1-Cobalt-silica nanocomposite via thermal calcination-reduction of gel precursors.pdf

Materials Chemistry and Physics 128 (2011) 70–76 Contents lists available at ScienceDirect Materials Chemistry and Physics...

Materials Chemistry and Physics 128 (2011) 70–76 Contents lists available at ScienceDirect Materials Chemistry and Physics journal homepage: www.elsevier.com/locate/matchemphys Cobalt/silica nanocomposite via thermal calcination-reduction of gel precursors Osama A. Fouad a,∗ , Salah A. Makhlouf b , Gomaa A.M. Ali c , A.Y. El-Sayed c a Central Metallurgical Research and Development Institute, CMRDI, P.O. Box 87, Helwan 11421, Egypt b Physics Department, Faculty of Science, Assiut University, Assiut 71516, Egypt c Chemistry Department, Faculty of Science, Al-Azhar University, Assiut branch, Assiut 71524, Egypt a r t i c l e i n f o a b s t r a c t Article history: Well dispersed and thermally stable Co/SiO2 nanocomposite powders have been synthesized successfully Received 26 August 2010 via thermal calcination-reduction process of the prepared gel precursors. Co-gelation method for the Received in revised form 11 February 2011 cobalt and silica sources in the same solution mixture was found to be the best way to prepare these Accepted 16 February 2011 materials. Single (Co3 O4 ) and (Co) phase formation could be confirmed by XRD and FTIR techniques for the products after calcination and reduction stages at the optimum process conditions, respectively. Keywords: Surface analysis and TEM investigations of the reduced sample containing 25 wt.% Co revealed that highly Co/SiO2 nanostructures dispersed cobalt nanoparticles embedded in a mesoporous inorganic polymeric silica matrix could be Nanocomposites Mesoporous obtained. The silica matrix played a key role in protecting metallic cobalt nanoparticles from oxidation Thermal stability upon heating in air up to 400 ◦ C. Addition of ethylene glycol to the gel mixture during gelation resulted in the formation of an additional organic polymeric protective layer. This layer promotes the formation of a regular pore structure upon its removal through calcinations and reduction steps. The current study could help in understanding the parameters affecting the dispersion and phase formation of cobalt species in silica matrix which indeed affect the activity of Co/SiO2 nanocomposite as a catalyst in different processes and reactions. © 2011 Elsevier B.V. All rights reserved. 1. Introduction hydrogenation of toluene , oxidation of ␣-pinene and alkenes [7,8], hydroformylation of ethene and as a magnetic separa- Magnetic transition metal (TM) nanoparticles have a great tor for specific biological entities in biomedical applications. affinity to oxygen and form oxides. This property causes sponta- In addition, it can provide the active materials for humidity [11,12] neous ignition in atmospheric air and consequently limits their and gas sensors. Silica (SiO2 ) has been regarded as a very effec- applications in open atmosphere especially at relatively higher tive inorganic polymer matrix for stabilizing the metal nanocrystals temperatures. Capping may provide a protective layer surround- against oxidation, tailoring a uniform particle size distribution and ing a magnetic core enhances the resistance of core materials to controlling the homogeneous dispersion of the ultrafine particles oxidation. In addition, the protective layer on the surface of in the matrix [1,2]. a particle may also prevent dipole–dipole interactions between Several methods and approaches have been explored for the closely spaced magnetic nanoparticles, grain growth and agglomer- synthesis of magnetic nanoparticles such as hydrogen arc plasma, ation during subsequent heat treatment. The process of capping reduction of metal salts, polyol process, -ray irradiation, metal therefore can control grain size and provide a better overall size carbonyl pyrolysis, sol–gel, water-in-oil microemulsion, template distribution [1,3]. synthesis and reversemicelles, ball milling and sonochemical either Nanocomposites of TM species embedded-in or capped-with in powder or thin film forms [14–19]. It is possible to obtain inorganic oxide matrices are considered effective solution for pro- nanocomposite materials at low temperatures through the incor- tecting TM active nanoparticles from oxidation. These composites poration of small amounts of different components in the matrix have gained much attention as promising materials for various. Cobalt oxide/silica nanocomposite thin films have been pre- applications as it posses the properties of metals, the matrices pared by dip-coating technique. Multi cobalt oxide phases and and the modified composites. It can be applied in high density cobalt silicate hydroxide crystalline phase were obtained after cal- recording media , as catalyst in Fischer–Tropsch synthesis , cination at 400–500 ◦ C. In general, there are two main approaches for obtaining dis- persed nanoparticles in silica matrix namely: (i) post-synthesis ∗ Corresponding author. Tel.: +20 02 25010640; fax: +20 02 25010639. treatment methods which is based on synthesis of cobalt E-mail address: [email protected] (O.A. Fouad). nanoparticles in pre-prepared silica matrix, such as ion-exchange, 0254-0584/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.matchemphys.2011.02.072  O.A. Fouad et al. / Materials Chemistry and Physics 128 (2011) 70–76 71 impregnation, grafting techniques and (ii) simultaneous synthe- 2.4. Samples characterization sis methods which involving in situ synthesis of cobalt and silicon 2.4.1. Crystal structure and phase analysis oxides from suitable solution or gel precursors, such solution- Phase identification, purity, relative crystallinity and crystallite size of the prod- based method as sol–gel is performed with or without template ucts were performed at room temperature using X-ray diffractometer equipped with molecules. The sol–gel method followed by thermal treatment an automatic divergent slit (XRD; Philips PW1700 diffractometer, Netherlands). appears preferably simple, allows a better control of the textural Diffraction patterns were obtained using CuK␣ radiation ( = 0.15418 nm) and a properties of the silica matrix and provides more effective nano- graphite monochromator in the 2 range from 10 to 80◦. XRD patterns are fitted using pseudo lorentzian line shapes for accurate determination of lattice constant metric scale dispersion of cobalt species in the matrix. To and apparent crystallite size. A standard Si powder sample is used to monitor the disperse the cobalt nanoparticles in the silica matrix, a solution- instrumental broadening of the diffraction peaks. Infrared spectra were measured based route promoted by noble metals and followed by thermal in the range 4000–400 cm−1 using a Fourier transform infrared spectrometer (FTIR; treatment is proposed by other research groups. Nanocompos- JASCO-480 Plus, Japan). The samples were well mixed with dried KBr powder and pressed to form disc tablets and used for measurements. ite cobalt/silica materials prepared by the simultaneous synthesis of the two precursor components usually possess high surface area 2.4.2. Surface area and homogeneous chemical composition. The specific surface area, pore volume and pore size distribution of the calcined In the present work, nanocomposites of Co/SiO2 were pre- samples were measured by nitrogen adsorption–desorption technique (NOVA- pared by simultaneous synthesis of the cobalt species and silica 3200, USA). The samples were degassed at 250 ◦ C for 3 h before analysis, the N2 isotherms were obtained at −196 ◦ C. The specific surface area was calculated by precursors followed by thermal decomposition (calcination) and using the BET (Brunauer–Emmett–Teller) method. The pore size distributions were hydrogen reduction. A comparison between co-gelation and post- determined by applying the BJH (Barrett–Joyner–Halenda) model to the nitrogen gelation methods, and the effect of some additives on the physical desorption branch. and chemical characteristics of the prepared samples were also explored. The effect of various parameters on the dispersion and 2.4.3. Morphology Morphology of the synthesized nanocomposites was analyzed using scanning phase formation of cobalt species in such composite material was electron microscopy (SEM, JEOL-JSM-5410, Japan) operating at 30.0 kV and trans- also studied. mission electron microscopy (TEM, JEOL-1230, Japan) operating at a maximum of 200 kV. 2.4.4. Thermal analysis 2. Experimental Thermal behavior of the precursor, calcined and reduced samples was investi- gated using thermal gravimetric analysis (TGA) and differential thermal analysis 2.1. Materials (DTA), (Thermal Analyst NETZSCH STA 409C/CD, Germany). A sample weighing 10–20 mg was loaded in a ceramic sample holder. Data were recorded upon heating Cobalt(II) nitrate hexahydrate (Co(NO3 )2 ·6H2 O, 98%, Riedel, UK) and TEOS up to 700 ◦ C at a heating rate of 10 ◦ C min−1 in an air stream of 50 ml min−1. (Si(C2 H5 O)4 ,99%, Merck, Germany) were used as cobalt and silica precursors respectively. Ethylene glycol (C2 H4 (OH)2 , 99% Sigma–Aldrich, Switzerland) and dimethylformamide, (HCON(CH3 )2 , 99%, Fluka, Germany) were used as capping 3. Results and discussion and reducing agents, respectively. Ammonium hydroxide solution (NH4 OH, 33%, Adwic, Egypt) was used for pH adjustment. Absolute ethanol (C2 H5 OH, 99.8%, BDH, When cobalt and silica precursors were added to each other Germany) and bidistilled water were used as solvents and/or washing solutions. All chemicals were used as received without any further treatments. before or after gelation in presence or absence of other additives, the color of the mixture changes gradually by time and upon heating and stirring from pink to blue color. The change in the 2.2. System setup and process stages color is mainly due to either the formation of cobalt hydroxide or cobalt-ethylene glycol complex when it used as an additive. Upon The chemical synthesis was performed in a simple reactor consisting of a three calcination of these mixtures, after washing and drying, the color necked glass flask connected with water vapor condenser, hot plate with mag- changed to brownish black due to the formation of cobalt oxide netic stirrer and a pH meter. The stages involved in this study were: (i) precursor phase species. By H2 reduction, the color changed to dark gray synthesis; (ii) drying; (iii) calcination and (iv) H2 reduction. (black) due to the formation of cobalt metal particles. The visual intensity of the color depends mainly on the cobalt ratio present in the precursor mixtures. 2.3. Synthesis procedure Synthesis of Co/SiO2 nanocomposites was performed using different chemical 3.1. Crystal structure and phase analysis routes in presence of various additives starting from alcoholic solutions of cobalt and silica sources at the desired stoichiometric ratios. In general, precursor syn- Fig. 1 shows the XRD patterns of co-gelation Co/SiO2 samples thesis was started by dissolving calculated amounts of cobalt nitrate and TEOS prepared with different Co weight percentages from 5 to 50% and separately in absolute ethanol. The two solutions were then dispersed ultrasoni- cally for 30 min. Two routes of gelation process were performed. The first route was calcined at 400 ◦ C for 3 h. The XRD pattern of 50 wt.% Co sam- co-gelation process, which included formation of the gel by adding the dispersed ple precursor before calcination is also shown for comparison. TEOS alcoholic solution dropwise to the dispersed cobalt nitrate alcoholic solution Both Co(NO3 )2 ·6H2 O (JCPDS # 71-726) and ␤-Co(OH)2 (JCPDS # in the three necked flask while stirring. Then bidistilled water (H2 O:TEOS = 3:1) was 30-0443) phases are detected in the precursor due to the partial added dropwise to make hydrolysis. The hydrolysis reaction was base catalyzed by the addition of NH4 OH solution, which was also used to adjust the pH value up to 5. conversion of Co(NO3 )2 into Co(OH)2 by the effect of NH4 OH addi- The mixture was left for gelation at 60 ◦ C for 3 h under continuous stirring. tion used for pH adjustment. No diffraction peaks corresponding The second route was pre-gelation process which included formation of the two to silica matrix were observed indicating that silica presents in the gels separately (silica gel and cobalt gel). Then the two gels were gradually mixed amorphous form. For the calcined samples (5–50) wt.% Co repre- while stirring and processed with the same sequence as above. sented in Fig. 1b–e respectively, the Co3 O4 phase (JCPDS card # In some other experiments additives such as ethylene glycol (EG) and dimethyl- formamide (DMF) were added as capping, reducing and/or drying agents in the ratio 78-1970) is the only detected phase. In addition, the peaks are more of cobalt nitrate: EG or DMF = 1:2. broad implying the nanoscaled particles size. Subsequently, the formed gels from both routes were dried in an oven at 80 ◦ C Fig. 2 shows the XRD patterns of the calcined co-gelation sam- for 24 h. Precursors were calcined at 400 ◦ C for 3 h to obtain either a cobalt oxide ples that contain 25 wt.% Co without and with some other additives phase in silica matrix or a cobalt silicate phase. To obtain cobalt nanocrystals in silica matrix, the calcined samples were reduced by pure H2 gas in a tube furnace at 700 ◦ C and the pre-gelation sample. It is obvious that there is a noticeable for 4 h. Series of samples have been prepared in which Co:SiO2 weight ratios varied change in phase formation for samples prepared without or with from 0 to 50%. additives. The samples prepared with chemical additives show the 72 O.A. Fouad et al. / Materials Chemistry and Physics 128 (2011) 70–76 = Co =SiO 2 400 Normalized intensity (arb. units) d 300 c 200 b 100 a 0 10 20 30 40 50 60 70 80 90 100 2θ (degrees) Fig. 1. XRD patterns of sol–gel samples dried at 80 ◦ C for 24 h, (a) SGP50; and sam- Fig. 3. XRD patterns of 25 wt.% Co samples dried at 80 ◦ C for 24 h, calcined at 400 ◦ C ples calcined at 400 ◦ C for 3 h (b) SGO5, (c) SGO10, (d) SGO15, (e) SGO25 and (f) for 3 h and reduced by H2 at 700 ◦ C for 4 h (a) SGR25, (b) SG-EGR25, (c) SG-GMR25 SGO50. and (d) SG-DMF R25. formation of extra CoSiO3 phase (JCPDS card # 72-1508). The sam- ples prepared through gel mixing (pre-gelation) route (Fig. 2c) have or reduced forms related to the route of synthesis are summarized higher peak intensities than those with no additives and with EG in Table 1. and DMF additions. This might be due to the fact that the amount It is clear from Table 1 that the sol–gel oxide samples have of cobalt oxide phase species in this sample is higher than others small crystallite size 7.3–8.1 nm independent on the Co content; and/or due to large particles size. and lattice constant (a) equals 8.09–8.10 Å, slightly larger than that Fig. 3 shows the XRD patterns of Co/SiO2 samples prepared by of bulk Co3 O4 due to surface relaxation usually observed for such different routes after H2 reduction. The diffraction peaks corre- nanoparticles. For the reduced samples, on the other hand, the sponding to metallic cobalt are detected in all samples (JCPDS card sample prepared by sol–gel route (SGR25) has the smallest crys- # 15-0806). For the sample prepared with additives, low inten- tallite size for cobalt metal (15.5 nm) and the gel-mixing sample sity peaks assigned to SiO2 phase (JCPDS card # 86-680) are also has comparatively larger crystallite size (23.3 nm) due to particle retained due to crystallization of amorphous SiO2 at high temper- agglomeration in the precursor gel. The increase in particle size atures. upon reduction is due to the high reduction temperature (700 ◦ C) The apparent crystallite size (Dp ) of the Co particle species can which enhances grain size growth. The lattice constant of the Co be estimated by the Scherrer formula: phase in all reduced samples was found to be 3.54–3.55 Å, very close to that of fcc metallic cobalt. k Fig. 4 and 5 show the FTIR spectra of the calcined and reduced Dp = (1) ˇ cos samples, respectively. The spectra of 0 wt.% Co and 100 wt.% Co where k = 0.9,  is the Cu-target wavelength, ˇ is the full-width samples are also shown for comparison. It is obvious that the at half-maximum (FWHM) of diffraction peaks corrected for the absorption bands at 3440 cm−1 and 1635 cm−1 appeared in both instrumental broadening and  is the diffraction angle. The crystal- figures belong to the absorbed water molecules. However, the lite size and lattice constant of various Co species in either oxidized peak intensities are higher in case of calcined than reduced sam- ples. These bands are ascribed to the stretching of O–H mode of H-bonded and the bending of O–H mode of water molecules, 600 respectively. The narrow absorption band at 1385 cm−1 confirms =Co3O4 =CoSiO3 Normalized Intensity (arb. units) 500 d 400 300 c 200 b 100 a 0 10 20 30 40 50 60 70 80 2θ (degrees) Fig. 2. XRD patterns of 25 wt.% Co samples dried at 80 ◦ C for 24 h and followed Fig. 4. FTIR spectra of Co/SiO2 samples dried at 80 ◦ C for 24 h, (a) SG-EGP25 and by calcination at 400 ◦ C for 3 h (a) SGO25, (b) SG-EGO25, (c) SG-GMO25 and (d) calcined at 400 ◦ C for 3 h (b) SGO0, (c) SGO25, (d) SG-EGO25, (e) SG-GMO25, (f) SG-DMFO25. SG-DMFO25 and (g) SGO100.  O.A. Fouad et al. / Materials Chemistry and Physics 128 (2011) 70–76 73 Table 1 Variation of crystallite size (Dp ) and lattice constant (a) of pure cobalt species (Co and Co3 O4 ) with the route of synthesis for various samples. Sample code State Synthesis method Co (wt.%) Dp (nm) a (Å) SGP50 Precursor Sol–gel 50 – – SGO5 Oxide Sol–gel 5 7.50 8.103 SGO10 Oxide Sol–gel 10 7.70 8.094 SGO15 Oxide Sol–gel 15 7.48 8.083 SGO25 Oxide Sol–gel 25 8.10 8.092 SGO50 Oxide Sol–gel 50 7.30 8.103 SG-EGO25 Oxide Sol–gel (Ethylene glycol) 25 – – SG-GMO25 Oxide Sol–gel (gel mixing) 25 – – SG-DMFO25 Oxide Sol–gel (dimethylformamide) 25 – – SGR25 Reduced Sol–gel 25 15.50 3.550 SG-EGR25 Reduced Sol–gel (ethylene glycol) 25 21.20 3.543 SG-GMR25 Reduced Sol–gel (gel mixing) 25 23.30 3.547 SG-DMFR25 Reduced Sol–gel (dimethylformamide) 25 20.40 3.546 the presence of nitrate ions in the network of all dried gels. This results obtained from XRD analyses and the interpretation of all the band is related to the asymmetric stretching of N–O bonds in the detected bands is in agreement with the findings of other authors NO3 − group. The disappearance of this band in the calcined samples. Moreover, the presence of the Si–O–Si bonds as confirmed confirms that all NO3 − species have been completely decomposed from IR spectra together with the results obtained by XRD analysis in the network of all calcined gels. Moreover, the SG-EG dried implying that the formed phases (Co, Co3 O4 and CoSiO3 ) are mostly sample shows a weak band at 2925 cm−1 corresponding to C–H embedded in an amorphous silica matrix. aliphatic bond of methene group in EG-Co(II) complex. The strong absorption bands at 1090 cm−1 and 800 cm−1 agree 3.2. Surface area and porosity well with the SiO2 bond structure. The band at 1090 cm−1 can be assigned to the asymmetric stretching vibration of the bond Si–O–Si Fig. 6 and 7 show the nitrogen adsorption–desorption isotherms in the SiO4 tetrahedron. The band at 800 cm−1 is corresponding of some sol–gel calcined samples and those prepared with chemi- to the vibration of the Si–O–Si symmetric stretch. The band at cal additives respectively. The results reveal that the isotherms can 460 cm−1 is related to bending modes of Si–O–Si bonds. The be classified according to the IUPAC as reversible type IV which weak intensity band at 960 cm−1 can be assigned to either Si–OH characterizes a material that contains mesoporosity and has a high and/or Si–O− stretching vibrations. The intensity and forma- energy of adsorption. These often contain hysteresis attributed to tion of Si–O− groups appears to be due to the interaction between the mesoporosity and show hysteresis loop of type H1 with nearly Co2+ ions and the siloxane matrix in the other dried gel. The vertical and parallel adsorption and desorption branches. This indi- band intensity decreases upon calcination of samples, due to the cates the presence of regular even pores without interconnecting conversion of Si(OH)4 to SiO2. channels. The sample prepared by gel mixing route and containing The absorption band at 660 cm−1 that related to the vibrations 25 wt.% Co (SG-GMO25) has an isotherm of type I which charac- of Co(III)–O bonds in Co3 O4 is observed in Fig. 4 [21,22]. This band terizes a material that has extremely fine pores (micropores). In is also assigned to Co–O stretching in Si–O–Co network by other addition, this sample has a hysteresis loop of type H3 with sloping authors. The band at 560 cm−1 which also associated with the adsorption and desorption branches covering a large range of P/Po Co–O stretching is confirmed. These two bands present strongly indicating presence of slit-like pores. in SG samples calcined at 400 ◦ C whereas in SG-EG, SG-GM and As displayed in Table 2, SBET surface area values amount to 221.2, SG-DMF samples they are week. This might be due to different per- 167.1 and 126.5 m2 g−1 were obtained for the calcined samples centage amounts of the Co3 O4 phase in each sample. The absence of SGO5, SGO15 and SGO25, respectively. It is clear that SBET val- the 560 and 660 cm−1 bands shown in Fig. 5 is consistent with the ues decrease with increasing Co content due to occupying of SiO2 disappearance of Co–O bonds, due to H2 reduction of cobalt oxide species into cobalt metal. These results are in agreement with the 200 Pore Volume (cm nm g ) a -1 -1 -3 180 6x10 100 -1....................................................................460 cm 660 cm......................................................... b 3 160 4 -1..................................3440 cm 1635 cm................................................... 800 cm....................................................... -1 Adsorbed Volume (cm g ) 3 -1............................................................1090 cm 80 560 cm........................................... 140 2 Desorption d c Adsorption Transmittance ( %) 120 0 60 100 0 4 8 12 16 20 c Pore Diameter (nm) b 40 a 80 a 60 -1 -1 -1 b -1 20 40 c 20 0 4000 3500 3000 2500 2000 1500 1000 500 0 0.0 0.2 0.4 0.6 0.8 1.0 Wavenumber (cm-1) P/Po Fig. 5. FTIR spectra of 25 wt.% Co samples dried at 80 ◦ C for 24 h, calcined at 400 ◦ C for 3 h and reduced by H2 at 700 ◦ C for 4 h (a) SGR25, (b) SG-EGR25, (c) SG-GMR25 Fig. 6. N2 adsorption–desorption isotherms and pore size distribution (inset) of and (d) SG-DMFR25. sol–gel samples calcined at 400 ◦ C for 3 h (a) SGO5, (b) SGO15 and (c) SGO25. 74 O.A. Fouad et al. / Materials Chemistry and Physics 128 (2011) 70–76 Table 2 Surface area SBET , micropore area Smic , external surface area SExt along with the particle size for the test composite materials calcined at 400 ◦ C in air for 3 h. Sample code Preparation method Co (wt.%) SBET (m2 g−1 ) Smic (m2 g−1 ) SExt (m2 g−1 ) VP (cm3 g−1 ) Pw (nm) PS (nm) SGO5 Sol–gel 5 221.10 26.86 194.32 0.17 3.10 9.43 SGO15 Sol–gel 15 167.10 20.20 146.88 0.15 3.61 10.86 SGO25 Sol–gel 25 126.50 60.10 66.40 0.15 4.78 12.78 SG-EGO25 Sol–gel (ethylene glycol) 25 144.60 40.30 104.30 0.21 5.73 9.83 SG-GMO25 Sol–gel (gel-mixing) 25 49.00 13.00 36.00 0.04 3.49 29.00 network pores with cobalt species. The SBET values for samples pre- 3.3. Morphology pared in presence of EG (SG-EGO25) and by gel mixing (SG-GMO25) are 144.6 and 49 m2 g−1 , respectively. Table 2 also displays the val- Fig. 8 shows the SEM and TEM images of calcined (SGO25, ues of the micropore area (Smic ) and external surface area (SExt ) Fig. 8a and c) and reduced (SGR25, Fig. 8 b and d) samples prepared which are related together by the following relation : by Sol–gel route containing 25 wt.% Co. Bulk aggregated particles together with finer aggregates are observed in SEM images for cal- SBET = Smic + SExt. (2) cined and reduced samples. TEM image of the calcined sample Average pore width, Pw , is estimated from the relation: (SGO25, calcined at 400 ◦ C for 3 h) reveals the well dispersion of Co3 O4 phase in the silica matrix, as shown in Fig. 8c. The cobalt 4Vp Pw = (3) oxide (Co3 O4 ) phase has a nearly spherical shape with average SBET particles size of about 10 nm obtained from particle size distribu- where Vp is the pore volume. The Pw values are 3.10, 3.60, 4.78, tion for the TEM image for 200 particles Fig. 8d. The results are in 5.73 and 3.49 nm of the above mentioned samples, respectively. agreement with those calculated from XRD data, see Table 1. TEM The equivalent spherical particle size (PS) expressed in nm of image of the reduced sample, (SGR25, calcined at 400 ◦ C for 3 h and the composite primary particles was estimated assuming all the reduced in H2 at 700 ◦ C for 4 h), reveals that the sample is com- particles to have the same spherical shape by using the relation: posed of spherical cobalt particles with silica frame in a core/shell structure. The particles are ranged in size from about 30–70 nm. 6000 PS = (4) Increasing of the particles size for the reduced samples might be SBET ×  due to agglomeration of the reduced cobalt species due to solid where  is the density of bulk composite in g/cm3 calculated using state diffusion at such relatively high reduction temperature. the weight ratio of each phase in the sample and SBET is expressed in m2 /g. The obtained results are listed in Table 2. It is clear that the 3.4. Thermal properties and stability particles sizes calculated from SBET are in agreement with those obtained from XRD patterns. Fig. 9 shows the results of the thermogravimetric analysis (TGA The inset in Fig. 7 shows the pore size distribution curves cal- and DTG) of precursor sample (SGP50) containing 50 wt.% Co. The culated from the desorption branch of the isotherms. It is obvious 50 wt.% Co precursor sample is selected to explore the effect of that the curves show peaks representing pore diameters between thermal treatment (calcination) on the sample under a condition 6 and 10 nm. The sample prepared using EG as a capping agent has of equally available ratio of cobalt and silica. It is clear that the cal- the largest surface area and pore diameter. This can be ascribed cination step can be divided according to the gradual weight loss to the fact that, upon calcinations EG eliminates without causing in the temperature range of 25–650 ◦ C into four stages. The first shrinkage of the pores leaving behind numerous open pores. Simi- stage (25–160 ◦ C) with weight loss amounts to 7.96% and can be lar behavior has been observed by other authors upon using citric attributed to evaporation of the adsorbed water species on the sur- acid as a capping agent. From these results we can conclude face of the precursor. The second stage (160–280 ◦ C) has a weight that co-gelation route give higher surface area and pore diameter loss of 8.77%. The third stage (280–390 ◦ C) has a weight loss of 5.09%. and smaller particle size than pre-gelation route. The fourth one (390–650 ◦ C) has a weight loss of 2.24%. The weight loss in the first three stages (25–390 ◦ C) representing the major 200 -3 weight loss profile and amounts to 21.82%. 50x10 Pore Volume (cm nm g ) According to the above observations, the overall reactions that -1 -1 180 a 40 Desorption might be involved in the synthesis of Co/SiO2 nanocomposites can 3 160 30 Adsorption be summarized as follows: Adsorbed Volume (cm g ) -1 20 a 140 3 10 b 3.4.1. Precursor synthesis (Gelation) stage c 120 0 Co(No3 )2 · 6H2 O + 2NH4 OH → Co(OH)2 + 2NH4 NO3 + 6H2 O (5) 0 4 8 12 16 20 b 100 Pore Diameter (nm) Si(OC2 H5 )4 + 4H2 O → Si(OH)4 + 4C2 H5 OH (6) 80 60 3.4.2. Calcination stage 400◦ C Co(OH)2 −→ CoO + H2 O (7) 40 c 400◦ C 20 3CoO + 1/2O2 −→ Co3 O4 (8) 0 400◦ C 0.0 0.2 0.4 0.6 0.8 1.0 Si(OH)4 −→ SiO2 + 2H2 O (9) P/Po 3.4.3. Reduction stage Fig. 7. N2 adsorption–desorption isotherms and pore size distribution (inset) of 700◦ C/4 h samples calcined at 400 ◦ C for 3 h (a) SG-EGO25, (b) SGO25 and (c) SG-GMO25. Co3 O4 + 4H2 −→ 3Co + 4H2 O (10)  O.A. Fouad et al. / Materials Chemistry and Physics 128 (2011) 70–76 75 Fig. 8. SEM and TEM images of sol–gel samples containing 25 wt.% Co dried at 80 ◦ C for 24 h, calcined at 400 ◦ C for 3 h and reduced by H2 at 700 ◦ C for 4 h. (a) SEM of SGO25 (b) SEM of SGR25, (c) TEM of SGO25, (d) TEM of SGR25 and (e) particle size distribution for SGO25 sample. 76 O.A. Fouad et al. / Materials Chemistry and Physics 128 (2011) 70–76 100................................... by hydrogen reduction of the synthesized gel precursors. The............... WL=7.96 % results showed that the samples prepared by co-gelation of cobalt Derivative Weight (% ºC ) -2.4.................... and silica precursors in the same solution with cobalt content not WL=8.77 % -2.0............... 90 more than 50 wt.% has the best physical characteristics. It mainly Weight (%)..................... -1.6 gives single Co3 O4 phase upon calcination and a single Co phase.......... 80 WL=5.09 % upon reduction. Addition of EG as a protective agent, enhances.............................. TGA -1.2 the formation of regular pores upon calcinations and reduction. -0.8 Well dispersed cobalt nanoparticles in silica matrix with high ther- 70 mal stability up to 400 ◦ C could be obtained by the proposed DTG -0.4 route. -1 60 0.0 100 200 300 400 500 600 References Temperature (ºC) X.J. Yin, K. Peng, A.P. Hu, L.P. Zhou, J.H. Chen, Y.W. Du, J. Alloys Compd. 479 Fig. 9. TGA and DTG curves for SGP50 sample dried at 80 ◦ C for 24 h. (2009) 372. O.A. Fouad, Int. J. Nanosci. 7 (2008) 1. N.J. Tang, W. Zhong, X.L. Wu, H.Y. Jiang, W. Liu, Y.W. Du, Mater. Lett. 59 (2005) 1723. 112 V.G. Kravets, L.V. Poperenko, Opt. Spektrosk. 104 (2008) 677. 110 D. Song, J. Li, J. Mol. Catal. A: Chem. 247 (2006) 206. L.B. Backman, A. Rautiainen, A.O.I. Krause, M. Lindblad, Catal. Today 43 (1998) Weight (%) 108 11. L.S. Sales, P.A. Robles-Dutenhefner, D.L. Nunes, N.D.S. Mohallem, E.V. 106 Gusevskaya, E.M.B. Sousa, Mater. Charact. 50 (2003) 95. 104 M. Lakshmi Kantam, B. Purna Chandra Rao, R. Sudarshan Reddy, N.S. Sekhar, B. Sreedhar, B.M. Choudary, J. Mol. Catal. A: Chem. 272 (2007) 1. 102 T.A. Kainulainen, M.K. Niemela, A.O.I. Krause, Catal. Lett. 53 (1998) 97. Q.A. Pankhurst, J. Connolly, S.K. Jones, J. Dobson, J. Phys. D: Appl. Phys. 36 (2003) 100 167. 98 K.M.S. Khalil, S.A. Makhlouf, Sens. Actuators A148 (2008) 39. S.A. Makhlouf, K.M.S. Khalil, Solid State Ionics 164 (2003) 97. 100 200 300 400 500 600 N. Koshizaki, K. Yasumoto, T. Sasaki, Nanostruct. Mater. 12 (1999) 971. Temperature (ºC) K.H. Kim, H.C. Park, S.D. Lee, W.J. Hwa, S.S. Hong, G.D. Lee, S.S. Park, Mater. Chem. Phys. 92 (2005) 234. M.M. Rashad, O.A. Fouad, Mater. Chem. Phys. 94 (2005) 365. Fig. 10. TGA curve of SG-GMR25 sample dried at 80 ◦ C for 24 h, calcined at 400 ◦ C K.S. Abdel-Halim, M. Bahgat, O.A. Fouad, Mater. Sci. Technol. 22 (2006) for 4 h and reduced by H2 at 700 ◦ C for 4 h. 1396. H.T. Zhang, G. Wu, X.H. Chen, X.G. Qiu, Mater. Res. Bull. 41 (2006) 495. Under certain experimental conditions the produced cobalt J.W. Park, E.H. Chae, S.H. Kim, J.H. Lee, J.W. Kim, S.M. Yoon, J.Y. Choi, Mater. Chem. Phys. 97 (2006) 371. oxide will react with silica to form cobalt silicate phase according D.E. Zhang, X.M. Ni, H.G. Zheng, Y. Li, X.J. Zhang, Z.P. Yang, Mater. Lett. 59 (2005) to Eq. (11): 2011. V. Musat, E. Fortunato, A.M. Botelho do Rego, R. Monteiro, Thin Solid Films 516 400◦ C CoO + SiO2 −→ CoSiO3 (11) (2008) 1499. S. Esposito, M. Turco, G. Ramis, G. Bagnasco, P. Pernice, C. Pagliuca, M. Bevilac- Thermal stability of a reduced sample (SG-GMR25) has been veri- qua, A. Aronne, J. Solid State Chem. 180 (2007) 3341. J.-S. Girardon, A. Constant-Griboval, L. Gengembre, P.A. Chernavskii, A.Y. Kho- fied up to 400 ◦ C without significant weight gain due to oxidation dakov, Catal. Today 106 (2005) 161. of metallic cobalt as can be observed from the TGA curve of Fig. 10. S.A. Makhlouf, M.A. Kassem, M.A. Abdel-Rahim, J. Mater. Sci. 44 (2009) As the temperature is increased from 400 to 600 ◦ C, a gradual 3438. S.S. Kalyan Kamala, P.K. Sahooa, M. Premkumara, N.V. Rama Raoa, T. Jagadeesh increase in the sample weight up to 12% is observed. This value is in Kumara, B. Sreedharb, A.K. Singha, S. Ramc, K. Chandra, Sekhara, J. Alloys agreement with the theoretical value (12.7%) of weight gain due to Compd. 474 (2009) 214. complete oxidation of metallic cobalt in the sample to form cobalt G. Ortega-Zarzosa, C. Araujo-Andrade, M.E. Compeaı̌ N-Jasso, J.R. Martiı̌Nez, J. Sol–Gel Sci. Technol. 24 (2002) 23. oxide (CoO) phase. These results indicate that the metallic cobalt K.S.W. Sing, D.H. Everett, R.A.W. Haul, L. Moscou, R.A. Pierotti, J. Rouquerol, T. in the Co/SiO2 nanocomposite is thermally stable up to 400 ◦ C. It Siemieniewska, Pure Appl. Chem. 57 (1985) 603. also gives Co/SiO2 high priority to be used as a catalyst in many H. Pröbstle, M. Wiener, J. Fricke, J. Porous Mater. 10 (2003) 213. J-Yves Piquemal, C. Potvin, J-Marie Manoli, G. Djéga-Mariadassou, Catal. Lett. reactions occurring at high temperatures [6,9,30,31]. 92 (189) (2004). R. Takahashi, S. Sato, T. Sodesawa, M. Suzuki, N. Ichikuni, Micropor. Mesopor. 4. Conclusion Mater. 66 (2003) 197. A. Szegedi, M. Popova, V. Mavrodinova, C. Minchev, Appl. Catal. A 338 (2008) 44. In conclusion, Co/SiO2 mesoporous nanocomposites have been V.A. de la Pena O’Shea, J.M. Campos-Martin, J.L.G. Fierro, Catal. Commun. 5 successfully synthesized through thermal decomposition followed (2004) 635.