Simple Preparation of Porous Carbon-Supported Ruthenium: Propitious Catalytic Activity in the Reduction of Ferrocyanate(III) and a Cationic Dye.

The present study involves the synthesis, characterization, and catalytic application of ruthenium nanoparticles (Ru NPs) supported on plastic-derived carbons (PDCs) synthesized from plastic wastes (soft drink bottles) as an alternative carbon source. PDCs have been further activated with CO2 and characterized by various analytical techniques. The catalytic activity of Ru@PDC for the reduction of potassium hexacyanoferrate(III), (K3[Fe(CN)6]), and new fuchsin (NF) dye by NaBH4 was performed under mild conditions. The PDCs had spherical morphology with an average size of 0.5 μm, and the Ru NP (5 ± 0.2 nm) loading (4.01 wt %) into the PDC provided high catalytic performance for catalytic reduction of ferrocyanate(III) and NF dye. This catalyst can be recycled more than six times with only a minor loss of its catalytic activity. In addition, the stability and reusability of the Ru@PDC catalyst are also discussed.

Introduction

Over the past few years, biowastes and non-biodegradable plastic waste materials as low-cost feedstock have been utilized for the production of value-added carbon nanomaterials with a wide range of applications.[1] Until now, enormous amounts of plastic wastes were generated because of increased demand of plastic-related production.[2] These plastics are not biodegradable and thus generate extremely troublesome components for landfilling. Handling and managing these plastic wastes is a huge task, because a large amount of these wastes are generally dumped into landfill or disposed in the ocean, thereby causing a very serious environmental issue.[3] Therefore, a large number of methods have been investigated to convert plastic wastes into useful products.[4] For instance, they have been transformed into different kinds of carbon-based nanostructures including nanotubes,[5] spheres,[6] hollow spheres,[7] nanosheets,[8] activated carbons,[9] and graphene flake/foil[10] for sustainable energy applications.[11]

Carbon-based nanostructure materials, particularly biowaste activated carbon spheres (CSs), have gained a wide range of interests because of their excellent energy storage ability, good electrical conductivity, biocompatibility, and electrical properties, and thus, they have received a wide attention.[12] In recent years, active metals supported on the porous carbon substrate such as Rh,[13] Re,[14] Ru,[15] Os,[16] Ir,[17] and Pt[18] have been used as active catalysts for organic/inorganic transformations. Among them, Ru-based nanomaterials are widely explored as nanocatalysts[19] because of their low cost and attractive feature of superior catalytic activity and stability.

Generally, potassium hexacyanoferrate(III), (K3[Fe(CN)6]), is well-known as one of the most common pollutants, which is found in contaminated air, water, and soil in the environment. It can easily accumulate inside humans, aquatic animals, and other living organisms through food chains[20] and has been proved to have mutagenicity, acute toxicity, carcinogenicity, and high environmental mobility, even in a trace level.[21] In contrast, Fe(II) is considered as an essential nutrient required in metabolic pathways for humans and animals. The most common type of anemia is caused by the iron deficiency whereas some diseases such as hemochromatosis can be due to iron overload; the United States Recommended Daily Allowance (USRDA) for iron is 18 mg.[22] Besides, the conversion of Fe(III) into Fe(II) offers several important advantages and attractive applications including possibilities for (i) tin purification, (ii) separation of copper out of molybdenum ore, (iii) wine and citric acid in large-scale preparation, (iv) serving as a benchmark for electron transfer reaction, and (v) medical diagnosis for diabetic patients and designing of amperometric biosensor for electrochemical applications.[23]

Recently, Pastoriza-Santos et al.[24] carried out the reduction of Fe(CN)63– with NaBH4, using gold nanorods containing metallodielectric hollow shells of SiO2 or TiO2. However, these catalytic supports are expensive compared to the carbon nanostructure. Remarkably, these supports are difficult to remove from the active phase, but the carbon support can be easily burnt to separate the active metallic phase. Carregal-Romero’s group fabricated spherical Au NP heterostructures[25] and have used them for catalytic reduction of Fe(CN)63– by NaBH4 in aqueous solution; however, this catalyst rate was 4 times higher than that of the uncatalyzed reaction. Despite their successful results, low stability and inconvenient recovery restrict practical application of the materials. Jana et al.[26] prepared mesoporous Au-boehmite film catalyst for reduction Fe(CN)63–. However, this catalyst achieved large rate constant and was four times recycled but required high loadings, harsh conditions, and high-boiling-point aprotic solvents for the catalytic methods. Likewise, Miao et al.[27] used Fe3O4@GNSs as the catalyst for the reduction of Fe(CN)63– in aqueous media. This magnetic composite was used as an ideal catalyst and shows recyclability along with persistent catalytic activity even after being recycled six times. Chen et al.[28] developed the submicron-sized PEGDMA@Au NP microsphere catalyst for the reduction of Fe(CN)63– by NaBH4 in aqueous solution, but they failed to report recyclability and leaching experiments. Likewise, Yang et al.[29] used 2,6-pyridinedicarboxylic acid-protected Au NP as a catalyst for the same reduction. However, the catalyst exhibits poor rate constant and lack of recyclability and stability. Wu et al.[30] have demonstrated an ionic liquid-based synthesis of hollow and porous platinum nanotubes as a new catalyst, however required harsh preparation conditions, exhibited poor stability, and produced silica waste after etching. Recently, Jiang demonstrated a new approach for the one-pot synthesis of gold hollow nanospheres exhibiting excellent catalytic activity toward the reduction of Fe(CN)63– by NaBH4 in water.[31] Being stable in air and water, this catalyst could be reused 10 times. However, a long time period was required for completion of reduction. Therefore, the above-mentioned catalysts are less advantageous with limited practical applications in catalysis, owing to some unique natures, such as great chemical stability, low corrosive capability, high thermal stability, hydrophobic feature, easy recovery, and low price.[32]

On the other hand, clean water is the major topic of the current research because it is an important source for humans and the environment.[33] Discharge of effluents containing toxic dyes and heavy metal ions from manufacturing industries such as cosmetic, leather, paper, textile, pharmaceuticals, and so on into nearby water bodies is highly detrimental to the human health and environment.[34] Hence, the development of robust and smart functionalized nanomaterials with practical feasibility and biocompatibility is necessitated to act as an effective adsorbent for removal of toxic dyes.[35] In particular, metals or metal oxide-containing carbon nanomaterials have been reported as inexpensive nano-adsorbents for the adsorption/removal of heavy metals and dyes.[34,35] Thus, utilization of inexpensive nano-adsorbents for the treatment of industrial dye effluents could be helpful in resolving the human and environment problems.[36,37]

In this work, small-sized Ru NPs were decorated on the PDC support by the microwave-assisted (MW) reduction method, which has been popularly employed over the past few years for its advantageous nature, such as easy control of particle size and surface area, purity and high production yield, quick operation time, desirable temperature regulation, and tenability for the carbonaceous structure. Herein, we report a novel preparation to fabricate Ru NPs supported on porous PDC (Ru@PDC) by converting waste plastics into porous carbons. We adopted plastic wastes as the primary carbon rich precursor and ruthenium(III) acetylacetonate [Ru(acac)3] as a metal precursor. A more detailed description for the preparation procedure of Ru@PDC catalyst is shown in Scheme .

Scheme 1

Schematic Diagram for the Preparation and Application of the Ru@PDC Catalyst

Results and Discussion

Phase Structure

The phase structures of the as-prepared samples were examined using powder X-ray diffraction (PXRD) and Raman spectroscopy. Analyzing the XRD of carbon samples exhibits two broad peaks with low intensities at 2θ ≈ 23.5° and 43.6° corresponding to the (002) and (100) diffraction planes ascribed to the graphitic and amorphous carbon structure, respectively (Figure a). Upon increasing the temperature, the diffraction peaks of PDC-600, PDC-700, and PDC-800 samples become broader slightly with larger intensities, indicating that the carbon structure is characteristic of a more extent of graphitic nature implying a trend of polymer aggregation into large polyaromatic structures. Aggregation reactions proceed progressively with temperature at 600 °C and higher until the carbonaceous materials are formed showing nanometer-scale morphology containing highly organized carbons.[38] On the other hand, additional sharp diffraction peaks were observed for the Ru@PDC composite at 2θ ≈ 38.4°, 42.3°, 44.1°, 58.5°, 69.6°, and 78.2° corresponding to the (100), (002), (101), (102), (110), and (103) planes of the hexagonal close-packed Ru (JCPDS 06-0663), in excellent agreement with the reported values.[31] The XRD pattern of Ru@PDC shows a broad peak located at 44.1° composed of overlap between both C(100) and Ru(101) diffractions, thus suggesting that the catalyst contains smaller size ruthenium particles with diameters ≈5 nm. Additionally, the low C(002) peak intensity in the catalyst was observed, because of general lack of graphitic ordering. Scherrer formula (eq ) in the following was adopted to calculate the apparent crystallite size for a given reflection.where D denotes the mean size of the crystallite perpendicular to the planes (hkl) and K is a Scherrer parameter adopted as 1.84 for (100) and 0.94 for (002) for half-widths. λ is equal to 0.15406 nm, the used wavelength of the X-ray radiation, β is the breadth at half maximum intensity in radians, and θ is the Bragg angle for the reflection concerned. The average particle size of Ru NP was evaluated to be 5–6 nm, consistent with the high resolution transmission electron microscopy (HR-TEM) results (vide infra).

Figure 1

(a) PXRD patterns, (b) Raman spectra, (c) N2 sorption-isotherms, and (d) TGA curves of the as-prepared PDC and Ru@PDC materials.

Raman Analysis

Raman spectroscopy is used to inspect the structure defects and disorder nature of carbonaceous materials. Raman spectra for all the samples (Figure b) exhibited two marked peaks at around 1344 and 1601 cm–1, which were ascribed to the D band and G band, respectively.[39] The G band at 1592–1601 cm–1 is due to the E2g C–C stretching mode of sp2-bonded two-dimensional hexagonal lattice of graphite layers, while the D-band at 1323–1344 cm–1 is attributed to the A1g vibrational mode and is characteristic of the disorder nature. The intensity ratio of ID/IG is used to evaluate the defects of carbon-based samples; a smaller ratio suggests more significant defects on graphitic carbons. As such, the band intensity ratios (ID/IG) for PDC-600 and PDC-700 are 0.48 and 0.51, respectively, both indicative of low graphitization.[40] However, the peak area ratio of ID/IG for pristine PDC and Ru@PDC increased to 0.78 and 0.83, respectively, indicating graphitization enhancement of the PDC-800 after the high-temperature treatment; the fact agrees with the XRD measurements (see Figure a). The resulting ID/IG ratios of all samples were calculated and are listed in Table . This consequence is in agreement with those by the XPS and HR-TEM analyses (vide infra).

Table 1

Textural Properties of the As-Prepared PDC and Ru@PDC Materials

sample	STota (m2 g–1)	SMicrob (m2 g–1)	SMesoc (m2 g–1)	VTota (cm3 g–1)	DPd (nm)	Dme (%)	ID/IG
PDC-600	97.6	31.2	66.4	0.032	7.8		0.48
PDC-700	294.2	103.1	191.1	0.071	7.5		0.51
PDC-800	466.7	217.2	249.5	0.086	7.5		0.78
Ru@PDC	396.5	184.6	211.9	0.082	7.3	6.01	0.83

Surface area (STot) from the BET method and total pore volume (VTot) calculated at P/P0 = 0.99.

Microporous surface areas (SMicro) obtained from t-plot analyses.

SMeso (SMeso = STot – SMicro).

Average pore size (DP) derived by BJH adsorption branches of isotherms.

Metal dispersion measured by H2 chemisorption at 323 K.

Textural Property

The nitrogen adsorption/desorption isotherms at 77 K of PDC and Ru@PDC samples are shown in Figure c. The isotherms exhibited a type-IV curve with hysteresis loop associated with capillary condensation in the range of P/P0 from 0.45 to 0.99. This finding indicated that the porosity of the obtained PDC and Ru@PDC was essentially made up of micropores/mesopores, and it may be generated by the CO2 activation. The textural properties including BET total surface area (STot), micropore surface area (SMicro), mesopore surface area (SMeso), total pore volume (VTot), and average pore diameter (DP) are summarized in Table . The BET surface area of Ru@PDC (SBET = 396.5 m2 g–1) sample significantly decreased compared to PDC-800 (466.7 m2 g–1), indicating that Ru NPs blocked the pores of the CSs and thus diminished the surface area and the total pore volume of the Ru@PDC.[41] These results verified that the Ru NPs were impregnated on the surface of the PDC matrix. The Ru NPs were well dispersed on the surface and no obvious aggregation was observed, whereas unsupported Ru NPs were likely to aggregate immediately.[42a] The Ru@PDC catalysts were characterized and are listed in Table . It seems that a higher BET specific surface area tends to favor a higher Ru dispersion. Masthan et al. have reported the H2 dispersion measurement for Ru/γ-Al2O3, the H2 absorption equilibrium was much faster at a higher temperature (373 K) rather than ambient temperature.[42b] The value of Ru dispersion and average crystallite size for Ru-based catalysts based on H2 chemisorption method is provided in Table S1, (Supporting Information).

The Barrett–Joyner–Halenda (BJH) model was adopted to evaluate the pore size distributions in terms of the adsorption branches of the isotherms (Figure S1 of the Supporting Information).

Thermal Stability

Thermal properties of the samples thus prepared were further examined by thermogravimetric analysis (TGA), as displayed in Figure d. The weight loss below 200 °C (12–18% for all samples) can be attributed to the adsorbed water evaporation. The other weight losses began at around 558 °C, which are mainly due to the burning of the carbon structures. It indicated the high purity of the prepared PDC. However, a trace amount of residues was also observed after complete decomposition of carbon samples. Inductively coupled plasma–atomic emission spectroscopy (ICP–AES) was further employed to analyze the elemental contents, obtaining the Zn species present in the carbon material with a content of 1017 ppm (i.e., ca. 0.10 wt %). The impregnation of ZnCl2 during the process tended to cause dehydration of the carbon substrate and subsequently to result in charring and aromatization along with the creation of porosities. The mobile liquid ZnCl2 (mp ≈ 283 °C) was expected to occur in the earlier stage of the activation. Sometimes, the Zn ions are expected to be strongly intercalated between the carbon interlayers, when the activation temperature is increased beyond 700 °C (bp of ZnCl2 ca. 730 °C), and a strong interaction between carbon atoms and Zn species between the carbon interlayers might leave trace Zn residue unvaporized.[21]

The burning of PDC in the Ru@PDC sample began at around 561 °C because of the interaction between Ru NP and carbon atoms inducing defects in the graphitic carbon structure of PDC.[43] While further heating, no weight loss was found significantly, verifying that the particle crystallinity took place at 800 °C. Moreover, according to the data from TG measurements, the content of Ru in Ru@PDC was evaluated to be 4.01 wt %. This experiment evidences additionally that Ru NPs are successfully incorporated onto the PDC support.

Morphology and Microstructure

The CSs were prepared by using plastic wastes as carbon sources without any catalysts under hydrothermal conditions, as reported earlier.[44] According to this method, the SEM images of PDC-800 spheres exhibited uniform spherical shapes ranging from 300 to 500 nm in diameter, as displayed in the Figure S2, Supporting Information. As can be seen, a minor agglomeration bonding between spheres can be ascribed to polymerization interruption or structure collapse. Additionally, the microstructures of PDC-600 and PDC-700 were further examined using HR-TEM as displayed in Figure S3 Supporting Information, which shows appearance of smooth surface with perfect spheres. HR-TEM observation at different magnifications (Figure a,b) and enlarged portions (Figure c–g) shows the shape of the PDC-800 particle, which has sizes of few hundred nanometers along with porous microstructure. Moreover, the selected area electron diffraction (SAED) reveals that the PDC-800 has a typical amorphous carbon microstructure (Figure h), which preserves the structural integrity and spherical morphology.

Figure 2

(a–g) HR-TEM images of the as-prepared bare PDC-800 with different magnifications and (h) electron diffraction pattern of the representative image of PDC-800. The bars represent (a) 0.5 μm, (b) 200 nm, (c) 100 nm, (d) 50 nm, (e) 20 nm, (f) 10 nm, and (g) 5 nm.

As shown in Figure , HR-TEM images of the Ru@PDC catalyst showed that Ru NP had an average size of 5 ± 0.2 nm, which are apparently distributed smoothly on the surface of PDC. A representative histogram of particle size distribution of Ru NP in Ru@PDC catalyst is shown in Figure S4 (Supporting Information). Energy-dispersive spectrometry analysis evidenced the presence of Ru, C, and O elements in the Ru@PDC catalyst (Figure S5, Supporting Information). Again, it is believed that Ru NPs have been successfully planted on the surface of PDC matrix.

Figure 3

(a–f) Typical HR-TEM images of the Ru@PDC catalyst with different magnifications.

As displayed in Figure a–c, the additional filed emission TEM (FE-TEM) images of Ru@PDC verify a uniform distribution of the Ru NP on the surface of PDC carbon matrix. The Ru presence was confirmed from SAED pattern of Ru NPs (Figure d). It was found that some Ru NPs aggregate to form small clusters with a maximum size of 6 nm. The images also indicate that some Ru NPs are successfully planted in the PDC carbon matrix.

Figure 4

(a–c) Additional HR-TEM images of the Ru@PDC catalyst and (d) SAED pattern.

Surface Element Composition Analysis

The surface element compositions of PDC-800 and Ru@PDC samples were characterized using XPS. Figure a shows the XPS survey spectra of PDC-800 and Ru@PDC, containing C, O, and Ru elements. The C 1s XPS spectrum (Figure b) exhibits a strong peak at 284.3 eV, ascribed to the C–C/C=C bonds, and two relatively weaker peaks with the binding energies (B.E) at about 285.1 and 288.7 eV, attributed to the C–H and C=O species, respectively. It is notable that the two bands with B.E at 284.3 and 280.4 eV can be readily assigned to Ru 3d3/2 and 3d5/2, respectively, in the nanoparticles by referring to the values of Ru metal at 285 and 280 eV, respectively.[45] As shown in Figure c for the O 1s spectrum, a broad band is found and deconvoluted into four peaks with a B.E ca. 529.9 eV (C=O), 530.7 (COOH), 532.6 (O–C–O), and 534.2 eV (C–OH). The Ru 3p signal of Ru@PDC (Figure d) is fitted into a pair with the B.E of ca. 461.1 (Ru 3p3/2) and 483.2 eV (Ru 3p1/2), corresponding to the photoemission of metallic Ru.[46] Additionally, the elemental analysis by the ICP–AES technique evidenced the presence of the Ru element in the Ru@PDC catalyst material with a content of 4.01 wt % (see Table S4, Supporting Information).

Figure 5

(a) XPS survey spectra of the pristine PDC-800 and the Ru@PDC samples, and the corresponding core-level spectrum of (b) C 1s + Ru 3d, (c) O 1s, and (d) Ru 3p.

FT-IR Study

FT-IR spectra were recorded for PDC and Ru@PDC samples, and the results are shown in Figure S6, Supporting Information. The appearance of a weakly broad band at ∼3345 cm–1 is attributed to the hydroxyl group (−OH), and a weak band at 1638 cm–1 is attributed to the skeleton vibration of aromatic (−C=C−) rings. The band at 2926 cm–1 is assigned to CH2 asymmetric stretching, while the band at 1442 cm–1 is ascribed to the C–H bending mode. Other bands at 1110–1258 cm–1 are due to the C–O stretching mode.[11a] After carbonization, most peaks disappear with increase of temperature from 600 to 800 °C (Figure S6, Supporting Information). However, a small peak at 1594 cm–1 remains, implying that some aromatic rings in the carbonized samples still exist. Meanwhile, after the carbonization process in conjunction with Ru NP immobilization, the bands of −OH groups in the Ru@PDC nanocomposite decrease in intensity. This is due to a strong interaction of the functional group with Ru metal.[44]

In addition, different types of plastic materials were utilized as carbon sources, including high-density polyethylene (HDPE), low-density polyethylene (LDPE), and polyacrylate (PC). The produced solid CSs had smooth surfaces and the size of CSs is in the range from 3 to 8 μm (Figure S7, Supporting Information). In terms of TEM analysis (Figure S8, Supporting Information), the CSs produced with various plastic precursors are solid in nature, and their spherical images of CSs can be clearly seen in all the cases of carbon sources at 800 °C (Figure S8, Supporting Information). All samples are carbon-rich materials (∼68–69%) as revealed by the elemental analysis and XRD patterns (Figure S9, Supporting Information), showing carbon content on weight basis, and possession of trace amounts of heteroatoms (such as N, S, and Cl Table S2, Supporting Information), whose presence should be beneficial to the preparation of active carbons. Controlling of the carbonization conditions and activation process is not the only key factor to determine porous structure of CSs, which may also be affected by the structure and nature of the precursors. It is important for starting plastics to possess high content in hydrocarbons for the preparation of porous carbon by the self-assembly approach with which hydrocarbons are aggregated into higher poly-hydrocarbons resulting in the formation of carbon materials. The elemental analysis (Table S3, Supporting Information) reveals that the hydrogen content is much greater in LDPE than in HDPE and PC. As reported, it is well-known processes for the generation of CSs from aromatic hydrocarbons and from degradation of plastic materials to the mixture of hydrocarbons.[44]

Catalytic Study

Ruthenium-supported carbon materials have been vastly envisaged for catalytic applications over the last years.[45,46] Because carbon-supported Ru NPs containing mesoporous structures and large surface areas have been explored as superior catalysts in the inorganic and organic fields.[47] Therefore, the designed Ru@PDC catalysts can serve as a novel catalyst for inorganic reduction reactions with a highly efficient performance. The dispersibility of heterogeneous catalysts in solution should play a key role in order to enhance the catalytic activity. In this manner, the presence of functional groups on the surface renders the catalyst to disperse fabulously in the solution, as displayed in Figure S6b (Supporting Information).

The catalytic activities of Ru@PDC have been tested for the reduction of [Fe(CN)6]3– using NaBH4[48] as a reducing agent. Initially, we tested the blank experiment in the absence of either reducing agent or Ru@PDC catalyst and found that the concentration of [Fe(CN)6]3– did not change. It is a fact that the reaction does not occur significantly in the presence of either the reducing agent or Ru@PDC catalyst alone (Figures S10a,b, Supporting Information). Therefore, the combination of all these reagents is required for the reduction reaction. The catalytic activity of Ru@PDC with excess NaBH4 was performed, showing that the characteristic yellow color (i.e., high catalytic activity) in aqueous solution disappeared in inorganic reaction within seconds, as shown in Figure a. We verified such an efficient reaction through analysis of the catalyst effect on the kinetic reduction of K3[Fe(CN)6] in the presence of NaBH4, which is essentially based on an electron-transfer process, and the kinetic change can be easily monitored using UV–vis absorption spectroscopy. The K3[Fe(CN)6] aqueous solution in light yellow showed its absorption band at 420 nm. When NaBH4 was added, the absorption intensity at 420 nm decreased gradually because of the [Fe(CN)6]4– formation, indicating that Fe(III) ions were reduced to Fe(II), along with the solution color changed to colorless, as displayed in Figure b,c. Under the given conditions, we tested the same reaction by NaBH4 in the presence of either the Ru NP catalyst (Figure d) or available commercial Ru/C catalyst (Figure e); but both of them take a long period of 30 s (reaction was incomplete). These results indicated that both Ru NP and Ru/C catalysts possess a less catalytic activity, when compared with the Ru@PDC catalyst. This is due to insufficient porous features to allow diffusion of reactants and products. When the concentration of NaBH4 far exceeds [Fe(CN)6]3–, the kinetic reduction can be treated as a pseudo-first-order reaction, and then the kinetic rate R can be expressed as eq (49)where t is the reaction time; k is the pseudo-first-order reaction rate constant; and C and C0 denote the concentration of [Fe(CN)6]3– at time t and the initial time t0, respectively. The C/C0 is proportional to the relative intensity of A/A0, where A and A0 are the peak absorbance at time t and t0, respectively. Hence, the pseudo-first-order kinetic model is described as eq

Figure 6

UV–vis spectra for the reduction of K3[Fe(CN)6] by NaBH4 in the presence of (a) Ru@PDC recorded within 30 s, (b)1.0 mg, (c) 2.0 mg of Ru@PDC, (d) Ru NP, (e) Ru/C, and (f) conversions vs catalysts.

The pseudo-first-order rate constant (k) can be estimated from the linear plot of ln(A/A0) versus the reduction time (t). Accordingly, the rate constant (k) is calculated to be 0.0942, 0.1011, 0.0612, and 0.021 s–1, for Ru@PDC (1.0 mg; 2.0 mg), Ru NP, and ruthenium black (Ru/C) catalysts, respectively. This plot ln(A/A0) versus time reveals that the [Fe(CN)6]3– can be converted completely into [Fe(CN)6]4–, as displayed in Figure f. Conversion of the catalytic products is determined by eq (50)

The calculated results demonstrated that a good conversion (98%) was observed for the Ru@PDC catalyst, while lower conversions were obtained by unsupported Ru NP (65%) and commercial Ru/C (23%) catalysts; however, both the Ru NP and Ru/C catalysts show slightly a lower catalytic activity due to its lower surface area. The Ru@PDC was found to exhibit a much better catalytic activity for [Fe(CN)6]3– reduction than that of ruthenium black (kRu@PDCs: 0.1011 s–1 vs kRublack: 0.021 s–1). Meanwhile, the k value of Ru@PDC (0.1011 s–1) is also higher than those of Pd/GPDAP (kPd/GPDAP = 2.330 × 10–2 s–1),[48a] Au@IFMC-100 (kAu@IFMC-100 = 3.07 × 10–2 s–1),[49] ce-MoS2 (kce-MoS = 53 ± 7 × 10–2 s–1),[51] and other unsupported noble metal nanoparticles (see Table S7, Supporting Information) under the same reaction conditions, further demonstrating that the Ru@PDC have excellent reactivity for the [Fe(CN)6]3– reduction.

In addition, we examined the catalytic activity of the Ru@PDC catalyst using Na2S2O3 as an alternative reducing agent. At first, we tested the catalytic reduction reaction in the absence of a catalyst, but the result showed that the reduction did not proceed significantly in Na2S2O3 solution (Figure S11a, Supporting Information). Then, we added different dosages of catalyst into the reaction mixture and found the depletion of [Fe(CN)6]3– peaking at 420 nm (Figure S11b,c, Supporting Information), revealing that the [Fe(CN)6]3– is converted into [Fe(CN)6]4–; the corresponding kinetic plot is given in inset of Figure S11 Supporting Information. The linear plot of ln(A/A0) against time yielded the apparent rate constant to be 0.0932 s–1 (1.0 mg) and 0.1154 s–1 (2.0 mg) at ambient temperature (Figure S8, Supporting Information). A few important studies reported previously based on the Na2S2O3 reducing agent are selected for comparison with our results, as summarized in Table S7. For instance, Ajit et al.[52] have reported a porous platinum nanostructured catalyst for the catalytic reduction of [Fe(CN)6]3–, yielding a smaller rate constant, kPt NNs = 1.52 × 10–4 s–1 for 60 min. Likewise, the Au/boehmite[26] catalyst yielded the reduction rate constant, kAu/boehmite, = 5.16 × 10–5 s–1 and NiWO4 NP[53] yielded kNiWO = 1.06 × 10–4 s–1. These catalytic results are less efficient than those obtained with the Ru@PDC catalyst.

The reaction mechanism invoked for the catalytic reduction of [Fe(CN)6]3– in the presence of reducing agents over the Ru@PDC catalyst comprises two steps based on the earlier reports.[24] They are (i) rapid polarization of the metal NP by NaBH4 (fast) and (ii) transfer of excess surface electrons to [Fe(CN)6]3– complex ions (slow). That is, an electron-transfer process must be involved to form the reduced [Fe(CN)6]4– ions. Hence, the reduction reactions in an aqueous solution can be written as eq (48b)

To test the catalytic activity of Ru@PDC, [Fe(CN)6]3– reduction by Na2S2O3 was investigated involving the following reaction eq (54)

Catalytic Activity

Catalytic activity depends significantly on particle size and shape, and thus, how to synthesize colloidal nanoparticles with well-controlled size and shape is urgent and challenging.[31] Actually, the nanomaterial owns several catalytic merits including crystal plane, crystal phase, and small size. Among them, the size effect is an important parameter for both homogenous and heterogeneous catalysis. The particle size effect of metal nanoparticles on the catalysis has been thoroughly investigated.[55] Regarding the particle size effect, the Ru@PDC catalyst yielded higher activity because of the smaller size of Ru embedded, resulting in a faster reaction (k = 0.1011 s–1), probably due to the larger surface area (particle size 5 ± 0.2 nm), in contrast to the boehmite supported Au NP (15–40 nm),[26] which led to the k was 0.103 min–1 for the reduction of K3[Fe(CN)6]. Likewise, the cases of Fe3O4@Au hollow sphere[20] (Au NP: 25–30 nm, k: 36.55 × 10–3 s–1), graphene/Pd[47b] (Pd NP: 18.8 nm, k: 36.55 × 10–3 s–1) catalysts appear to have larger particle size than our catalyst systems. The detailed data of various catalysts with different sizes of metal nanoparticles and their rate constants are listed (Table S6, Supporting Information). The results indicated that the smaller particle size may lead to a larger surface area and subsequently was favorable for a good catalyst with high efficiency. In other words, the small nanoparticles are more effective catalysts than the larger NP, because an increase in the electron density for the small metal atoms leads to an increased reactivity on the surface of the catalysts. Moreover, PDC-supported metal (Mn, Fe, Co, Cu, and Ni) catalysts for the reduction of K3[Fe(CN)6] were investigated, and the results given in Table S6 (Supporting Information) shows that the reaction is sensitive to the change of metals. Turnover frequency (TOF) is used to quantify the catalytic activity of Ru@PDC and is defined as the number of [Fe(CN)6]3– molecules converted to [Fe(CN)6]4– with 1.0 mg of catalyst per s. Typically, 3.0 mmol of [Fe(CN)6]3– solution was completely reduced in the presence of Ru@PDC in 30 s, while Ru NPs and Ru/C exhibited lower catalytic activity within the same reaction time. We can simply calculate TOF values for a catalytic reaction using eq :[56]

The TOF of Ru@PDC was up to 5.0 × 10–5 s–1 for the [Fe(CN)6]3– reduction reaction, which was much larger than those of other nanocatalysts reported previously and was comparable with that of Ru NP and commercial Ru/C (see Table S7, Supporting Information).

Effect of Catalyst Dosage

Figure S12 Supporting Information shows the k result as a function of different amounts of catalyst (0.25–3.0 mg mL–1). The k value is proportion to the catalyst amount because of an increase in the number of reaction sites.[57] The obtained slope can be used to evaluate the pseudo-first-order of kinetic rate constant. It becomes feasible to understand why a large catalyst amount in the reaction may result in a rapid reduction of [Fe(CN)6]3–. For instance, Reddy et al.[58] have demonstrated that gum acacia-stabilized gold nanoparticles (GA/Au NPs) for catalytic reduction of [Fe(CN)6]3– showed a similar dependence of the k value on catalyst dosages. Moreover, the reaction orders and the rate constants were displayed in Table S8 (Supporting Information). Hence, the reaction constant and the order of the reaction are increasing linearly with increase of Ru@PDC catalyst dosage.

Stability and Reusability

Stability and reusability should be considered for the practical applications of catalysts.[59] The Ru@PDC catalyst may be facilely recovered by ultracentrifugation (10 000 rpm) after the reaction. As a result, the catalyst was inspected up to six consecutive cycles for the reduction of [Fe(CN)6]3– and monitored by UV–vis spectroscopy. As shown in Figure a, the apparent rate constants reveal that the catalyst preserves more than 85% of the initial catalytic ability after six consecutive cycles. The reaction mixtures were also analyzed by ICP–optical emission spectrometry to check if any Ru was leached, but no significant leaching was found. Hence, the reaction rate change cannot be caused by the loss of Ru from the catalyst. The rational decrease of catalytic ability might be due to the trace loss of catalyst after several times of use, as confirmed by HRTEM images (Figure b,c). The powder XRD pattern of the reused Ru@PDC catalyst demonstrates to understand retention of the crystallinity after the catalytic reaction as shown in Figure d. The energy-dispersive X-ray spectroscopy (EDX) analysis of the Ru NP proved the existence of an elemental ruthenium signal. Other EDX signals emitting from C, O, and Cu atoms were also noticed (Figure e). This result indicated that the Ru species were successfully immobilized on the nanocomposite, even after being recycled six times.

Figure 7

(a) Recycling test of the Ru@PDC catalyst toward the [Fe(CN)6]3– reduction, (b,c) FE-TEM images, (d) XRD pattern, (e) EDX spectrum, (f) N2 sorption-isotherm, and (g) corresponding pore size distribution of the reused Ru@PDC catalyst.

Figure f shows the nitrogen adsorption/desorption isotherms of the reused Ru@PDC catalyst and pore-size distributions (Figure g) derived from the adsorption branch of the isotherms according to the BJH method. We also noted that the surface area of the spent Ru@PDC catalyst had shown slightly lower surface area (SBET = 388.3 m2 g–1) than the fresh catalyst (SBET = 396.5 m2 g–1, Table ) after six uses. The catalytic activity decreased from the fourth to sixth cycles, probably because of the loss of catalyst during the recycling process. Notably, the reused catalyst exhibits a type-IV curve corresponding to a mesoporous structure with pore volume (VTot = 0.080 cm3 g–1); hence, these results demonstrated that the reused Ru@PDC catalyst remained stable without any change in its porous structure after six runs.

Reduction of New Fuchsin (NF)

In addition, we performed the reduction of a cationic triarylmethane dye (new fuchsin, NF); the chemical structure and characteristics of dye used in the study is shown in Table S9, (Supporting Information). The reduction process of NF dye was monitored using UV–vis spectrophotometry, as shown in Figure . Note that the reaction does not proceed significantly in the absence of catalyst, indicating the indispensable role of the catalyst for the NF reduction. As shown in Figure b–f, the absorbance peak of NF at 543 nm decreases to a different extent depending on the added amount of 0.5, 1.0, 1.5, 2.0, and 3.0 mg of Ru@PDC catalyst. The NF solution changes red color to colorless when the reduced NF is formed, confirming the catalytic activities (inset of Figure b–f). As revealed in Figure g, the absorbance at 543 nm disappears within 9 min after the introduction of 3.0 mg of Ru@PDC catalyst. In contrast, 30, 18, 16, and 13 min are required to complete the reduction by adding 0.5, 1.0, 1.5, and 2.0 mg of catalysts, respectively (see Figure h). It suggests that an increased catalyst dosage should result in fast diffusion of dye molecules/reagents on the catalyst surface and thus enhance the catalytic activity which may be reflected in the apparent rate constant kapp measurements (Figure h). The kapp value is evaluated to be 0.1911, 0.3061, 0.4019, 0.6903, and 0.7601 min–1 for addition of 0.5, 1.0, 1.5, 2.0, and 3.0 mg of catalyst, respectively (see Figure i). These results indicate that the reaction follows a pseudo-first-order kinetics because of the presence of excessive NaBH4. The kapp obtained for Ru@PDC (0.7601 min–1) turns out to be much larger than that of graphene quantum dots (kGQDs = 0.0263 min–1).[60] The fact may be caused by (i) a large surface area of Ru@PDC to effectively adsorb a more amount of NF molecules and (ii) fast electron transfer from NaBH4 to the adsorbed NF dye molecules via the catalyst. Thus, it facilitates the reduction of organic pollutants absorbed on the catalyst surface.

Figure 8

UV–vis spectra for the reduction of NF dye in aqueous medium in the (a) absence of catalyst, in contrast to the presence of (b) 0.5, (c) 1.0, (d) 1.5, (e) 2.0, and (f) 3.0 mg of Ru@PDC catalyst, (g) plot of ln(A/A0) vs time for catalytic reduction of NF at different catalyst dosages, (h) linear region of plot of ln(A/A0) vs time of catalytic reduction of NF used for calculation of kapp, and (i) dependence of kapp on catalyst dosage for reduction of NF.

Conclusions

In this work, a porous and heterogeneous stable catalyst, Ru@PDC, was successfully prepared and characterized. The catalyst exhibit enhanced catalytic activity for the reduction of inorganic complex and cationic dye with higher reaction kinetics as compared with other catalysts. Furthermore, the as-prepared nanocatalyst possesses several advantages: (i) it can be easily separated from the reaction mixture, (ii) not much loss of catalytic activity is found even after several cycles of reuse, (iii) it takes short time (∼30 s) to complete the reaction showing superior activity, (iv) the catalyst was fabricated using plastic as solid waste feedstock, and (v) moreover, the catalyst has been used for perspective applications. The catalysts were found to be stable and effective for more than 6 runs, with a conversion efficiency of ∼98%. The decrease of conversion can be probably attributed to a reduction in the surface active sites of the catalyst.

Experimental Section

Materials

Ruthenium(III) acetylacetonate (Ru(acac)3, ≥99.9%), ruthenium black (Ru/C, ≥98%), potassium ferricyanide (K3[Fe(CN)6], 99%), New Fuchsin (NF), sodium borohydride (NaBH4, 99.99%), and sodium thiosulphate (Na2S2O3, ≥98%) were purchased from Sigma-Aldrich. All other chemicals belonged to analytical grade, and all solutions were freshly prepared using Milli-Q water.

Preparation of Porous Carbon

Plastic-derived carbon (PDC) have been prepared from soft drink plastics collected from a local market at Taipei, which were utilized as carbon sources according to the methods reported previously.[7−9] Typically, waste plastic bottles shredded into small size of 30–50 mm pieces (∼2.0 g) were added in the 100 mL capacity Teflon-lined stainless steel autoclave. The sealed autoclave was then heated to the temperature (ramp at 20 °C min–1) in the muffle furnace at ∼300 °C for 6 h. Subsequently, the carbonaceous material was cooled slowly back to room temperature (RT), followed by thorough washing with copious amounts of benzene and ethanol, and then air-dried at 100 °C overnight. The PDC samples thus obtained were represented to be PDC-x, where x denotes the final carbonization temperature (in °C) used. To enhance their porosities, the PDC substrates were pyrolyzed again under the condition of flowing CO2 (flow rate 30 mL min–1) at 400 °C for 30 min. In comparison, the carbon sources generated by other plastic materials such as HDPE, LDPE, PC, and polypropylene showed the maximum extent of oil formation containing aromatic hydrocarbons to form CSs. The mechanism of char formation in the thermal process of polymer degradation was interpreted appreciably elsewhere.[3,4] Furthermore, we have analyzed the chemical composition of plastic wastes and their elemental composition of plastic-derived char (before activation), PDC (after activation), and Ru@PDC as displayed in Tables S2–S4, Supporting Information. In addition, the yields of carbon after carbonization and activation under different conditions were presented in Table S5, Supporting Information.

Physical Activation

In general, the endothermic reactions for physical activation of the carbonaceous material by using water vapor steam and CO2 are shown in eqs and 9(61)

Besides, the C + H2O reaction in eq is accompanied by the formation of CO2 + H2, while catalyzed on the carbon surface as shown in eq . Because of endothermic reactions for eqs and 9, the activation process may be controlled accurately in the heating furnace. External heating to remain high temperature is required to drive. In contrast to the reactions 8 and 9 which can be driven accurately at high temperature, the reaction 11 is extremely exothermic and its progression is difficult to control. A decrease in the average particle size and the product yield may happen as a result of over-heating the external carbon surface.[62]

Preparation of the Ru@PDC Catalyst

Ru@PDC nanocomposite was obtained by immobilization of Ru NPs on the PDC-800 support. In brief, 0.2 g of the as-prepared PDC-800 powdered sample was mixed with Ru(acac)3 for 4.01 wt % loading in 5.0 mL tetrahydrofuran. Then, the mixture was removed into a 50 mL Teflon-coated microwave reactor for microwave irradiation at power of 300 W for 1 h. The Ru0 state was reduced effectively from the Ru3+ ions upon irradiation and was dispersive on the mesoporous PDC carbon support.

Catalyzed Reduction of Ferrocyanate(III)

To conduct the reduction reaction, we prepared 3.0 mL of 3 × 10–3 M [K3Fe(CN)6] into 1.0 mg mL–1 Ru@PDC catalyst and then added rapidly 0.2 mL of 0.04 M ice-cold fresh NaBH4 or Na2S2O3. As the reaction proceeded, the solution in yellow faded to be colorless. Subsequently, the kinetic measurements were carried out for K3Fe(CN)6 by using UV–vis spectroscopy fixed at 420 nm to monitor the reduction reaction in a quartz cuvette. When the reaction was complete, the catalyst was removed by using an ultracentrifuge for further investigation of its reusability.

Catalyzed Reduction of the Cationic Dye

Experimental procedure for the cationic dye reduction is based on an earlier report.[63] Typically, 1.0 mg mL–1 of solid catalyst (Ru@PDC) was added into 3.0 mL of NF (5 mM) dye solution at RT under vigorous magnetic stirring in dark condition for 5 min to allow the dye molecules adsorbed physically onto the Ru@PDC composite. Then, 0.1 mL of 0.04 M ice-cold NaBH4 was added and the mixture was kept stirring under ambient conditions. Approximately, 3.0 mL of sample was taken from the mixture and the dye solution was analyzed by UV–vis spectrophotometric measurements (at 553 nm) to follow the catalytic reduction reaction.