Effect of Molten Salt Synthesis Processing Duration on the Photo- and Radioluminescence of UV-, Visible-, and X-ray-Excitable La2Hf2O7:Eu3+ Nanoparticles.

Ln3+-ion-doped nanomaterials possess excellent properties because of their high color purity, longer excited state lifetime, narrow emission, and large Stokes shifts. In this work, we studied the correlation between the luminescence properties of La2Hf2O7:Eu3+ pyrochlore nanoparticles (NPs) synthesized by a molten salt synthesis (MSS) method at a relatively low temperature and several MSS processing durations (from 1 to 12 h). We synthesized these NPs with different sizes just by changing the MSS processing time without subjecting to high temperature. Raman spectroscopy confirmed the stabilization of the ideal pyrochlore structure of the La2Hf2O7:Eu3+ NPs at various MSS processing durations. The synthesized NPs exhibited bright red emission under UV, visible, and X-ray excitations, highlighting their potential applications as a red phosphor and scintillator. As the MSS processing time was increased from 1 to 12 h, a spectral change in the position of the charge transfer state in the La2Hf2O7:Eu3+ NPs was observed. The sample processed by the MSS with a duration of 3 h exhibited the highest luminescence intensity, which was attributed to its optimum crystals with least surface defects and less agglomeration. The obtained results strongly and unambiguously indicate the brighter side of this new type of pyrochlore-based NPs in the fast growing field of solid-state lighting and scintillator materials.

Introduction

Materials with A2B2O7-type composition fall into a special class of structure called pyrochlore, in which A represents a trivalent rare-earth (RE) ion and B represents a tetravalent transition metal ion such as Sn4+, Ti4+, Zr4+, and Hf4+. Recently, they have gained a significant interest as ceramics because of their unique properties such as high radiation stability, ability to accommodate actinides and lanthanides at both AIII and BIV sites,[1] and ability to dissipate excess radiation energy by forming antisite defects. For the same reason, they have been used for immobilization of high-level nuclear waste.[2] Other than nuclear energy applications, these materials have also been found to be an excellent host for lanthanide-doped phosphor materials because of their other favorable properties, such as suitable refractive index value, high chemical and thermal stability, wide band gap, etc.[3−5] In addition, because of their high chemical stability, high catalytic activity, high melting point, and excellent oxide ion conductivity,[6] they have innumerable applications in many other scientific and technologic areas, such as magnetic materials,[7] gas sensors,[8] catalysts,[9] solid oxide fuel cells,[10] lithium ion batteries,[11] structural materials in fusion reactors,[12] etc. Moreover, A2B2O7 pyrochlore nanocrystals possess excellent luminescent properties, including intense photoluminescence (PL) in the infrared region, efficient pump-power dependence of light emission, and excellent cathodoluminescence.[13]

Structurally, A2B2O7-type compounds belong to the fluorite–pyrochlore phase family.[14] Preference for either phase is contingent on the ionic radius of the RE3+ ion in the A-site and the B4+ ion in the B-site. The fluorite structure is favorable if the ionic radii of A3+ and B4+ ions are similar (rA/rB < 1.46). The oxygen vacancies are randomly distributed on the anion sites. The similar ionic radii of the A3+ and B4+ cations give them the capability to swapping sites in the lattice, which leads to antisite defects.[14] On the other hand, the pyrochlore structure has D3 symmetry and differs from the fluorite in that it has a more ordered structure, which causes the cations to prefer the site most suitable for their size. The A3+ ions have an eightfold coordination with oxygen, and the B4+ ions have a sixfold coordination with oxygen. The anion Wyckoff sites (8b and 48f) are entirely occupied.

Among pyrochlores, La2Hf2O7 stands out with unique properties such as high density, high refractive index, wide band gap, and high thermal/chemical stability, which are very important from a technical perspective. It has been envisaged to be a very good dielectric material because of its rather low defect densities and less Fermi-level pinning than those of HfO2.[15] Also, because of its high stopping power for X- and γ-rays with ZHf = 72 and its high density of 7.9 g/cm3, it was found to be very attractive for novel high-energy radiation detectors.[16] La2Hf2O7 is also known to exhibit defect-induced luminescence on irradiation with ultraviolet photons, and the oxygen vacancies were found to be the most probable reason for the recombination centers in their photoluminescence process.[17]

Doped RE2Hf2O7 nanoceramics with metal ions have proven to be very successful in tailoring their electrical,[18] optical,[19,20] nuclear,[21] and order–disorder phenomena;[22] scintillation;[23] etc. Our group recently has done some work on photo- and radioluminescence (RL) of the La2Hf2O7:Eu3+ nanoparticles (NPs).[4,16]

There are various ways to induce changes in the particle size of nanophasic materials, such as changing the pH value of the precursor solution, varying the annealing temperature, altering the synthesis duration, differing the reaction kinetics, and modifying the thermodynamics of reactions. The changes of synthesis duration and temperature may indirectly affect the obtained particle size, crystallinity, coordination number, and geometry around dopant ions in an inorganic phosphor. This also leads to change in the type, size, and density of defects. It was reported that the synthesis temperature and duration have profound effects on the size, morphology, defect density, and surface energy of obtained products,[24−26] which ultimately alter the optical properties of nanomaterials. Melato and his group investigated the effect of annealing time on the structure, morphology, and photoluminescent properties of MgAl2O4:In3+ nanophosphor.[27] On the other hand, Motloung et al. investigated the effects of the annealing time on the optical properties of ZnAl2O4.[28] There are a few more reports on the effect of annealing time on optical properties of YAG:Ce3+ and ZnAl2O4:Cr3+, etc.[29−31] Not many studies in this direction have been carried out to explore the effect of synthesis processing time on the optical properties of nanophosphors, not to mention those during the molten salt synthesis (MSS) process, until the current study.

Exploitation of a luminescence probe that is sensitive to the local structure is expected to give useful information on its local site, site symmetry, and distribution ratio at the A3+ and B4+ sites in A2B2O7-type compounds, which are very important for designing efficient optoelectronic devices. A 4f electron exhibits special photophysical characteristics compared with 3d and 5f electrons and therefore exploring nanomaterials doped with luminescent lanthanide ions has been found to be an indispensable approach for designing high-quality nanophosphors for future applications. Once doped with the Ln3+ ion, NPs exhibit photophysical properties that are different from those of their bulk counterparts because of their small grains. Meltzer et al. reported increased photoluminescence efficiency of nanophasic materials compared to that of their bulk counterparts.[32]

It is well documented that Eu3+ ions form an ideal spectroscopic probe for lanthanide sites due to their distinct spectroscopic signature, integer J-numbers, and the ease of assessing their site symmetry. Their ground-state 7F0 and the most informative excited state 5D0 are nondegenerate. Therefore, they are not split by crystal-field effects.[33,34] Eu3+ ions with the characteristic orange red emission at 592 and 614 nm are highly sensitive to the local surroundings.[35] The Eu3+ photoluminescence and spectral data can give highly reliable information on whether the dopant ions are localized on the surface of A2B2O7 NPs or they have percolated either interstitially or substitutionally inside the A2B2O7 lattice. Although closeness in size and charge guarantees most of the Eu3+ ions to occupy the A3+ site, there may be lattice strain to mismatch in ionic radii between eight-coordinated Eu3+ and RE3+ ions, which may affect the optical properties of Eu3+ ions.

Despite the wide array of applications, investigations on RE2Hf2O7-type compositions are limited. To the best of our knowledge, among the available literature, only Papan and his group have carried out luminescence spectroscopy and Judd–Ofelt (JO) analyses on their combustion-synthesized europium-doped Y2Hf2O7, Gd2Hf2O7, and Lu2Hf2O7.[20] In this study, we have correlated the changes in optical properties of europium-doped lanthanum hafnate La2Hf2O7:Eu3+ NPs synthesized by an environmentally friendly MSS method as a function of processing duration with the Judd–Ofelt parameter and other related photophysical parameters.[36] Judd–Ofelt analysis was carried out to decipher various photophysical information, such as radiative and nonradiative transition rates, quantum efficiency, Judd–Ofelt parameters, branching ratio, etc. In addition, systematic studies were carried out to explore the potential applicability of these La2Hf2O7:Eu3+ NPs as phosphor materials. The focus was to synthesize highly efficient red phosphors for multifunctional applications that have the ability to get excited by UV light (mid and near UV), blue light-emitting diodes (LEDs), green LEDs, and X-ray. We have also explored the effect of the MSS processing duration of these NPs on their X-ray-excited optical luminescence for the first time.

Results and Discussion

Raman Spectroscopy

Raman spectroscopy is one of the most sensitive techniques to probe metal–oxygen (M–O) vibrational modes, so it is one of the most sought-out techniques to differentiate between ordered pyrochlore and disordered fluorite phases (DFP), which is very difficult to achieve by X-ray diffraction (XRD) most of the time. The disordered fluorite phase (DFP) exists in the Fm3̅m space group wherein all of the cationic ions (A3+ and B4+) are randomly distributed. On the other hand, the ordered pyrochlore phase (OPP) known to exist in the space group exhibits close structural resemblance to the fluorite phase except that there are two cationic sites and three anionic sites (i.e., 48f(OI), 8a(OII), and 8b(OIII)) with one-eighth of the oxygen ions (OIII) at the 8b site absent in the OPP. Therefore, Raman spectroscopy provides unambiguous information in determining whether a synthesized A2B2O7 sample exists in an OPP or a DFP structure.

On the basis of the group theory, it is well documented in the literature that there are a total of six Raman-active vibrational modes for OPP existing in the wavenumber range of 200–1000 cm–1, specifically, ΓOPP = A1g + Eg + 4F2g. On the other hand, the DFP has mainly one active Raman mode, that is ΓDFP = F2g, because the seven O2– ions are randomly oriented over the eight anionic sites in this particular phase, leading to a high level of structural disordering.[37,38] Phase transition from the OPP (Fd3̅m (A2B2O6O′, space group, Z = 8) to DFP (AO2, Fm3̅m, Z = 4)) proceeds by the disappearance of A1g and Eg Raman modes and decrease in the number of F2g mode from 4 to 1. The ionic radius ratio (rA/rB) plays an important role in determining the type of structure that the A2B2O7 composition is going to attain.[39] It is reported that if rA/rB is less than 1.46 the fluorite phase is more likely to form and if rA/rB exceeds 1.46 the ordered pyrochlore phase is more likely to get stabilized at room temperature. It was proposed that rA/rB follows this trend for different A2B2O7 compositions: DFP rA/rB < 1.21 < Δ-phase rA/rB < 1.42–1.44 < OPP rA/rB < 1.78–1.83 < monoclinic pyrochlore rA/rB < 1.92.[40] La2Hf2O7 is one of the most favorable candidates with an ionic radius ratio of 1.45 to be stabilized in OPP. As observed from the Raman spectra (Figure ), except for the LHO-1h and LHOE-1h samples, six vibrational modes related to the La–O and Hf–O vibrational frequency were clearly identified for all other La2Hf2O7 samples. It may happen that these two samples have not been sufficiently processed for the pyrochlore network to be fully evolved, even though some of the characteristic spectral signatures of an ideal pyrochlore have started to appear around 320, 400, and 520 cm–1. The peak positions for the other La2Hf2O7 NPs are around 304, 321, 401, 506, 520, and 610 cm–1, which correspond to F2g, Eg, F2g, A1g, F2g, and F2g.[41] The vibrational Raman bands of F2g, Eg, and F2g modes at a low wavenumber region (300–400 cm–1) originate from vibrations of the La–O and Hf–O bonds, whereas the F2g modes at high wavenumbers (522 and 641 cm–1) come into picture due to stretching of the Hf–O bonds. There is a small kink around 750 cm–1, which is attributed to the distortion of HfO6 octahedral geometry.[41] The Raman spectroscopy data acquired suggested that the as-synthesized La2Hf2O7:Eu3+ NPs are indeed in the pyrochlore phase based on the observed six typical Raman-active vibrational modes existing in the wavenumber range of 200–1000 cm–1. On the other hand, a defect fluorite phase only processes one active vibrational mode. This mode usually occurs at ∼300 cm–1 similar to the naturally occurring defect fluorite rare-earth hafnate.[42] There are not much appreciable differences in the Raman spectra of the Eu3+-doped La2Hf2O7 samples (Figure b) from their undoped counterparts, indicating that europium doping does not alter the basic pyrochlore network of lanthanum hafnate.

Figure 1

Raman spectra of (a) LHO and (b) LHOE NPs as a function of the MSS processing duration varying from 1 to 12 h. Insets show the deconvoluted Raman peaks for the samples with the MSS processing duration of 3 h.

XRD Patterns and Refinement

Differentiation of the structures of DFP (Fm3®m) and OPP (Fd3®m) using X-ray diffraction (XRD) is very difficult because of their phase resemblance within the cubic structure family. XRD in most cases using a Cu Kα source in common research labs fails to detect weaker superlattice reflections of the pyrochlore phase. A stronger source such as synchrotron radiation is needed to resolve the very weak lattice reflection. As reported in the literature[43,44] pertaining to the Rietveld refinement of the pyrochlore phase La2Zr2O7 and La2Hf2O7, a synchrotron X-ray source is really needed. More specifically, these superlattice reflections are relatively strong in the neutron patterns, likely a consequence of the greater relative sensitivity of neutrons to O atoms in the presence of Hf and Ln atoms. For this reason, our lab XRD instrument identified only the defect fluorite phase from our LHO and LHOE NPs, which should not be considered as a discrepancy with the Raman data (Figure ). In our case, we have used XRD to confirm that our samples are free of impurities and to estimate their particle size.

Figure shows the XRD patterns of the La2Hf2O7 and La2Hf2O7:5.0%Eu3+ powder samples synthesized at 650 °C by the MSS method with various processing durations. No visible impurity phases such as La2O3, Eu2O3, and HfO2 were detected. Meanwhile, it is well known that the presence of weak reflections due to the ordered pyrochlore phase (OPP) cannot be ruled out, as they are sometimes not detected in Cu Kα-based XRD. Therefore, the ideal scenario would be probing the crystal phase of these samples using laser-based Raman spectroscopy, synchrotron X-ray diffraction, or neutron diffraction. Here, all of the XRD patterns and the corresponding 2θ angles and (hkl) indexes are in agreement with reported XRD pattern of La2Hf2O7 (JCPDS No. 78-1292). When comparing the undoped and europium-doped samples, the results indicate that doping does not change the crystal structure of the La2Hf2O7 host. In addition, changing the MSS processing duration did not change the crystal structure of the prepared NPs.

Figure 2

XRD patterns of (a) undoped and (b) 5% Eu3+-doped La2Hf2O7 NPs as a function of the MSS processing duration varying from 1 to 12 h.

The determined particle size is given in Table , which displays the lattice parameter and particle size as a function of the MSS processing duration for both the undoped and doped La2Hf2O7 samples. As far as the lattice parameter is concerned, as the MSS processing duration increases, there is not much change in the case of the undoped La2Hf2O7 samples, but the lattice parameter decreases in the case of the La2Hf2O7:Eu3+ samples.

Table 1

Calculated Lattice Parameters and Particles Sizes of the LHO and LHOE NPs

samples	2θ (deg)	FWHM (β)	lattice parameter (Å)	particle size (nm)
LHO-1h	28.67	0.74	10.77	10.7 ± 1.7
LHO-3h	28.68	0.36	10.77	22.1 ± 1.2
LHO-6h	28.67	0.34	10.77	23.4 ± 1.2
LHO-9h	28.68	0.31	10.77	25.6 ± 1.1
LHO-12h	28.70	0.29	10.77	27.4 ± 1.1
LHOE-1h	28.63	0.47	10.79	14.9 ± 0.9
LHOE-3h	28.66	0.43	10.78	18.5 ± 0.8
LHOE-6h	28.68	0.36	10.77	22.1 ± 0.7
LHOE-9h	28.70	0.34	10.77	23.4 ± 0.7
LHOE-12h	28.67	0.33	10.77	24.1 ± 0.7

Regarding the particle size, it can be seen that there is an increase with an increase in the MSS processing duration. The increased particle size is a result of the longer MSS processing time to allow the continuous growth of the NPs. Furthermore, it can be seen that the particle size decreases after doping La2Hf2O7 with 5% Eu3+ at the same MSS processing duration. On doping a foreign ion into a crystal lattice, the crystallite size is reduced because of the lattice distortion caused by the difference in the ionic radius of the dopant and the original ion replaced.[45] There is also a possibility that a fraction of the Eu3+ dopants substitutes the Hf4+ sites, which leads to the creation of oxygen vacancies and consequently significant strain and reduction in particle size in the La2Hf2O7 lattice.[46]

To rule out the formation of any metastable state in the La2Hf2O7:Eu3+ NPs by the MSS at a relatively low temperature of 650 °C, the Rietveld refinement of the XRD data and thermogravimetric analysis (TGA) were carried out using LHO-6h and LHOE-6h NPs as representatives. The Rietveld-refined XRD patterns of these two samples are shown in Figure a,b. The lattice parameter, cell volume, space group, and bond angle estimated from the refinement along with other refinement parameters are tabulated in Table S1. Our original XRD data were fitted well by the Rietveld refinement with the fluorite structure instead of the pyrochlore phase with superlattice reflection. The crystal structure of the LHO-6h and LHOE-6h NPs (Figure S1) is in agreement with previous reports for La2Hf2O7.[47] The estimated uncertainty on the lattice parameter is 10.773 ± 0.001 from the LHO-6h NPs and 10.765 ± 0.001 from the LHOE-6h NPs (Table S1). It can also be seen that doping Eu3+ into the La2Hf2O7 lattice does not distort its basic fluorite network, which can be easily visualized from Figures and S1b. The chemical content of the unit cell is 16.0000 La + 48.0002 O + 16.0000 Hf + 8.0001 O. The normalized site occupation numbers in percentage are 100.0000 La1:100.0004(2) O2:100.0000 Hf3:100.0012(4) O4. A summary of the obtained results from the Rietveld refinements is shown in Table S2.

Figure 3

Rietveld-refined XRD patterns of the (a) LHO-6h and (b) LHOE-6h NPs.

Figure 4

Scanning electron microscopy (SEM) micrographs of the LHOE NPs as a function of the MSS processing duration: (a) 1 h, (b) 3 h, (c) 6 h, (d) 9 h, and (e) 12 h.

As the MSS processing duration increased from 1 to 12 h, SEM images (Figure ) confirmed that the particle size of the LHOE NPs kept increasing, which is consistent with the calculated values from the XRD data (Figure and Table ). Moreover, the particles look spherical at a short MSS processing duration but they tend to form large clusters and finally agglomerate at longer MSS processing times of 9 and 12 h. The particle size distribution of the LHOE NPs was calculated by ImageJ from these SEM images (Figure S2). There is a progressing enhancement of the obtained mean diameter of these LHOE NPs from 19 to 50 nm as the MSS processing time increased from 1 to 12 h. The representative transmission electron microscopy (TEM) image of the LHOE-6h NPs clearly shows that the synthesized NPs have well-defined morphology (Figure S3) and their size is in a nanometer range with uniform distribution.

TGA Measurements

TGA data (Figure ) were also collected to rule out the possible formation of any metastable state in the LHO and LHOE NPs synthesized by the MSS method. There is barely any weight change from the LHO-6h and LHOE-6h NPs other than the initial weight loss due to physisorbed water molecules and residual nitrate ions from the molten salt. Both samples are very stable throughout the temperature range up to 1000 °C. This again suggested the thermal stability of the La2Hf2O7 NPs for potential high-temperature applications such as nuclear waste host, thermal barrier coatings, and luminescence host.

Figure 5

TGA graphs of the (a) LHO-6h and (b) LHOE-6h NPs.

Excitation and Emission Spectra of Photoluminescence

From the excitation spectra of the La2Hf2O7:5%Eu3+ NPs processed by MSS at 650 °C for different time intervals (Figure a), we could see a broad band ranging from 220 to 320 nm, and there are several very fine features from 350 to 500 nm. Such a broad band is attributed to the allowed charge transfer state (CTS) with three possible origins: (a) host absorption band due to electronic transition from O2– → Hf4+, (b) intervalence charge transfer between Hf4+ and Eu3+, and (c) the predominant contribution would be due to electron transfer from O2– → Eu3+.[48−50] The fine structure in the region of 350–500 nm is attributed to the intra f–f transition of Eu3+ ions with peaks at 362, 383, 395, 414, and 464 nm assigned to 7F0 → 5D4, 7F0 → 5G2–4, 7F0 → 5L6, 7F0 → 5D3, and 7F0 → 5D2, respectively. Among them, the intensities of the 395 nm (near UV) and 464 nm (blue light) peaks are the highest. This indicated that our La2Hf2O7:5%Eu3+ NPs have the ability to be effectively pumped by far UV, mid UV, near UV, blue light, and green light for being used as a phosphor. From all of our La2Hf2O7:5%Eu3+ NPs synthesized by the MSS route, we could see that the photoluminescence excitation (PLE) intensity of the f–f band is higher than that of the CTS. This is very unusual and highly desirable because intuitively the f–f transition is forbidden in nature, whereas the CTS is an allowed transition. This can be attributed to two reasons: (i) overlap of the CTS and the absorption band of the host[51] and (ii) the forbidden 4f–4f transition “steals” some intensity from the allowed CTS transition.[52] Excitation lines originating from the 7F0 → 5L6 (≈395 nm) and 7F0 → 5D2 (≈464 nm) transitions are highly suitable as color converters in light-emitting diodes (LEDs) because they overlap very well with the emission spectra of efficient near-UV and blue LEDs, respectively.[53] However, there is no perfect correlation between the MSS processing duration and the variation in the wavelength position of the CTS maxima of our LHOE NPs (Figure b). Initially, the LHOE-1h sample with the smallest particle size displayed the maximum blue shift compared to that in the LHOE-3h sample, indicating the widening of band gap. From the LHOE-3h sample to the LHOE-6h sample and then to the LHOE-9h sample, the CTS maxima exhibited a blue shift although there is a continuous increase in particle size. Finally, there was a red shift in the CTS maxima from the LHOE-9h sample to the LHOE-12h sample. The lack of perfect correlation between the CTS maxima and the MSS processing duration can be attributed to the fact that there were accompanying structural changes along with the MSS processing duration and particle size variation. Although the reason is unknown, further control of the identity of these particles by proper synthesis procedures and detailed investigation by various characterization techniques are guaranteed. Moreover, such observation in this case cannot be attributed to quantum confinement, which is normally observed when nanoparticles are sufficiently small, typically 10 nm or less. Such a phenomenon is more prevalent in systems with simple energy-level diagrams, such as metal or semiconductor nanoparticles. Pyrochlore oxides, such as La2Hf2O7, have complex band diagrams: the valence band (VB) is mainly contributed from O 2p, whereas the conduction band (CB) has contributions from Hf 4d, La 4d, and La 4f states.

Figure 6

(a) Excitation spectra and (b) variation of the CTS maxima of the La2Hf2O7:Eu3+ NPs as a function of the MSS processing duration varying from 1 to 12 h.

The La2Hf2O7:5%Eu3+ NPs processed by MSS for different durations displayed similar emission spectral features of the Eu3+ ion after being excited at 265 nm (Figure a). The peaks at 579, 592, 612, 655, and 710 nm are related to the 5D0 → 7F0, 5D0 → 7F1 (magnetic dipole transition, MDT, ΔJ = ±1), 5D0 → 7F2 (hypersensitive electric dipole transition, EDT, ΔJ = ±2), 5D0 → 7F3, and 5D0 → 7F4 transitions, respectively. A few important observations can be made from these emission spectra: (i) appearance of the 5D0 → 7F0 transition, (ii) significant stark splitting of the 5D0 → 7F1 (MDT), 5D0 → 7F2 (hypersensitive EDT), and 5D0 → 7F4 (EDT) transitions, and (iii) higher intensity of the 5D0 → 7F2 transition than that of the 5D0 → 7F1 transition. The intensity of the 5D0 → 7F1 MDT peak at 592 nm is independent of the local environment of the Eu3+ ion. On the other hand, the intensity of the hypersensitive EDT peak is easily affected by the local symmetry/environment of Eu3+ ions and the crystal field induced by surrounding ligands. The ratio of integral areas of the 5D0 → 7F2 and 5D0 → 7F1 transitions, known as asymmetry factor, gives information related to the local symmetry of Eu3+ ions in hosts. The origins of certain transitions, such as 5D0 → 7F0 and 5D0 → 7F3, are governed by neither MDT nor EDT. Their appearance in an emission spectrum indicated a highly asymmetric environment around the europium ion. Therefore, we concluded that the Eu3+ ions in the La2Hf2O7:5%Eu3+ NPs are localized in a highly asymmetric and distorted environment.

Figure 7

(a) Emission spectra and (b) variation of the integral emission intensity of the La2Hf2O7:5%Eu3+ NPs as a function of the MSS processing duration varied from 1 to 12 h.

In fact, on the basis of the selection rule allowed for electric dipole, the 5D0 → 7F0 transition is allowed only with point group symmetry designated as C, C1, C2, C3, C4, C6, C2, C3, C4, and C6.[54] On the basis of the splitting pattern in the emission characteristics of our La2Hf2O7:Eu3+ NPs, there is no splitting observed in the 5D0 → 7F0 transition. By Gaussian deconvolution of the 5D0 → 7F1, 5D0 → 7F2, 5D0 → 7F3, and 5D0 → 7F4 transitions, the numbers of stark components were found to be 3, 2, 2, and 3, respectively (Figure ). This phenomenon suggests the C4 point group symmetry for Eu3+ ions in our doped La2Hf2O7 NPs, whereas the actual symmetry of both A and B sites of A2B2O7 compounds is D3.[55] The reduction of the symmetry of Eu3+ ions in the La2Hf2O7 host of these NPs could be induced by the lattice strain due to the mismatch of the ionic radii of La3+ and Eu3+ ions. It is also possible that the 5.0% Eu3+ ions are not entirely stabilized at the La3+ site in the host lattice. In other words, some of the Eu3+ ions may get localized at the Hf4+ site, which invokes the need for charge compensation by oxygen vacancies. Such defects in the vicinity of Eu3+ ions further reduce the local structural symmetry of the Eu3+ ions. Moreover, the hypersensitive EDT band is more intense than the MDT band, which also indicates the lower symmetry of the Eu3+ ion. In fact, the LHOE-3h sample has the maximum asymmetry ratio and maximum emission intensity (Table and Figure b). By correlating the asymmetry ratio with the particle size/surface defects of our La2Hf2O7:Eu3+ NPs, there are two ways to look into this phenomenon. The highest asymmetry ratio from the LHOE-3h sample suggests a highly asymmetric local environment around the Eu3+ ions compared with other LHOE samples. Therefore, Laporte’s selection rules are more relaxed for the LHOE-3h sample compared with the other samples, which means higher transition probability as well as oscillator strength for the f–f transition, leading to enhanced emission output.[56] Second, this phenomenon can also be explained in terms of size/surface defects. The LHOE-3h sample has minimal surface defects and least agglomeration that favor high radiative transition and low scattering of excited and emitted light, respectively, leading to high luminescence output.[25,57] The PL emission enhancement up to 3 h MSS processing duration can be ascribed to the reduction in nonradiative pathways as a result of reduction in surface defects with the increased particle size. Beyond 3 h MSS processing duration, the La2Hf2O7:5%Eu3+ NPs tend to agglomerate, which may cause scattering of the emitted light and therefore reduction in emission intensity. These observations could be explained on the basis of our SEM results (Figure ). Figure schematically shows the progressive increase in particle size along the increasing MSS processing duration, which reflects the increased particle size but reduced surface defects.

Figure 8

Stark splitting patterns of the 5D0 → 7F (J = 0–4) transitions of the LHOE-3h NPs.

Table 2

Asymmetry Ratio of the La2Hf2O7:5%Eu3+ NPs as a Function of the MSS Processing Time Varied from 1 to 12 h

	asymmetry ratio
samples	PL	RL
LHOE-1h	1.57
LHOE-3h	3.99	2.57
LHOE-6h	3.19
LHOE-9h	3.49
LHOE-12h	3.34

Figure 9

Schematic showing the increase in the particle size of the La2Hf2O7:5%Eu3+ NPs as a function of the MSS processing duration.

Furthermore, the application of X-ray-excited luminescence (XEL) has been explored for scintillating and bioimaging purposes, which require bright luminescent particles. Studies related to the variation of the radioluminescence output of the La2Hf2O7:5%Eu3+ NPs as a function of the MSS processing duration will give highly useful information with respect to the optimized processing time of Eu3+ ions for X-ray luminescence. The emission spectra under X-ray excitation with the power of 12 W (Figure a) display typical spectral features of the Eu3+ ions similar to those obtained under UV excitation (Figure a). The X-ray luminescence from the LHOE-3h sample was the brightest among all of the samples we studied (Figure b).

Figure 10

(a) X-ray-excited luminescence of the La2Hf2O7:5%Eu3+ NPs, (b) corresponding XEL intensity expressed as an integrated area under the XEL spectra in the range of 550–750 nm as a function of the MSS processing time.

High-energy X-rays are produced by events that occur to disrupt the nuclear stability of atoms. Radiation having lower energy, such as ultraviolet, originates from the electron clouds that surround the nucleus or from the interaction of one atom with another. These forms of radiation occur due to the fact that electrons moving in orbits around the nucleus of an atom are arranged in different energy levels within their probability distribution functions. When a sample is excited by energetic X-ray beams, a large number of free charge carriers or excitons (bound state of e– and h+ pairs) are generated in the crystal lattice because of the photoelectric effect. In the case of La2Hf2O7, there are two crystallographic sites: eight-coordinated La3+ and six-coordinated Hf4+ sites. On the basis of the UV luminescence decay measurement, it was found that the Eu3+ ions are localized at both La3+ (EuLa) and Hf4+ sites (EuHf′). The Eu3+ ions occupying Hf4+ sites may lead to the generation of negative antisite defects (EuHf′) and positive oxygen vacancy defects (VÖ) as indicated below by the Kroger–Vink notation.UV-excited PL spectra displayed both EDT and MDT (Figure a). EDT is attributed to EuHf′, whereas MDT is attributed to EuLa. As discussed earlier, the Eu3+ ions get distributed at both the La3+ and Hf4+ sites in the La2Hf2O7:Eu3+ NPs. For La2Hf2O7, the valence band (VB) is composed of O 2p orbitals hybridized with Hf 5d orbitals along with a minor contribution from La 4f states.[25,58] On the other hand, the conduction band (CB) is mainly composed of La 4d states (in majority spin component), 4f states (in minority spin component), and Hf 5d states. Hf 5d states contribute solely to the lower part of the conduction band.

The EuHf′ defects are XEL-active and amenable to emission under X-ray excitation. This is similar to what Dobrowolska et al. have observed in europium-doped barium hafnate.[59] Excitation with an energetic 12 W X-ray beam leads to the formation of excitons (bound electron–hole pairs) between electrons in the lower part of the CB comprising Hf 5d states (5d1 Hf3+) and holes in the VB, if they possess similar momentum. The energy absorbed by excitons can be transported through the La2Hf2O7 lattice. They will be trapped in the EuHf′ defect sites, and then radiative recombination is followed. The 5D0 → 7F2/5D0 → 7F1 ratio (IAR) indicates a local crystal structure surrounding the doped Eu3+ ions. For the same LHOE-3h sample, the X-ray-excited luminescence shows an IAR value of 2.57, which is much lower as compared to that for the UV-excited luminescence (3.99). This indicated that the crystal symmetry surrounding the Eu3+ ions under X-ray excitation was predominantly from a high-symmetry site compared with that under UV excitation. This is due to the contribution of both EuLa and EuHf′ to the UV-excited luminescence, whereas only EuHf′ contributes to X-ray-excited luminescence.

Furthermore, the emitted photon counts corresponding to green and red emission under the UV excitation (Figure ) are much larger than the ones under X-ray excitation (Figure ). La2Hf2O7 is a material with high dielectric constant, and excitons formed in these cases on exposure to ionizing radiation such as X-ray are weakly bonded and are known as Wannier–Mott-type excitons.[60] They break easily on collision with phonons. The number of excitons thus available for energy transfer to Eu3+ ions is small and hence low RL output. On the other hand, under UV excitation of 265 nm, the La2Hf2O7 host absorbs the energy and transfers it to radiative Eu3+ centers directly. This is known as host-sensitized energy transfer.

Potential Red-Emitting Phosphor Excitable by Various Lights

Commercial red phosphors for near-ultraviolet (NUV)-based white light-emitting diodes are excited by NUV or blue LEDs. On the basis of the Laporte selection rules, the intra f–f transitions are forbidden in nature; therefore, Eu3+-based phosphors have poor absorptivity in the near-ultraviolet/blue region. As a result, they exhibit weaker emission under similar excitation. Because the excitation spectra of our samples are very rich in near-, mid-, and far-UV regions as well as blue and green regions, they exhibit intense red emission corresponding to the 5D0 → 7F2 transition of Eu3+ ions under excitations by far UV (at 200 nm), mid UV (at 265 nm), near UV (at 393 nm), blue light (at 463 nm), and green light (at 534 nm) (Figure ). Such intense red-emitting NPs can be explored for white LEDs based on phosphor-converted LEDs, wherein blue LEDs are combined with red and green phosphors.

Figure 11

(a) Emission spectra and (b) schematic of the LHOE-3h NPs under various excitations, including far UV, mid UV, near UV, blue light, and green light.

Luminescence Decay Behavior

We studied the PL lifetime decay behavior corresponding to the 5D0 state of Eu3+ ions in the La2Hf2O7:5%Eu3+ NPs, which is shown in Figure S4. The LHOE-1h sample exhibited monoexponential decay. The unusual behavior of the LHOE-1h sample is attributed to the fact that Eu3+ ions must not have entered the lattice site (either La or Hf site) at all and were sitting on the surface of the formed NPs, as we suggested on the basis of the PLE data. All other LHOE samples demonstrated bi-exponential behavior based on the exponential fitting equation given belowwhere I(t) is the intensity at time t, T1, and T2 are luminescence lifetimes, and A1 and A2 are their relative magnitudes. For all other samples, it is possible that Eu3+ ions have been stabilized at two different lattice sites in the La2Hf2O7 host. The fitted lifetime values and their relative percentage for all samples are tabulated in Table . Unlike in Nd2Zr2O7:Eu3+, we could not find any splitting in the 5D0–7F0 transition at room temperature.[61] This may get resolved at a very low temperature close to that of liquid N2. The presence of more than one spectral peak in the 5D0 → 7F0 transition shows that more than one site or species is present, but it does not allow the determination of the exact number of sites or species because the sites or species with a symmetry other than C, C, or C do not give an observable 5D0 → 7F0 transition.[62]

Table 3

Photoluminescence Lifetime Values of the LHOE NPs at λex = 265 nm and λem = 612 nm

samples	τ1	τ2	τ1%	τ2%
LHOE-1h	1.6		100	0.00
LHOE-3h	1.0	2.6	29.0	71.0
LHOE-6h	1.0	2.7	29.0	71.0
LHOE-9h	1.2	2.9	35.0	65.0
LHOE-12h	1.0	2.9	27.0	73.0

In a pyrochlore structure, there are two lattice sites, i.e., the eight-coordinated La3+ ion site with scalenohedra geometry and the six-coordinated Hf4+ ion site with distorted octahedral geometry. The stabilization of the Eu3+ ions at the Hf4+ sites invokes the need for charge compensation by forming negatively charged antisite defect EuHf′ and positively charged oxygen vacancies VÖ according to eq The long lifetime is mostly associated with the symmetric La3+ sites, whereas the short one is due to the asymmetric Hf4+ sites where the f–f transition rules are relaxed and they become allowed. In our case, the long lifetime having a higher percentage could be ascribed to the Eu3+ ions sitting at the La3+ sites, whereas the short-lifetime species is because the Eu3+ ions are localized at the Hf4+ sites. The reason is obviously the closeness of the ionic radii of the Eu3+ and La3+ ions, and furthermore, there is no need for charge compensation. Another interesting observation from Table is that there is a monotonous increase in the lifetime values for both short- and long-lived species as a function of the MSS processing duration. The increase in the lifetime values with the MSS processing time is ascribed to the reduction of surface defects, which ultimately results in the reduction of nonradiative transitions in the LHOE NPs.

Judd–Ofelt Analysis

Judd–Ofelt (JO) analysis is one of the most useful techniques to get information on rare-earth-doped phosphors by correlating the emission spectrum with the local structure, polarizability, and covalency. In the case of the europium ion, this hypothesis is valid only when its doped samples show a pure magnetic dipole transition. In case of other ions, an invariable absorption spectrum is used to carry out such analysis. The details of these calculations have already been reported in one of our earlier studies.[63] There are two phenomenal parameters, i.e., Ω2 and Ω4, which are known as short- and long-range JO parameters. Normally, Ω2 gives information related to the symmetry, covalency, and polarizability, whereas Ω4 gives information related to the rigidity and viscosity of the metal–oxygen bond.

Among all of the LHOE samples, the LHOE-3h sample was found to have the highest Ω2 and Ω4 values. The large Ω2 value reflects the high covalency around the Eu–O bond and maximum distortion, and the large Ω4 value indicates the rigid structure of the Eu–O bond.[64] The trend from all of the La2Hf2O7:5%Eu3+ NP samples is that Ω2 is greater than Ω4, which confirms the highly asymmetric environment around the Eu3+ ions in the La2Hf2O7:5%Eu3+ NPs, which is consistent with the discussed luminescence data. Therefore, the Judd–Ofelt analysis also supports our previous interesting observation that the LHOE-3h sample possesses the best optical properties.

Furthermore, as can be easily seen from Table , the radiative transition probability was maximum in the case of the LHOE-3h sample. Accordingly, its nonradiative transition rate is minimal. This may be attributed to the fact that the 3 h MSS processing duration is optimum to lower the surface defect density but not to create agglomerates. From the branching ratio measurements, we can see that the maximum photon emission from all of the LHOE samples comes from the red emission at 612 nm due to the 5D0 → 7F2 transition. Interestingly, β2 is the highest for LHOE-3h (Table ), which suggests that the red emission intensity is maximum for the LHOE-3h sample. On the other hand, its β1 is the lowest, which indicated that the optical purity of red emission for this sample would be very high due to the least contribution from orange emission at 593 nm.

Table 4

Calculated Judd–Ofelt and Photophysical Parameters of the La2Hf2O7:5%Eu3+ NPsa

samples	AR (s–1)	ANR (s–1)	η (%)	Ω2 (×10–20)	Ω4 (×10–20)	β1 (%)	β2 (%)	β3 (%)
LHOE-1h	189	437	30.2	1.02	0.99	26.5	41.8	20.4
LHOE-3h	294	186	61.2	2.37	1.40	17.0	62.4	18.4
LHOE-6h	269	177	60.3	2.06	1.38	18.6	59.3	19.7
LHOE-9h	282	151	65.1	2.25	1.32	17.7	61.9	18.0
LHOE-12h	274	148	64.8	2.16	1.28	18.3	61.2	18.1

AR, radiative rate; ANR, nonradiative rate; Ω, Judd–Ofelt parameter; and β, branching ratio.

Conclusions

In summary, we demonstrated that the europium-doped lanthanum hafnate (La2Hf2O7:Eu3+) nanoparticles with an ordered pyrochlore phase structure could be successfully prepared at 650 °C by a molten salt synthesis method for various durations. The emission output and quantum yield were found to be the highest for the MSS-processed sample for 3 h owing to its least surface defect and least agglomeration. The emission spectrum revealed the predominant asymmetric environment of Eu3+ ions with large spectral splitting and the presence of the forbidden 5D0 → 7F0 transition. The optical purity of red emission was also very high for the LHOE-3h NPs because of the least contribution among all of the LHOE samples from orange emission at 593 nm as seen from the branching ratio calculations. The large spectral splitting could be seen in the reduction of point group symmetry from D2 to C4 because of the created lattice strain and distortion due to the localization of the Eu3+ ions at La3+/Hf4+ sites. The synthesized nanoparticles displayed a unique ability of emitting bright red light under far-UV to green light excitations. These La2Hf2O7:Eu3+ nanoparticles were further explored for their applications as radioluminescent phosphors. Interestingly, the synthesized La2Hf2O7:Eu3+ NPs have the ability to converting highly energetic X-ray into red light, which highlights its suitability for X-ray scintillators. Our results indicate the important role that the molten salt synthesis processing duration plays in optimizing nanophosphors for optoelectronic applications. This work opens a new pathway to optimize the molten salt synthesis conditions suitably for getting highly efficient luminescent materials for phosphor and scintillator applications.

Experimental Details

Synthesis of La2Hf2O7:Eu3+ NPs

In this study, we synthesized the La2Hf2O7:5.0%Eu3+ NPs through the facile MSS method as reported before but with various processing durations (from 1 to 12 h) instead of the standard 6 h.[4,13,16,36] The chemicals used are all of analytical grade and were used without any further purification. Lanthanum nitrate hexahydrate (La(NO3)3·6H2O, 99.0%), hafnium dichloride oxide octahydrate (HfOCl2·8H2O, 99.0%), and europium(III) nitrate hexahydrate (Eu(NO3)3·6H2O, 99.9%) were first measured in stoichiometric ratio and dissolved in water (Millipore, 18.2 mΩ at 25 °C). After 30 min of stirring, 200 mL of ammonium hydroxide was titrated into the mixture solution dropwise within ∼2 h. This first step allows the formation of a single-source precursor via a co-precipitation technique. In the next step, potassium nitrate (KNO3, 99.9%), sodium nitrate (NaNO3, 98%), and the formed single-source precursor were mixed in a ratio of 30:30:1 and grinded together into fine powder. The fine powder was then MSS-processed at 650 °C for various hours, more specifically, 1, 3, 6, 9, and 12 h, as shown in Scheme . The synthesized La2Hf2O7:5%Eu3+ NPs were finally washed multiple times with deionized water to remove any residual salt from the surface of the obtained NPs. On the basis of the MSS processing durations, the La2Hf2O7:5%Eu3+ NPs were denoted LHOE-1h, LHOE-3h, LHOE-6h, LHOE-9h, and LHOE-12h, respectively. Similarly, undoped La2Hf2O7 NP counterparts were synthesized and labeled as LHO-1h, LHO-3h, LHO-6h, LHO-9h, and LHO-12h, respectively. On the basis of the authors’ experience, the MSS method allows for a size-controllable synthesis of these A2B2O7 refractory metal oxide NPs at a relatively low temperature while ensuring no fluorite–pyrochlore phase transformation.[4,13,16,36]

Scheme 1

Steps Involved in the Synthesis of the La2Hf2O7:5%Eu3+ NPs Using the MSS Method at 650 °C with Various Processing Durations

Characterization

All synthesized La2Hf2O7:5%Eu3+ NPs were systematically characterized using X-ray diffraction (XRD), Raman spectroscopy, scanning electron microscopy (SEM), photoluminescence (PL), radioluminescence (RL), fluorescence decay, quantum yield, and time-resolved emission spectroscopy. Powder XRD was used to distinguish any unwanted impurities, if any, remained from the synthesized La2Hf2O7:5.0%Eu3+ NPs. This characterization was measured by a BRUKER D8 Advance X-ray diffractometer with a Cu Kα1 radiation (λ = 0.15406 nm, 40 kV, 40 mA), with a scanning mode in the 2θ range of 10–90° and a scanning step size of 0.04° at a scanning rate of 2.0°/min. Rietveld refinement was carried out using PROGRAM FullProf.2k (version 6.00, Mar2017-ILL JRC) by varying the cell parameters; position of the La, Hf, and O atoms; occupancy; and thermal parameters. We calculated the crystallite size of these NPs from their XRD data using the Debye–Scherer formulawhere d is the crystallite size, λ is the wavelength of the used X-ray, and θ is the angle of the corresponding Bragg reflection, which was fitted to calculate the full width at half-maximum (FWHM). B is the FWHM in radian, and k is a Scherer constant. The value of the Scherer constant was calculated earlier by Scherer as 0.94 based on cubic crystallites. Later, Klug and Alexander had done a simplified calculation of the Scherer equation, which gives the value 0.89.[65] Here, the BM was calculated using the FWHM value of the nearby highest-intensity peak of pure silicon. FWMH of Bs was determined based on the XRD peak with the highest intensity at 29.7°, and fitting was done using the pseudo-Voigt function. Thermogravimetric analysis (TGA) was done with a TA instruments SDT-Q600 thermogravimetric analyzer with a ramp rate of 10 °C/min in air. To confirm the crystal structure of the NPs, we used a Raman spectrometer (Renishaw-2000, Renishaw, Inc.) with the 514 nm line of an Ar ion laser having a power of about 5 mW. Emission and excitation spectra were recorded using an Edinburgh Instrument FLS 980 fluorometer system having a steady-state xenon lamp source. On the other hand, the lifetime was measured with a pulsed microsecond xenon lamp source having a frequency range of 1–100 Hz using the time-correlated single-photon counting technique. A 150 mm BenFlect coated integrating sphere was employed to measure the absolute quantum yield. The spectral sensitivity of the spectrometer and the sphere was modified using a calibrated lamp for spectral light throughput. Finally, RL spectra were acquired with a silver X-ray source tube built in to the Edinburgh Instruments FLS 980 fluorometer system at a power of 12 W (60 kV and 200 μA).