Functionalization of Tailored Porous Carbon Monolith for Decontamination of Radioactive Substances

As the control over radioactive species becomes critical for the contemporary human life, the development of functional materials for decontamination of radioactive substances has also become important. In this work, a three-dimensional (3D) porous carbon monolith functionalized with Prussian blue particles was prepared through removal of colloidal silica particles from exfoliated graphene/silica composite precursors. The colloidal silica particles with a narrow size distribution were used to act a role of hard template and provide a sufficient surface area that could accommodate potentially hazardous radioactive substances by adsorption. The unique surface and pore structure of the functionalized porous carbon monolith was examined using electron microscopy and energy-dispersive X-ray analysis (EDS). The effective incorporation of PB nanoparticles was confirmed using diverse instrumentations such as X-ray diffraction (XRD), Fourier-transform infrared (FT-IR), and X-ray photoelectron spectroscopy (XPS). A nitrogen adsorption/desorption study showed that surface area and pore volume increased significantly compared with the starting precursor. Adsorption tests were performed with 133Cs ions to examine adsorption isotherms using both Langmuir and Freundlich isotherms. In addition, adsorption kinetics were also investigated and parameters were calculated. The functionalized porous carbon monolith showed a relatively higher adsorption capacity than that of pristine porous carbon monolith and the bulk PB to most radioactive ions such as 133Cs, 85Rb, 138Ba, 88Sr, 140Ce, and 205Tl. This material can be used for decontamination in expanded application fields.


Introduction
Recently, the need for intensive energy sources has increased dramatically, because human life and activities are becoming complex and energy-consuming. At the same time, a significant interest has been drawn to nuclear energy and technology dealing with dangerous radioactive substances to increase energy production capacity/efficiency, leading to a rapid increase in nuclear waste production. Therefore, it becomes even more important for us to control the potentially hazardous nuclear wastes effectively. Even if several methods have been developed to remove the nuclear wastes, it is still a challenging task to remove the radioactive substances or produce functional media materials for contamination control [1][2][3][4][5][6].
Since the Fukushima accidents, a desire for the development of environmentally benign and functional materials for decontamination of radioactive substances is ever

Preparation of the Porous Carbon Monolith with Tailored Pore Structure
The overall production procedure to obtain the PB@PCM is presented in Figure 1. In general, it is important to disperse the silica particles into the exfoliated graphene (EG) network. This was difficult to achieve because the EG-silica interactions were unfavorable and uniform mixing was almost impossible due to mild self-aggregation of the silica particles. To improve the EG-silica interactions, the surface of the silica particles was modified with amino silane coupling agent. The zeta potential value of silica particles increased from −4.03 to +6.17 mV after surface treatment (Supplementary Information, S1). Hence, the aggregation of silica particles could be suppressed and dispersion was promoted. The SEM image in Figure 1a shows no apparent trace of silica particle aggregation. the aggregation of silica particles could be suppressed and dispersion was promoted. The SEM image in Figure 1a shows no apparent trace of silica particle aggregation. The particle locations were random and any regular texture was not observed. The external appearance of the EG/silica composite precursor was monolithic, because stacking of the EG sheets occurred spontaneously. In addition, this shape could be retained in the final product, PB@PCM.
Subsequently, the silica particles were removed by etching with HF and the PCM material was produced as shown in Figure 1b. A silica trace was clearly observed in the center of the FE-SEM image. Note that the PCM possessed an interconnected open pore structure and relatively small window pores. The formation of the window pores might be associated with local dewetting of the thin EG sheets, because the EG-silica particle interface was still less friendly after silane treatment. However, it could be inferred that the transport of radioactive ion species could be facilitated due to the presence of the window pores. The irregular pores might have originated from the inherent internal voids. It could be expected that the surface area increased after the silica etching.
The PB decoration into the PCM materials led to the formation of the PB@PCM monolithic material, which would be used for removal of radioactive species. After the decorating with the PB nanoparticles, the rough surface of PCM material becomes smooth compared with that before the decoration, but there is no obvious difference in the structure or pore shape (Figure 1c). It was known that the PB decoration process was spontaneous and the average size of the PB nanoparticles was indiscernible [40][41][42][43]. The detailed internal structure of the PB@PCM is observed further in Figures 2 and 3. Figure 2 shows the macro and microscopic morphology of the PCM, which was uniform, and variation in appearance was negligible. This was because the components, EG and silica, in the precursor were homogeneous. It was difficult to observe any regular pore pattern and the pores were open and interconnected. The features shown in Figure 1b were observed analogously at the center and outer regions of the PCM monolith. It was concluded that the internal and external structure of the PCM was identical. This seemed desirable, because adsorption performance would be constant and independent. The particle locations were random and any regular texture was not observed. The external appearance of the EG/silica composite precursor was monolithic, because stacking of the EG sheets occurred spontaneously. In addition, this shape could be retained in the final product, PB@PCM.
Subsequently, the silica particles were removed by etching with HF and the PCM material was produced as shown in Figure 1b. A silica trace was clearly observed in the center of the FE-SEM image. Note that the PCM possessed an interconnected open pore structure and relatively small window pores. The formation of the window pores might be associated with local dewetting of the thin EG sheets, because the EG-silica particle interface was still less friendly after silane treatment. However, it could be inferred that the transport of radioactive ion species could be facilitated due to the presence of the window pores. The irregular pores might have originated from the inherent internal voids. It could be expected that the surface area increased after the silica etching.
The PB decoration into the PCM materials led to the formation of the PB@PCM monolithic material, which would be used for removal of radioactive species. After the decorating with the PB nanoparticles, the rough surface of PCM material becomes smooth compared with that before the decoration, but there is no obvious difference in the structure or pore shape (Figure 1c). It was known that the PB decoration process was spontaneous and the average size of the PB nanoparticles was indiscernible [40][41][42][43]. The detailed internal structure of the PB@PCM is observed further in Figures 2 and 3. Figure 2 shows the macro and microscopic morphology of the PCM, which was uniform, and variation in appearance was negligible. This was because the components, EG and silica, in the precursor were homogeneous. It was difficult to observe any regular pore pattern and the pores were open and interconnected. The features shown in Figure 1b were observed analogously at the center and outer regions of the PCM monolith. It was concluded that the internal and external structure of the PCM was identical. This seemed desirable, because adsorption performance would be constant and independent.

Characterization of the Functionalized Porous Carbon Monolith
In this work, effective introduction of the PB nanoparticles and retention of porous structure were critical for PB@PCM production. A high-angle annular dark-field scanning TEM (HAADF-STEM) image of the PB@PCM indicates that the porous structure remained intact after PB addition ( Figure 3). In addition, successful introduction of the PB nanoparticles was examined with EDS mapping analysis. It was verified that a coating layer of PB nanoparticles was introduced along the surface of the PCM. Several major elements such as C, N, O, and Fe were observed and traces of K and Cl also appeared ( Supplementary  Information, S2). The EDS mapping also provided the elemental compositions of the PB and PB@PCM. The Fe composition was high for PB nanoparticles, whereas C ratio increased for PB@PCM due to the presence of EG. The N and O compositions were almost constant regardless of the PB nanoparticle addition.

Characterization of the Functionalized Porous Carbon Monolith
In this work, effective introduction of the PB nanoparticles and retention of porous structure were critical for PB@PCM production. A high-angle annular dark-field scanning TEM (HAADF-STEM) image of the PB@PCM indicates that the porous structure remained intact after PB addition ( Figure 3). In addition, successful introduction of the PB nanoparticles was examined with EDS mapping analysis. It was verified that a coating layer of PB nanoparticles was introduced along the surface of the PCM. Several major elements such as C, N, O, and Fe were observed and traces of K and Cl also appeared ( Supplementary  Information, S2). The EDS mapping also provided the elemental compositions of the PB and PB@PCM. The Fe composition was high for PB nanoparticles, whereas C ratio increased for PB@PCM due to the presence of EG. The N and O compositions were almost constant regardless of the PB nanoparticle addition.

Characterization of the Functionalized Porous Carbon Monolith
In this work, effective introduction of the PB nanoparticles and retention of porous structure were critical for PB@PCM production. A high-angle annular dark-field scanning TEM (HAADF-STEM) image of the PB@PCM indicates that the porous structure remained intact after PB addition ( Figure 3). In addition, successful introduction of the PB nanoparticles was examined with EDS mapping analysis. It was verified that a coating layer of PB nanoparticles was introduced along the surface of the PCM. Several major elements such as C, N, O, and Fe were observed and traces of K and Cl also appeared ( Supplementary Information, S2). The EDS mapping also provided the elemental compositions of the PB and PB@PCM. The Fe composition was high for PB nanoparticles, whereas C ratio increased for PB@PCM due to the presence of EG. The N and O compositions were almost constant regardless of the PB nanoparticle addition.
The PB@PCM was further characterized using diverse instrumentations. The FT-IR spectrum of the EG showed weak peaks for C-O and C-C bending because there was a small amount of carbonyl groups on the EG (Figure 4a). A sharp peak for C≡N stretching appeared in the PB spectrum. All those features were seen in the PB@PCM spectrum, indicating that PB nanoparticles were successfully introduced to the surface of PCM. A similar trend was observed from the XRD profiles in Figure 4b. Peaks for highly crystalline PB nanoparticles were observed at 17(200), 25(220), 36(400), and 40(420) degrees, whereas that for crystalline EG was seen at 27(002) degrees [44]. The PB@PCM was further characterized using diverse instrumentations. The FT-IR spectrum of the EG showed weak peaks for C-O and C-C bending because there was a small amount of carbonyl groups on the EG (Figure 4a). A sharp peak for C≡N stretching appeared in the PB spectrum. All those features were seen in the PB@PCM spectrum, indicating that PB nanoparticles were successfully introduced to the surface of PCM. A similar trend was observed from the XRD profiles in Figure 4b. Peaks for highly crystalline PB nanoparticles were observed at 17(200), 25(220), 36(400), and 40(420) degrees, whereas that for crystalline EG was seen at 27(002) degrees [44]. Those representative diffraction peaks of PB and EG could also be assigned in the PB@PCM profile. The formation of a PB nanoparticle layer on the PCM was further confirmed by XPS. The survey spectrum of the PB@PCM in Figure  The second one was highly associated with the presence of cyanide (C≡N) groups in the PB nanoparticles. The N(1s) peak was also fitted using the Gaussian method ( Figure 4e) and showed the presence of two peaks: C-N (399.31 eV, 57.1 at%) and C=N (391.2 eV, 42.9 at%). In the fitted curve of the Fe(2p) in Figure 4f, two characteristic peaks of Fe (2p1/2) and Fe (2p3/2) were observed at 724.6 and 710.9 eV, respectively, both indicating the presence of α-FeOOH, whereas the peak at 708.2 eV was attributed to [Fe(CN)6] [45,46].
As an increase in surface area was important to achieve a high removal capacity, porous structure was introduced to the PB@PCM. N2 adsorption/desorption isotherms (BET) provided critical information regarding surface characteristics of the PB@PCM for practical application (Supplementary Information, S3). Note that the PB@PCM showed a stable adsorption behavior, indicating that it could adsorb a significantly higher amount of volume at relatively low pressure than the PCM. The structural parameters closely related to the surface properties were measured and are summarized in Table 1. In the fitted curve of the Fe(2p) in Figure 4f, two characteristic peaks of Fe (2p1/2) and Fe (2p3/2) were observed at 724.6 and 710.9 eV, respectively, both indicating the presence of α-FeOOH, whereas the peak at 708.2 eV was attributed to [Fe(CN)6] [45,46].
As an increase in surface area was important to achieve a high removal capacity, porous structure was introduced to the PB@PCM. N 2 adsorption/desorption isotherms (BET) provided critical information regarding surface characteristics of the PB@PCM for practical application ( Supplementary Information, S3). Note that the PB@PCM showed a stable adsorption behavior, indicating that it could adsorb a significantly higher amount of volume at relatively low pressure than the PCM. The structural parameters closely related to the surface properties were measured and are summarized in Table 1. The specific area of the PB@PCM (128.76 m 2 /g) was approximately three times larger than that of PCM (43.49 m 2 /g), owing to the PB nanoparticle addition. Average pore diameter decreased from 0.35 to 0.27 µm after the PB addition. Even if the pore structure was irregular and wide open, the decrease in the average pore size was meaningful for adsorption capacity improvement. A significant change also occurred to other parameters except the total pore area after the PB introduction, suggesting that the change was attributed to the PB nanoparticle addition.

Adsorption/Decontamination Performances of the Functionalized Porous Carbon Monolith
To investigate the radioactive ion removal capacity of the PB@PCM, extensive adsorption tests were performed with the stable 133 Cs isotope instead of the radioactive 137 Cs. The 133 Cs ion uptake behavior by the PB@PCM was monitored at pH 7.0 and 20 • C with varying 133 Cs equilibrium concentration, as shown in Figure 5a. The specific area of the PB@PCM (128.76 m 2 /g) was approximately three times larger than that of PCM (43.49 m 2 /g), owing to the PB nanoparticle addition. Average pore diameter decreased from 0.35 to 0.27 μm after the PB addition. Even if the pore structure was irregular and wide open, the decrease in the average pore size was meaningful for adsorption capacity improvement. A significant change also occurred to other parameters except the total pore area after the PB introduction, suggesting that the change was attributed to the PB nanoparticle addition.

Adsorption/Decontamination Performances of the Functionalized Porous Carbon Monolith
To investigate the radioactive ion removal capacity of the PB@PCM, extensive adsorption tests were performed with the stable 1 33 Cs isotope instead of the radioactive 137 Cs. The 133 Cs ion uptake behavior by the PB@PCM was monitored at pH 7.0 and 20 °C with varying 133 Cs equilibrium concentration, as shown in Figure 5a. In these experiments, the adsorption equilibrium was gradually established for 24 h after the addition of a designated amount of the PB@PCM to the test solutions. Accordingly, the adsorption capacity of the PB@PCM for 133 Cs ion was measured after 24 h contact time. The 133 Cs uptake by adsorption was calculated by the following equation: In these experiments, the adsorption equilibrium was gradually established for 24 h after the addition of a designated amount of the PB@PCM to the test solutions. Accordingly, the adsorption capacity of the PB@PCM for 133 Cs ion was measured after 24 h contact time. The 133 Cs uptake by adsorption was calculated by the following equation: where qe is the equilibrium adsorption capacity (mmol/g) of adsorbent, Co and Ce the initial and equilibrium concentration (mg/L) of the 133 Cs ion, V the total volume (L) of solution, A the atomic weight (g/mol) of 133 Cs, and W is the weight (g) of the PB@PCM.
To analyze the adsorption behavior quantitatively, the equilibrium adsorption data was fitted linearly according to the Langmuir and Freundlich isotherms. Note that the Langmuir isotherm is dependent on an assumption that all the surface adsorption sites have identical affinity toward the adsorbate; hence, adsorption at one site does not affect the adsorption at an adjacent site [47]. In addition, each adsorbate molecule tends to occupy a single site; hence, monolayer formation is promoted on the adsorbent surface. The linear and nonlinear form of the Langmuir equations are as follows: where q m is the maximum 133 Cs uptake (mmol/g) and b the constant (L/mg) that refers to the bonding energy of adsorption related to free energy and net enthalpy. On the other hand, the Freundlich model includes the reversible adsorption at a heterogeneous surface; therefore, it is not restricted to monolayer formation [48]. The linear and nonlinear forms of the Freundlich adsorption isotherms are as follows: where K f is the constant (mmol/g) related to the adsorption capacity of the adsorbent, and 1/n the intensity of the adsorption constant. Compared with the Langmuir isotherm, K f stands for analogous physical meaning to q m , even if those parameters are fundamentally different. Linear fitting was conducted according to the linear equation forms; consequently, the fitting results are shown in Figure 5b,c. Additional regression analysis was carried out by computer software (OriginPro 8.0, Originlab corporation, Northampton, MA, USA) to obtain the accurate parameters, and values are summarized in Table 2. It is clear that the fitting to the Langmuir isotherm showed a higher consistency with the adsorption data from the experiments with a regression coefficient of R 2 = 0.999 and reasonable errors for parameters. A supplementary parameter was suggested to further characterize the 133 Cs adsorption behavior by the PB@PCM, which was separation factor (R L ), a dimensionless constant [49]. R L value was calculated by the following equation: where b and C o were the Langmuir constant and initial 133 Cs concentration. The type of adsorption isotherm can be classified as unfavorable (R L > 1), linear (R L = 1), and favorable (R L = 0), depending on its value. The parameter obtained in this study was approximately 0.068, supporting the favorable adsorption of 133 Cs on the PB@PCM. Therefore, it was concluded that the PB@PCM possessed desirable structural and surface features for 133 Cs ion adsorption. The adsorption behavior was also investigated as a function of contact time between 133 Cs ions and PB@PCM. The curve in Figure 5d shows that the adsorption capacity increased rapidly during the initial 4 h and reached a maximum value (saturation) after 24 h. For quantitative analysis of the adsorption mechanism, the experimental data were fitted to the pseudo-first-order [50] and pseudo-second-order [51] kinetics and the results are exhibited in Figure 5e,f. The kinetics equations are as follows: where q e and q t are the 133 Cs uptake at equilibrium and time t, respectively, K 1 the constant (1/min) for first-order adsorption, and K 2 the rate constant (g/mmol·min) for second-order adsorption. The parameter values are summarized in Table 3 and it shows that the pseudosecond-order model described the adsorption mechanism better than the pseudo-first-order model with regression coefficients R 2 = 0.999. The pseudo-second-order model is dependent on the assumption that the adsorption occurs on the adsorbent without interactions between adsorbates, and the desorption rate is negligible compared to the adsorption rate. It was inferred that the interactions between PB nanoparticles in PB@PCM and incoming radioactive ions could be facilitated due to the permanent/induced electrostatic forces between partially negative cyano groups (-C≡N) of PB nanoparticles and positive 133 Cs ions. On the other hand, the interactions between the ions are supposed to be inconspicuous due to electrostatic repulsion. Table 3. The kinetic parameters obtained using pseudo 1st-order and 2nd-order models at room temperature for the adsorption of 133 Cs to the PB@PCM. [a] R 2 = regression coefficient. The K 1 , K 2 , q e1 , q e2 values and the nonlinear regression correlations for pseudofirst-order and pseudo-second-order models were measured by nonlinear regression analysis using OriginPro 8.0. [b] Three sets of valid data were collected, and the q e and q t values for 133 Cs were calculated to be in the error range of ±5%.

Pseudo-First-Order [a] Pseudo-Second-Order [a]
To reveal the effectiveness of synergistic interaction between PB nanoparticles and PCM material, the 133 Cs ion uptake test was performed using the PCM, bulk PB and PB@PCM ( Supplementary Information, S4). It was important that the PB@PCM provided an improved uptake by adsorption compared with the PCM and bulk PB (A compilation of investigations carried out by using various PB-based adsorbents is also presented in Supplementary Information, S5 [52][53][54][55][56][57]). The maximum uptake value of PCM, bulk PB, and PB@PCM were ca. 0.0018, 0.0069, and 0.7225 mmol/g, respectively. It is noteworthy that the maximum 133 Cs adsorption capacity of the PB@PCM is approximately 400 and 100 times higher than that of pristine PCM and bulk PB, respectively. This striking difference could be interpreted by the enhanced porosity and permeability of PCM and high surface area and large amount of PB nanoparticles in the synthesized PB@PCM, providing abundant active sites for 133 Cs adsorption. Compared to the bulk PB, the PB@PCM containing both interconnected macroporous structure and nanometer-sized PB particles have enhanced adsorption properties due to increased mass transport through the material and maintenance of a specific surface area.
It was also expected that the PB@PCM can adsorb various radioactive species; hence, uptake tests were performed at pH 7.0 and 20 • C with other radioactive ion species such as 133 Cs, 85 Rb, 138 Ba, 88 Sr, 140 Ce, and 205 Tl (Figure 6a). The adsorption capacity value varied from 0.72 to 0.13 mmol/g, with a decreasing order of 133 Cs (0.72), 85 Rb (0.63), 138 Ba (0.32), 88 Sr (0.26), 140 Ce (0.18), and 205 Tl (0.13). This trend might have originated from the degree of interaction between PB nanoparticles in PB@PCM and individual radioactive ion species. More positively charged ions could interact more favorably with the PB nanoparticles. It can be inferred that 133 Cs might have a highly positive nature, because its electronegativity and ionization energy was lowest compared with other ion species ( Supplementary Information, S6). However, the underlying correlation needs to be examined systematically in a separate research project.  In addition, the effect of the pH on the 133 Cs adsorption capacity was also investigated ( Figure 6b). Interestingly, the removal efficiency of the PB@PCM increased in acidic conditions and decreased in basic media, and the PB@PCM exhibited the highest uptake capacity at pH 7.0. This phenomenon can be ascribed to the competitive interaction between hydrogen (H+) and hydroxyl (OH−) ions during the adsorption process. In acidic condition (<pH 7.0), the H+ ion competes with the 133 Cs ions towards the active sites of PB@PCM, which results in a relatively lower adsorption capacity. As the pH increased (≈pH 7.0), the decrease of the hydrogen ion concentration made it possible to increase the concentration of 133 Cs ions to bind electrons, leading to a high adsorption uptake. However, when the basification was in progress (>pH 7.0), the stabilization attributed to the reaction between 133 Cs ions and increased OH− ions occurred, and then the adsorption capacity decreased.

Materials
Tetraethyl orthosilicate (TEOS, 98%), 3-aminopropyltrimethoxysilane (APS), iron(III) chloride (FeCl3, 97%), potassium hexacyanoferrate(III) (K3Fe (CN)6, ≥99%), cesium, barium, strontium, and rubidium standard solution (1000 ppm) for inductively coupled plasmamass spectrometer (ICP-MS) analyses were purchased from Sigma Aldrich (Milwaukee, WI, USA). Sodium chloride (NaCl), Ammonia solutions (NH4OH, 28.0~30.0%), hydrochloric acid (HCl), Ethanol (C2H5OH, ≥99%), and N-Methyl-2-pyrrolidone (NMP) were supplied from Samchun Chemical (Seoul, Korea). Hydrofluoric acid (HF, 40%) was purchased from J. T. Baker (Phillipsburg, NJ, USA). All chemicals were used as received without further purification. Electrochemically exfoliated graphene (EG) was produced according to the method reported previously [52]. In addition, the effect of the pH on the 133 Cs adsorption capacity was also investigated ( Figure 6b). Interestingly, the removal efficiency of the PB@PCM increased in acidic conditions and decreased in basic media, and the PB@PCM exhibited the highest uptake capacity at pH 7.0. This phenomenon can be ascribed to the competitive interaction between hydrogen (H+) and hydroxyl (OH−) ions during the adsorption process. In acidic condition (<pH 7.0), the H+ ion competes with the 133 Cs ions towards the active sites of PB@PCM, which results in a relatively lower adsorption capacity. As the pH increased (≈pH 7.0), the decrease of the hydrogen ion concentration made it possible to increase the concentration of 133 Cs ions to bind electrons, leading to a high adsorption uptake. However, when the basification was in progress (>pH 7.0), the stabilization attributed to the reaction between 133 Cs ions and increased OH− ions occurred, and then the adsorption capacity decreased.

Synthesis of Monodispersed Silica Microparticles
A certain amount of TEOS (36 g) was vigorously mixed with ethanol (200 mL) (Solution 1). An amount of ammonia solution (54 mL) and NaCl (0.1 g) was dissolved in a mixture of water (40 mL) and ethanol (190 mL) (Solution 2). Subsequently, solution 1 was slowly added to solution 2 at a speed of 1.2 µL/min using a micro syringe pump at 30 • C. Then the mixture solution was stirred for 1 h at 800 rpm. The produced microparticles were purified by centrifugation and washed with ethanol and distilled water 3 times. Finally, the obtained homogeneous silica particles were stored at room temperature.

Surface Modification of the Silica Particles
The obtained silica particles (1 g) were dispersed in ethanol (10 mL) by ultra-sonication for 10 min, resulting in a milky suspension. A small amount of glacial acetic acid (60 µL) was added to a mixture of ethanol (95 mL) and water (5 mL) to make the solution acidic (pH 5.0) and then APS (4.2 g, 19 mM) was added dropwise to the acidic solution. The resulting clear colorless solution was stirred (400 rpm) at RT for 15 min in order to form reactive silanol groups by hydrolysis.

Preparation of Porous Exfoliated Graphene Monolith
A certain amount of exfoliated graphene (EG, 240 mg) was dispersed in 20 mL of NMP and sonicated for 90 min to achieve a homogeneous mixing, which was critical for this experiment. Then, the silica particles were added to the EG solution, and the EG to silica weight ratio was 10:90. The obtained solution was filtered with a nylon membrane filter (47 mm in diameter, 0.2 µm pore size, Whatman) under vacuum suction to obtain an EG/silica composite precursor. The film type EG/silica composite was retrieved by peeling it off from the nylon membrane filter. A film type porous EG was obtained by removing the silica particles with HF solution (10%) for a few minutes. The PCM was dried in a vacuum oven at 120 • C overnight.

Fabrication of Prussian Blue Decorated Porous Carbon Monolith
The PB@PCM was manufactured by a redox reaction in an aqueous mixture solution of FeCl 3 , K 3 [Fe(CN) 6 ], and PCM. Typically, PCM was soaked in 10 M FeCl 3 solution for 15 min and the solution in sample was completely eliminated with a filter paper. Then, the PCM decorated with Fe 3+ ions was dried in a convection oven at 80 • C for at least 3 h. Then, the preprocessed PCM was immersed in a K 3 [Fe(CN) 6 ] (0.64 g, 100 mL) solution and stirred at 55 • C for 12 h. The final PB@PCM product was obtained and washed with distilled water 3 times and dried under vacuum at 80 • C overnight to complete evaporation of the solvent. The mass of the finally obtained PB@PCM sample was 2.4 g, which can be scaled up by increasing the amount of the EG/silica composite precursor. The bulk PB was fabricated by direct mixing of equimolar FeCl 3 and K 4 [Fe(CN) 6 ] solutions at room temperature for 6 h. The final products were washed with ethanol several times to remove residual reagent. Subsequently, the resulting bulk PB was dried under a vacuum oven and pressed into pellets.

Radioactive Material Extraction Test
The adsorption behavior of the PB@PCM for 133 Cs ion was monitored as follows: the cesium standard solutions with varying concentration from 0.01 to 100 mg/L (ppm) were produced in 100 mL of distilled water. To control the pH, ammonia or hydrochloric acid solution was injected into the aqueous solutions. Then, 0.1 g of adsorbents (PB@PCM) was introduced into the solution and stirred at 300 rpm, which would be suitable for 133 Cs ion adsorption. Adsorption experiments were carried out at 20 • C and pH 7.0, and the adsorption capacity for 133 Cs ion was monitored from 0 min to 24 h. After adsorption, 3 mL of the solution was filtered through a syringe filter (PTFE, pore diameter 0.2 µm), and then an ICP-MS test was performed to determine the 133 Cs ion concentration as a function of adsorption time. To obtain accurate results, the 133 Cs ion concentration was measured three times and the average values were collected. The adsorption capacity to the other ions-rubidium, barium, strontium, cerium, and titanium-was also measured identically.

Characterization
The morphology was confirmed using a field emission scanning electron microscopy at 10 kV (Tescan Mira-3 FEG, Brno, Czech Republic) and transmission electron microscopy (FEI, Tecnai G2-20, Hillsboro, OR, USA). A particle size analyzer (Malvern, Zetasizer Nano-ZS, Worcestershire, United Kingdom) was used to determine the silica particle size and zeta potential. The FT-IR spectra were recorded on Alpha-P (Bruker, Ettlingen, Germany). The XRD patterns were collected using an Ultima IV with Cu Kα radiation (Rigaku, Tokyo, Japan). The BET analyses were performed with an ASAP 2020 (Micromeritics, Gwinnett County, GA, USA) using the N2 adsorption-desorption isotherms. The pore size and distribution in the porous carbon was analyzed with an AutoPore IV 9500 mercury porosimeter at room temperature (Micromeritics, Gwinnett County, GA, USA). The XPS data were obtained using an Axis Nova (KRATOS, Tokyo, Japan) with monochromatic Al-Kα X-ray source under 10-8 Torr vacuum analysis chamber. The ICP-MS was used to analyze and confirm the cesium concentration in solution (iCAP RQ, Thermo Fisher Scientific, Waltham, MA, USA).

Conclusions
In this study, convenient manufacture of monolithic porous carbon material embedded with Prussian blue particles was demonstrated and its radioactive substance removal performance was also examined using various radioactive ion species. The functional porous carbon was prepared by etching of silica template and subsequent Prussian blue nanoparticle decoration. The obtained porous carbon possessed an increased surface area compared with the precursor and the intermediate. In addition, its pore structure was open and interconnected with small window pores, which was advantageous for ion transport and movement. The radioactive substance removal was achieved by the incorporation of Prussian blue nanoparticles, which can adsorb radioactive ion species effectively. The 133 Cs adsorption behavior could be described by both Langmuir and Freundlich isotherms, and fitting to the Langmuir isotherm provided a closer correlation with the experimental data. The porous carbon showed a better adsorption capacity than the precursor and an intermediate capacity for radioactive ion species such as 133 Cs, 87 Rb, 88 Sr, 137 Ba, 140 Ce, and 205 Tl. The maximum and minimum adsorption capacity values were obtained for 133 Cs and 205 Tl, respectively. As a porous carbon material was obtained by a relatively simple, cost-effective, and environmentally benign method, an interest in analogous functional porous materials for nuclear waste control is expected to increase significantly. Therefore, this research can offer important information for future relevant research and, moreover, development of advanced materials and devices.