Synthesis, characterization and photothermal analysis of nanostructured hydrides of Pd and PdCeO2

Hyperthermia was shown to be an important co-adjuvant therapy to conventional cancer treatments. Nanoparticles can be used in the hyperthermia therapy to improve the localized absorption of energy imposed by external sources, in order to kill tumor cells solely by the effect of heat and with minimum thermal damage to surrounding healthy cells. Nanoparticles can also serve as carriers of drugs that specifically act on the tumor when heated, including hydrogen that can be desorbed to locally promote an antioxidant effect and reduce the viability of cancer cells. In this context, palladium hydride nanoparticles emerge as promising materials for the hyperthermia therapy. In this study, palladium nanocubes (PdNC) and PdCeO2 nanoparticles were synthesized. Nanofluids produced with these nanomaterials were hydrogenated and then tested to examine their photothermal effects. Nanofluids made of PdHx nanoparticles presented significant temperature increases of more than 30 °C under 3 min of diode-laser irradiation. On the other hand, nanofluids with PdCeO2H nanoparticles presented temperature increases around 11 °C under the same experimental conditions. The behavior observed with the PdCeO2H nanofluids can be attributed to the effect of H+ in reducing Ce+4 to Ce+3.


Scientific Reports
| (2020) 10:17561 | https://doi.org/10.1038/s41598-020-74378-1 www.nature.com/scientificreports/ When compared to their bulk versions, nanoscale metallic hydrides present faster hydrogen absorption and release reaction kinetics, due to their larger surface area and lower desorption temperature and activation energies. Thus, nanocrystalline palladium reacts fast for the formation of palladium hydride (PdH x ). It can also release hydrogen and regenerate the metal at a high rate, as well as be recharged afterwards 24 .
Palladium nanocubes and nanoparticles of PdCeO 2 were manufactured in this work. The hydrogen diffusivity of PdCeO 2 is smaller than those of pure palladium in crystalline or nanocube forms. This is due to the hydrogen trap effect of CeO 2 in the Pd matrix, which delays the hydrogen desorption. Therefore, nanoparticles made of PdCeO 2 can potentially increase the period between production and its practical application, such as in the hyperthermia treatment. This material is also non-toxic, presents good bio-affinity and is a potent antioxidant for reactive oxygen species (ROS) 25,26 . Azambuja et al. 27 performed the synthesis of nanostructured Pd soaked with CeO 2 in its matrix through internal oxidation.
We note, however, that the use of palladium nanoparticles to treat diseases still has its limitations. This is due to the fact that the produced nanoparticles may contain materials that are not biocompatible, like metallic powders, salts (acetate, chloride) and other contaminants derived from the synthesis 28 .
The present work aims at synthesizing and characterizing palladium nanocubes (PdNC) and the nanostructured PdCeO 2 alloy. It also aims at evaluating the photothermal effects of nanofluids obtained with the dispersion of these nanoparticles in distilled water. Some of the nanofluids were bubbled with hydrogen in order to form hydrides of Pd nanocubes and PdCeO 2 nanoparticles. Nanofluids with different concentrations were continuously heated with a diode-laser (829.1 nm) during experiments with controlled powers of 154 mW, 218 mW and 439 mW. Figure 1a,b show the hydrogen permeation curves for pure Pd and PdCeO 2 , respectively, obtained at room temperature under a current density of 5 mA/cm 2 . These figures provide the hydrogen diffusivity of Pd as 3.3 × 10 -11 m 2 s −1 (Fig. 1a), which is almost six times larger than that of PdCeO 2 (0.6 × 10 -12 m 2 s −1 -see Fig. 1b). Thus, PdCeO 2 absorbs less hydrogen during the hydrogenation period. The hydrogen diffusivity decreases with the addition of CeO 2 in Pd, due to the accommodation of this oxide in the Pd matrix. The Ce-oxide lattice has  In this case, the interface between CeO 2 and Pd is semicoherent and can effectively trap hydrogen atoms. Our objective with the addition of CeO 2 in Pd was to reduce the hydrogen diffusivity using a biocompatible material, in order to allow longer time periods between production and use of the hydrides. Besides that, its valence also plays an important role in relation to the hydrogenation, as discussed below. Figure 1c,d show the X-ray diffraction patterns obtained for pure Pd and PdCeO 2 , respectively, before and after hydrogenation. The displacement of the peaks to the left, which represents a significant absorption of hydrogen, can be observed in Fig. 1c,d. The intensity of the diffraction peaks in the hydrogenated PdCeO 2 alloy is lower than in the original material. Hydride precipitation, when it occurs, decreases the diffraction intensity of the primary phase. Although the hydride peaks are not clear, the shift to the left and the broadening of Pd peaks reveal the beginning of the phase transformation.

Results and discussion
From the diffraction patterns of Fig. 1c,d it was possible to determine with Bragg's Law the lattice parameters that are presented in Table 1. These lattice parameters reveal that the amount of hydrogen absorbed was 0.15% H/Pd for PdNC and 0.11% H/Pd for PdCeO 2 . Hydrogen at low concentrations is sufficient to form hydrides but does not necessarily form the extra peaks related to the reflection of the PdH x phase. These results obtained here are in accordance with those available in the literature 3,29 .
Figure 2a-f show the microstructural analysis for both materials obtained via SEM and TEM, respectively. The formation of Pd nanocubes with mean size 20 nm can be observed by SEM (Fig. 2a) and by TEM (Fig. 2b,c). The nano Ce oxides precipitated in the Pd matrix present nanospheres and large needles 27 . After milling, only these nanoprecipitates were observed. The CeO 2 was always incorporated within the Pd matrix (Fig. 2e,f).
The use of CTAB in the synthesis of the nanocubes is important to stabilize the cubic morphology in the presence of bromide 30 . In addition, in a synthesis starting from Na 2 PdCl 4 the CTAB has another important role, by reacting with the precursor and forming a complex as shown by reaction I. The addition of an excess of ascorbic acid to the colloidal complex reduces the dissociative organic salt and shifts the reaction balance slowly toward dissolution (reaction II) 31 . After mixing CTAB with Na 2 PdCl 4 , the bromide ions (Br − ) dissociated from the surfactant can replace the chloride ions (Cl − ) and bind to Pd 2+ , which then reacts with CTAB + to form an organic salt (reaction I). Figure 2g shows the EDS mapping of Pd nanocubes, where the presence of only palladium can be observed, free of contaminants from the organic reagents used in the reaction. Figure 2g shows the presence of clusters of Pd nanocubes. Under higher magnification, the cubic morphology becomes well defined. Figure 2h shows a CeO 2 particle bound to Pd. In the nanostructured PdCeO 2 alloy, the presence of iron due to the milling process was observed. Although being undesirable in the alloy, Fe would not be a concern for the hyperthermia cancer treatment application of the nanocubes, because it is a biocompatible element.
Zhao et al. 3 presented the absorption spectrum of the palladium nanocubes and their respective hydrides up to a concentration of 0.06 g/L, in the UV-VIS-NIR, showing that the palladium nanocubes absorb more strongly in the ultra-violet (UV) range. In the presence of hydrogen, there is an increase in the absorption curve in the visible and near infrared (VIS-NIR) ranges, starting from 500 nm. Thus, the nanostructured palladium hydride has great potential for the hyperthermia therapy, since it has a large absorptivity in the NIR range and allows a controlled release of hydrogen.
In order to evaluate the photothermal effects of the nanoparticles produced in this work, distilled water nanofluids were prepared and tested under diode-laser heating, as described in the next section. Figure 3a   www.nature.com/scientificreports/ of 0.2 g/L under 154 mW heating presented maximum ∆T = 9 °C) than those presented by the nanofluids of Pd nanocubes of high concentration (Pd nanocubes with concentration of 0.4 g/L under 154 mW presented maximum ∆T = 8 °C). This demonstrates that the effect of hydrogen desorption on the temperature variation is predominant over that of the absorption cross section resulting from the nanoparticles in the fluid. At the highest power of 439 mW, the concentration of nanoparticles has a more significant effect; the larger concentrations resulted in larger temperature variations, regardless the nanoparticles (PdNC or PdH x ) in the nanofluids. Therefore, for larger concentrations the effects of the larger absorption cross sections of the nanofluids are more important for the temperature increase than the effects of the hydrogen desorption. In addition, the curves for the nanofluids of Pd hydride nanocubes exhibited different profiles, presenting a steeper temperature variation at small times due to the presence of hydrogen, followed by profiles similar to those of the nanofluids of PdNC.    Table 2 shows that the presence of hydrogen in the nanofluids of PdCeO 2 had the opposite effect. A reduction in the energy absorbed by the nanofluid can be noticed, which reduced the temperature variation from 27.1 °C for the PdCeO 2 nanofluid to 11.2 °C for the PdCeO 2 hydrogenated nanofluid. This result might have been caused by the agglomeration of the nanoparticles after hydrogenation, in addition to possible changes in valence for cerium from Ce +4 to Ce +3 . In fact, there was a modification of the color of the PdCeO 2 nanofluid when it was hydrogenated, as shown by Fig. 5.
As an oxide, the most stable form of cerium is Ce +4 , where oxygen atoms occupy the tetrahedral positions and a portion of Ce +3 has the positive charge deficiency compensated by oxygen gaps 25 . The concentration of Ce +3 increases with the decrease in particle size 25 and cerium oxide nanoparticles exhibit a significant amount of this ion. Therefore, the reacting medium will be the determining factor for the oxidation or reduction effects of cerium oxide nanoparticles. The excess of hydrogen causes Ce +4 to react with the hydrogen ions that were dissociated at the palladium surface, thus forming H + that reduces Ce +4 to Ce +3 (reaction III) 32 . This fact modifies the absorption of the material in the NIR, resulting in a lower temperature increase than that of its non-hydrogenated form.
Das et al. 32 demonstrated with XPS that Ce +4 and Ce +3 coexist in cerium oxide nanoparticles and that the presence of these two valence states on the nano-Ceria surface act as an antioxidant. Nanoparticles then eliminate the free radicals from the culture system, which gives the cerium oxide nanoparticles unique properties for biomedical use. In addition, they revealed that nano-Ceria exhibit auto-catalytic or cyclic regeneration, that is, in the presence of H 2 O 2 it is possible to regenerate the Ce +3 → Ce +4 → Ce +3 system. Therefore, this material has great potential for biological applications with antioxidant activity and pseudo-infinite half-life. Hydrogenation induces the change from Ce +4 to Ce +3 by increasing the hydrogen volume 33 and modifying material properties.
The photothermal conversion efficiencies of the nanofluids produced in this work are presented in Table 3. These values were obtained with the solution of an inverse problem by using the transient temperature variations with the laser power of 439 mW. These efficiencies follow the same trend of the temperature variations (see also Table 2). For PdNC, the efficiencies increased with the hydrogenation of the nanofluids, but the opposite behavior was observed with PdCeO 2 nanoparticles. Also, the efficiencies increased with the concentration of PdNC. The maximum efficiency (70%) was obtained with the hydrogenated PdNC nanofluid with concentration of 0.4 g/L (III)   Table 3 for PdNC nanofluids are in accordance with those reported in the literature 3 . The photothermal conversion efficiencies obtained with the other laser powers were similar to those presented in Table 3, but exhibited larger standard deviations; they are not reported here for the sake of brevity. Figure 6 summarizes the temperature variations for nanofluids of PdNC, PdCeO 2 and their hydrogenated forms, obtained with the laser heating power of 439 mW. Although the absorption of the diode-laser energy in the hydrogenated PdCeO 2 nanofluid was smaller than for the other nanofluids (Table 3 and Fig. 6), the resulting temperature increase might still be appropriate for the hyperthermia treatment of cancer depending on the aimed application, which may involve mild temperature increases to make the tumor cells more susceptible to chemotherapy or radiotherapy treatments.
The results obtained in the present work demonstrate the immense potential of Pd nanocubes and nanoparticles of PdCeO 2 , as well as their hydrogenated forms, in applications such as the hyperthermia therapy of cancer.

Methods
Hydrogen permeation tests were performed at room temperature with foils about 150 µm thick, of pure palladium and PdCeO 2 . A cathodic current density of 5 mA cm −2 , which was enough to produce hydride on the surface of the cathodic loading side, was applied to each material. In order to calculate the apparent hydrogen diffusion coefficient, electrochemical permeation tests were performed using a double-cell experiment separated by the sample 34 . An oxidation cell (detection side) was filled with NaOH 0.1 M and the electrochemical potential used was obtained from open potentiometric circuit. A reduction cell (cathodic charging side) was filled with 0.1 M H 2 SO 4 + 2 mg/l As 2 O 3 solution. The anodic current was detected on the cell oxidation side. Both currents were generated or detected by an AUTOLAB PGSTAT100N potentiostat.
The apparent hydrogen diffusion coefficient (D app ) was calculated according to 18 : where D app is the apparent diffusion coefficient, L is the thickness of the sample and t b is the breakthrough time, which corresponds to the time when the first hydrogen atoms permeating through the sample are detected. For single-phase samples without hydride formation, the permeation curve is sigmoidal. When hydride reaction takes place, the hydrogen permeation curves exhibit a different evolution, which corresponds to hydride nucleation and growth during the test.
Palladium nanocubes (PdNC) were prepared through a precipitation technique using 10 mM aqueous of Na 2 PdCl 4 (disodium tetrachloropalladate, Sigma-Aldrich 98%) and CTAB (cetyltrimethylammonium bromide, Sigma-Aldrich 98%) added to a solution 0.1 M of ascorbic acid (Sigma-Aldrich 99%). The final solution was maintained under magnetic agitation for 1 h at room temperature. After the reaction, the sample was washed several times and then air dried.
The alloy containing Pd and 3 wt % of Ce was melted in an arc-furnace under inert atmosphere. The alloy was then cold-rolled up to 200 µm of thickness and 15 mm × 300 mm ribbon plates were cut. These were subjected to thermal treatment at 800 °C for 24 h to promote oxygen diffusion in the Pd matrix and the formation of Ce-oxides. The oxide grows in needle or plates shapes, in accordance with the [110] direction of Pd, due to the perfect accommodation between the lattice parameters of CeO 2 (5.42 Å) and the diagonal of the Pd cubic cell (5.50 Å) in the (110) plane, as shown in Fig. 7. After the oxidation step, the PdCeO 2 alloy was hydrogenated www.nature.com/scientificreports/ and submitted to mechanical milling using a ball mill at 300 rpm for 12 h. Hydrogenation was performed to facilitate the comminution of the alloy. The hydrogen absorbed by the alloy was then desorbed during milling, due to the heating caused by friction between the balls of the mill. An ultrasonic probe was used for dispersion of the PdNC and PdCeO 2 nanoparticles in distilled water for 1 min. PdNC nanofluids with concentrations of 0.2 g/L and 0.4 g/L, as well as a PdCeO 2 nanofluid with concentration of 0.4 g/L, were prepared. Half of the produced nanofluid samples were hydrogenated to form nanoparticles of metal hydride. These nanofluids were bubbled with hydrogen gas at the pressure of 2 bar and constant flow for 1 h at room temperature.
Both nanocrystalline materials were characterized by X-ray diffraction using the Cu-K α radiation with wavelength of λ = 1.5406 Å. SEM analyses for morphological characterization of the nanoparticles were conducted with a SEM-FEG equipped with a FEI QUANTA detector FEG 250. TEM analyses with a FEI TECNAI G2F20 HRTEM and a TITAN G2 80-200a were also performed to determine the morphology of the materials. Identification of the chemical composition of the materials was performed via EDS analysis of fields with a TITAN G2 80-200a.
Tests were performed in order to measure the temperature variation in each nanofluid and in distilled water, under the irradiation of a diode-laser. The fluids were pipetted into one well (diameter of 6.75 mm and height of 11.1 mm) of 96-well plates, with volume of 260 µL. Heat was provided during 180 s by a diode-laser (NEWPORT, model 525B at a wavelength of 829.1 nm), with collimated beam (Collimator THORLABS CFC-2X-B) and for three different powers (154 mW, 218 mW and 439 mW). The laser power was measured with an optical meter (THORLABS PM100D with Standard Photodiode Power sensor S121C) and the laser beam exhibited a Gaussian profile with 3.1 mm of diameter (THORLABS-Compact USB 2.9 CMOS Camera). The collimator was perpendicular to the 96-well plate, coaxial with the well containing the fluid and at a distance of 135 mm from the fluid surface. Temperature measurements of the surface of the fluids were taken with an infrared camera (FLIR A325), with a frequency of one measurement per second and pixel size of 0.24 mm. Figure 8a shows the apparatus used for the photothermal measurements, while Fig. 8b presents a thermal image during the laser irradiation of the well with the nanofluid. Other neighboring wells in the 96-well plate containing distilled water, which were not heated by the diode-laser but served as reference for the temperature measurements, also appear in the figure.
The heating experiments were carried out in triplicates to ensure reproducibility. The curves presented for the temperature variations correspond to the mean values of the three experiments, which were conducted in a room with controlled ambient temperature (23.5 ± 0.5 °C).
The measured temperatures were also used for the calculation of the photothermal conversion efficiencies of the produced nanofluids, by following a procedure similar to that proposed by Roper et al. 35 . While these authors used the transient temperature variation during the cooling period and the steady-state temperature after the heating period to calculate the photothermal conversion efficiency, in this work we used an inverse analysis with the transient temperature measurements during the heating period. The model parameters were estimated with the temperature measurements in experiments with distilled water and with the produced nanofluids. The inverse parameter estimation problems were solved within the Bayesian framework of statistics by using the Markov Chain Monte Carlo (MCMC) method 36 . This method allows for the estimation of mean values and the related uncertainties of the model parameters, through stochastic simulation of the posterior probability distribution function. The values reported in this work for the photothermal conversion efficiencies are the means of the

Conclusions
Palladium nanocubes were synthesized via chemical precipitation in the presence of CTAB as a surfactant. The synthesis method was effective, resulting in nanocubes free of contamination and about 20 nm in size. Internal oxidation and mechanical grinding were also used to produce a nanostructured PdCeO 2 alloy, which presented the formation of nano-Ceria in the palladium matrix. The materials were tested in experiments involving the heating of nanofluids with a diode-laser. Temperature variations of the nanofluids were substantially larger than that of distilled water, demonstrating the photothermal effect of the nanoparticles developed in this work, including those that were hydrogenated. Our results reveal a great potential for the application of these nanoparticles in the hyperthermia cancer therapy.