- Email: [email protected]
Facile radiolytic synthesis of ruthenium nanoparticles on graphene oxide and carbon nanotubes
Materials Science and Engineering B 205 (2016) 28–35
Contents lists available at ScienceDirect
Materials Science and Engineering B journal homepage: www.elsevier.com/locate/mseb
Facile radiolytic synthesis of ruthenium nanoparticles on graphene oxide and carbon nanotubes J.V. Rojas a,∗ , M. Toro-Gonzalez a , M.C. Molina-Higgins a , C.E. Castano b a
Mechanical and Nuclear Engineering Department, Virginia Commonwealth University, 401 West Main Street, Richmond, Virginia, 23284, USA Nanomaterials Core Characterization Facility, Chemical and Life Science Engineering Department, Virginia Commonwealth University, 601 West Main Street, Richmond, Virginia, 23284, USA b
a r t i c l e
i n f o
Article history: Received 27 September 2015 Received in revised form 4 December 2015 Accepted 15 December 2015 Available online 24 December 2015 Keywords: Carbon nanotubes Graphene oxide Nanoparticles Ruthenium Radiation chemistry Radiolysis Gamma irradiation
a b s t r a c t Ruthenium nanoparticles on pristine (MWCNT) and functionalized carbon nanotubes (f-MWCNT), and graphene oxide have been prepared through a facile, single step radiolytic method at room temperature, and ambient pressure. This synthesis process relies on the interaction of high energy gamma rays from a 60 Co source with the water in the aqueous solutions containing the Ru precursor, leading to the generation of highly reducing species that further reduce the Ru metal ions to zero valence state. Transmission electron microscopy and X-Ray diffraction revealed that the nanoparticles were homogeneously distributed on the surface of the supports with an average size of ∼2.5 nm. X-ray Photoelectron spectroscopy analysis showed that the interaction of the Ru nanoparticles with the supports occurred through oxygenated functionalities, creating metal-oxygen bonds. This method demonstrates to be a simple and clean approach to produce well dispersed nanoparticles on the aforementioned supports without the need of any hazardous chemical. © 2015 Elsevier B.V. All rights reserved.
1. Introduction The synthesis of supported metal nanoparticles is of significant importance for various applications in nanotechnology [1]. Interestingly enough, this area of research is approached from two complementary points of view. First, the need to synthesize nanoparticles with a homogeneous size and shape. Nanoparticles tend to aggregate due to their high surface area, causing deleterious effects in their physical and chemical properties, directly impacting their overall performance. Therefore, the use of organic and inorganic supports has been implemented during the nanoparticle synthesis to control the growth and hinder the formation of nanoparticle aggregates. The second point of view resides in the potential to enhance the properties of certain materials by adding metallic nanoparticles on their surface, that may result in a synergistic effect between the support and the metallic nanoparticles [2]. In this context, various types of supports such as polymers, carbon based materials, and oxides have been reported in the
∗ Corresponding author. Mechanical and Nuclear Engineering Department, Virginia Commonwealth University, 401 West Main Street, Richmond, Virginia 23284-3067301, Tel.: +1 804 8284267; fax: +1 804 8277030. E-mail addresses: [email protected] (J.V. Rojas), [email protected] (C.E. Castano). http://dx.doi.org/10.1016/j.mseb.2015.12.005 0921-5107/© 2015 Elsevier B.V. All rights reserved.
literature. The use of supports such as carbon nanotubes and, most recently, graphene oxide have stimulated their interest in catalysis, biosensors, and electronics, among others. These supports offer significant advantages due to their high chemical and thermal stability, high surface area, and outstanding mechanical, physical and chemical properties [2–4]. Furthermore, the surface of the carbon supports can be modified with different functionalities to trigger nanoparticle-support interactions [5,6]. The synthesis of transition metal nanoparticles on carbon nanotubes and graphene oxide has been reported with substantial interest in catalytic applications. Among those, ruthenium nanoparticles on carbon supports have been investigated for hydrogenation reactions [3,7–22]. Ruthenium is a transition metal occurring in the Group VIIIA of the periodic table with an electronic configuration of [Kr](4d)7(5s)1, displaying different oxidation states from −2 to +8 in its compounds. The most common precursor in the chemistry of Ru is the hydrated Ru(III) chloride, RuCl3 ·xH2 O, which is a compound that exhibits in solution a complex mixture of Ru(IV) and Ru(III) species rather than a pure source of Ru(III) [23]. Ruthenium is a very stable metal at low temperatures but usually oxidizes at high temperatures. However, in solid phase it only forms one stable oxide, RuO2 . Other oxide forms such as RuO, RuO3 and RuO4 are only stable at temperatures above 900 K, being volatile and toxic [24]. Various approaches have been reported in the literature with regards to the synthesis of Ru nanoparticles on carbon supports.
J.V. Rojas et al. / Materials Science and Engineering B 205 (2016) 28–35
Microwave synthesis of Ru on carbon nanotubes (CNTs) [8,25] and graphene [16,17] has been previously achieved at a controlled pressure, temperature, time, and atmosphere, leading to Ru nanoparticle with sizes of 2–4 nm. Another method for the production Ru nanoparticles on CNTs is through the reduction of ruthenium salt in polyol at temperatures above 100 ◦ C, this method resulted in ∼5 nm Ru nanoparticles on the CNTs [14]. Furthermore, the decoration of CNTs with Ru nanoparticles using supercritical water has been reported. In this approach, the particles are synthesized in an autoclave at ∼450 ◦ C for about 2 h giving 5 nm Ru nanoparticles on the CNTs [15]. Supercritical water has also been reported to produce Ru on graphene oxide. In this case, the reduction of Ru precursors takes place in a autoclave maintained at a temperature of ∼400 ◦ C and a pressure of 23.5 MPa for 2 h under nitrogen atmosphere, producing Ru nanoparticles of ∼3 nm [11]. Carbon nanotubes have also been decorated with Ru nanoparticles through the impregnation method. In this approach, the support is initially mixed with the metal precursor and the solvent and homogeneously dispersed through ultrasonication. Then, the samples are stirred and aged for a period of time, followed by drying and reduction with hydrogen flow [9,26]. Besides the various physical and chemical methods to synthesize Ru nanoparticles on carbon supports, alternative synthesis methods that minimize the use of additional reagents, as well as the need for high temperature and pressure, or controlled environment are always sought. This would allow for potential scalability and reproducibility of the nanocomposites. Radiation chemistry has recently made significant contribution to the field of synthesis of nanomaterials and the modification of their properties for specific applications [27]. The interaction of ionizing radiation, such as gamma rays, with aqueous solutions leads to the generation of highly reducing species that can bring metal ions present in the solution down to zero valence state. Furthermore, due to the mechanisms of interaction of gamma rays with matter, the reducing species are homogeneously distributed in the solution. Consequently, the metallic atoms will also be generated homogeneously throughout the solution followed by coalescence and growth [28]. This synthesis method eliminates the need for reducing agents that may sometimes poison the nanomaterial and affect their behavior. Moreover, the synthesis is carried out in aqueous environments which is desirable for certain applications. Nonetheless, it can also be implemented with other solvents such as ammonia or alcohols [29]. Radiation synthesis has previously been used for the synthesis of a wide variety of nanomaterials, such as metallic nanoparticles, core-shell nanostructures, metal oxides, and alloys [28,30–32]. In this work, ruthenium nanoparticles have been synthesized on pristine multi-walled carbon nanotubes, carboxylic acid functionalized carbon nanotubes, and reduced graphene oxide through a radiolytic method using 60 Co as the gamma source. The resulting Ru-C support nanostructures were characterized with XRD, TEM, and XPS, to study the morphology of the nanoparticles and their interaction with the support. The effect of the surface modification of the nanotubes on the yield of Ru nanoparticles is evidenced and a mechanism of synthesis and interaction of the particles with the support is presented here. This work evidences the potential use of gamma irradiation as a facile and novel method for the synthesis of well distributed Ru nanoparticles on carbon supports.
2. Materials and methods 2.1. Materials Multi-walled carbon nanotubes (MWCNTs), functionalized multi-walled carbon nanotubes (f-MWCNTs), graphene oxide (GO), ruthenium (III) chloride hydrate (RuCl3 ·xH2 O), and sodium
29
dodecyl sulfate (SDS; CH3 (CH2 )11 OSO3 Na) were obtained from Sigma–Aldrich and used without any further purification. Multiwalled Carbon nanotubes were > 98% carbon basis, length within 2.5 and 20 m and outside diameter within 6 and 13 nm. Functionalized multi-walled carbon nanotubes had an extent of labeling of carboxylic acid of > 8%, an average diameter of 9.5 nm, and average length of 1.5 m. Deionized water (18 M) obtained from an EMD Millipore Direct-QTM 3 UV water purification system was used to prepare the solutions along with high purity (99.5%) isopropyl alcohol (C3 H8 O) from ACROS Organic which was used as a radical scavenger. 2.2. Sample preparation Samples containing the carbon-based support (MWCNTs, fMWCNTs or GO), ruthenium precursor, and surfactant were prepared by modifying a methodology described in the literature [31,33]. As an initial step, a solution of water and isopropanol (2:1 v/v) was prepared, vortexed and stirred for ∼5 min. Subsequently, a solid mixture of RuCl3 ·xH2 O, SDS, and either MWCNTs, f-MWCNT or GO was prepared and stored in a glass vial. Next, 20 ml of waterisopropanol solution was added to the mixtures and sonicated for ∼30 min using a probe sonicator to break up the aggregates. This procedure lead to a homogeneous black suspension containing 5 mM RuCl3 (0.1 mmol), 5 mM SDS (0.1 mmol), and 1 mg of MWCNTs, f-MWCNTs, or GO per milliliter of solvent. 2.3. Irradiation procedure The aqueous solutions prepared by the procedure mentioned above were irradiated with a Cobalt-60 Gammacell 220 Excel (MDS, Nordion) that consists of a cylindrical stainless steel cage containing double sealed source pencils of Co-60. For irradiation purposes, the samples were placed into the chamber through a drawer mechanism and positioned in the outermost location in order to ensure a uniform absorbed dose distribution. The samples were irradiated at a dose rate of ∼7 kGy/h (0.7 Mrad/h) until an absorbed dose of 60 kGy was achieved. Then, the suspensions were cleaned to remove any excess of surfactant and/or unreacted species. For the cleaning process, the samples were initially centrifuged followed by removal of the supernate. Next, the precipitate was redispersed in deionized water and repeated several times. Finally, the samples were dispersed in acetone for further characterization. 2.4. Materials characterization The morphology, particle size, and particle size distribution of the resulting nanoparticles were studied by transmission electron microscopy (TEM) with a Zeiss Libra 120 Plus operating at 120 kV. For TEM imaging, a drop of the purified nanomaterial was diluted in acetone and sonicated for ∼10 min. The samples were prepared by immersing a formvar-carbon coated copper grid into the diluted suspension and allowed to dry at room temperature. Particle size analysis was carried out using the Image J 4.18 v software. The crystalline structure of the nanoparticles deposited on the carbon supports was investigated through X-ray diffraction (XRD) using a PANalytical X’Pert Pro MPD X-ray diffractometer with a ˚ The samples were prepared by copper anode (Cu K␣, = 1.5401 A). depositing drops of the sample dispersion on a silicon wafer and allowing it to dry. This process was repeated until a thick layer was obtained. X-ray photoelectron spectroscopy (XPS) was carried out using a ThermoFisher Escalab 250 X-ray photoelectron spectrometer with a monochromated aluminum X-ray source. For the analysis, few drops of the sample were deposited on a silicon wafer and allowed
30
J.V. Rojas et al. / Materials Science and Engineering B 205 (2016) 28–35
Fig. 1. TEM image and particle size distribution of MWCNT decorated with Ru nanoparticles at 60 kGy and 5 mM SDS.
Fig. 2. TEM image and particle size distribution of f-MWCNT decorated with Ru nanoparticles at 60 kGy and 5 mM SDS.
to dry. The software CASAXPS 2.3.16 v was used for profile fitting and analysis of the spectra. 3. Results and discussion 3.1. Synthesis of the nanoparticles on the supports Various methods to synthesize Ru nanoparticles on carbon supports have been previously reported in the literature. Among those, chemical reduction [13], electrochemical reduction [34], sol–gel [35], microwave assisted synthesis [8], and hydrothermal reduction [36] have been widely explored. In this work, a method to synthesize Ru nanoparticles on carbon supports such as carbon nanotubes and graphene oxide in aqueous solution using gamma irradiation was developed. The use of gamma rays to produce nanoparticles
has significant advantages that can lead to the generation of well dispersed nanoparticles without the need for aggressive reducing agents nor the use of high pressure or temperature. The synthesis of the Ru nanoparticles on the carbon supports using gamma irradiation relies on the radiolysis of water, process of high importance in radiation chemistry. When an aqueous solution is irradiated with high energy gamma rays, several species such as e− aq , OH• , H• , H+ , OH- , and H2 O2 are generated and the yield is determined by the absorbed dose in the aqueous solution. The species OH• and H2 O2 are highly oxidizing species whereas e− aq and H• are reducing agents [37]. Therefore, the environment can be triggered towards an oxidizing or reductive environment through the addition radical scavengers. The species e− aq and H• have sufficient redox potential to bring Ru (III) and Ru(IV) species down to a zero valence state. In order to prevent the oxidation of Ru(0)
Fig. 3. TEM image and particle size distribution of GO decorated with Ru nanoparticles at 60 kGy and 5 mM SDS.
J.V. Rojas et al. / Materials Science and Engineering B 205 (2016) 28–35
due to the presence OH• and H2 O2 , isopropanol was used as radical scavenger. This secondary alcohol reacts with the oxidizing species resulting in the generation of more reducing species. The generation of Ru(0) subsequently leads to the formation of Ru nuclei that continue to the growth process. Since the reduction of the Ru precursor was carried out in the presence of the carbon supports, the formation and growth of the nanoparticles takes place on their surface. Functional groups on the carbon nanotubes and the graphene oxide act as anchoring points for the nanoparticles [38]. Therefore, in this synthesis method, homogeneous dispersion of the support in order to expose a high number of functional groups before irradiation is imperative. The SDS was added to facilitate the dispersion of the carbon support in the aqueous solution since the surfactant promotes homogeneous distribution of the support in the solution through steric and electrostatic interactions [39].
31
Fig. 4. XRD pattern of MWCNTs, f-MWCNTs, and GO.
3.2. Size and distribution After gamma irradiation, the suspensions containing the nanomaterial were characterized using TEM, XRD, and XPS. Homogeneously distributed Ru nanoparticles on the various carbon supports were observed in the TEM images shown in Figs. 1–3. The nanoparticles are deposited and strongly bound to the supports since they stayed after the cleaning process of the samples which included several steps of sonication and centrifugation. Comparing the TEM images of the different supports, it can be observed that the nanoparticles are finely dispersed on the walls of functionalized carbon nanotubes and on the surface of the graphene oxide. However, fewer nanoparticles and relatively large aggregates on the walls of the non-functionalized nanotubes are observed with respect to the f-MWCNTs. The superior amount of well dispersed nanoparticles on the carboxylic acid functionalized carbon nanotubes relative to that observed on the non-functionalized may be explained by the larger number of nucleation sites homogeneously distributed on their surface. The average particle size of Ru obtained for the MWCNTs, f-MWCNTs, and GO were 2.5 ± 0.5 nm, 2.2 ± 0.6 nm, and 2.4 ± 0.7 nm, respectively. The size distribution histograms of the Ru nanoparticles on the different supports are presented in Figs. 1–3. The formation of the metallic nanoparticles with narrow particle size distribution is explained by the mechanism of nuclei formation due to gamma irradiation and the presence of the surfactant SDS. Gamma irradiation of water leads to the formation of homogeneously distributed reducing species such as e− aq and H• and the number of species can be controlled with the absorbed dose. Given the large absorbed dose used in this synthesis (60kGy), a large concentration of reducing species are formed in solution ([e− aq ] = 1.67 × 10−2 M, [H• ] = 3.79 × 10−3 M), consuming most of the Ru ions available in solution during the nucleation process. The presence of a solid support (CNT or GO) providing nucleation sites along with the surfactant generating electrostatic forces and steric hindrance limit the diffusion of Ru and control the nanoparticle growth. Therefore, small nanoparticles with a narrow size distribution are produced and deposited on the surface of the carbon support. 3.3. Structure The XRD patterns for the three carbon supports are shown in Fig. 4. The spectra for the non-functionalized and functionalized multi-walled carbon nanotubes have the characteristic peaks of carbon at 25.68◦ (0 0 2) and 42.21◦ (0 1 0) (PAN-ICSD Ref. code 98008-5678). On the other hand, the graphene oxide pattern has the most intense peak at 11.4◦ that corresponds to the plane (0 0 1) of GO [5,11,40]. The spectra for the different carbon supports decorated with ruthenium nanoparticles are presented in Fig. 5. The peaks at 2 of 38.3◦ ,42.0, 43.8◦ , 58.0◦ , and 69.0◦ correspond to
Fig. 5. XRD patterns of Ru nanoparticles supported on MWCNTs, f-MWCNTs, and GO.
the planes of hexagonal closed packed metallic ruthenium (010), (0 0 2), (0 1 1), (0 1 2), and (1 1 0), respectively (PAN-ICSD Ref. code 98-009-0997, with 2 of 38.45◦ ,42.27, 44.14◦ , 58.49◦ , and 69.64◦ ). Additionally, the peaks at 2 angles of 11.4◦ , 22.6◦ , and 25.9◦ correspond to graphene oxide (0 0 1), reduced graphene oxide (0 0 2), and graphite (0 0 2) plane reflections of CNTs, respectively [40,41]. Moreover, the broadening observed on the ruthenium and carbon peaks is characteristic of nanocrystalline materials. Therefore, the XRD patterns were used to calculate the average particle size by using the Scherrer Equation: L = K/(ˇ cos) Where K corresponds to a shape factor related with the shape of the crystallite (which is equivalent to 0.89, assuming a circular grain), is the X-ray wavelength (1.5406 A˚ for Cu),  is the broadening of diffraction line measured at half its maximum intensity and corresponds to the peak angle [42]. Using Scherrer equation on the ruthenium peak at 43.8◦ , the average particle size was 2.5 nm, 2.5 nm, and 2.7 nm for MWCNTs, f-MWCNTs, and GO, respectively. These results are in agreement with the results obtained from the TEM analysis. The slight difference between the results obtained from the two techniques resides on the fact that TEM analysis leads to a local measurement of the average NP diameter, whereas XRD provides a more statistical representation of the particle size due to the larger amount of sample that contribute to the diffraction pattern. 3.4. Surface composition and chemical states XPS analysis was performed to study the surface composition and chemical state of the different carbon supports before and after decoration with ruthenium nanoparticles. Fig. 6 shows the high resolution (HR) C1s and O1s XPS spectra for the different carbon supports. In the case of MWCNTs, Fig. 6 a), C1s was deconvoluted into five peaks and O1s was deconvoluted into three peaks.
32
J.V. Rojas et al. / Materials Science and Engineering B 205 (2016) 28–35
Fig. 6. High resolution XPS spectra of C1s and O1s for (a) MWCNTs, (b) f-MWCNTs, and (c) GO.
Regarding the C1s, the peaks located at 284.5 eV and 284.8 eV corresponds to C C bonds with sp2 hybridization and C C bonds with sp3 hybridization. The peak observed at 285.7 eV associates to the contribution of both C O and C OH functionalities. The following peak at 289.4 eV represents carboxyl groups [43]. Also, a peak at 291.1 eV is observed in the pattern, this peak is associated with a shake-up feature typical of aromatic structures or –* shake-up. This feature is related to the interaction of emitted photoelectrons with electrons [44,45]. For the O1s, the peak at 532.3 reveals the presence of C O/C O bonds. It is worth to notice that the position of this peak above 532 eV may indicate a larger contribution from C O than from C O bonds. The next peak at 533.3 eV corresponds to C-OH bonds, while the peak at 534.7 eV is associated to oxygen in water molecules [46]. The HR XPS spectra of the C1s and O1s in the f-MWCNTs shown in Fig. 6 b) present similar features to the ones observed in MWCNTs. However, a broader C1s peak as well as more
pronounced peaks within 285 and 290 eV are remarkable in the fMWCNTs. Furthermore, the O1s peak is significantly broader in the f-MWCNTs. These characteristics can be associated to carbonoxygen interactions such as hydroxyl, carbonyl, and carboxyl groups. Also, the peak below 532 eV indicates a significant presence of carboxyl functionalities obtained from the functionalization through carboxylic acid [47]. Fig. 6 c) shows the XPS analysis of C1s and O1s peaks for the graphene oxide. The double peak observed in the C1s is characteristic of the graphene oxide and it was deconvoluted into six peaks. The first three components at BE of 284.6, 285.7, and 286.7 eV correspond to C sp2 hybridization, hydroxyl and epoxide groups, being the first and third component prominent. The last three are related to carbonyl, carboxyl and –* shake-up, respectively [5]. In addition, the two peaks at 531.1, 532.3, and 533.1 eV in the O1s peak are associated with single and double bonds between oxygen and carbon.
J.V. Rojas et al. / Materials Science and Engineering B 205 (2016) 28–35
33
Fig. 7. High resolution XPS spectra of C1s, O1s, and Ru3p for MWCNTs/Ru.
The composites formed by the carbon supports and ruthenium nanoparticles were also analyzed through XPS. The peaks corresponding to oxygen, carbon, and ruthenium can be distinctly detected in the survey scans of the different composites. Furthermore, the composites and nanoparticles are remarkably clean of byproducts according to the survey spectra, where the only element detected other than O, C, and Ru, came from the Si substrate used to prepare the sample. The Ru3p region was chosen instead of the Ru3d due to the strong overlapping of these peaks with the C1s, which increases the complexity of defining the electronic states of Ru. The HR XPS spectra of C1s, O1s, and Ru3p doublet for MWCNTs decorated with Ru NPs are shown in Fig. 7. To start with, the HR spectra for the C1s shows similar features to the C1s in the support alone. The peak observed at 284.6 eV corresponds to sp2 hybridized graphite-like carbon atoms. The second peak located at 284.8 eV corresponds sp3 hybridization, and the third and fourth peaks located at 285.8 and 290.1 eV correspond to hydroxyl and carboxyl groups, respectively. The HR O1s was deconvoluted into 3 peaks as shown in Fig. 7. The first one at 531.48 eV originates from the double bond between carbon and oxygen whereas the second one at 532.13 eV reveals C O bonds and it is also related with weak interactions between oxygen and ruthenium [48]. Furthermore, the third peak located around 529 eV reveals Ru O bonds [49]. Lastly, the Ru 3p core-level spectrum was analyzed and the doublet, with 3p3/2 and 3p1/2 components, was identified. The XPS fitting reveals the presence of 4 peaks, the two located at 462.2 and 484.4 eV correspond to Ru in the zero valence state whereas the other two at 466.1 eV and 488.6 eV evidence Ru at a higher valence state, Ru(IV). The Ru (IV) peaks in both states are shifted to higher binding energies respect to the anhydrous oxides species with the same oxidation state. As a result. these Ru (IV) peaks for the 3p3/2 and 3p1/2 states may correspond to some hydrated form Ru (IV) such as RuO2 ·xH2 O or RuOx Hy [50–52]. The HR spectra for C1s, O1s and Ru3p for the functionalized carbon nanotubes decorated with ruthenium nanoparticles are shown in Fig. 8. The C1s peak was deconvoluted into 4 bands at 284.5, 284.8, 286.1, and 288.7 eV. Similar to the features previously observed in the f-MWCNTs substrate, the peaks correspond to the sp2 and sp3 hybridization as well as carbon-oxygen interactions. Likewise, the fitting in the O1s peak also confirmed the presence of C-O interactions at BE of 531.8, and 532.7 eV. In addition, the peak at BE of 529.9 reveals the formation of Ru-O bonds [48] such as that of ruthenium (IV) dioxide. Finally, the Ru3p was fitted taking into account the 3p3/2 and 3p1/2 components. Similar to the results observed in the Ru-MWCNTs, the main contribution to the spectrum is given by the Ru (0), whose peaks are observed at BE of 462.4 and 484.6 eV. Furthermore, the contribution, identified as that of Ru (IV) at 466.3 and 488.2 eV, is also observed [51]. Fig. 9 shows the HR XPS spectra of C1s, O1s, and Ru3p for the ruthenium nanoparticles supported on graphene oxide. The C1s
does show a significant difference compared to the C1s observed for the graphene oxide in Fig. 6c. There is a substantial reduction in the contribution of the C O peak after the irradiation process which indicated the transformation of GO into reduced graphene oxide, rGO. Nonetheless, the peaks that represent sp2 and sp3 hybridizations as well as carbon-oxygen interaction are present within the BE interval from 286 to 290 eV. The peaks at 531.6 and 532.5 eV in the O1s, once again, are linked to carbon-oxygen bonds, and the presence of Ru-O interaction is also revealed in this case at 529.9 eV. The Ru3p spectrum was fitted for the 3p3/2 and 3p1/2 where the Ru (0) and ruthenium (IV) are observed, being Ru(0) the major contribution to the pattern. It has been demonstrated that the more carboxylic groups in the surface of the carbon supports the better the metal dispersion due to the promotion of anchoring sites through metal cation exchange. Since the samples were prepared by a ruthenium precursor that contains Ru(III) and Ru(IV) species, it is expected that both species are adsorbed at the acidic functional groups. Experimental and theoretical results have shown that Ru species are extremely stable when two of the Ru bonds are with the oxygenated functional groups of carbon [38,53]. Therefore, since Ru(IV) is more prone to form solid stable oxides, it may explain the appearance of this species in the results obtained by XPS. Under gamma irradiation, the generation of reduced Ru(0) from the Ru(III) and Ru(IV) species leads to the formation of the Ru nanoparticles stabilized in the surface of the carbon supports anchored by Ru(IV)-OOC bonds. Although, adsorbed species Ru(II) and Ru(III) on the carbon supports with partial charge transfer could be proposed, the presented deconvolution of the Ru 3p HRXPS peak only matched with Ru(0) and Ru(IV) species. Ru(IV) species may have also been formed on the Ru nanoparticles due to the presence of O2 molecules in aqueous solution and/or when exposure to air during XPS sample preparation [38]. This single step production method of supported Ru nanoparticles on carbon nanotubes and reduced graphene oxide represents a novel approach that lead to the formation of well dispersed nanoparticles. The amount of nanoparticles deposited on the carbon nanotubes, as it has been reported in the literature, is clearly affected by the concentration of nucleation sites available on the nanotubes surface. In this case, both TEM and XPS analysis evidenced this fact. In other words, the large concentration of carbon-oxygen functionalities on the carboxylic acid functionalized carbon nanotubes demonstrated by XPS led to a higher yield of Ru nanoparticles deposited on their surface. Furthermore, the larger concentration of functional groups behaving as nucleation sites causes a more homogeneous particle size distribution and avoids the formation of aggregates. Similarly, the large concentration of functionalities on the GO resulted in < 3 nm Ru nanoparticles well distributed on their surface. Gamma irradiation represents a potential method for large scale production of
34
J.V. Rojas et al. / Materials Science and Engineering B 205 (2016) 28–35
Fig. 8. High resolution XPS spectra of C1s, O1s, and Ru3p for f-MWCNTs/Ru.
Fig. 9. High resolution XPS spectra of C1s, O1s, and Ru3p for GO/Ru.
this type of nanocomposites in a clean environment at room temperature. Gamma rays produce a uniform distribution of reducing species that results in a homogeneous nucleation and growth of the nanoparticles. 4. Conclusions In this work, the feasibility to produce ruthenium nanoparticles on carbon nanotubes, carboxylic acid functionalized carbon nanotubes and graphene oxide supports through a facile aqueous route using a gamma irradiation was demonstrated. This method is based on the hydrolysis of water that result in the generation of highly reducing and oxidizing species upon the interaction with the gamma rays. The use of isopropanol in this work allowed to scavenge the oxidizing species and created a highly reducing environment that brought Ru precursor species to a zero valence state. This synthesis approach lead to the production of nanoparticles homogeneously distributed on the surface of the supports with an average size of ∼2.5 nm. As expected, more nanoparticles and negligible aggregates were observed on the carboxylic acid functionalized MWCNTs compared to the pristine nanotubes. In the GO support, the large concentration of oxygenated sites led to the formation of homogeneous distributed nanoparticles. Furthermore, XPS revealed that the interaction of the Ru nanoparticles with the support was through oxygenated functionalities. This work evidences the potential use of gamma irradiation as a simple approach to make well dispersed nanoparticles on the aforementioned supports. Moreover, the fact that no reducing agents are required makes of this a clean approach to produce these materials. Acknowledgements This work was funded by the Virginia Commonwealth University with the support of the Mechanical and Nuclear Engineering department and the NRC-HQ-84-14-FOA-002, Faculty
Development Program in Radiation Detection and Health Physics at VCU. The authors would like to thank the Nanomaterials Core Characterization Facility (NCC) from the School of Engineering. Finally, we also would like to acknowledge the assistance received from Dr. Stoyan Toshkov at the Nuclear Radiation Lab, University of Illinois Urbana Champaign during the sample irradiation.
References [1] B.R. Cuenya, Thin Solid Films 518 (2010) 3127–3150. [2] J.M. Campelo, D. Luna, R. Luque, J.M. Marinas, A.A. Romero, ChemSusChem 2 (2009) 18–45. [3] C. Du, Q. Ao, N. Cao, L. Yang, W. Luo, G. Cheng, Int. J. Hydrogen Energy 40 (2015) 6180–6187. [4] G.G. Wildgoose, C.E. Banks, R.G. Compton, Small 2 (2006) 182–193. [5] T. Kuila, S. Bose, A.K. Mishra, P. Khanra, N.H. Kim, J.H. Lee, Prog. Mater. Sci. 57 (2012) 1061–1105. [6] V. Datsyuk, M. Kalyva, K. Papagelis, J. Parthenios, D. Tasis, A. Siokou, I. Kallitsis, C. Galiotis, Carbon 46 (2008) 833–840. [7] D.K. Mishra, A.A. Dabbawala, J.-S. Hwang, J. Mol. Catal. A Chem. 376 (2013) 63–70. [8] C. Antonetti, M. Oubenali, A.M. Raspolli Galletti, P. Serp, G. Vannucci, Appl. Catal. A Gen. 421–422 (2012) 99–107. [9] Y. Wang, Z. Rong, Y. Wang, P. Zhang, Y. Wang, J. Qu, J. Catal. 329 (2015) 95–106. [10] D.V. Jawale, E. Gravel, C. Boudet, N. Shah, V. Geertsen, H. Li, I.N.N. Namboothiri, E. Doris, Chem. Commun. 51 (2015) 1739–1742. [11] J. Zhao, W. Hu, H. Li, M. Ji, C. Zhao, Z. Wang, H. Hu, RSC Adv. 5 (2015) 7679–7686. [12] X. Zhang, Y. Ye, H. Wang, S. Yao, Radiat. Phys. Chem. 79 (2010) 1058–1062. [13] V. Selvaraj, M. Vinoba, M. Alagar, J. Colloid Interface Sci. 322 (2008) 537–544. [14] J. Lu, Carbon 45 (2007) 1599–1605. [15] Z. Sun, Z. Liu, B. Han, Y. Wang, J. Du, Z. Xie, G. Han, Adv. Mater. 17 (2005) 928–932. [16] R. Marcos Esteban, K. Schütte, D. Marquardt, J. Barthel, F. Beckert, R. Mülhaupt, C. Janiak, Nano-Struct. Nano-Objects 2 (2015) 28–34. [17] D. Marquardt, C. Vollmer, R. Thomann, P. Steurer, R. Mülhaupt, E. Redel, C. Janiak, Carbon 49 (2011) 1326–1332. [18] M. Jahjah, Y. Kihn, E. Teuma, M. Gómez, J. Mol. Catal. A Chem. 332 (2010) 106–112. [19] Y. Ma, Y. Huang, Y. Cheng, L. Wang, X. Li, Appl. Catal. A Gen. 484 (2014) 154–160. [20] Y. Huang, Y. Ma, Y. Cheng, L. Wang, X. Li, Catal. Commun. 72 (2015) 20–23. [21] M. Gopiraman, S. Ganesh Babu, Z. Khatri, W. Kai, Y.A. Kim, M. Endo, R. Karvembu, I.S. Kim, J. Phys. Chem. C 117 (2013) 23582–23596.
J.V. Rojas et al. / Materials Science and Engineering B 205 (2016) 28–35 [22] G. Li, H. Nagasawa, M. Kanezashi, T. Yoshioka, T. Tsuru, J. Mater. Chem. A 2 (2014) 9185. [23] E.A. Seddon, K.R., Seddon, Elsevier, Amsterdam; New York, 1984, pp. 1373. [24] H. Over, Appl. Phys. A Mater. Sci. Process. 75 (2002) 37–44. [25] X. Ni, B. Zhang, C. Li, M. Pang, D. Su, C.T. Williams, C. Liang, Catal. Commun. 24 (2012) 65–69. [26] P.P. Upare, M. Lee, S.-K. Lee, J.W. Yoon, J. Bae, D.W. Hwang, U.H. Lee, J.-S. Chang, Y.K. Hwang, Catal. Today (2015). [27] A.V. Krasheninnikov, K. Nordlund, Nucl. Instrum. Methods Phys. Res. Sect. B: Beam Interactions with Materials and Atoms 216 (2004) 355–366. [28] J. Belloni, M. Mostafavi, H. Remita, J.-L. Marignier, a.M.-O. Delcourt, New J. Chem. 22 (1998) 1239–1255. [29] J. Belloni, Catal. Today 113 (2006) 141–156. [30] J. Belloni, M. Mostafavi, Radiation chemistry of nanocolloids and clusters, in: C.D. Jonah, B.S.M. Rao (Eds.), Radiation Chemistry: Present Status and Future Trends, Elsevier, 2001, pp. 411–452. [31] J.V. Rojas, C.H. Castano, J. Nanopart. Res. 16 (2014). [32] J.V. Rojas, C.H. Castano, J. Radioanal. Nucl. Chem. 302 (2014) 555–561. [33] J.V. Rojas, C.H. Castano, Radiat. Phys. Chem. 81 (2012) 16–21. [34] Y. Zhao, L. Fan, J. Ren, B. Hong, Int. J. Hydrogen Energy 39 (2014) 4544–4557. [35] Y. Chen, Y. Liu, J. Mater. Chem. A 2 (2014) 9193. [36] S.H. Lee, N. Kakati, S.H. Jee, J. Maiti, Y.-S. Yoon, Mater. Lett. 65 (2011) 3281–3284. [37] J.V. Rojas, C.H. Castano, Production of metal nanoparticles on carbon nanotubes by gamma irradiation, in: B.I. Kharisov, O.V. Kharissova, U.O. Mendez (Eds.), Radiation Synthesis of Materials and Compounds, CRC Press, 2013, pp. 491–503.
35
[38] B.F. Machado, M. Oubenali, M. Rosa Axet, T. Trang Nguyen, M. Tunckol, M. Girleanu, O. Ersen, I.C. Gerber, P. Serp, J. Catal. 309 (2014) 185–198. [39] A.G. Hsieh, C. Punckt, S. Korkut, I.A. Aksay, J. Phys. Chem. B 117 (2013) 7950–7958. [40] M. Sun, G. Wang, X. Li, C. Li, J. Power Sources 245 (2014) 436–444. [41] K.-P. Lee, A.I. Gopalan, P. Santhosh, S.H. Lee, Y.C. Nho, Compos. Sci. Technol. 67 (2007) 811–816. [42] B.D. Cullity, S.R. Stock, Elements of X-Ray Diffraction, Prentice Hall, 2001. [43] T.I.T. Okpalugo, P. Papakonstantinou, H. Murphy, J. McLaughlin, N.M.D. Brown, Carbon 43 (2005) 153–161. [44] R.H. Bradley, K. Cassity, R. Andrews, M. Meier, S. Osbeck, A. Andreu, C. Johnston, A. Crossley, Appl. Surf. Sci. 258 (2012) 4835–4843. [45] B.C. Pirlot, I. Willems, A. Fonseca, J.B. Nagy, J. Delhalle, Adv. Eng. Mater. 4 (2002) 1–11. [46] Y. Lei, G. Gao, W. Liu, T. Liu, Y. Yin, Appl. Surf. Sci. 317 (2014) 49–55. [47] H. Gong, S.-T. Kim, J.D. Lee, S. Yim, Appl. Surf. Sci. 266 (2013) 219–224. [48] J.Y. Shen, A. Adnot, S. Kaliaguine, Appl. Surf. Sci. 51 (1991) 47–60. [49] C. Mun, J.J. Ehrhardt, J. Lambert, C. Madic, Appl. Surf. Sci. 253 (2007) 7613–7621. [50] K.C. Park, I.Y. Jang, W. Wongwiriyapan, S. Morimoto, Y.J. Kim, Y.C. Jung, T. Toya, M. Endo, J. Mater. Chem. 20 (2010) 5345. [51] D.R. Rolison, P.L. Hagans, K.E. Swider, J.W. Long, Langmuir 15 (1999) 774–779. [52] C. Roth, M. Goetz, H. Fuess, Appl. Electrochem. 31 (2001) 793–798. [53] F.J. Derbyshire, V.H.J.D. Beer, G.M.K. Abotsi, A.W. Scaroni, J.M. Solar, D.J. Skrovanek, Appl. Catal. 27 (1986) 117–131.
Contents lists available at ScienceDirect
Materials Science and Engineering B journal homepage: www.elsevier.com/locate/mseb
Facile radiolytic synthesis of ruthenium nanoparticles on graphene oxide and carbon nanotubes J.V. Rojas a,∗ , M. Toro-Gonzalez a , M.C. Molina-Higgins a , C.E. Castano b a
Mechanical and Nuclear Engineering Department, Virginia Commonwealth University, 401 West Main Street, Richmond, Virginia, 23284, USA Nanomaterials Core Characterization Facility, Chemical and Life Science Engineering Department, Virginia Commonwealth University, 601 West Main Street, Richmond, Virginia, 23284, USA b
a r t i c l e
i n f o
Article history: Received 27 September 2015 Received in revised form 4 December 2015 Accepted 15 December 2015 Available online 24 December 2015 Keywords: Carbon nanotubes Graphene oxide Nanoparticles Ruthenium Radiation chemistry Radiolysis Gamma irradiation
a b s t r a c t Ruthenium nanoparticles on pristine (MWCNT) and functionalized carbon nanotubes (f-MWCNT), and graphene oxide have been prepared through a facile, single step radiolytic method at room temperature, and ambient pressure. This synthesis process relies on the interaction of high energy gamma rays from a 60 Co source with the water in the aqueous solutions containing the Ru precursor, leading to the generation of highly reducing species that further reduce the Ru metal ions to zero valence state. Transmission electron microscopy and X-Ray diffraction revealed that the nanoparticles were homogeneously distributed on the surface of the supports with an average size of ∼2.5 nm. X-ray Photoelectron spectroscopy analysis showed that the interaction of the Ru nanoparticles with the supports occurred through oxygenated functionalities, creating metal-oxygen bonds. This method demonstrates to be a simple and clean approach to produce well dispersed nanoparticles on the aforementioned supports without the need of any hazardous chemical. © 2015 Elsevier B.V. All rights reserved.
1. Introduction The synthesis of supported metal nanoparticles is of significant importance for various applications in nanotechnology [1]. Interestingly enough, this area of research is approached from two complementary points of view. First, the need to synthesize nanoparticles with a homogeneous size and shape. Nanoparticles tend to aggregate due to their high surface area, causing deleterious effects in their physical and chemical properties, directly impacting their overall performance. Therefore, the use of organic and inorganic supports has been implemented during the nanoparticle synthesis to control the growth and hinder the formation of nanoparticle aggregates. The second point of view resides in the potential to enhance the properties of certain materials by adding metallic nanoparticles on their surface, that may result in a synergistic effect between the support and the metallic nanoparticles [2]. In this context, various types of supports such as polymers, carbon based materials, and oxides have been reported in the
∗ Corresponding author. Mechanical and Nuclear Engineering Department, Virginia Commonwealth University, 401 West Main Street, Richmond, Virginia 23284-3067301, Tel.: +1 804 8284267; fax: +1 804 8277030. E-mail addresses: [email protected] (J.V. Rojas), [email protected] (C.E. Castano). http://dx.doi.org/10.1016/j.mseb.2015.12.005 0921-5107/© 2015 Elsevier B.V. All rights reserved.
literature. The use of supports such as carbon nanotubes and, most recently, graphene oxide have stimulated their interest in catalysis, biosensors, and electronics, among others. These supports offer significant advantages due to their high chemical and thermal stability, high surface area, and outstanding mechanical, physical and chemical properties [2–4]. Furthermore, the surface of the carbon supports can be modified with different functionalities to trigger nanoparticle-support interactions [5,6]. The synthesis of transition metal nanoparticles on carbon nanotubes and graphene oxide has been reported with substantial interest in catalytic applications. Among those, ruthenium nanoparticles on carbon supports have been investigated for hydrogenation reactions [3,7–22]. Ruthenium is a transition metal occurring in the Group VIIIA of the periodic table with an electronic configuration of [Kr](4d)7(5s)1, displaying different oxidation states from −2 to +8 in its compounds. The most common precursor in the chemistry of Ru is the hydrated Ru(III) chloride, RuCl3 ·xH2 O, which is a compound that exhibits in solution a complex mixture of Ru(IV) and Ru(III) species rather than a pure source of Ru(III) [23]. Ruthenium is a very stable metal at low temperatures but usually oxidizes at high temperatures. However, in solid phase it only forms one stable oxide, RuO2 . Other oxide forms such as RuO, RuO3 and RuO4 are only stable at temperatures above 900 K, being volatile and toxic [24]. Various approaches have been reported in the literature with regards to the synthesis of Ru nanoparticles on carbon supports.
J.V. Rojas et al. / Materials Science and Engineering B 205 (2016) 28–35
Microwave synthesis of Ru on carbon nanotubes (CNTs) [8,25] and graphene [16,17] has been previously achieved at a controlled pressure, temperature, time, and atmosphere, leading to Ru nanoparticle with sizes of 2–4 nm. Another method for the production Ru nanoparticles on CNTs is through the reduction of ruthenium salt in polyol at temperatures above 100 ◦ C, this method resulted in ∼5 nm Ru nanoparticles on the CNTs [14]. Furthermore, the decoration of CNTs with Ru nanoparticles using supercritical water has been reported. In this approach, the particles are synthesized in an autoclave at ∼450 ◦ C for about 2 h giving 5 nm Ru nanoparticles on the CNTs [15]. Supercritical water has also been reported to produce Ru on graphene oxide. In this case, the reduction of Ru precursors takes place in a autoclave maintained at a temperature of ∼400 ◦ C and a pressure of 23.5 MPa for 2 h under nitrogen atmosphere, producing Ru nanoparticles of ∼3 nm [11]. Carbon nanotubes have also been decorated with Ru nanoparticles through the impregnation method. In this approach, the support is initially mixed with the metal precursor and the solvent and homogeneously dispersed through ultrasonication. Then, the samples are stirred and aged for a period of time, followed by drying and reduction with hydrogen flow [9,26]. Besides the various physical and chemical methods to synthesize Ru nanoparticles on carbon supports, alternative synthesis methods that minimize the use of additional reagents, as well as the need for high temperature and pressure, or controlled environment are always sought. This would allow for potential scalability and reproducibility of the nanocomposites. Radiation chemistry has recently made significant contribution to the field of synthesis of nanomaterials and the modification of their properties for specific applications [27]. The interaction of ionizing radiation, such as gamma rays, with aqueous solutions leads to the generation of highly reducing species that can bring metal ions present in the solution down to zero valence state. Furthermore, due to the mechanisms of interaction of gamma rays with matter, the reducing species are homogeneously distributed in the solution. Consequently, the metallic atoms will also be generated homogeneously throughout the solution followed by coalescence and growth [28]. This synthesis method eliminates the need for reducing agents that may sometimes poison the nanomaterial and affect their behavior. Moreover, the synthesis is carried out in aqueous environments which is desirable for certain applications. Nonetheless, it can also be implemented with other solvents such as ammonia or alcohols [29]. Radiation synthesis has previously been used for the synthesis of a wide variety of nanomaterials, such as metallic nanoparticles, core-shell nanostructures, metal oxides, and alloys [28,30–32]. In this work, ruthenium nanoparticles have been synthesized on pristine multi-walled carbon nanotubes, carboxylic acid functionalized carbon nanotubes, and reduced graphene oxide through a radiolytic method using 60 Co as the gamma source. The resulting Ru-C support nanostructures were characterized with XRD, TEM, and XPS, to study the morphology of the nanoparticles and their interaction with the support. The effect of the surface modification of the nanotubes on the yield of Ru nanoparticles is evidenced and a mechanism of synthesis and interaction of the particles with the support is presented here. This work evidences the potential use of gamma irradiation as a facile and novel method for the synthesis of well distributed Ru nanoparticles on carbon supports.
2. Materials and methods 2.1. Materials Multi-walled carbon nanotubes (MWCNTs), functionalized multi-walled carbon nanotubes (f-MWCNTs), graphene oxide (GO), ruthenium (III) chloride hydrate (RuCl3 ·xH2 O), and sodium
29
dodecyl sulfate (SDS; CH3 (CH2 )11 OSO3 Na) were obtained from Sigma–Aldrich and used without any further purification. Multiwalled Carbon nanotubes were > 98% carbon basis, length within 2.5 and 20 m and outside diameter within 6 and 13 nm. Functionalized multi-walled carbon nanotubes had an extent of labeling of carboxylic acid of > 8%, an average diameter of 9.5 nm, and average length of 1.5 m. Deionized water (18 M) obtained from an EMD Millipore Direct-QTM 3 UV water purification system was used to prepare the solutions along with high purity (99.5%) isopropyl alcohol (C3 H8 O) from ACROS Organic which was used as a radical scavenger. 2.2. Sample preparation Samples containing the carbon-based support (MWCNTs, fMWCNTs or GO), ruthenium precursor, and surfactant were prepared by modifying a methodology described in the literature [31,33]. As an initial step, a solution of water and isopropanol (2:1 v/v) was prepared, vortexed and stirred for ∼5 min. Subsequently, a solid mixture of RuCl3 ·xH2 O, SDS, and either MWCNTs, f-MWCNT or GO was prepared and stored in a glass vial. Next, 20 ml of waterisopropanol solution was added to the mixtures and sonicated for ∼30 min using a probe sonicator to break up the aggregates. This procedure lead to a homogeneous black suspension containing 5 mM RuCl3 (0.1 mmol), 5 mM SDS (0.1 mmol), and 1 mg of MWCNTs, f-MWCNTs, or GO per milliliter of solvent. 2.3. Irradiation procedure The aqueous solutions prepared by the procedure mentioned above were irradiated with a Cobalt-60 Gammacell 220 Excel (MDS, Nordion) that consists of a cylindrical stainless steel cage containing double sealed source pencils of Co-60. For irradiation purposes, the samples were placed into the chamber through a drawer mechanism and positioned in the outermost location in order to ensure a uniform absorbed dose distribution. The samples were irradiated at a dose rate of ∼7 kGy/h (0.7 Mrad/h) until an absorbed dose of 60 kGy was achieved. Then, the suspensions were cleaned to remove any excess of surfactant and/or unreacted species. For the cleaning process, the samples were initially centrifuged followed by removal of the supernate. Next, the precipitate was redispersed in deionized water and repeated several times. Finally, the samples were dispersed in acetone for further characterization. 2.4. Materials characterization The morphology, particle size, and particle size distribution of the resulting nanoparticles were studied by transmission electron microscopy (TEM) with a Zeiss Libra 120 Plus operating at 120 kV. For TEM imaging, a drop of the purified nanomaterial was diluted in acetone and sonicated for ∼10 min. The samples were prepared by immersing a formvar-carbon coated copper grid into the diluted suspension and allowed to dry at room temperature. Particle size analysis was carried out using the Image J 4.18 v software. The crystalline structure of the nanoparticles deposited on the carbon supports was investigated through X-ray diffraction (XRD) using a PANalytical X’Pert Pro MPD X-ray diffractometer with a ˚ The samples were prepared by copper anode (Cu K␣, = 1.5401 A). depositing drops of the sample dispersion on a silicon wafer and allowing it to dry. This process was repeated until a thick layer was obtained. X-ray photoelectron spectroscopy (XPS) was carried out using a ThermoFisher Escalab 250 X-ray photoelectron spectrometer with a monochromated aluminum X-ray source. For the analysis, few drops of the sample were deposited on a silicon wafer and allowed
30
J.V. Rojas et al. / Materials Science and Engineering B 205 (2016) 28–35
Fig. 1. TEM image and particle size distribution of MWCNT decorated with Ru nanoparticles at 60 kGy and 5 mM SDS.
Fig. 2. TEM image and particle size distribution of f-MWCNT decorated with Ru nanoparticles at 60 kGy and 5 mM SDS.
to dry. The software CASAXPS 2.3.16 v was used for profile fitting and analysis of the spectra. 3. Results and discussion 3.1. Synthesis of the nanoparticles on the supports Various methods to synthesize Ru nanoparticles on carbon supports have been previously reported in the literature. Among those, chemical reduction [13], electrochemical reduction [34], sol–gel [35], microwave assisted synthesis [8], and hydrothermal reduction [36] have been widely explored. In this work, a method to synthesize Ru nanoparticles on carbon supports such as carbon nanotubes and graphene oxide in aqueous solution using gamma irradiation was developed. The use of gamma rays to produce nanoparticles
has significant advantages that can lead to the generation of well dispersed nanoparticles without the need for aggressive reducing agents nor the use of high pressure or temperature. The synthesis of the Ru nanoparticles on the carbon supports using gamma irradiation relies on the radiolysis of water, process of high importance in radiation chemistry. When an aqueous solution is irradiated with high energy gamma rays, several species such as e− aq , OH• , H• , H+ , OH- , and H2 O2 are generated and the yield is determined by the absorbed dose in the aqueous solution. The species OH• and H2 O2 are highly oxidizing species whereas e− aq and H• are reducing agents [37]. Therefore, the environment can be triggered towards an oxidizing or reductive environment through the addition radical scavengers. The species e− aq and H• have sufficient redox potential to bring Ru (III) and Ru(IV) species down to a zero valence state. In order to prevent the oxidation of Ru(0)
Fig. 3. TEM image and particle size distribution of GO decorated with Ru nanoparticles at 60 kGy and 5 mM SDS.
J.V. Rojas et al. / Materials Science and Engineering B 205 (2016) 28–35
due to the presence OH• and H2 O2 , isopropanol was used as radical scavenger. This secondary alcohol reacts with the oxidizing species resulting in the generation of more reducing species. The generation of Ru(0) subsequently leads to the formation of Ru nuclei that continue to the growth process. Since the reduction of the Ru precursor was carried out in the presence of the carbon supports, the formation and growth of the nanoparticles takes place on their surface. Functional groups on the carbon nanotubes and the graphene oxide act as anchoring points for the nanoparticles [38]. Therefore, in this synthesis method, homogeneous dispersion of the support in order to expose a high number of functional groups before irradiation is imperative. The SDS was added to facilitate the dispersion of the carbon support in the aqueous solution since the surfactant promotes homogeneous distribution of the support in the solution through steric and electrostatic interactions [39].
31
Fig. 4. XRD pattern of MWCNTs, f-MWCNTs, and GO.
3.2. Size and distribution After gamma irradiation, the suspensions containing the nanomaterial were characterized using TEM, XRD, and XPS. Homogeneously distributed Ru nanoparticles on the various carbon supports were observed in the TEM images shown in Figs. 1–3. The nanoparticles are deposited and strongly bound to the supports since they stayed after the cleaning process of the samples which included several steps of sonication and centrifugation. Comparing the TEM images of the different supports, it can be observed that the nanoparticles are finely dispersed on the walls of functionalized carbon nanotubes and on the surface of the graphene oxide. However, fewer nanoparticles and relatively large aggregates on the walls of the non-functionalized nanotubes are observed with respect to the f-MWCNTs. The superior amount of well dispersed nanoparticles on the carboxylic acid functionalized carbon nanotubes relative to that observed on the non-functionalized may be explained by the larger number of nucleation sites homogeneously distributed on their surface. The average particle size of Ru obtained for the MWCNTs, f-MWCNTs, and GO were 2.5 ± 0.5 nm, 2.2 ± 0.6 nm, and 2.4 ± 0.7 nm, respectively. The size distribution histograms of the Ru nanoparticles on the different supports are presented in Figs. 1–3. The formation of the metallic nanoparticles with narrow particle size distribution is explained by the mechanism of nuclei formation due to gamma irradiation and the presence of the surfactant SDS. Gamma irradiation of water leads to the formation of homogeneously distributed reducing species such as e− aq and H• and the number of species can be controlled with the absorbed dose. Given the large absorbed dose used in this synthesis (60kGy), a large concentration of reducing species are formed in solution ([e− aq ] = 1.67 × 10−2 M, [H• ] = 3.79 × 10−3 M), consuming most of the Ru ions available in solution during the nucleation process. The presence of a solid support (CNT or GO) providing nucleation sites along with the surfactant generating electrostatic forces and steric hindrance limit the diffusion of Ru and control the nanoparticle growth. Therefore, small nanoparticles with a narrow size distribution are produced and deposited on the surface of the carbon support. 3.3. Structure The XRD patterns for the three carbon supports are shown in Fig. 4. The spectra for the non-functionalized and functionalized multi-walled carbon nanotubes have the characteristic peaks of carbon at 25.68◦ (0 0 2) and 42.21◦ (0 1 0) (PAN-ICSD Ref. code 98008-5678). On the other hand, the graphene oxide pattern has the most intense peak at 11.4◦ that corresponds to the plane (0 0 1) of GO [5,11,40]. The spectra for the different carbon supports decorated with ruthenium nanoparticles are presented in Fig. 5. The peaks at 2 of 38.3◦ ,42.0, 43.8◦ , 58.0◦ , and 69.0◦ correspond to
Fig. 5. XRD patterns of Ru nanoparticles supported on MWCNTs, f-MWCNTs, and GO.
the planes of hexagonal closed packed metallic ruthenium (010), (0 0 2), (0 1 1), (0 1 2), and (1 1 0), respectively (PAN-ICSD Ref. code 98-009-0997, with 2 of 38.45◦ ,42.27, 44.14◦ , 58.49◦ , and 69.64◦ ). Additionally, the peaks at 2 angles of 11.4◦ , 22.6◦ , and 25.9◦ correspond to graphene oxide (0 0 1), reduced graphene oxide (0 0 2), and graphite (0 0 2) plane reflections of CNTs, respectively [40,41]. Moreover, the broadening observed on the ruthenium and carbon peaks is characteristic of nanocrystalline materials. Therefore, the XRD patterns were used to calculate the average particle size by using the Scherrer Equation: L = K/(ˇ cos) Where K corresponds to a shape factor related with the shape of the crystallite (which is equivalent to 0.89, assuming a circular grain), is the X-ray wavelength (1.5406 A˚ for Cu),  is the broadening of diffraction line measured at half its maximum intensity and corresponds to the peak angle [42]. Using Scherrer equation on the ruthenium peak at 43.8◦ , the average particle size was 2.5 nm, 2.5 nm, and 2.7 nm for MWCNTs, f-MWCNTs, and GO, respectively. These results are in agreement with the results obtained from the TEM analysis. The slight difference between the results obtained from the two techniques resides on the fact that TEM analysis leads to a local measurement of the average NP diameter, whereas XRD provides a more statistical representation of the particle size due to the larger amount of sample that contribute to the diffraction pattern. 3.4. Surface composition and chemical states XPS analysis was performed to study the surface composition and chemical state of the different carbon supports before and after decoration with ruthenium nanoparticles. Fig. 6 shows the high resolution (HR) C1s and O1s XPS spectra for the different carbon supports. In the case of MWCNTs, Fig. 6 a), C1s was deconvoluted into five peaks and O1s was deconvoluted into three peaks.
32
J.V. Rojas et al. / Materials Science and Engineering B 205 (2016) 28–35
Fig. 6. High resolution XPS spectra of C1s and O1s for (a) MWCNTs, (b) f-MWCNTs, and (c) GO.
Regarding the C1s, the peaks located at 284.5 eV and 284.8 eV corresponds to C C bonds with sp2 hybridization and C C bonds with sp3 hybridization. The peak observed at 285.7 eV associates to the contribution of both C O and C OH functionalities. The following peak at 289.4 eV represents carboxyl groups [43]. Also, a peak at 291.1 eV is observed in the pattern, this peak is associated with a shake-up feature typical of aromatic structures or –* shake-up. This feature is related to the interaction of emitted photoelectrons with electrons [44,45]. For the O1s, the peak at 532.3 reveals the presence of C O/C O bonds. It is worth to notice that the position of this peak above 532 eV may indicate a larger contribution from C O than from C O bonds. The next peak at 533.3 eV corresponds to C-OH bonds, while the peak at 534.7 eV is associated to oxygen in water molecules [46]. The HR XPS spectra of the C1s and O1s in the f-MWCNTs shown in Fig. 6 b) present similar features to the ones observed in MWCNTs. However, a broader C1s peak as well as more
pronounced peaks within 285 and 290 eV are remarkable in the fMWCNTs. Furthermore, the O1s peak is significantly broader in the f-MWCNTs. These characteristics can be associated to carbonoxygen interactions such as hydroxyl, carbonyl, and carboxyl groups. Also, the peak below 532 eV indicates a significant presence of carboxyl functionalities obtained from the functionalization through carboxylic acid [47]. Fig. 6 c) shows the XPS analysis of C1s and O1s peaks for the graphene oxide. The double peak observed in the C1s is characteristic of the graphene oxide and it was deconvoluted into six peaks. The first three components at BE of 284.6, 285.7, and 286.7 eV correspond to C sp2 hybridization, hydroxyl and epoxide groups, being the first and third component prominent. The last three are related to carbonyl, carboxyl and –* shake-up, respectively [5]. In addition, the two peaks at 531.1, 532.3, and 533.1 eV in the O1s peak are associated with single and double bonds between oxygen and carbon.
J.V. Rojas et al. / Materials Science and Engineering B 205 (2016) 28–35
33
Fig. 7. High resolution XPS spectra of C1s, O1s, and Ru3p for MWCNTs/Ru.
The composites formed by the carbon supports and ruthenium nanoparticles were also analyzed through XPS. The peaks corresponding to oxygen, carbon, and ruthenium can be distinctly detected in the survey scans of the different composites. Furthermore, the composites and nanoparticles are remarkably clean of byproducts according to the survey spectra, where the only element detected other than O, C, and Ru, came from the Si substrate used to prepare the sample. The Ru3p region was chosen instead of the Ru3d due to the strong overlapping of these peaks with the C1s, which increases the complexity of defining the electronic states of Ru. The HR XPS spectra of C1s, O1s, and Ru3p doublet for MWCNTs decorated with Ru NPs are shown in Fig. 7. To start with, the HR spectra for the C1s shows similar features to the C1s in the support alone. The peak observed at 284.6 eV corresponds to sp2 hybridized graphite-like carbon atoms. The second peak located at 284.8 eV corresponds sp3 hybridization, and the third and fourth peaks located at 285.8 and 290.1 eV correspond to hydroxyl and carboxyl groups, respectively. The HR O1s was deconvoluted into 3 peaks as shown in Fig. 7. The first one at 531.48 eV originates from the double bond between carbon and oxygen whereas the second one at 532.13 eV reveals C O bonds and it is also related with weak interactions between oxygen and ruthenium [48]. Furthermore, the third peak located around 529 eV reveals Ru O bonds [49]. Lastly, the Ru 3p core-level spectrum was analyzed and the doublet, with 3p3/2 and 3p1/2 components, was identified. The XPS fitting reveals the presence of 4 peaks, the two located at 462.2 and 484.4 eV correspond to Ru in the zero valence state whereas the other two at 466.1 eV and 488.6 eV evidence Ru at a higher valence state, Ru(IV). The Ru (IV) peaks in both states are shifted to higher binding energies respect to the anhydrous oxides species with the same oxidation state. As a result. these Ru (IV) peaks for the 3p3/2 and 3p1/2 states may correspond to some hydrated form Ru (IV) such as RuO2 ·xH2 O or RuOx Hy [50–52]. The HR spectra for C1s, O1s and Ru3p for the functionalized carbon nanotubes decorated with ruthenium nanoparticles are shown in Fig. 8. The C1s peak was deconvoluted into 4 bands at 284.5, 284.8, 286.1, and 288.7 eV. Similar to the features previously observed in the f-MWCNTs substrate, the peaks correspond to the sp2 and sp3 hybridization as well as carbon-oxygen interactions. Likewise, the fitting in the O1s peak also confirmed the presence of C-O interactions at BE of 531.8, and 532.7 eV. In addition, the peak at BE of 529.9 reveals the formation of Ru-O bonds [48] such as that of ruthenium (IV) dioxide. Finally, the Ru3p was fitted taking into account the 3p3/2 and 3p1/2 components. Similar to the results observed in the Ru-MWCNTs, the main contribution to the spectrum is given by the Ru (0), whose peaks are observed at BE of 462.4 and 484.6 eV. Furthermore, the contribution, identified as that of Ru (IV) at 466.3 and 488.2 eV, is also observed [51]. Fig. 9 shows the HR XPS spectra of C1s, O1s, and Ru3p for the ruthenium nanoparticles supported on graphene oxide. The C1s
does show a significant difference compared to the C1s observed for the graphene oxide in Fig. 6c. There is a substantial reduction in the contribution of the C O peak after the irradiation process which indicated the transformation of GO into reduced graphene oxide, rGO. Nonetheless, the peaks that represent sp2 and sp3 hybridizations as well as carbon-oxygen interaction are present within the BE interval from 286 to 290 eV. The peaks at 531.6 and 532.5 eV in the O1s, once again, are linked to carbon-oxygen bonds, and the presence of Ru-O interaction is also revealed in this case at 529.9 eV. The Ru3p spectrum was fitted for the 3p3/2 and 3p1/2 where the Ru (0) and ruthenium (IV) are observed, being Ru(0) the major contribution to the pattern. It has been demonstrated that the more carboxylic groups in the surface of the carbon supports the better the metal dispersion due to the promotion of anchoring sites through metal cation exchange. Since the samples were prepared by a ruthenium precursor that contains Ru(III) and Ru(IV) species, it is expected that both species are adsorbed at the acidic functional groups. Experimental and theoretical results have shown that Ru species are extremely stable when two of the Ru bonds are with the oxygenated functional groups of carbon [38,53]. Therefore, since Ru(IV) is more prone to form solid stable oxides, it may explain the appearance of this species in the results obtained by XPS. Under gamma irradiation, the generation of reduced Ru(0) from the Ru(III) and Ru(IV) species leads to the formation of the Ru nanoparticles stabilized in the surface of the carbon supports anchored by Ru(IV)-OOC bonds. Although, adsorbed species Ru(II) and Ru(III) on the carbon supports with partial charge transfer could be proposed, the presented deconvolution of the Ru 3p HRXPS peak only matched with Ru(0) and Ru(IV) species. Ru(IV) species may have also been formed on the Ru nanoparticles due to the presence of O2 molecules in aqueous solution and/or when exposure to air during XPS sample preparation [38]. This single step production method of supported Ru nanoparticles on carbon nanotubes and reduced graphene oxide represents a novel approach that lead to the formation of well dispersed nanoparticles. The amount of nanoparticles deposited on the carbon nanotubes, as it has been reported in the literature, is clearly affected by the concentration of nucleation sites available on the nanotubes surface. In this case, both TEM and XPS analysis evidenced this fact. In other words, the large concentration of carbon-oxygen functionalities on the carboxylic acid functionalized carbon nanotubes demonstrated by XPS led to a higher yield of Ru nanoparticles deposited on their surface. Furthermore, the larger concentration of functional groups behaving as nucleation sites causes a more homogeneous particle size distribution and avoids the formation of aggregates. Similarly, the large concentration of functionalities on the GO resulted in < 3 nm Ru nanoparticles well distributed on their surface. Gamma irradiation represents a potential method for large scale production of
34
J.V. Rojas et al. / Materials Science and Engineering B 205 (2016) 28–35
Fig. 8. High resolution XPS spectra of C1s, O1s, and Ru3p for f-MWCNTs/Ru.
Fig. 9. High resolution XPS spectra of C1s, O1s, and Ru3p for GO/Ru.
this type of nanocomposites in a clean environment at room temperature. Gamma rays produce a uniform distribution of reducing species that results in a homogeneous nucleation and growth of the nanoparticles. 4. Conclusions In this work, the feasibility to produce ruthenium nanoparticles on carbon nanotubes, carboxylic acid functionalized carbon nanotubes and graphene oxide supports through a facile aqueous route using a gamma irradiation was demonstrated. This method is based on the hydrolysis of water that result in the generation of highly reducing and oxidizing species upon the interaction with the gamma rays. The use of isopropanol in this work allowed to scavenge the oxidizing species and created a highly reducing environment that brought Ru precursor species to a zero valence state. This synthesis approach lead to the production of nanoparticles homogeneously distributed on the surface of the supports with an average size of ∼2.5 nm. As expected, more nanoparticles and negligible aggregates were observed on the carboxylic acid functionalized MWCNTs compared to the pristine nanotubes. In the GO support, the large concentration of oxygenated sites led to the formation of homogeneous distributed nanoparticles. Furthermore, XPS revealed that the interaction of the Ru nanoparticles with the support was through oxygenated functionalities. This work evidences the potential use of gamma irradiation as a simple approach to make well dispersed nanoparticles on the aforementioned supports. Moreover, the fact that no reducing agents are required makes of this a clean approach to produce these materials. Acknowledgements This work was funded by the Virginia Commonwealth University with the support of the Mechanical and Nuclear Engineering department and the NRC-HQ-84-14-FOA-002, Faculty
Development Program in Radiation Detection and Health Physics at VCU. The authors would like to thank the Nanomaterials Core Characterization Facility (NCC) from the School of Engineering. Finally, we also would like to acknowledge the assistance received from Dr. Stoyan Toshkov at the Nuclear Radiation Lab, University of Illinois Urbana Champaign during the sample irradiation.
References [1] B.R. Cuenya, Thin Solid Films 518 (2010) 3127–3150. [2] J.M. Campelo, D. Luna, R. Luque, J.M. Marinas, A.A. Romero, ChemSusChem 2 (2009) 18–45. [3] C. Du, Q. Ao, N. Cao, L. Yang, W. Luo, G. Cheng, Int. J. Hydrogen Energy 40 (2015) 6180–6187. [4] G.G. Wildgoose, C.E. Banks, R.G. Compton, Small 2 (2006) 182–193. [5] T. Kuila, S. Bose, A.K. Mishra, P. Khanra, N.H. Kim, J.H. Lee, Prog. Mater. Sci. 57 (2012) 1061–1105. [6] V. Datsyuk, M. Kalyva, K. Papagelis, J. Parthenios, D. Tasis, A. Siokou, I. Kallitsis, C. Galiotis, Carbon 46 (2008) 833–840. [7] D.K. Mishra, A.A. Dabbawala, J.-S. Hwang, J. Mol. Catal. A Chem. 376 (2013) 63–70. [8] C. Antonetti, M. Oubenali, A.M. Raspolli Galletti, P. Serp, G. Vannucci, Appl. Catal. A Gen. 421–422 (2012) 99–107. [9] Y. Wang, Z. Rong, Y. Wang, P. Zhang, Y. Wang, J. Qu, J. Catal. 329 (2015) 95–106. [10] D.V. Jawale, E. Gravel, C. Boudet, N. Shah, V. Geertsen, H. Li, I.N.N. Namboothiri, E. Doris, Chem. Commun. 51 (2015) 1739–1742. [11] J. Zhao, W. Hu, H. Li, M. Ji, C. Zhao, Z. Wang, H. Hu, RSC Adv. 5 (2015) 7679–7686. [12] X. Zhang, Y. Ye, H. Wang, S. Yao, Radiat. Phys. Chem. 79 (2010) 1058–1062. [13] V. Selvaraj, M. Vinoba, M. Alagar, J. Colloid Interface Sci. 322 (2008) 537–544. [14] J. Lu, Carbon 45 (2007) 1599–1605. [15] Z. Sun, Z. Liu, B. Han, Y. Wang, J. Du, Z. Xie, G. Han, Adv. Mater. 17 (2005) 928–932. [16] R. Marcos Esteban, K. Schütte, D. Marquardt, J. Barthel, F. Beckert, R. Mülhaupt, C. Janiak, Nano-Struct. Nano-Objects 2 (2015) 28–34. [17] D. Marquardt, C. Vollmer, R. Thomann, P. Steurer, R. Mülhaupt, E. Redel, C. Janiak, Carbon 49 (2011) 1326–1332. [18] M. Jahjah, Y. Kihn, E. Teuma, M. Gómez, J. Mol. Catal. A Chem. 332 (2010) 106–112. [19] Y. Ma, Y. Huang, Y. Cheng, L. Wang, X. Li, Appl. Catal. A Gen. 484 (2014) 154–160. [20] Y. Huang, Y. Ma, Y. Cheng, L. Wang, X. Li, Catal. Commun. 72 (2015) 20–23. [21] M. Gopiraman, S. Ganesh Babu, Z. Khatri, W. Kai, Y.A. Kim, M. Endo, R. Karvembu, I.S. Kim, J. Phys. Chem. C 117 (2013) 23582–23596.
J.V. Rojas et al. / Materials Science and Engineering B 205 (2016) 28–35 [22] G. Li, H. Nagasawa, M. Kanezashi, T. Yoshioka, T. Tsuru, J. Mater. Chem. A 2 (2014) 9185. [23] E.A. Seddon, K.R., Seddon, Elsevier, Amsterdam; New York, 1984, pp. 1373. [24] H. Over, Appl. Phys. A Mater. Sci. Process. 75 (2002) 37–44. [25] X. Ni, B. Zhang, C. Li, M. Pang, D. Su, C.T. Williams, C. Liang, Catal. Commun. 24 (2012) 65–69. [26] P.P. Upare, M. Lee, S.-K. Lee, J.W. Yoon, J. Bae, D.W. Hwang, U.H. Lee, J.-S. Chang, Y.K. Hwang, Catal. Today (2015). [27] A.V. Krasheninnikov, K. Nordlund, Nucl. Instrum. Methods Phys. Res. Sect. B: Beam Interactions with Materials and Atoms 216 (2004) 355–366. [28] J. Belloni, M. Mostafavi, H. Remita, J.-L. Marignier, a.M.-O. Delcourt, New J. Chem. 22 (1998) 1239–1255. [29] J. Belloni, Catal. Today 113 (2006) 141–156. [30] J. Belloni, M. Mostafavi, Radiation chemistry of nanocolloids and clusters, in: C.D. Jonah, B.S.M. Rao (Eds.), Radiation Chemistry: Present Status and Future Trends, Elsevier, 2001, pp. 411–452. [31] J.V. Rojas, C.H. Castano, J. Nanopart. Res. 16 (2014). [32] J.V. Rojas, C.H. Castano, J. Radioanal. Nucl. Chem. 302 (2014) 555–561. [33] J.V. Rojas, C.H. Castano, Radiat. Phys. Chem. 81 (2012) 16–21. [34] Y. Zhao, L. Fan, J. Ren, B. Hong, Int. J. Hydrogen Energy 39 (2014) 4544–4557. [35] Y. Chen, Y. Liu, J. Mater. Chem. A 2 (2014) 9193. [36] S.H. Lee, N. Kakati, S.H. Jee, J. Maiti, Y.-S. Yoon, Mater. Lett. 65 (2011) 3281–3284. [37] J.V. Rojas, C.H. Castano, Production of metal nanoparticles on carbon nanotubes by gamma irradiation, in: B.I. Kharisov, O.V. Kharissova, U.O. Mendez (Eds.), Radiation Synthesis of Materials and Compounds, CRC Press, 2013, pp. 491–503.
35
[38] B.F. Machado, M. Oubenali, M. Rosa Axet, T. Trang Nguyen, M. Tunckol, M. Girleanu, O. Ersen, I.C. Gerber, P. Serp, J. Catal. 309 (2014) 185–198. [39] A.G. Hsieh, C. Punckt, S. Korkut, I.A. Aksay, J. Phys. Chem. B 117 (2013) 7950–7958. [40] M. Sun, G. Wang, X. Li, C. Li, J. Power Sources 245 (2014) 436–444. [41] K.-P. Lee, A.I. Gopalan, P. Santhosh, S.H. Lee, Y.C. Nho, Compos. Sci. Technol. 67 (2007) 811–816. [42] B.D. Cullity, S.R. Stock, Elements of X-Ray Diffraction, Prentice Hall, 2001. [43] T.I.T. Okpalugo, P. Papakonstantinou, H. Murphy, J. McLaughlin, N.M.D. Brown, Carbon 43 (2005) 153–161. [44] R.H. Bradley, K. Cassity, R. Andrews, M. Meier, S. Osbeck, A. Andreu, C. Johnston, A. Crossley, Appl. Surf. Sci. 258 (2012) 4835–4843. [45] B.C. Pirlot, I. Willems, A. Fonseca, J.B. Nagy, J. Delhalle, Adv. Eng. Mater. 4 (2002) 1–11. [46] Y. Lei, G. Gao, W. Liu, T. Liu, Y. Yin, Appl. Surf. Sci. 317 (2014) 49–55. [47] H. Gong, S.-T. Kim, J.D. Lee, S. Yim, Appl. Surf. Sci. 266 (2013) 219–224. [48] J.Y. Shen, A. Adnot, S. Kaliaguine, Appl. Surf. Sci. 51 (1991) 47–60. [49] C. Mun, J.J. Ehrhardt, J. Lambert, C. Madic, Appl. Surf. Sci. 253 (2007) 7613–7621. [50] K.C. Park, I.Y. Jang, W. Wongwiriyapan, S. Morimoto, Y.J. Kim, Y.C. Jung, T. Toya, M. Endo, J. Mater. Chem. 20 (2010) 5345. [51] D.R. Rolison, P.L. Hagans, K.E. Swider, J.W. Long, Langmuir 15 (1999) 774–779. [52] C. Roth, M. Goetz, H. Fuess, Appl. Electrochem. 31 (2001) 793–798. [53] F.J. Derbyshire, V.H.J.D. Beer, G.M.K. Abotsi, A.W. Scaroni, J.M. Solar, D.J. Skrovanek, Appl. Catal. 27 (1986) 117–131.