Accepted Manuscript Title: Synthesis and structural features of resorcinol–formaldehyde resin chars containing nickel nanoparticles Author: M.V. Galaburda V.M. Bogatyrov J. Skubiszewska-Zi˛eba O.I. Oranska D. Sternik V.M. Gun’ko PII: DOI: Reference:
S0169-4332(15)02748-8 http://dx.doi.org/doi:10.1016/j.apsusc.2015.11.053 APSUSC 31770
To appear in:
APSUSC
Received date: Revised date: Accepted date:
11-6-2015 2-10-2015 6-11-2015
Please cite this article as: M.V. Galaburda, V.M. Bogatyrov, J. SkubiszewskaZi˛eba, O.I. Oranska, D. Sternik, V.M. Gun’ko, Synthesis and structural features of resorcinolndashformaldehyde resin chars containing nickel nanoparticles, Applied Surface Science (2015), http://dx.doi.org/10.1016/j.apsusc.2015.11.053 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Synthesis and structural features of resorcinol–formaldehyde resin chars containing nickel nanoparticles M.V. Galaburdaa, V.M. Bogatyrova, J. Skubiszewska-Ziębab, O.I. Oranskaa, D. Sternikb, V.M. Gun’ko a a
Chuiko Institute of Surface Chemistry, 17 General Naumov Str., Kyiv 03164, Ukraine ; Faculty of Chemistry, Maria Curie-Skłodowska University, Maria Curie-Sklodowska Sq.3, 20-031 Lublin, Poland;
ip t
b
M
an
us
cr
Abstract A series of meso- and microporous carbons containing magnetic Ni nanoparticles (Ni/C) with a variety of Ni loadings were synthesized by a simple one-pot procedure through carbonization of resorcinol-formaldehyde polymers containing various amounts of nickel(II) acetate. Such composite materials were characterized by N2 sorption, Raman spectroscopy, X-ray diffraction (XRD), scanning electron microscopy (SEM) and Transmission electron microscope (TEM). The XRD patterns reveal peaks corresponding to face centered cubic nickel with the average size of crystallites of 17-18 nm. SEM and TEM results reveal that the formation of the nanoparticles took place mainly in the carbon spheres (1-2 µm in size) and on the outer surface as well. The as-prepared composites are characterized by a core-shell structure with well-crystallized graphitic shells about 8-15 nm in thickness. The Raman spectra show that Ni content influences the structure of the carbon. It was also shown that the morphology (particle shape and sizes) and porosity (pore volume and pore size distribution) of the chars are strongly dependent on water and nickel contents in the blends. One of the applications of Ni/C was demonstrated as a magnetically separable adsorbent.
Ac ce pt e
d
Introduction Porous carbon materials are of great interest due to their well-controlled pore architecture, high surface area, electrical conductivity, thermal stability, chemical inertness, biocompatibility, and specific surface properties. These features contribute to their high performance in various applications such as separation, catalyst supports, adsorption, energy storage/conversion, and biomedical engineering [1-3]. However, carbon powders dispersed in water can cause secondary pollution. The conventional separation approach normally involves a filtration or centrifugation procedure, which is rather complex and expensive. Because of easy separation and controlled placement of magnetosensitive nanoparticles by means of an external magnetic field, the magnetic nanoparticles have been widely investigated [4, 5]. In addition, the mesoporous carbon materials with incorporated magnetic nanoparticles can also act as selective catalysts and adsorbents with further modification of the carbon surface. Porous high-quality magnetic carbon nanocomposites have so far been prepared using several popular methods including template-based synthesis [6], chemical vapor deposition [7, 8], filling process [9], sol–gel process [10], hydrothermal/solvothermal method [11], pyrolysis procedure [12], detonation induced reaction [13] etc. Recently most frequently encountered methods are CVD and templatebased synthesis because they allow obtaining high-quality products. However, the quality is much dependent on the structure of the template used, which is difficult to obtain according to the desired structures and properties. Besides, the removal of the template, without destruction of the product, is still one of the main challenges for the synthesis. The main merit of the CVD method is that it allows a precise control on the carbon nanostructure, but the main disadvantages of this method are the high cost, complex equipment and energy consummation. The pyrolysis procedure is based on the heat treatment of a mixture containing the soluble metal salt (magnetic metal source) and a type of organic compounds enriched with carbon. Pyrolysis provides an available way to fabricate nanostructured materials, due to its high efficiency and low cost. 1 Page 1 of 18
Ac ce pt e
d
M
an
us
cr
ip t
Carbons have been widely used as good adsorbents for various applications and, therefore, they must exhibit a high volume of pores of a variety of pore size to be able to adsorb a wide range of molecules, coupled with an adequate proportion of meso- and macropores to facilitate the access to micropores. As applications become more specific, a better control of the pore size distribution is required [14, 15]. For instance, the recovery of gasoline vapor requires carbon with a homogeneous microporosity allowing the adsorption and desorption of light hydrocarbons found in the gasoline vapor, or separation of gas mixtures where the micropore width is able to differentiate the kinetics of adsorption of two molecules with similar dimensions, and the storage of methane where a well define micropore size is required. Thus, simple techniques to control pore size, structure, morphology and configuration (films and powder) are required for the practical applications. Consequently, a fast and simple synthesis of magnetic meso/microporous carbons with a high surface area and large pore volume is quite required for the practical applications. The porosity of chars is affected by several factors such as the structure of polymeric precursors, carbonization temperature, heating rate and residence time in the reaction zone, particle size, etc. Resorcinol–formaldehyde resins (RFR) are of interest as porous polymers or precursors of chars, which can be used to prepare highly porous carbon materials [16, 17]. Moreover, the formation of the resorcinol-formaldehyde resin took place due to the condensation of resorcinol and formaldehyde in the aqueous solution where water can play a role of a solvent and a porogen. Thus, the structure and properties of the RFR depend strongly on the conditions, at which the process occurs, and concentrations of the reagents. These factors influence the morphological and textural characteristics of the chars [18, 19]. Therefore, the aim of the work is the synthesis of porous carbon nanostructured materials by pyrolysis of resorcinol-formaldehyde polymers filled with nickel(II) acetate, and to study the influence of the amount of the latter, as well as water content and temperature treatment on the morphological, structural, textural features and adsorption properties of chars. Acetate was chosen as the filler because it can be dissolved easily and kept dissolved during the gel synthesis, as well as its degradation during pyrolysis treatment leaves no ashes inside the carbon material and mainly produces water and carbon oxides. 2. Experimental 2.1. Materials Nickel(II) acetate ((CH3COO)2Ni·4H2O, Ni(ac)2, Reachim Company, Russia), resorcinol (99.9%, Chimlaborreativ, Ukraine) and 37% aqueous solution of formaldehyde (Chimlaborreativ, Ukraine) were used in the synthesis of Ni/resorcinol–formaldehyde resin compounds. In a typical preparation of the doped polymers, 10 g of resorcinol, 15 g of formaldehyde, 50 g of distilled water and 8 g (RF50-Ni1), 14 g (RF50-Ni2) or 20 g (RF50-Ni3) of Ni(ac)2 were added under stirring at room temperature. All the systems were hermetically closed and placed into the thermostatic oven at 85 °C during 24 h. The formation of the Ni/resorcinol-formaldehyde polymer gels took place with creation of chelate complexes (Scheme I). During this process, the creation of coordinate bonds between nickel and oxygen from the OH groups of the same aromatic ring (Scheme I,(1)) or from the different aromatic rings in the same or different polymer chains (Scheme I,(2)) may occur. Then this composite was dried at room temperature, ground and dried at 90°C for 10 h. Carbonization of the samples was carried out in a tubular furnace under argon atmosphere upon heating from room temperature to 700 °C (first series) and 800 °C (second series) at a heating rate of 5 °C/min and kept at the maximum temperature for 2 h. The materials of the first series were labeled as RFC50-Ni1-700, RFC50-Ni2700, RFC50-Ni3-700 and the second one: RFC50-Ni1-800, RFC50-Ni2-800 and RFC50-Ni3-800. The samples with different water content were synthesized in the same way using 35 g and 20 g of distilled water and 8 g of Ni(ac)2 with subsequent pyrolysis under argon atmosphere at 800 °C for 2 hours (third series). These composites were labeled as RFC35-Ni1-800 and RFC20-Ni1-800. The main characteristics of the samples are given in Table 1. 2 Page 2 of 18
All MB (C16H18ClN3S·3H2O, molecular weight 373.9 g/mol, Sigma-Aldrich) solutions used in this study were prepared by weighing and dissolving the required amounts of MB in distilled water. Other chemicals were used without further purification.
Ac ce pt e
d
M
an
us
cr
ip t
2.2. Characterization methods X-ray diffraction (XRD) patterns were recorded at 2θ=10-80° (in 2θ range 10-80° or from 10 to 80° on 2θ) using a DRON UM1 diffractometer (Burevestnik, St.-Petersburg, Russia) with Co Kα radiation and a graphitic monochromator in geometry of Bragg-Brentano. The average sizes of crystallites were estimated using the Sherrer equation [20]. Chars were characterized using a JSM-6700F (GEOL) field emission scanning electron microscope (FESEM) and Transmission electron microscope (TEM) Titan G2 60-300, FEI. Thermal Analysis, including Thermogravimetry and Differential Thermal Analysis (TG-DTG-DTA) (Derivatograph C, MOM, Hungary) was used to obtain data on decomposition processes, thermal stability and temperature of phase transformations of the prepared materials via heating the samples (20-22 mg) in a static air atmosphere from 20 to 1200°C at the heating rate of 10 °C/min. To analyze the structural characteristics of carbon/Ni composites, low-temperature (77.4 K) nitrogen adsorption–desorption isotherms were recorded using a Micromeritics ASAP 2405N adsorption analyzer. The specific surface area (SBET) was calculated according to the standard BET method [21]. The total pore volume Vp was evaluated from the nitrogen adsorption at p/p0 = 0.98–0.99 (p and p0 denote the equilibrium and saturation pressure of nitrogen at 77.4 K, respectively). The nitrogen desorption data were used to compute the pore size distributions (PSDs, differential fV(R)~dVp/dR and fS(R)~dS/dR) using a self-consistent regularization procedure under nonnegativity condition (fV(R)≥0 at any pore radius R) at a fixed regularization parameter α=0.01 with a complex pore model with slit-shaped pores [22]. The differential PSDs SCV/SCR with respect to pore volume fV(R)~dV/dR, ∫fV(R)dR~Vp were recalculated as incremental PSD (IPSD, ∑ΦV,i(R)=Vp). The fV(R) and fS(R) functions were also used to calculate contributions of micropores (Vmicro and Smicro at R<1 nm), mesopores (Vmeso and Smeso at 125 nm) to the total pore volume and the specific surface area. The Raman spectra were recorded using an inVia Reflex Microscope DMLM Leica Research Grade, Reflex, Renishaw, UK (excitation at 514 nm). The adsorption isotherms were measured from aqueous solutions of methylene blue (MB). For determination of the adsorption capacity, 50 mL of aqueous solution of methylene blue at concentration of 68.58 mg/l was added into flasks containing 0.2 g of Ni/C composites that were previously heated at 120°C for 2 hours and kept in desiccator. The flasks were placed in a thermostatic shaker at 25 °C for 12 h in order to establish the adsorption equilibrium. At the end of the equilibrium contact times, the adsorbent and the solution were separated by a magnet and the fixed amount from the supernatant of solution was determined at maximum working wavelengths (663 nm). Dilution of dye solutions was required if there absorbances were higher than 0.8. The solutions were analyzed for residual MB concentration using a Lambda 35 (Perkin-Elmer) UV-visible spectrophotometer (1 cm quartz cell). Each experiment was duplicated under identical conditions. The concentration of MB was determined using a calibration curve, which was linear over the concentration range used in this study. The amount of methylene blue adsorbed on chars per gram of carbon, q (mg/g) was determined from equation:
(
)
(1) q = V Co − Ceq / m where Co and Ceq (mg/l) are the initial and equilibrium concentrations of MB, respectively, V is the volume of solution (l), and m is the amount of carbon (g).
3 Page 3 of 18
Ac ce pt e
d
M
an
us
cr
ip t
3. Results and discussion 3.1. Characteristics of the composites The XRD patterns (Fig 1) show that Ni(II) was completely reduced to metallic state after pyrolysis at 700 and 800 °C. This assumption is based on the presence of three peaks observed at 2θ = 52°, 61° and 92° which can be assigned to the 111, 200, and 220 reflections of metallic nickel with the facecentered cubic structure according to the JCPDS Card No. 04-0850. The average size of Ni crystallites of 17-18 nm, estimated using the Scherrer equation from the broadening of (111) XRD diffraction line, does not change with increasing nickel content. In addition, an intensive diffraction peak at 2θ =30.55° with interlayer spacing of 0.33-0.34 nm observed in the resulting composites can be attributed to graphite plane (002) [23]. Thus, it can be assumed that the materials are well graphitized during carbonization due to the effect of nickel nanoparticles. It is well known that the nanosized transition metals such as Fe, Co, and Ni can accelerate the development of graphitic structure (called catalytic graphitization) of carbon when they are heated together with carbon materials in an inert atmosphere [24, 25]. To further test the structure and graphitic character of the Ni/C nanocomposites, they were characterized using Raman spectroscopy. The spectra were collected in the range of 100–3200 cm-1. As it can be seen from the XRD data all chars consist of small crystallites and amorphous fragments; however, according to the Raman spectra (Fig. 1c, Table 2), their crystallinity differs. Deconvolution of the G and D bands (Fig. 1b) gives the main G1 and D1 bands and additional lines G2 (at 1513 cm-1 corresponds to the graphite-like structures [26]) and D2 (at 1195 cm-1 related to the defect structures with sp3 C atoms or to mixed sp2–sp3 bonding). The intensity of the G1 band (Fig. 1c), which corresponds to the stretching vibrations of the basal graphene layers in all carbon materials [27], decreases with increasing Ni content in comparison with D band that is associated with disorder structures and decreased crystallinity, for example in the presence of defects and finite size effects [28]. The ideal single-crystal graphite or diamond shows only a peak at 1580 or 1350 cm-1, respectively. Apart from the D and G peaks, a so-called 2D peak appears at about 2700 cm-1 as an overtone of the D band [24]. This peak is thought to originate from finite-size disordered structures of graphite in the surface layers of the nanocomposites [29]. The values of AD/AG and the peak full-width at half maximum of the G1 band (FWHMG) were calculated by deconvolution of the spectra using Lorentzian functions. The integral intensity ratio of the D1 and G1 peaks, FWHMG and the peak position change with the Ni content are presented in Table 2. The ratio between the integral intensities of the D1 and G1 peaks (AD/AG) decreases with increasing crystallinity of a graphite-like fraction. Notice that the D/G intensity ratio for the RFC50Ni3-800 is greater than that for the RFC50-Ni1-800 (Fig. 1c, Table 1). A relatively high AD/AG ratio indicates that amorphous carbon is produced under these conditions. The increasing of D peak and its shift to a lower wavenumber may indicate a decrease in size of the graphite clusters and disordering of carbon structure [30]. Catalytic graphitization resulted in the decrease in D1 peak widths. The position of the G1 line depends on domain size. The shift toward the high frequency is observed (shift by 20 cm-1 in RFC50-Ni3-800, which is higher than that of the G line of graphite at 1582 cm-1), reflecting changes in crystallite sizes and decrease in contribution of defects in the structures with decreasing FWHMG (Table 2). In the composites obtained at different temperatures, the G1 peak position shifted slightly toward higher frequencies after annealing at 800 °C (Table 2). The ratio AD1/AG1 also increases (since the G band becomes narrower). As a rule, the maximum of component D1 exhibits a tendency to a lowfrequency shift under annealing. All these data indicate that annealing brings structural ordering of the sp2 coordinated atomic rings and an increase in the graphene-like clusters in size. According to the Raman spectra (Fig. 1c), there is no linear correlation between the water content in the reaction medium and the band deconvolution results (Table 2). It may be because of the
4 Page 4 of 18
cr
ip t
differences in the carbon structures of the chars since water influences their formation and porosity, and then the graphitization process. To understand the thermal stability and anti-oxidation behavior of the nanocomposites TG and DTG curves were measured for the Ni/C samples in the air atmosphere. The TG and DTG curves (Fig. 2) can be divided into several sections vs. temperature. First, from room temperature to about 100░°C, the weight loss is due to desorption of physisorbed water. Second, the weight increases at T > 100°C due to adsorption of oxygen on metal and carbon phases. Third, the weight decreases after treatment at 450 °C independently on the Ni content that suggest carbon combustion with CO2 removal. The weight increase phenomena are visible above 750 °C due to oxidation of metallic Ni nanoparticles in air with the formation of NiO. The wide temperature range of oxidation can be explained by the presence of nanoparticle aggregates of various densities. The greater the density of nickel and carbon aggregates, the higher is the temperature of complete oxidation of nickel particles. Furthermore, in samples with a high content of carbon (Table 2), there are barriers for oxygen penetration to dense particles of the chars.
d
M
an
us
The TG data in the range of 100–1200°C were used to estimate the carbon and Ni contents (Table 2). Since the TG measurements were carried out under the air atmosphere, therefore the carbon phase was gasified to CO2, whilst metal-related particles were oxidized to metal oxides. Based on the weight of the studied sample and the weight loss after TG measurements it was possible to calculate the weight of formed NiO and, thus, the content of Ni in it. The carbon content was calculated by the difference between the initial weight of the samples and calculated weight of the nickel. Additionally, the combustion temperature of chars is shifted to lower temperatures by 60 °C for RFC50-Ni1-800 compared to RFC35-Ni1-800 and RFC20-Ni1-800. This may be explained by the formation of more porous chars since the pores in carbons can facilitate the diffusion of air to burn carbons and removal the products (CO and CO2). It is clearly seen that the burning process starts at different temperatures for the studied samples.
Ac ce pt e
The microscopic analysis (SEM) was undertaken to elucidate a microstructural arrangement of nickel and carbon and to define the morphology of the Ni/C nanocomposites. Images (Fig. 3) shows embedded and highly dispersed nickel nanoparticles (bright light dots) in the carbon matrix (light gray pellets) and big Ni clusters of irregular shape are located at the outer surfaces of the nanocomposites (Fig. 3a, b, and h) and formed the aggregates (Fig. 3c, i). The formation of large Ni aggregates on the outer surface decreases the content of Ni inside the carbon particles. The presence of large amounts of nuclei of the Ni particles during carbonization is the reason of the formation of numerous small Ni nanoparticles (15-20 nm in size). The formed Ni/C aggregates have the sizes of 12 µm. With increasing Ni content highly inhomogeneous dispersion of Ni is observed and its clusters are clearly discernible that results in the formation of denser particles (see the supplementary file). An inhomogeneous dispersion of Ni can be attributed to the following reasons: inhomogeneous distribution of Ni(ac)2 during polymerization, and redistribution of Ni nanoparticles during heating caused by their migration and gathering effect. Moreover, SEM images of the composites show carbon nanotube growing on the Ni particles (Fig. 3b, f and i). Fig. 4 shows HRTEM image of the Ni/C composite with spherical-like Ni nanoparticles of approximately 18 nm in size (Fig. 4a,b). It confirms the Ni particles agglomeration. All Ni/C composites are characterized by a core-shell structure with well-crystallized graphitic shells (Fig. 4a, area A) about 8-15 nm in thickness (Fig. 4c), which corresponds to 24 - 32 graphene layers (Fig. 4d). The interlayer distance between the graphitic layers surrounding the metallic particles was found to 0.34–0.46 nm. These values are larger than that of bulk graphite (0.3354 nm) due to the presence of defects in the carbon phase (penta- and heptagonal rings, etc.). The carbon shell corresponds to (002) graphitic basal planes having a separation of 0.36 nm. The value of the spacing between nickel planes 5 Page 5 of 18
is determined to be 0.19 nm (Fig. 4d), which matches closely with the d spacing of (111) planes of nickel lattice (0.2 nm). All the results obtained from HRTEM are in a good agreement with the results based on the XRD data. Despite the crystallinity of the shells, they possess some degree of disorder in carbon between the Ni nanoparticles (Fig. 4, area B). The generation of the graphitic shells is attributed to the catalytic activity of Ni nanoparticles in the formation of graphite.
Ac ce pt e
d
M
an
us
cr
ip t
The structural characteristics of nanocomposites were studied using low-temperature nitrogen adsorption/desorption isotherms (Fig. 5). All of these materials exhibit isotherms of type IV (H3 type of hysteresis loops) according to the IUPAC classification [31]. Capillary condensation occurs in a relative wide range of pressure at p/p0 = 0.45-1.0 indicating the presence of mesoporosity. However, there is a decrease in the BET surface area and total pore volume as the loading of Ni increases. This is due to increased density as a result of incorporating Ni nanoparticles and/or the blockage of mesopores by Ni nanoparticles. According to the pore size distributions, the main contribution into the structural porosity of the nanocomposites is due to the existence of micropores and, in some extent of mesopores (Fig. 5, Table 3, Vmicro, Smicro at R < 1 nm and Smeso, Vmeso at 1 < R < 25 nm). The BET surface area and pore volume (Table 3) of the microporous carbon/nickel nanocomposite with a low Ni loading (RFC50-Ni1-800) are as high as 319 m2/g and 0.219 cm3/g, respectively. These values gradually decrease with increasing Ni content (samples RFC50-Ni2-800 and RFC50-Ni3-800, Table 3). This is associated with high density of Ni and mesostructural degradation. The same concerns the first series of studied nanocomposites carbonized at 700░°C (Table 3). For carbons RFC50-Ni1-700 and RFC50-Ni3-700 from the first series of samples the PSD peaks (Fig. 5) in the range of micropores have the same position in relation to axis of pore radius but for RFC50Ni2-700 sample this peak shifts toward smaller pore sizes. In the range of wider pores one can see two wide peaks with maxima at about 15 and 80 nm for samples RFC50-Ni2-700 and RFC50-Ni3700 and at 20 and 55 nm for sample RFC50-Ni1-700. For the second series of samples, the PSD peaks in the range of micropores retain practically the same position for all studied samples but they differ in their intensity. However these samples differ in their mesoporosity (Fig. 5b). Here we can see a number of peaks at various pore radii in the range of mesoand macropores (20 - 100 nm). Indeed sample RFC50-Ni3-800 (Fig. 5b, Table 3) is characterized by a minimal porosity among all the chars studied (Table 3) but the contribution of mesopores in all samples of second series is almost the same (Smeso/SBET ≈ 27-29%). It is noteworthy that for all three series of nanocomposites the contribution of slit-shaped pores is lower in comparison with cylindrical pores and voids between particles (Table 3, cslit, ccyl, and cvoid). The morphology and the porosity of chars depend strongly on the water content during the polymer formation because water promotes a high cross-linking degree of the polymer and thereby increases the porosity of the formed chars [32, 33]. With increasing content of water (third series of saples), the mesoporosity and macroporosity decrease, but the microporosity increases, and the bulk density decreases (Fig. 5, Table 3). According to the shapes of the nitrogen adsorption–desorption isotherms and PSDs (Fig. 5) RFC20-Ni1-800 prepared at minimal water content (20 ml H2O) is nano/mesoporous but RFC35-Ni1-800 (35 ml H2O) is mainly microporous (Table 3). It should be taken into account that nitrogen adsorption isotherms give incomplete information about macropores (Table 3, Vmacro) due to incomplete of their filling even at p/p0 > 0.99. However, according to SEM images (Fig. 3) the contribution of macropores is significant. Moreover, the bulk density (ρ) decreases with increasing water content in the reaction media (Table 1). The deviations of the pore shape from slitlike one of the SCV/SCR model (Table 3, ∆w) [22] are smaller for samples with different Ni content carbonized at 700 and 800°C (first and second series) than that for the third series with different contents of water. This that can be explained by changes in the particle morphology (Fig. 5) with an increase in contributions of mesopores (RFC20-Ni1-800) or micropores (RFC35-Ni1-800). 3.2. Measurements and analysis of adsorption capacity 6 Page 6 of 18
us
cr
ip t
Magnetically separable porous carbon materials are highly attractive for many applications associated with liquid-phase processes. In this study, MB was chosen as a model pollutant for adsorption experiment (Fig. 6). Note that a rather fast uptake of the dye occurs during the first 0.5 hour of the adsorption process (40% of MB was adsorbed). Next stage is slower since the additional adsorbed amount of dye reaches 25 % during following 3 hours and 35 % during next 9 hours. This is due to textural features of the composites, especially the micropore volume, which reaches more than 55 % of the total pore volume for RFC50-Ni1-800. It is likely related to the size of the adsorbate molecule, since the nitrogen molecule is much smaller in comparison with the methylene blue and can be adsorbed within the micropores, which dominate in the Ni/C sample. Mesoporous RFC20-Ni1-800 and RFC35-Ni1-800 composites with larger pore size have a small stress effect and bulky dye molecules can be sufficiently adsorbed in mesopores thanks to of easy transportation. Moreover, large Ni aggregates on the outer surface may also limit the diffusion of guest molecules into pores. Notice that the amount of adsorbed MB decreases with decreasing content of carbon (Fig. 6). Therefore, an optimum choice is supposed to fabricate activated carbons, which enable high surface areas and efficient adsorption properties that will be discussed in the next work.
M
an
The high usefulness of carbons is due to their high textural characteristics and chemical/mechanical stability. Considering the magnetic properties, the nanocomposites studied can be good adsorbents to remove various pollutants. The advantage of metal-containing carbon nanocomposites is the high mobility of small particles in the external magnetic fields. This makes easier their separation from liquid media and eliminates the time-consuming procedures typical for non-magnetic carbon sorbents. The chars can be easily separated (Fig. 6b) by placing a lab magnet near the glass bottle. The clear solution can be decanted off or easily removed. This simple attempt demonstrates the suitability of the material as magnetically separable adsorbent to remove dyes or other pollutants from the aqueous media.
Ac ce pt e
d
4. Conclusions In summary, magnetically separable micro-mesoporous chars with the specific surface area ≈180-300 m2/g and the total pore volume ≈0.13-0.21 cm3/g were successfully synthesized by an effective simple procedure, which employs the carbonization (in an inert atmosphere) of resorcinol-formaldehyde polymers containing nickel (II) acetate. Ni nanoparticles with a size of about 18 nm are found to be embedded in the carbon matrix as aggregates of 1-2 µm in size, and also located on the outer surface in the form of large clusters. It was shown that Ni content, as well as water volume and temperature treatment, strongly influence the formation of Ni doped chars of various particle morphology and porosity. The Raman spectra analysis indicates a partial crystallization of carbon in the Ni/C chars. It also shows that the increase in heating temperature from 700░°C to 800 °C cannot result in an obvious improvement in carbon crystallinity, which is accordant with the results of XRD analysis. The content of graphitic structures in the chars after pyrolysis is enlarged with decreasing nickel content. SEM and HRTEM analysis revealed the formation of carbon layers over nickel nanoparticles. The thickness of the carbon shells over nickel cores is in the range of 8-15 nm. Additionally, the fibrous carbon structures are present in the composites. The carbon structures formed over highly reactive nickel nanoparticles protected them from oxidation. Practically all nickel nanoparticles are effectively encapsulated by carbon layers. Despite a variety of the structural characteristics of the chars, the total adsorption capacity extends in proportion to the specific surface area and carbon content. Acknowledgements MVG is grateful to the International Visegrad Fund (Visegrad 4 Eastern Partnership Program, contract Visegrad/V4EaP Scholarship No51400029) for the support of this research. 7 Page 7 of 18
The authors are grateful to European Community, Seventh Framework Programme (FP7/2007–2013), Marie Curie International Research Staff Exchange Scheme (IRSES grant No 612484) for financial support of this project. References
Ac ce pt e
d
M
an
us
cr
ip t
T. Yu, Y. Deng, L. Wang, R. Liu, L. Zhang, B. Tu and D. Zhao, Ordered mesoporous nanocrystalline titanium-carbide/carbon composites from in-situ carbothermal reduction, Adv. Mater. 19 (2007) 2301–2306 –. A. H. Lu, W. Schmidt, N. Matoussevitch, H. Bonnemann, B. Spliethoff, B. Tesche, E. Bill, W. Kiefer and F. Schuth, Nanoengineering of a magnetically separable hydrogenation catalyst, Angew. Chem. Int. Ed. 43 (2004) 4303–4306. J. Lee , S. Jin , Y. Hwang , Je-G. Park , H. M. Park, T. Hyeon, Simple synthesis of mesoporous carbon with magnetic nanoparticles embedded in carbon rods, Carbon 43 (2005) 2536–2543. J. Lee, D. Lee, E. Oh, J. Kim, Y. P. Kim, S. Jin, H. S. Kim, Y. Hwang, J. H. Kwak, J. G. Park, C. Shin and T. Hyeon, Preparation of a magnetically switchable bio-electrocatalytic system employing cross-linked enzyme aggregates in magnetic mesocellular carbon foam, Angew. Chem. Int. Ed. 44 (2005) 7427–7436. V. V. Baranauskas, M. A. Zalich, M. Saunders, T. G. St Pierre and J. S. Riffle, Poly(styrene-b-4vinylphenoxyphthalonitrile)-cobalt complexes and their conversion to oxidatively-stable cobalt nanoparticles, Chem. Mater. 17 (2005) 5246–5254. J. Lee, S. Jin, Y. Hwang, J. G. Park, H. M. Park and T. Hyeon, Simple synthesis of mesoporous carbon with magnetic nanoparticles embedded in carbon rods, Carbon 43 (2005) 2536–2543. D. Wei, Y. Liu, L. Cao, L. Fu, X. Li, Y. Wang, and G. Yu, A magnetism-assisted chemical vapor deposition method to produce branched or iron-encapsulated carbon nanotubes, J. Amer. Chem. Soc. 129 (2007) 7364–7368. J. Phillips, T. Shiina, M. Nemer and K. Lester Graphitic structures by design Langmuir 22 (2006) 9694–9703. G. Korneva, H. Ye, Y. Gogotsi, D. Halverson, G. Friedman, J. C. Bradley and K. G. Kornev, Carbon nanotubes loaded with magnetic particles, Nano Lett. 5 (2005), 879–884. D.Yang, J. H. Hu, S. K. Fu, Controlled synthesis of magnetite-silica nanocomposites via a seeded solgel approach, J. Phys. Chem. C 113 (2009) 7646–7651. N. He, Y. F. Guo, Y. Deng, Z. F. Wang, S. Li and H. Na Liu, Carbon encapsulated magnetic nanoparticles produced by hydrothermal reaction, Chin. Chem. Lett. 18 (2007) 487–490. S.V. Pol, V.G. Pol, A. Frydman, G.N. Churilov, A. Gedanken, Fabrication and magnetic properties of Ni nanospheres encapsulated in a fullerene-like carbon, J. Phys. Chem. B 19 (2005), 9495–9498. N. Luo, X. Li, X. Wang, H. Yan, C. Zhang, and H. Wang, Synthesis and characterization of carbonencapsulated iron/iron carbide nanoparticles by a detonation method, Carbon 48 (2010) 3858–3863. X.Q. Wang, D.-E. Jiang, S. Dai, Surface modification of ordered mesoporous carbons via 1,3-dipolar cycloaddition of azomethine ylides, Chem. Mater. 20 (2008) 4800–4802. M. Inagaki, New Carbons. Control of Structure and Functions, Elsevier, Oxford, 2000. A. Léonard, N. Job, S. Blacher, J.-P. Pirard, M. Crine, W. Jomaa, Suitability of convective air drying for the production of porous resorcinol–formaldehydeand carbon xerogels, Carbon 43 (2005) 1808– 1811. S. Tanaka, T. Yasuda, Y. Katayama, Y. Miyake, Pervaporation dehydrationperformance of microporous carbon membranes prepared from resorci-nol/formaldehyde polymer, J. Membr. Sci. 379 (2011) 52–59. V. Gun’ko, V. Bogatyrov, O. Oranska, I. Urubkov, R. Leboda, B. Charmas, J. Skubiszewska-Zieba, Synthesis and characterization of resorcinol–formaldehyde resinchars doped by zinc oxide, Appl. Surf. Sci. 30-31 (2014) 263–271. 8 Page 8 of 18
Ac ce pt e
d
M
an
us
cr
ip t
V. Gun’ko, V. Bogatyrov, V. Turov, R. Leboda, J. Skubiszewska-Zieba, I. Urubkov, Structural features of resorcinol–formaldehyde resin chars andinterfacial behavior of water co-adsorbed with low-molecular weight organics, Appl. Surf. Sci. 283 (2013) 683– 693. P. Scherrer, Bestimmung der Grösse und der inneren Struktur von Kolloidteilchen mittels Röntgenstrahlen, Nachr. Ges. Wiss. Göttingen 26 (1918) 98–100. S.J. Gregg and K.S.W. Sing, Adsorption, surface area and porosity, Academic Press, London, 1982. V.M. Gun’ko, S.V. Mikhalovsky, Evaluation of slitlike porosity of carbon adsorbents, Carbon 42 (2004) 843–849. T. Hiraoka, T. Kawakubo, J. Kimura, R. Taniguchi, A. Okamoto, T. Okazaki, et al., Selective synthesis of double-wall carbon nanotubes by CCVD of acetylene using zeolite supports, Chem. Phys. Lett. 382 (2003) 679–685. T. Livneh, T. L. Haslett and M. Moskovits, Distinguishing disorder-induced bands from allowed Raman bands in graphite, Phys. Rev. B: Condens. Matter 66 (2002) 195110–21. H. Song, X. Chen, X. Chen, S. Zhang, H. Li, Influence of ferrocene addition on the morphology and structure of carbon from petroleum residue, Carbon 41 (2003) 3037–46. W. Weisweiler, N. Subramanian, B. Terwiesch, Catalytic influence of metal melts on the graphitization of monolithic glasslike carbon, Carbon 9 (1971) 755–8. A. Yoshida, Y. Kaburagi, Y. Hishiyama, Full width at half maximum intensity of the G band in the first order Raman spectrum of carbon material as a parameter for graphitization, Carbon 44 (2006) 2333–2335. F. Tuinstra, J.L. Koenig, Raman spectrum of graphite, J. Phys. Chem., 53 (1970) 1126–30. T.-H. Ko, W.-S. Kuo, Y.-H. Chang, Raman study of the microstructure changes of phenolic resin during pyrolysis. Polym. Compos. 21 (2000) 745–750. R. Sergiienko, E. Shibata, K. Sunghoon, K. Takuya, N. Takashi, Nanographite structures formed during annealing of disordered carbon containing finely-dispersed carbon nanocapsules with iron carbide cores, Carbon 47 (2009) 1056 –1065. J. Rouquerol, D. Avnir, C.W. Fairbridge, D.H. Everett, J.M. Haynes, N. Pernicone, J.D.F. Ramsay, K.S.W. Sing, K.K. Unger, et. al., Recommendations for the characterization of porous solids (Technical Report), Pure Appl. Chem. 66 (1994) 1739–1758. V. Gun’ko, V. Bogatyrov, V. Turov, R. Leboda, J. Skubiszewska-Zieba, I. Urubkov, Structural features of resorcinol–formaldehyde resin chars andinterfacial behavior of water co-adsorbed with low-molecular weightorganics, Appl. Surf. Sci. 283 (2013) 683– 693. V.M. Gun’ko, I.N. Savina, S.V. Mikhalovsky, Cryogels: morphological, structuraland adsorption characterization, Adv. Colloid Interface Sci. 187 (2013) 1–46. Scheme 1. Synthesis and suggested structure of Ni/resorcinol-formaldehyde polymers. Fig. 1 XRD patterns of Ni/C nanocomposites of the first (inset) and second series (a), deconvolution of the Raman spectrum of RFC50-Ni2-700 sample (b) and Raman spectra of the all chars (c). Fig. 2 TG and DTG curves of the nanocomposites of the first (a, b), second and third series (c, d) of the chars. Fig. 3 SEM images of: RFC50-Ni1-800 (a, b and c), RFC50-Ni2-800 (d), RFC50-Ni3-800(e), RFC50-Ni1-700 (f), RFC50-Ni2-700 (g), RFC50-Ni3-700 (h), RFC35-Ni1-800 (i). Fig. 4 HRTEM image of the RFC50-Ni3-800 composite. Carbon shells and disordered carbon areas are marked by A and B, respectively. Fig. 5 Nitrogen adsorption/desorption isotherms (a, c) and pore size distributions (b, d) of the nanocomposites of the first (a, b), second and third series (c, d). Fig. 6 Adsorption capacities of the chars for methylene blue (a) and separating of the RFC50-Ni3-800 nanocomposite from aqueous solution of MB by the magnet (b).
9 Page 9 of 18
Table 1 The initial ratio of the components used to prepare RFR /Ni compounds (R/F=1: 2).
Second series Third series
RFC50-Ni1-700 RFC50-Ni2-700 RFC50-Ni3-700 RFC50-Ni1-800 RFC50-Ni2-800 RFC50-Ni3-800 RFC35-Ni1-800 RFC20-Ni1-800
R/Ni (mol/mol) 1:0.35 1:0.62 1:0.89 1:0.35 1:0.62 1:0.89 1:0.35 1:0.35
ip t
First series
Components ratio R/W (w/w) 1:5 1:5 1:5 1:5 1:5 1:5 1:3.5 1:2
Samples
cr
Series
17 7 17 7 18 7 18 7 18 7 18 8 18 7 18 7
d(002) D1, graphite, (cm-1) nm 0.341 1350
an
dcr, (nm)
Ac ce pt e
d
M
Table 2 Structural characteristics of carbon chars. Samples Carbon Ni Phase content content (wt %) RFC50-Ni1-700 74.7 25.3 Ni Cgraphite RFC50-Ni2-700 67 33 Ni Cgraphite RFC50-Ni3-700 50.8 49.2 Ni Cgraphite RFC50-Ni1-800 71.1 28.9 Ni Cgraphite RFC50-Ni2-800 61.9 38.1 Ni Cgraphite RFC50-Ni3-800 46.1 53.9 Ni Cgraphite RFC35-Ni1-800 75 25 Ni Cgraphite RFC20-Ni1-800 77 23 Ni Cgraphite
us
Note: R/F and R/Ni are the mole ratio of resorcinol (R), formaldehyde (F) and Ni acetate (Ni); and R/W is the weight ratio of resorcinol and water (W) in the reaction mixture.
Table 3 Textural characteristics of Ni-doped chars. SBET, Smicro, Smeso, Smacro, Samples m2/g m2/g m2/g m2/g First series RFC50-Ni1-700 384 288 95 0.9 RFC50-Ni2-700 327 280 46 0.5 RFC50-Ni3-700 331 249 87 0.6 Second series RFC50-Ni1-800 319 232 86 1.0 RFC50-Ni2-800 253 180 72 0.6
G1, (cm-1)
AD1/AG1 FWHMG (cm-1)
1573
1
66.1
0.339
1344
1591
1.7
61.4
0.339
1337
1598
2.4
57.4
0.340
1353
1583
1.3
66.8
0.337
1350
1601
1.8
61.2
0.337
1343
1602
2.8
54
0.337
1336
1597
1.9
59.2
0.337
1347
1580
1.2
66.7
Vp, cm3/g
Vmicro, cm3/g
Vmeso, cm3/g
Vmacro, cm3/g
∆w
cslit
0.218 0.20 0.196
0.113 0.094 0.096
0.086 0.091 0.088
0.019 0.015 0.011
0.164 0.117 0.11
0.256 0.226 0.225
0.219 0.181
0.101 0.079
0.094 0.09
0.024 0.012
0.194 0.181
0.249 0.227
10 Page 10 of 18
RFC50-Ni3-800 Third series RFC35-Ni1-800 RFC20-Ni1-800
188
131
56
0.4
0.137
0.074
0.055
0.008
0.115
0.216
388 397
295 269
91 125
1.6 3.6
0.231 0.329
0.112 0.104
0.08 0.146
0.039 0.079
0.212 0.206
0.101 0.164
an
us
cr
ip t
Note: Micropores (Smicro, Vmicro) at radius or half-width R < 1 nm, mesopores (Smeso, Vmeso) at 1 nm < R < 25 nm, and macropores (Smacro, Vmacro) at 25 nm < R < 100 nm; ∆w is a criterion showing the deviation of the pore model from the real pore in respect to the specific surface area. Weight coefficients (cslit, ccyl and cvoid) in the SCV/SCR model show contributions of different pores [22].
HO
Ni
2+
CH 2
OH HO
M
+ 2n CH O 2
H 2O , Ni 2+
Ac ce pt e
HO
OH
Ni 2+
HO
d
n HO
HO
OH
CH 2
OH
Scheme I. Synthesis and suggested structure of Ni/resorcinol-formaldehyde polymers.
11 Page 11 of 18
ip t cr us
Ac ce pt e
d
M
an
Fig. 1. XRD patterns of Ni/C nanocomposites of the first (inset) and second series (a), deconvolution of the Raman spectrum of RFC50-Ni2-700 sample (b) and Raman spectra of the all chars (c).
12 Page 12 of 18
ip t cr
Ac ce pt e
d
M
an
us
Fig. 2. TG and DTG curves of the nanocomposites of the first (a, b), second and third series (c, d) of the chars.
13 Page 13 of 18
ip t cr us an
Ac ce pt e
d
M
Fig. 3. SEM images of: RFC50-Ni1-800 (a, b and c), RFC50-Ni2-800 (d), RFC50-Ni3-800(e), RFC50-Ni1-700 (f), RFC50-Ni2-700 (g), RFC50-Ni3-700 (h), RFC35-Ni1-800 (i).
14 Page 14 of 18
ip t cr us an
Ac ce pt e
d
M
Fig. 4. HRTEM image of the RFC50-Ni3-800 composite. Carbon shells and disordered carbon areas are marked by A and B, respectively.
15 Page 15 of 18
ip t cr us an M d Ac ce pt e
Fig. 5. Nitrogen adsorption/desorption isotherms (a, c) and pore size distributions (b, d) of the nanocomposites of the first (a, b), second and third series (c, d).
16 Page 16 of 18
ip t cr us an
(b)
Ac ce pt e
HIGHLIGHS
d
M
Fig. 6. Adsorption capacities of the chars for methylene blue (a) and separating of the RFC50-Ni3800 nanocomposite from aqueous solution of MB by the magnet (b).
Facile synthesis of the Ni-doped carbon sorbents via carbonization of resorcinolformaldehyde polymers/nickel(II) acetate mixtures in the inert atmosphere. Effects of Ni content, as well as water volume and temperature treatment on the morphology and texture of the chars. Ni/C composites are characterized by a core-shell structure with well-crystallized graphitic shells. Ni content influences the structure of the carbon. Nickel-doped carbon nanocomposites were used as a magnetically separable adsorbent.
17 Page 17 of 18
an M ed pt ce Ac
Page 18 of 18