Tetrahydrofuran, hydrogen peroxide, and cesium chloride were obtained from Sisco Research Laboratories Pvt. Ltd. (Mumbai, Maharashtra, India). Potassium ferricyanide was purchased from Sigma India (Bengaluru, Karnataka, India). The remaining chemicals (analytical grade) were obtained from commercial suppliers such as Sigma-Aldrich India (Bengaluru, Karnataka, India). Milli-Q water was utilized throughout the experiments.

A total of 120 mg of Mesoporous silica Nanoparticles (MSN) (200 nm) was suspended in 300 µL of an aqueous solution of K₃ [Fe (CN)₆] (0.65 M) and mixed over a cyclomixer for 3 h, followed by the collection of potassium hexacyanoferrate-inserted MSN by centrifugation. Potassium hexacyanoferrate-incorporated MSN were suspended in a reaction mixture containing 170 µL of THF (12 M) and 340 µL of H₂O₂ (0.7 M) under stirring over a cyclomixer and allowed to stand for 1.5 h, followed by incubation in an oven at 65°C for 6–7 h. The yellow mesoporous silica nanoparticles (MSNPs) turned deep blue, indicating the pore-inserted PBNPs. Furthermore, PBNP-inserted MSNPs were harvested by repeated centrifugation and washing with double distilled water.

The single precursor synthesis of PBNPs using potassium hexacyanoferrate in a homogeneous medium has been described previously (Pandey and Pandey, 2012; Pandey and Pandey, 2013a; Pandey P. C. and Pandey A. K., 2013; Pandey and Pandey, 2013c; Pandey and Pandey, 2016; Pandey et al., 2018; Pandey et al., 2020; Pandey et al., 2021a; Pandey et al. 2021b). Accordingly, the synthesized PBNPs were characterized using UV-Vis spectrometry (Hitachi U 2900, Tokyo Japan), TEM Tecnai G2 20 S TWIN (FEI, Hillsboro, Oregon, United States), X-ray diffraction (Rigaku MiniFlex 600 desktop XRD system, Japan), TGA (TGA 50, Simadzu Asia Pacific Pte Ltd, Singapore), and XPS (XPS K-alpha, ThermoFisher Scientific, United States). For TEM characterization, the PBNPs were diluted in Milli-Q water, followed by drop casting on a carbon-coated copper grid having a mesh size of 300, and visualized. A thin film of homogenous PBNPs casted on glass slide and powder of PBNP-2 incorporated mesoporous silica were processed for recording of X-Ray diffractogram. Additional characterization of PBNP-incorporated mesoporous silica was performed using a Nova NanoSEM 450 HR-SEM (FEI, Hillsboro, Oregon, United States) and X-ray fluorescence (XRF). Finally, a quantitative analysis of the specific surface area and porosity distribution of the PBNP-incorporated mesoporous silica was performed by using the Brunauer–Emmett–Teller (BET) analysis (BELLSORP MAX II and BELCAT-II, MicrotracBEL Corp., Osaka, Japan).

The PBNP-1, PBNP-2, and PBNP-3–modified graphite paste were made by mixing PBNP-1/PBNP-2/PBNP-3, graphite powder, and Nujol oil. The process of paste fabrication involved mixing 5 µL of PBNP-1/PBNP-2/PBNP-3 with 9 mg spectroscopic grade graphite powder under sonication for 30 min, followed by drying overnight at 60°C in a vacuum oven. Furthermore, active PBNP-modified graphite paste was made by mixing graphite powder [68% (w/w)], PBNP-adsorbed graphite [2.5% (w/w)], and Nujol oil [30% (w/w)]. The electrode body had a well with a recessed depth of 2 mm, which was packed with an active paste.

The stepwise construction of the PB-modified electrode was characterized using electrochemical impedance spectroscopy (EIS) in the absence and presence of cesium ions; this approach provided an efficient mechanism for investigating the interference properties of surface-limited electrodes. The impedance measurement was recorded on a CHI660B electrochemical workstation (CH Instruments, Inc., Austin, TX, United States); recordings were obtained between 1 Hz and 10 kHz at a suitable polarization potential. A semicircle section of the Nyquist plot representing charge transfer resistance and a linear section representing a diffusion-limited process are present in recordings obtained in the absence and presence of cesium ions. The almost straight line that the bare PBNP-modified electrode displayed in the electrochemical impedance spectroscopy (EIS) data can identify a diffusion-limiting stage of an electrochemical process. In this approach, the semicircle diameter corresponds to the electron transfer resistance (Ret). The usage of PBNP-mediated electrochemical sensing of cesium ions was justified by the semicircle diameter of the PBNP-modified electrode in the presence of cesium ions. The Nyquist diagram was plotted on the R(CR)(CR) circuit.

Cesium ion sensing was studied by fluorescence spectroscopy since PBNPs act as quenchers. Our previous study has demonstrated that functional PBNPs synthesized from 3-APTMS and cyclohexanone quench the in situ emission of fluorescence radiation within the visible range, demonstrating the quenching capability of PB (Pandey et al., 2020; Pandey et al., 2021a). However, the stability in the nano-geometry of PB is an important requirement for fluorescent quenching, as PB tends to agglomerate with time, converting the material to a photochemically inactive state (Pandey et al., 2020; Pandey et al., 2021b). Accordingly, the PBNPs, as reported herein as PBNP-1, PBNP-2, and PBNP-3, can be used as a fluorescence quencher of standard fluorophore: fluorescein. It is significant to note that PB refers to the formation of PBNPs from the use of a single precursor that is potassium ferricyanide, thus a double precursor is avoided, which allows an uncontrolled ratio of ferrous/ferric ions and hexacyanoferrate ions to yield a definite proportion of Fe II–C–N–Fe III within the polymeric network to precisely allow the interaction of cesium ions during the sensing process. A Hitachi F-7000 fluorescence spectrophotometer (Tokyo, Japan) was applied to measure the fluorescence in the samples. Since PBNPs behave as light quenchers, it is expected that the emission spectra of known luminescent/fluorescent materials may be significantly quenched in the presence of PBNPs. Fluorescein is chosen for such measurements since it is a well-known fluorescent material. Aqueous solutions of fluorescein of known concentration (10^–7 M) were used under variable reaction conditions. Fluorescence sensing of cesium ions was recorded by monitoring the fluorescence spectra of (i) fluorescein only, (ii) fluorescein and metal PBNPs, (iii) fluorescein, PBNPs, and a variable concentration of cesium ions. The results showed that PBNP-1 made from the use of the cationic polymer polyethyleneimine yielded a cesium ion concentration–dependent fluorescence response when compared with that associated with PBNP-2 and PBNP-3, which is mainly attributed to fluorescence energy transfer in the presence of cesium ions with a cationic polymer as a spacer around PBNPs. The other two PBNPs did not show significant cesium ion concentration–dependent fluorescence quenching.

The PBNPs inserted into a mesoporous support (50 mg) were incubated overnight with different concentrations of cesium ions between 1 ppm and 50 ppm. The adsorbed cesium ions over PBNP-incorporated mesoporous support were separated by centrifugation; the presence of cesium ions in the supernatant was monitored by using ICP spectroscopy. The Kd value was calculated to be 3.5 × 10⁴ even at 3% PB content within the heterogeneous matrix. The PBNP-incorporated mesoporous support acted as a biocompatible adsorbent for cesium ion removal from contaminated water.

The cesium uptake performance of the material was evaluated in a batch that was made by equilibrating 0.1 g of the sorbent and 10 mL of feed solution for 2 h at room temperature, followed by overnight standing. The batch distribution coefficient for Cs was measured using a test solution containing 5 g/L NaNO₃ spiked with ¹³⁷Cs radiotracer. To understand the ion exchange isotherm for Cs, test solutions of varying Cs concentrations ranging from 1 to 1,000 ppm were prepared by adding a requisite amount of CsNO₃ solution in 5 wt% NaNO₃ solutions and spiking them with ¹³⁷Cs tracer. The kinetics of the adsorption was studied using a feed solution having 10 ppm Cs spiked with ¹³⁷Cs tracer. In each case, after equilibration, the solution was filtered using a 0.2-micron syringe filter. The filtrate was analyzed for ¹³⁷Cs by counting with a NaI/Tl scintillation detector that was coupled to a single-channel analyzer.

The major objective of the current study was the nano-geometry–controlled synthesis of PBNPs with three different sizes depending on the reducing agent and to understand their impact on the sensing and removal of cesium ions. Since the traditional synthesis of PBNPs includes the agglomeration of nuclei into large particles, which is associated with poor processability, a good control over the synthetic strategy for PBNP formation is the first stage of the investigation (Pandey and Pandey, 2012; Pandey and Pandey, 2013a; Pandey P. C. and Pandey A. K., 2013; Pandey and Pandey, 2013c). Accordingly, the role of three different reagents, (i) PEI, (ii) THF and H₂O₂, and (iii) 2-(3,4-epoxycyclohexyl) ethyltrimethoxysilane and cyclohexanone, was evaluated. These reagents allowed for the controlled dissociation of potassium ferricyanide into a Fe⁺⁺ derivative that subsequently interacted with potassium hexacyanoferrate, enabling the controlled nucleation of PB in two steps, followed by functionality as an efficient stabilizer to protect PB nano-geometry. The use of these reagents in PBNPs has already been demonstrated (Pandey and Pandey, 2012; Pandey and Pandey, 2013a; Pandey P. C. and Pandey A. K., 2013; Pandey and Pandey, 2013c). However, the use of these materials for the sensing and removal of cesium radionuclides has not been studied to this point. Accordingly, an investigation of the synthesis of PBNP-1, PBNP-2, and PBNP-3 as a homogeneous formulation for the sensing of cesium ions has been demonstrated in this study. Further efficient removal of cesium radionuclides from contaminated samples essentially requires heterogeneous formations of PBNPs; by maintaining the nano-geometry of the PBNP-2 in the heterogeneous phase has been described. TEM characterization confirmed the size and shape of the as-made PBNPs, as shown in Figure 1. PBNP-1 was ~ 6.40 nm in size and spherical in shape (Figures 1A, D, G). PBNP-2 exhibited a rectangular shape and was ~ 60 nm in size (Figures 1B, E, H). PBNP-3 also exhibited a rectangular shape and size of ~ 28.8 nm (Figures 1C, F, I). Accordingly, the XRD results of the synthesized PB plane in both homogeneous and heterogeneous matrices were studied. Figures 2A–C show the XRD patterns for PBNP-1, PBNP-2, and PBNP-3, respectively. Figure 2D shows the XRD result for the PBNP-incorporated mesoporous silica support. According to Figures 2A–C, the strong and distinct peak was shown at 2θ values of 17.42, 24.96, 35.43, 39.54, 43.54, 50.72, and 54.03. The (200), (220), (400), (420), (422), (440), and (600) planes were allocated to these values, indicating an FCC lattice structure (JCPDS no. 73-0687). Since the XRD of the PBNP thin film on a glass surface was obtained, the diffractogram (as shown in Figures 2A–C ) also indicates the presence of silica, as evidenced by a plateau at 2θ values between 20 and 25⁰.

These reagents, particularly THF and H₂O₂, also enable the synthetic incorporation of PBNPs within mesoporous silica since they possess sufficient stability for efficient cesium ion removal. The XRD spectrum, as shown in Figure 2D, again confirms the presence of a sharp peak that was indexed with 2θ values of 17.42, 24.96, 35.43, 39.54, 43.54, 50.72, and 54.03⁰, which are attributed to PB. The HR-SEM images of PBNP-2–incorporated mesoporous silica are shown in Figure 3; the presence of PBNP-2 in mesoporous silica was confirmed by using X-ray fluorescence (XRF), as shown in Table 1.

The thermogravimetric analysis (TGA) profiles for Prussian blue PBNP-1, PBNP-2, and PBNP-3 are shown in Figure 4. The thermal stability of the synthesized PBNPs was investigated using TGA. Figures 4A–C show the TGA results and depict four regions for PBNP-1 and three regions for PBNP-2 and PBNP-3 in which the PB lost weight. For PBNP-1, the first weight loss started at approximately 100°C and ended at 245°C; for PBNP-2, the weight loss occurred at 140°C and ended at 288°C. For PBNP-3, the weight loss started at 235°C and ended at 405°C; these findings can be related to intrinsically absorbed water molecules in the PB moiety being eliminated. The second weight loss for PBNP-1, PBNP-2, and PBNP-3 occurred at 260°C–430°C, 312°C–477°C, and 425°C to 572°C, respectively. The third weight loss for PBNP-1, PBNP-2, and PBNP-3 occurred at 441°C–554°C, 482°C–713°C, and 592°C–700°C, respectively. The fourth weight loss for PBNP-1, which occurred between 562°C and 774°C, was moderated and could be attributed to the presence of soluble and insoluble PBNPs. In PBNP-1, the mass losses at steps I, II, III, and IV were 3.73%, 10.77%, 0.92%, and 0.94%, respectively. In PBNP-2, the mass losses at steps I, II, and III were 2.30%, 3.74%, and 1.71%, respectively. In PBNP-3, the mass losses at steps I, II, and III were 0.53%, 0.34%, and 0.22%, respectively. In step I, mass reductions were attributed to water loss in the PB structure; mass losses in steps II, III, and IV were correlated with the release of cyanide groups from the PB structure.

XPS analysis was conducted on PBNP-1, PBNP-2, and PBNP-3 under ambient conditions. The analysis of the XPS data indicated the existence of both Fe (II) and Fe (III) species in the PBNP samples that were synthesized. The binding energies of PBNP-1, PBNP-2, and PBNP-3 were measured to be 721.04 eV and 707.89 eV, 721.29 eV and 708.25 eV, and 721.18 eV and 708.25 eV, respectively; these values correspond to the Fe 2p1/2 and Fe 2p3/2 orbitals. The observed peaks serve as empirical support for the existence of the distinctive Fe²⁺ moiety within PB. Furthermore, the identification of peaks at binding energies (BEs) of 712.82 eV for PBNP-2, 711.87 eV for PBNP-2, and 712.25 eV for PBNP-3 provides evidence for the existence of Fe⁺³ entities. The X-ray photoelectron spectroscopy (XPS) analysis of all three PBNPs demonstrates that PBNP-1 exhibits a noticeably higher intensity in comparison to PBNP-2 and PBNP-3 [as seen in Figures 5A−C].

The surface area and pore diameter of the as-synthesized catalyst were examined via BET analysis. The N₂ adsorption–desorption isotherms of PBNP-2–incorporated mesoporous silica are shown in Figure 6. The PBNP-incorporated mesoporous silica exhibited a type-IV isotherm; a 40.3 m²/g specific surface area and 7.14 nm pore width were determined according to the adsorption–desorption curve. These results justify controlling the nanoscale features of chemically synthesized PB and the formation of PBNP-incorporated mesoporous silica supports for subsequent removal applications.

The optical properties of PB show broad optical adsorption over 550 nm–800 nm, corresponding to the blue region. It should be noted that fully oxidized PBNP thin films exhibit a lower absorption peak at 760 nm. However, in its fully reduced state, PW absorbs quite weakly in the visible region, resulting in a transparent state. The optical absorption coefficient value of a semiconductor may be represented by its direct/indirect optical band gap. The band gaps previously recorded by Qiu et al. (2020) for PB, PY, and PW have been described in the order of 1.75 eV, 2.02 eV, and 3.53 eV, respectively; these parameters indicate that the optoelectronic properties of PB-related compounds should be controlled for cesium ion sensing. We have previously reported that PBNPs act as fluorescence quenchers (Pandey et al., 2020; Pandey et al., 2021a). Accordingly, we investigated the fluorescence quenching ability of the well-known fluorophore fluorescein, which exhibits excitation and emission spectra at 494 and 521 nm, respectively. Figure 10 shows the fluorescence quenching ability of PBNP-1, PBNP-2, and PBNP-3 on fluorescein emission (Figures 10A–C). This finding clearly predicts excellent fluorescence quenching of PBNPs with a quenching constant as calculated from the Stern–Volmer (S-V) plot (Mitra and Pandey, 2022). The values of the quenching constant were found to be 10,283 M⁻¹ s⁻¹, 15,737 M⁻¹ s⁻¹, and 21,181 M⁻¹ s⁻¹ for PBNP-1, PBNP-2, and PBNP-3, respectively, as indicated in Figure 10D; these results confirm the excellent quenching functionality of the as-made PBNPs. PBNPs cause efficient quenching of fluorescein fluorescence; this phenomenon is likely due to the energy transfer process between fluorescein as the donor and PBNPs as the acceptor. In stable functional PBNPs made using PEI/organotrialkoxysilane/THF + H₂O₂ as the reagent, the reagent protects the nano-geometry of PBNPs for specific applications and behaves as a specific spacer to meet the requirement of energy transfer, as has been previously discussed (Pandey et al., 2020; Pandey et al., 2021b). Accordingly, we further investigated the variation of the fluorescence quenching functionality of the as-made PBNPs in the presence of cesium ions. Notably, PBNP-2 and PBNP-3 showed very poor variation in the fluorescence quenching of fluorescein as a function of cesium ions; on the other hand, PBNP-1 displayed cesium ion concentration–dependent quenching functionality, as shown in Figure 11. The quenching ability of PBNP-1 decreased with increasing cesium ion concentrations. The reason behind this variation might be associated with the charge available around the spacer that might alter the quenching ability of PBNPs in the presence of charged cesium ions. The neutral spacer may not participate in altering the field quenching ability of PBNP; on the other hand, the charged spacer-like cationic polymer PEI may alter the quenching ability of PBNPs. Since PBNP-1 is derived from a cationic polymer, polyethyleneimine, an excellent result of the decrease in quenching ability of PBNP-1 has been recorded as a function of cesium ion concentration (Figure 11A). A linear relationship between cesium ion concentration and quenching extent in terms of percentage quenching of PBNPs was recorded, as shown in Figure 10D, justifying the innovative results from cesium ion sensing via fluorescence measurements. We further examined the difference in the quenching ability of PBNPs and K₃[Fe (CN)₆], if any, to understand the impact of making PBNPs from the single precursor potassium ferricyanide (Figure 11B). Figure 11B shows the results of this study, which indicate that K₃[Fe (CN)₆] is not a fluorescent quencher, and confirms the requirement of both MII and MIII Fe ions for fluorescence quenching functionality.

The name “Prussian blue” refers to polynuclear hexa-cyanometallates, which have a cubic crystal structure with a three-dimensional ….N-MIII-N-C-MII-C…. configuration and form a supramolecular network. If MII and MIII are paramagnetic transition metal ions, then they introduce magnetic properties such as ferri-3,4 and ferromagnetism (Kohler and Storcheva, 2015). Accordingly, PBNPs may display superparamagnetic characteristics as a function of temperature, as shown in Figure 12. The synthesized PBNPs displayed superparamagnetism with high saturation, zero coercivity, and remanence, as shown in Figure 12A; M(T) for zero field cooling (ZFC) is shown in Figure 12B; and the field cooling (FC) program is shown in Figure 12C at different field strengths (i.e., 25, 50, and 100 Oe) with blocking temperatures of 4.3 K, 4.4 K, and 4.2 K as appropriate for a superparamagnetic material. Magnetization hysteresis loops at 2 K, 5 K, and 10 K are shown in Figure 12D, again justifying the superparamagnetic characteristics.

Even at room temperature, the paramagnetic behavior is relatively low. Cesium ions residing in the cubic holes of the lattice may balance out the charge differences between MII and MIII, altering the paramagnetic properties. Accordingly, we investigated the change in the paramagnetic ability of PBNPs as a function of cesium ion concentration. Notably, the paramagnetic functionality of PBNPs in both homogeneous and heterogeneous phases (within the mesoporous matrix) is altered dramatically in the cesium ions; this result provides excellent information on the sensing and removal of cesium ions based on PBNP formulation in the homogeneous and heterogeneous phases. Figure 13 displays the variation in the magnetic property of PBNP powder as a function of cesium ion concentration, indicating the dependence of the paramagnetic character of PBNPs on cesium ion concentration. Furthermore, the PBNPs present in heterogeneous phases, especially within mesoporous silica nanoparticles, may serve as a potential adsorbent for cesium ion removal; accordingly, we investigated the ability of PBNP-incorporated mesoporous silica for cesium ion sensing based on magnetic measurements. Cesium ions of variable concentrations of 10–600 ppm were incubated with PBNP-incorporated mesoporous silica, followed by collection of the heterogeneous matrix, washing, and drying. The magnetic measurement of cesium-adsorbed PBNPs within mesoporous silica was examined, as shown in Figure 14. The results clearly indicate that the magnetic properties of PBNPs vary as a function of cesium ion concentration, which is associated with an increase in unpaired electrons in PBNPs. These findings provide insights into the effective adsorbent behavior of PBNP-incorporated mesoporous silica in radioactive cesium removal from contaminated water.

One of the major tasks of the current study was to evaluate the materials for use as an efficient biocompatible matrix for ¹³⁷Cs removal from contaminated water. Adsorption is a cost-effective approach to remove cesium ions from contaminated water. The transport of cesium from contaminated water to the surface of a solid matrix is implied by this separation process (i.e., PBNP-incorporated mesoporous silica as an adsorbent). A material for this application should possess a tailored surface chemistry and porosity to obtain effective cesium ion separation (as evident from Figure 3); this approach justifies the synthetic incorporation of PBNPs within mesoporous silica. The results recorded in Figure 2 on XRD and PBNP-incorporated mesoporous silica and the BET results shown in Figure 6 clearly confirm the formation of a PBNP-incorporated mesoporous silica matrix as a potential ¹³⁷Cs ion adsorbent. After the magnetic measurements were obtained, we examined the binding ability of PBNP-incorporated mesoporous silica–based materials using energy dispersive X-ray analysis. Figure 15 shows the findings from the adsorption of cesium ions at different concentrations that were present in contaminated water. The data were obtained in the absence of cesium ions (Figure 15) and in the presence of 300 ppm cesium ions. Figure 15 clearly confirms the efficient binding ability of PBNP-incorporated mesoporous silica for cesium ion removal. These findings are further confirmed by the measurement of cesium ions in the supernatant of these four samples by Inductive coupled plasma Emission spectroscopy (ICPE). The results have clearly confirmed the efficiency of PBNP-incorporated mesoporous silica, which was noted to be highly suitable for the removal of cesium ions.

We have further examined the functionality of PBNP-incorporated mesoporous silica for the removal of cesium ions based on the measurement of radioactivity of ¹³⁷Cs ions. The adsorption coefficient of PBNP-incorporated mesoporous silica based on the radioactivity measurement of ¹³⁷Cs was calculated to be 3.5 × 10⁴ even at 3% PB content within the heterogeneous matrix, justifying the efficiency of the material for radioactive cesium removal, as shown in Table 3. Figure 16A shows the cesium adsorption isotherm of PBNP-incorporated mesoporous silica over an equilibrium Cs concentration range of 1–1,000 ppm. Almost all of the Cs was taken up by the material at lower cesium concentrations, followed by a sharp increase beyond 0.001 M Cs, indicating occupancy of all of the sites by Cs. Flattening of the curve beyond 0.001 M Cs signifies the attainment of equilibrium Cs uptake capacity. The maximum Cs sorption capacity (qe mmol/g) of the sorbent was calculated to be 0.1 meq/g, which is in good agreement with previously reported values, as compared in Table 4. The previously described adsorption isotherm was fitted using both the Freundlich and Langmuir adsorption isotherm models, shown in Figures 16B, C. The fitting parameters were obtained to provide a mechanistic understanding of the surface properties and affinity of the sorbent. The Freundlich and Langmuir isotherm model equations can be represented as follows: I n   t h e   F r e u n d l i c h   m o d e l : ln ⁡ q e = ln ⁡ K F + 1 n ln ⁡ C e   . . . . . . . . . . . . . . . . (1) where K_F denotes the sorption capacity when the metal ion equilibrium concentration is equal to 1, and 1/n denotes the degree of sorption dependency with the equilibrium concentration. I n   t h e   L a n g m u i r   m o d e l : 1 q e = 1 q m + 1 K L C e q m   . . . . . . . . . . . . (2) where q_m and K_L are the maximum monolayer capacity and sorption coefficient, respectively.

As per the fitting with the Freundlich model, the adsorption data show a linear relationship (R ² = 0.98); the values of K_F and “n,” as obtained from the log–log plot, were found to be 5.23 and 1.34, respectively. This “n” value indicates a monolayer adsorption of Cs taking place on the sorbent surface. Similarly, the fitting with the Langmuir model (R ² = 0.99) culminated in q_e and K_L values of 0.037 mmol/g and 2,280, respectively.

A study on the kinetics of adsorption is important for optimizing uptake performance. Hence, the sorbent was in contact with 10-ppm Cs solution for different time intervals; the uptake capacity of the sorbent as a function of time is shown in Figure 16D. It can be inferred that equilibrium is attained within 20 min of maximum uptake capacity.