Surface-Doped Zinc Gallate Colloidal Nanoparticles Exhibit pH-Dependent Radioluminescence with Enhancement in Acidic Media

As abnormal acidic pH symbolizes dysfunctions of cells, it is highly desirable to develop pH-sensitive luminescent materials for diagnosing disease and imaging-guided therapy using high-energy radiation. Herein, we explored near-infrared-emitting Cr-doped zinc gallate ZnGa2O4 nanoparticles (NPs) in colloidal solutions with different pH levels under X-ray excitation. Ultrasmall NPs were synthesized via a facile hydrothermal method by controlling the addition of ammonium hydroxide precursor and reaction time, and structural characterization revealed Cr dopants on the surface of NPs. The synthesized NPs exhibited different photoluminescence and radioluminescence mechanisms, confirming the surface distribution of activators. It was observed that the colloidal NPs emit pH-dependent radioluminescence in a linear relationship, and the enhancement reached 4.6-fold when pH = 4 compared with the colloidal NPs in the neutral solution. This observation provides a strategy for developing new biomaterials by engineering activators on the nanoparticle surfaces for potential pH-sensitive imaging and imaging-guided therapy using high-energy radiation.

pH value plays a pivotal role in modulating cellular behaviors, including cell metabolism, proliferation, apoptosis, vesicle trafficking, etc. An abnormal acidic pH symbolizes the dysfunctions of cells and diseases, such as cancer, Alzheimer's, and other neurodegenerative diseases. 1−4 Acid pH also increases the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection. 5,6 Sensitive acidity detection and accurate pH measurements are highly desirable in molecular and biomedical research.
Fluorescence-based approaches for pH detection have been developed, which depend on fluorescence signals using nanoprobes. 7−10 However, their in vivo applications are limited due to the low penetration depths compared with clinically approved high-energy excitations such as X-or γ-rays for imaging or therapeutics.
Recently, nanophosphors have been explored as agents in radiological applications, especially in vivo tumor therapeutics. 11−13 The combination of insignificant scattering of X-rays in tissues and the high tissue penetration of near-infrared (NIR) optical photons emitted from the phosphors open a pathway of achieving deep tissue optical imaging in vivo with unprecedented high spatial resolution. 14 Chen et al. designed pH-responsive luminescent Gd 2 O 3 S:Tb core/shell nanocapsules that could release doxorubicin drug in an acidic media, which is applicable for cancer therapy due to the low-pH environment in tumors. 15 With an almost free background, Xray radioluminescence (RL) showed enhanced intensity in the low-pH medium, possibly because of optical absorption of the luminescent material or energy transfer. Using NaGdF 4 :Eu 3+ NPs, Sudheendra et al. found that the X-ray RL decreased with a decrease in pH of the NP solution, possibly due to the surface-coated molecules. 16 It has been reported that pH sensor films could be used at specific locations to indicate the pH level using X-rays, 17,18 while a pH-responsive solution is of practical significance for in vivo application, ideally with RL enhancement at a low pH value using X-rays, since the enhanced RL would be an indicative measure of pH and a guide to locate the target for therapeutic imaging. In this work, we report the first observation of pH-dependent enhanced Xray RL using Cr-doped zinc gallate ZnGa 2 O 4 (ZGO) colloidal NPs in low-pH acidic media.
The zinc gallate, a cubic spinel oxide, is an appealing host material for a broad range of optical and biological applications. 19 Cr-doped ZGO (ZGO:Cr) NPs are persistent luminescent nanoparticles (PLNPs) with long-lasting emissions, emitting NIR around 700 nm with high brightness that is suitable for in vivo imaging, as it corresponds to a transmission maximum for biological tissues. 20,21 The ZGO-based materials produced by traditional solid-state reactions or sol−gel methods require subsequent calcination, causing large particles that are unsuitable for biological applications. In recent years, solution synthesis appears promising, since the samples are only heated in the moderate temperature range of 150−300°C , with controllable sub 10 nm size with narrowed size distribution. 22−25 High-energy excitation is used to study the RL of these materials, although limited compared to PL. Song et al. synthesized Cr-and W-codoped ZGO via a hydrothermal method and observed higher luminescence intensity upon Xray excitation since W improved the X-ray photon absorption efficiency and provided additional electrons to Cr 3+ . 26 On the other hand, Beke et al. synthesized a ZGO/SiC core/shell structure with a size of 9−9.5 nm and observed that the inclusion of SiC in the core could enhance X-ray RL, as SiC was also excited by X-rays that provided electrons transferred to Cr 3+ to enhance the luminescence. 27 Despite the high performances of the solution-synthesized ultrasmall NPs, the detailed structure of the NPs regarding the dopant distribution is unknown, and research on their luminescence properties related to particle size and processing requires much exploration.
In this work, we present a well-controlled synthesis of ZGO:Cr NPs with activator Cr 3+ ions located on the nanoparticle surface (Figure 1a), and we demonstrated that such ultrasmall NPs have the potential for biomedical applications (Figure 1b). When X-rays interact with NPs, different X-rays (transmitted, Compton scattered, coherent scattered, and fluorescent), electrons (secondary and Auger), holes, and other signals are produced (Figure 1c). We found that the luminescence mechanism by this interaction confirms the activator distribution on surface of NPs. Further, our prepared colloidal NP solutions form smaller aggregates in a neutral solution while larger aggregates form in acidic solutions (Figure 1d,e). Experimentally, the RL from the colloidal solutions exhibits a novel pH dependence with a linear relationship, and the acidic medium enhances the luminescence intensity by 4.6-fold at pH = 4 in comparison to a neutral solution with pH = 7.2. Since a low pH value of the acidic medium is an indication of cell dysfunction or disease, such as tumors, this observation indicates potential applications of the PLNPs with activators on the surface for bioimaging or imaging-guided therapy upon high-energy activation.
The hydrothermal method yields high-crystallinity and highly luminous NPs with easy control of the NP size. With the low addition of 1.75 vol % of NH 4 OH precursor in the synthesis, heating at 180°C for 4, 8, and 20 h yields NPs with diameters of 6.1, 8.1, and 11.5 nm, respectively, while with 2.55% and 3.29% of NH 4 OH in the synthesis, heating at the same temperature for 20 h yields NPs with increased size of 14.2 and 18.5 nm, respectively (Figures S1 and S2). The higher concentration of NH 4 OH promoted the growth of the ZGO lattice according to eqs S2−S4.
The ZnGa 2−x Cr x O 4 sample synthesized with a low NH 4 OH concentration of 1.75% for a short time of 4 h has the smallest diameter, which exhibits the highest performance in the RL when x = 0.01 (denoted as ZGO:0.01Cr thereafter). We first present a characterization of this sample using transmission electron microscopy (TEM) and first-principles density functional theory (DFT) calculations. An X-ray diffraction (XRD) pattern of the as-prepared NPs is shown in Figure 2a, which matches the standard 04-19-5774. The Rietveld refinement reveals a pure phase with a lattice parameter of a = 0.8386(3) nm. Figure 2b shows a selected-area electron diffraction (SAED) with poly rings. Intensity profiles of the reflections are produced from the center beam, and after subtracting the high background, the profiles are similar to the XRD pattern. 29 Figure 2c shows the compositional analysis by X-ray energy-dispersive spectroscopy (EDS). An evident Cr peak is identified, and the quantitative analysis is consistent DFT calculations are conducted to confirm the formation of the Cr dopant on the surface. A recent work indicated that the Cr−O bond length is significantly longer than those of Ga−O and Zn−O, although Cr 3+ is slightly smaller than Ga 3+ . 28 Here, we evaluate the stability of Cr on the location of the crystal from the surface to volume by selecting a slab of 1 × 1 × 5 supercells with a vacuum space in 1.6672 nm isolating the crystal lattices (inset in Figure 2f). Ga is on the termination of the crystal surface. The formation energy E f of the Cr dopant is given by where E ZGO:Cr , E ZGO , E Ga , and E Cr represent the total energies of Cr-doped ZGO, undoped ZGO, pure element Ga, and pure element Cr, respectively. The total energies are calculated by using the Quantum ESPRESSO program, and the detailed results are given in Table S1. We only replace one Ga with Cr at the 21 locations, as indicated in the inset in Figure 2f.
The computed E f is plotted in Figure 2f. It is found that, on the crystal surface, E f is significantly lower than the locations in the volume, indicating that Cr is more stable on the surface compared with its location in the volume. With a low concentration of NH 4 OH which leads to a slow growth rate of ZGO, during the slow furnace cooling process, Cr 3+ ions attend the reactions last at lower temperatures so that they   Figure 3a with different Cr 3+ doping levels (x = 0.005, 0.01, 0.02, and 0.05), while the excitation spectra are shown in Figure S4a. The undoped ZGO sample showed a broad band at 350−620 nm with a maximum peak around 455 nm ( 2 E B → 4 A 2 ), which is caused by self-activation originating from the partial substitution of the Zn 2+ site by Ga 3+ in the spinel structure. 30 The predominant Cr 3+ emission comes from the 2 E → 4 A 2 transition. The optimum concentration is found to be 0.01Cr, and beyond this composition, the emission intensity is reduced by concentration quenching.
The PL spectra of ZGO:0.01Cr, synthesized under varied NH 4 OH concentrations and heating times, are presented in Figure 3b, showing the size effect. Their excitation spectra are shown in Figure S4b. It is found that as the particle size increases, the PL emission also increases, possibly because of reduced surface areas where surfactants of hydroxyl radicals are attached, acting as quenchers. 31 From the Fourier transform infrared (FTIR) spectrum in Figure S5, the presence of hydroxyls is evident. It was recently reported that subsequent annealing of as-synthesized 10 nm NPs at 800°C increased the PL emission intensity, possibly related to the removal of surfactants at the high temperature and redistribution of Cr ions for optimum emission. 25 However, after high-temperature annealing, the NPs are severely aggregated.
The RL spectra of these samples are shown in Figure 3c. These emission peaks still show the characteristic emissions of Cr 3+ , while the shape of the emission appears as the N2 and phonon sideband (PSB). Note that the emission from the host disappears, indicating a thorough energy transfer from the host to the Cr 3+ ions. The RL and PL have different luminescence mechanisims. 32 As shown in Figure 1c, within the NPs, electrons and holes are generated by incoming X-rays. The electrons/holes can reach the surface to excite the activators, while they cannot travel a long distance in the lattice due to their limited energy (estimated to be less than 1 nm for energy up to 200 eV). Thus, as the particle size increases, a larger portion of electrons and holes are annihilated in the volume, and they do not contribute to the excitation; thus, these NPs, with the same 1.75% NH 4 OH addition, show reduced emissions as the size increases. However, with higher concentrations of NH 4 OH, as the NH 4 OH promoted the crystal growth, the activators incorporated in the particle volume instead of on the surface; more activators could be excited by X-ray excited electrons and holes in the volume to enhance the RL. Note that the X-rays and electrons that escaped from the particles can also excite the particles nearby by multiple scattering, causing even higher emission if the activators are on the surface. Therefore, both the PL and RL data support the activator surface distribution on the NPs synthesized with a low NH 4 OH concentration.
The emission stability of ZGO:0.01Cr NPs (6.1 nm in diameter) irradiated with X-rays at a maximum fluence for 10 min is shown in Figure 3d. No emissions appear in the range of 360−600 nm, while emissions from the 2 E → 4 A 2 transition of Cr 3+ (∼696 nm) are almost steady without any peak shifts over

Nano Letters pubs.acs.org/NanoLett
Letter the tested period. A stability test of emission spectra of the ZGO host is shown in Figure S6. Figure 3e plots the integral intensities (area under the curve), which are almost stable with a small reduction from 2.3−4.1% that is better than the stability of a halide. 33 Photos of the tested samples are inserted in Figure 3c, where undoped ZGO exhibits a blue color and ZGO:Cr exhibits a red color, over the long-term irradiation. Colloidal solutions are prepared by dispersing ZGO:0.01Cr (6.1 nm synthesized under 1.75% NH 4 OH and 4 h heating) NPs in phosphate-buffered saline (PBS) solutions with different pH values at an optimum concentration of 3 mg/ mL. The UV−vis absorption spectra are shown in Figure S7. The PL emission spectra of colloidal solutions are recorded as shown in Figure 4a, and excitation spectra are shown in Figure  S8. Note that the surfactants on the NP surfaces absorb energy and the absorption peak energy in Figure S7 is higher than that of the excitation energy in Figure S8, while the solution with pH = 4 shows the highest absorption in Figure S7. All luminescence measurements are performed using a fresh aliquot of NPs in the PBS at pH = 7.2, 6.5, 6, 5, and 4. The PL shows emission at 696 nm due to the 2 E → 4 A 2 transition, whereas the emission from water is negligible. In addition, selfactivation is pronounced. The process involves an energy transfer from the Ga 3+ ion placed in the octahedral sites toward its first six neighbors. 34 This blue emission exists in all specimens at different pH values but is lower than that of the NPs in the water-like medium. As shown in the decay curves in Figure S9, the emission lifetime is in the millisecond range. With an increase in acidity, the average lifetime increases, indicating more surface defects on larger aggregates. Lowering the pH from 7.2 to 4 with increased acidity yields a slight increase in the red emission. A close observation indicates a slight enhancement of the emission as the acidity increases (inset in Figure 4a). This observation suggests that the UV energy is not sufficient to excite the colloidal NPs, as the emission from the host is significant and is not thoroughly transferred to the Cr 3+ ions as it is in the powders, and thus energetic X-rays may be used to examine the colloidal solutions.
Dynamic light scattering (DLS) and zeta (ζ) potential experiments are performed to study the colloidal solutions. As shown in Figure 4b and Table S2, hydrodynamic sizes are found to be in the range from 38.2 to 73.5 nm. At neutral pH = 7.2, the measured distribution data have a very intense peak around 38.2 nm. The polydispersity index (PDI) is around 0.15, suggesting monodispersed colloidal particles. As shown in Figure 4b, the hydrodynamic diameter increases in acidic media when the pH is reduced from 7.2 to 4, indicating the formation of larger aggregates causing multiple scattering effects to enhance the emission. In the acidic medium, the surface negative charges of nanoparticles with hydroxyl radicals are reduced, affecting the aggregate equilibrium formed in the neutral solution because of reduced repulses, resulting in a larger aggregation. The PDI also increases with the reduction of pH. The ζ potential measurements reveal the role of surface charge and interactions at different PBS-buffer solutions. 35 The ζ potentials are given in Table S2. The ζ potential indicates the dispersion stability of particles in colloids, ranging from 36.1 to

Nano Letters
pubs.acs.org/NanoLett Letter 52.2 mV when the pH is reduced from 7.2 to 4. The pH has a pronounced effect on the ζ potentials. The intense charge on the surface increases the charge−charge repulsions between the particles, thus maintaining a stable and monodisperse suspension.
The colloidal solutions are further tested under X-ray excitation. The background emissions are shown in Figure S10, which are negligible to the intensities from the dispersed NPs. The RL spectra are shown in Figure 5a, which display an intense peak at around 693 nm with a few broad shoulders on the side of the near-infrared region. The emission from the host was significantly reduced compared with the PL in Figure  4a, although it is still present. The integral intensities are plotted vs the pH values, as shown in the inset of Figure 5a. The emission increases by 4.6-folds with the acidity when the pH decreases from 7.2 to 4. As illustrated in Figure 1d,e, larger aggregates will show higher emission than smaller aggregates due to the multiple scattering effects and possibly favorable energy transfers between the linked NPs.
The stability of the colloidal solution with pH 5 is tested, as shown in Figure 5b. Each spectrum was collected every 10 s, and the dispersed NPs in the cuvette were exposed for 10 min. A close observation depicts invariable changes in the emission intensity at 696 nm, clearly suggesting that NPs remained dispersed and emission centers were not settled. The integrated intensity measurements show that the colloidal intensity is reduced by only 1.02% (Figure 5c), which is even less than that of the powder samples (Figure 3e). This experiment demonstrated that the experimental materials have the potential for applications in pH-sensitive cells, where pH is a crucial factor in developing novel therapeutics such as nanoparticle-assisted X-ray photodynamic therapy. The sensitive detection of pH through the X-ray emissions can provide a general guide to distinguish local disease intracellular environment through the pH-responsive agents.

■ ASSOCIATED CONTENT Data Availability Statement
The data that support the plots in the manuscript are available from the corresponding authors upon reasonable request.