure 5

a Emission spectra of the fabricated pc-LEDs at varied current over 10–800 mA; b NIR images and thermal images of the fabricated pc-LEDs at varied driving currents; c Photos of the lighting pc-LEDs taken by a normal camera and a NIR camera

Full size image

Discussions

In conclusion, we have successfully synthesized the YAG:2%Ce3+, 20%Mn2+ phosphor, which exhibits a dominant emission within the NIR region, specifically spanning the range of 700 to 900 nm. Preliminary validation suggests the feasibility of incorporating YAG:2%Ce3+, 20%Mn2+ into NIR pc-LED devices, underscoring its promising potential for future applications. A thorough analysis confirms that within the garnet structure, Mn exists in its divalent (+2) form, and the NIR emission arises from Mn2+ occupying 6-coordinate octahedral AlO6 sites (\({\rm{Mn}}^{2+}_{\rm{oct}}\)). Our hypothesis, which posits that Mn2+ ions occupying octahedral sites (\({\rm{Mn}}^{2+}_{\rm{dod}}\)) within a robust crystal field initiates NIR emission, provides a compelling explanation for the NIR phosphors observed in other Mn2+-activated materials. This, in turn, offers valuable insights into the underlying mechanisms responsible for NIR emission in Mn2+ ions. It is noteworthy that, while the NIR emission of YAG:Ce3+, Mn2+ only slightly encroaches into the far-red region, the NIR emission exhibited by Mn2+ demonstrates significant potential for spectral redshift through modulation of the local crystal field. A comprehensive understanding of the NIR emission capabilities generated by Mn2+ bodes well for the future advancement of NIR phosphors.

Materials and methods

Synthesis

A series of samples, Y3Al5O12(YAG):2%Ce3+, xMn2+ (x = 0%–20%), were synthesized by substituting Mn2+–Si4+ for Y3+–Al3+ in YAG:2%Ce3+. To conduct a more in-depth study on the occupancy of Mn2+ and its NIR emission characteristics, we also successfully prepared three materials: YAG:20%Mn2+, Y2Mg0.92Mn0.08Al4SiO12, and Lu3Al5O12:2%Ce3+, 20%Mn2+ (referred to as LuAG:Ce3+, Mn2+). These materials are prepared using the traditional high-temperature solid-state reaction. The raw materials, Y2O3(99.99%), Lu2O3(99.99%), Al2O3(99.9%), SiO2(99.99%), MgO(99.9%), CeO2(99.99%) and MnCO3(99.9%), were weighed according to the stoichiometric ratio, and ground in an agate mortar adequately to get the mixture. Then, put the mixture into a tube furnace and sintered at 1550 °C for 4 h in the 10% H2 + 90% N2 reducing atmosphere. The final product was ground into fine powders for the subsequent investigations.

Characterizations

The Rietveld refinement of X-ray diffraction (XRD) pattern was performed by FullProf program using data collected in multifunctional horizontal X-ray diffractometer (Ultima IV, Rigaku, Japan) with Cu Kα radiation (λ = 1.5406 Å). The counting time for each step was set at 1 s with a step size of 0.02. Electron paramagnetic resonance (EPR) spectra was recorded at room temperature on a Bruker Elexsys E580 spectrometer operating at 9.48 GHz. The high-resolution transmission electron microscopy (HRTEM) images and elemental distributions were measured on a transmission electron microscope (JEM-2100F, UHR) and JSM-7600F (JEOL), respectively. The photoluminescence (PL) and photoluminescence excitation (PLE) were recorded by a Hitachi F-7000 and an Edinburg FLS-980 fluorescence spectrophotometer. The temperature-dependent PL spectra were obtained by the measurement system containing using a heating stage (Linkam THMS-600) and a QEPro high performance spectrometer (Ocean Optics) which gives the time-integrated intensities. A pulsed laser from an optical parametric oscillator (OPO) and the electric signal detected by a Tektronix digital oscilloscope TDS 3052 were used to measure the fluorescence decay of Mn2+ emission. The IQE was obtained with an absolute photoluminescence quantum yield measurement system (Quantaurus-QY Plus C13534-12, Hamamatsu Photonics).

X-ray absorption spectroscopy (XAS) measurement and analysis

Data reduction, data analysis, and EXAFS fitting were performed and analyzed with the Athena and Artemis programs of the Demeter data analysis packages that utilizes the FEFF6 program to fit the EXAFS data59. The energy calibration of the sample was conducted through a standard Mn foil, which as a reference was simultaneously measured. A linear function was subtracted from the pre-edge region, then the edge jump was normalized using Athena software. The χ(k) data were isolated by subtracting a smooth, third-order polynomial approximating the absorption background of an isolated atom. The k3-weighted χ(k) data were Fourier transformed after applying a Hanning window function (Δk = 1.0). For EXAFS modeling, the global amplitude EXAFS (CN, R, σ2 and ΔE0) were obtained by nonlinear fitting, with least-squares refinement, of the EXAFS equation to the Fourier-transformed data in R-space, using Artemis software, EXAFS of the Mn foil is fitted and the obtained amplitude reduction factor S02 value (0.718) was set in the EXAFS analysis to determine the coordination numbers (CNs) in the Mn–O scattering path in sample.

Computational details

All the calculations are performed in the framework of the density functional theory with the projector augmented plane-wave method, as implemented in the Vienna ab initio simulation package60. The generalized gradient approximation proposed by Perdew-Burke-Ernzerhof (PBE) is selected for the exchange-correlation potential61. The cut-off energy for plane wave is set to 500 eV. The energy criterion is set to 10−5 eV in the iterative solution of the Kohn-Sham equation. All the structures are relaxed until the residual forces on the atoms have declined to less than 0.02 eV/Å. The Franck-Condon approximation is commonly used to interpret the excitation spectrum, that is, assuming that the electronic transition is very fast compared with the motion of nuclei in the lattice62.

Fabrication of the pc-LED devices

The pc-LED devices were fabricated by the GaN-based blue LEDs coated with the mixture which prepared by the synthesized phosphor and transparent silicon resin. The curing process was performed in an oven at 120 °C for 120 min. The photoelectric properties of the fabricated pc-LEDs, such as NIR output power and photoelectric efficiency, were measured using an integrated test system (LHS-1000, EVERFINE) equipped with a high accuracy array spectrophotometer (350–1100 nm, HAAS-2000).