Synthesis, Characterization, and Adsorption Studies of

---

1352

Langmuir 2002, 18, 1352-1359

Synthesis, Characterization, and Adsorption Studies of Nanocrystalline Copper Oxide and Nickel Oxide Corrie L. Carnes, Jennifer Stipp, and Kenneth J. Klabunde* Department of Chemistry, Kansas State University, Manhattan, Kansas 66506

John Bonevich National Institute of Standards and Technology, Metallurgy, Stop 8554, 100 Bureau Drive, Gaithersburg, Maryland 20899 Received May 10, 2001. In Final Form: September 10, 2001

Nanocrystals of CuO and NiO have been produced by an alkoxide-based synthesis involving the corresponding metal chlorides, ethanol, and water. The resulting oxides are in the form of powders, with the CuO having crystallites in the size range of 7-9 nm and the NiO having crystallites in the size range of 3-5 nm. These crystallites aggregate together to form larger spherical particles, which have been studied by transmission electron microscopy and Brunauer-Emmet-Teller methods and were found to contain many pores and tunnels. It is because of this that an uncharacteristically high surface area is found, averaging about 135 m2/g for CuO and 375 m2/g for NiO. As seen with other metal oxides, once they are made as nanoparticles their reactivity is greatly enhanced. This is thought to be due to morphological differences, whereas larger crystallites have only a small percentage of reactive sites on the surface, smaller crystallites will possess much higher surface concentration of such sites. Elemental analysis, X-ray diffraction, and infrared spectroscopy have been used to characterize this nanoparticles, and reactions with CCl4, SO2, and paraoxon have demonstrated significantly enhanced reactivity and/or capacity compared with common commercial forms of the oxide powders.

I. Introduction Copper oxide nanoparticles have been of considerable interest due to the role of CuO in catalysis, in metallurgy, and in high-temperature superconductors. 1-4 Various preparation methods to ultrafine CuO have been reported. In 1993 Hai-Yan, Yu-Ling, Jing-Kui, and Si-Shen published a preparative method for ultrafine CuO particles utilizing deposition of an aqueous CuO sol onto a SrTiO3 substrate.5 Once annealed, this produced a CuO-coated substrate that acted as a YBCO superconducting thin film. In 1997 Kakihata, Usami, Yamamoto and Shibata published a route to CuO powder from copper chloride by a precipitation-stripping method.6 The average particle size for the resultant CuO was found by laser light scattering to be about 600 nm. In 1998 Dianzeng, Jianqun, and Xi reported on a solid-state reaction between copper chloride and sodium hydroxide to prepare copper oxide.7 The CuCl2‚ 2H2O was directly mixed with the NaOH and ground together and, after washing and drying, yielded a crystallite size of about 23 nm. Nickel oxide nanoparticles have been widely studied for many years because of their useful electronic and magnetic properties.8-10 NiO has also become very important due to its role in catalysis and as a p-type (1) Larsson, P.; Andersson, A. J. Catal. 1998, 179, 72. (2) Chika´n, V.; Molna´r, A Ä .; Bala´zsik, K. J. Catal. 1999, 184, 134. (3) Raveau, B.; Michel, C.; Herview, M.; Groult, D. Crystal Chemistry of High-Tc Superconducting Copper Oxides; Springer-Verlag: Berlin, 1991. (4) Poole, C. P.; Datta, T.; Farach, H. A.; Rigney, M. M.; Sanders, C. R. Copper Oxide Superconductors; John Wiley & Sons: New York, 1988. (5) Hai-Yan, D.; Yu-Ling, Z.; Jing-Kui, L.; Si-Shen, X. J. Mater. Sci. 1993, 28, 5176-5178. (6) Kakihata, T.; Usami, K.; Yamamoto, H.; Shibata, J. Technol. Rep. Kansai Univ. 1998, 40, 67-79. (7) Dianzeng, J.; Jianqun, Y.; Xi, X. Chin. Sci. Bull. 1998, 43, 7, 571-573.

semiconductor.11-14 Many preparation methods have been known and used for many years. In 1965 Yao published data on microcrystals of NiO.15 This was prepared by mixing NiO powder with Na4B2O7 in a platinum crucible and heating at 1130 °C. Later in 1976 Cronan, Micale, Topic, Leidheiser, and Zettlemoyer reported on high surface area NiO prepared by many methods.16 One such method was by precipitation of nickel hydroxide from a nickel chloride solution by adding sodium hydroxide. In 1986 Fernandez and co-workers used an aqueous Ni(NO3)2 sol to form Ni(OH)2 and then heat treating to form NiO.17 More recently in 1993 Hooker and Klabunde published data where metal vapor synthesis (MVS) was used to prepare nanoscale NiO.18 The surface area was found to be 230 m2/g, and crystallite size was found to be 36 nm. This paper describes improved synthesis and isolation of multigram amounts of pure nanocrystalline CuO and NiO using modified sol-gel approaches. Properties of these materials, including chemical reactivities, are described herein. (8) Samsonov, G. V. The Oxide Handbook; IFI/Plenum: New York, 1973. (9) Xuping, Z.; Guoping, C. Thin Solid Films 1997, 298, 53. (10) Richardson, J. T.; Milligan, W. O. Phys. Rev. 1956, 102, 1289. (11) Richardson, J. T. J. Catal. 1966, 6, 328. (12) Fievet, F.; Figlarz, M. J. Catal. 1975, 39, 350. (13) Hannay, N. B. Semiconductors; Reinhold Publishing Corp.: New York, 1959. (14) Smith, R. A. Semiconductors; Cambridge University Press: London, 1978. (15) Yao, Y. J. Phys. Chem. 1965, 69, 3930. (16) Cronan, C. L.; Micale, F. J.; Topic, M.; Leidheiser, H.; Zettlemoyer, A. C. J. Colloid Interface Sci. 1976, 55, 546-557. (17) Fernandez Rodriguez, J. M.; Morales, J.; Tirado, J. L. J. Mater. Sci. 1986, 21 (10), 3668-3672. (18) Hooker, P. D.; Klabunde, K. J. Chem. Mater. 1993, 5, 10891093.

10.1021/la010701p CCC: $22.00 © 2002 American Chemical Society Published on Web 01/15/2002

Nanocrystalline CuO and NiO

II. Experimental Section

Langmuir, Vol. 18, No. 4, 2002 1353

The preparation consists of two main steps: 1. Synthesis of the Copper Hydroxide Powder. The chemicals used in the synthesis were directly from the company without further purification. Under argon 1.500 g (0.0112 mol) of copper(II) chloride (Aldrich) was added to a 250 mL roundbottom flask. This was dissolved with 70 mL of absolute ethanol (McCormick) to form a clear green solution. Then 0.0224 mol of sodium hydroxide (Fisher) was dissolved in absolute ethanol (McCormick) and was added dropwise to form the copper hydroxide gel. The reaction was then stirred at room temperature for 2 h. During this time the reaction mixture forms a blue-green gel. After the reaction was complete, the solution was filtered and washed with water to remove the sodium chloride. The copper hydroxide was then air-dried on the frit, to give a 90% yield. 2. Conversion of Copper Hydroxide to Copper Oxide. Data from thermal gravimetric analysis (TGA) confirmed the copper hydroxide to copper oxide conversion occurs between 190 and 220 °C. The dry copper hydroxide powder was then placed into a Schlenk tube, connected to a flow of argon and surrounded by a furnace, and was heated at 250 °C for 15 min, and after heat treatment the copper oxide powder is black. Commercial copper oxide was purchased from Aldrich (CMCuO). B. Preparation of Nanocrystalline (NC) NiO. The reactions involved in the preparation are shown below.

Quantachrome NOVA 1200 instrumentation. The samples were first outgassed at the desired temperature and then allowed to cool to room temperature. Next they were further cooled to 77 K and exposed to nitrogen (30% N2, 70% He) where the adsorption of nitrogen molecules occurs. Here the amount of nitrogen adsorbed as a monolayer was measured. From the number of molecules adsorbed and the area occupied by each, the surface area was directly calculated. (3) Powder X-ray Diffraction (XRD). For XRD studies, the metal oxide samples were heat treated under argon, directly before being placed onto the sample holder. The instrument used was a Scintag XDS 2000 spectrometer. Cu KR radiation was the light source used with applied voltage of 40 kV and current of 40 mA. Two θ angles ranged from 20° to 85° with a speed of 2°/min. The crystallite size was then calculated from the XRD spectra using the Scherrer equation. (4) Infrared Spectroscopy (FT-IR). FT-IR was used to observe solvent removal during the heat-treatment process. These experiments were conducted on an RS-1 FTIR spectrometer from Mattson with a liquid nitrogen cooled MCT detector. Heat-treated samples were made into KBr pellets and studied. (5) Thermogravimetric Analysis (TGA). TGA was used to determine the conversion of the metal hydroxide to the metal oxide during heat treatment. These studies were conducted under a nitrogen flow. To measure the weight loss, the samples were placed onto a basket and heated at a rate of 10°/min from room temperature to 700 °C. The instrument used was a thermogravimetric analyzer TGA-50 from the Shimadzu Company. (6) Elemental Analysis. Oxide samples after heat treatment was transferred to glass vials, under argon atmosphere, and sent to Galbraith Laboratories for analysis. Elemental analysis was conducted for the metal, C, and H. The amount of oxygen was obtained by subtracting the sum of the metal, C, and H from 100. D. Adsorption Studies. 1. Reaction of CuO, and NiO with CCl4.

NiCl2‚6H2O + 2NaOH f Ni(OH)2 + 2NaCl + 6H2O

2CuO + CCl4 f CO2 + 2CuCl2

Ni(OH)2 f NiO + H2O

2NiO + CCl4 f CO2 + 2NiCl2

A. Preparation of Nanocrystalline (NC) CuO. The reactions involved in the preparation are shown below.

CuCl2 + 2NaOH f Cu(OH)2 + 2NaCl Cu(OH)2 f CuO + H2O

The preparation consists of two main steps: 1. Synthesis of the Nickel Hydroxide Powder. The chemicals used in the synthesis were directly from the company without further purification. Under argon 1.500 g (0.0063 mol) of nickel(II) chloride hexahydrate (Aldrich) was added to a 250 mL round-bottom flask. This was dissolved with 70 mL of absolute ethanol (McCormick) to form a clear green solution. A solution of 0.0126 mol of sodium hydroxide (Fisher) dissolved in 100 mL of absolute ethanol (McCormick) and was added dropwise to form the nickel hydroxide gel. The reaction was then stirred at room temperature for 2 h. During this time the reaction mixture forms a green gel. After the reaction was complete, the solution was filtered and washed with water to remove the sodium chloride. The nickel hydroxide was then air-dried on the frit, to give a 92% yield. 2. Conversion of Nickel Hydroxide to Nickel Oxide. Data from TGA confirmed the nickel hydroxide to nickel oxide conversion occurs between 230 and 250 °C. The dry nickel hydroxide powder was then placed into a Schlenk tube, connected to a flow of argon, and it was heated at 250 °C for 15 min. After the heat treatment was complete, it was allowed to cool to room temperature, yielding dark green nickel oxide. Commercial nickel oxide was purchased from Aldrich (CMNiO). C. Characterization. (1) Transmission Electron Microscopy (TEM). Medium-resolution TEM studies were carried out by adding dry ethanol to the heat-treated metal oxide and sonicating this slurry. A drop of this slurry was then placed onto a carbon-coated copper grid. TEM experiments were performed by Dan Boyle at the KSU Microscopy and Image Processing Facility using a Philips 201 TEM and by Ryan Richards in the chemistry department using a Philips CM12 TEM. HRTEM was investigated by John Bonevich at NIST. (2) Brunauer-Emmet-Teller (BET). Surface area measurements were done by using BET methods. These were conducted using both Micromeritics Flowsorb II 2300 and

The reaction between the metal oxides and CCl4 was carried out to characterize the chemical reactivity of the metal oxides. These reactions were conducted in a U-tube that was connected to a gas chromatograph (GOW-MAC gas chromatograph series 580).19 The U-tube was made of Pyrex and connected between the injector port and the column (Alltech Chromosorb W-HP). An oxide sample (0.100 g) was placed in the U-tube between two small plugs of quartz wool. The most favorable temperatures for the reaction were found to be 250 °C (CuO) and 300 °C (NiO), and the U-tube was heated to the respective temperature. The injector port was kept at 100 °C. Injections of 2 µL of CCl4 were made every 7 min. Any CO2 coming off the sample, or CCl4 that was not destroyed, was then sent via helium (20 cm3/min) through the column (90 °C) to be separated. They were then detected by a thermal conductivity detector (120 °C), and peak areas were recorded. Injections of CCl4 were made until the oxide bed had been exhausted. 2. Sulfur Dioxide Adsorption. A quartz spring balance was used to measure the adsorption of SO2 onto CuO and NiO. The apparatus consists of a basket, which is used to hold the sample; this is attached to a quartz spring.20 The basket and the spring are closed within the vacuum line. The SO2 gas tank is also attached to the vacuum line. As the SO2 adsorbs onto the metal oxide in the basket, the weight change causes the spring to move, and this movement is noted by the telescope. Once the telescope is calibrated, it is accurate to (0.1 mg. Due to the electrostatic properties of the fine powders, it was found to be much easier and more accurate to work with granules. Using the granules helped to eliminate losses during transfers, (19) Koper, O. B.; Lagadic, I.; Volodin, A.; Klabunde, K. J. Chem. Mater. 1997, 9, 2468. (20) (a) Stark, J. V.; Park, D. G.; Lagadic, I.; Klabunde, K. J. Chem. Mater. 1996, 8, 1904. (b) Klabunde, K. J.; Stark, J.; Koper, O.; Mohs, C.; Park, D. G.; Decker, S.; Jiang, Y.; Lagadic, I.; Zhang, D. J. Phys. Chem. 1996, 100, 12142.

1354

Langmuir, Vol. 18, No. 4, 2002

weighing, and dynamic vacuum. Therefore, the samples were first pressed into pellets at 1000 pounds load and then crushed and sifted through a mesh to achieve granules of uniform size of 0.250-1.17 mm. By use of a relatively low pelletization pressure, only a small change in surface area resulted. Granules (100 mg) of oxides were placed into the basket on the spring balance. The samples were placed under dynamic vacuum for 1 h at room temperature. The SO2 gas was allowed to fill the vacuum line, including the spring balance, to the desired pressure. The spring position was noted over the next hour. This was followed by 100 min of evacuation to remove all physisorbed species. After the evacuation, the spring position was noted again indicating the presence of remaining strongly chemisorbed species. 3. Destructive Adsorption of Diethyl 4-Nitrophenyl Phosphate. The destructive adsorption of diethyl 4-nitrophenyl phosphate (DNPP, also called paraoxon) was carried out to determine the capacity of the oxide to dissociatively chemisorb a polar organophosphorus compound that is considered a chemical warfare mimic. A 0.100 g sample of nanoparticles was placed into a 250 mL round-bottom flask that had been flushed with argon, and 100 mL of dry pentane was then added to the flask, and stirring commenced. Then 8 µL of paraoxon was added to the flask, and ultraviolet/visible spectroscopy (SIM Aminco Milton Roy 3000 array) was used to monitor the disappearance of paraoxon at 270 nm, by extracting samples at desired intervals. This reaction was monitored every 20 min for 3 h, and then at 20 h. The powder was then filtered, and FTIR was then used to detect adsorbed species on the solid. The used solid was also washed with 10-mL portions of CH2Cl2, IR spectra of the extracted CH2Cl2 showed that no adsorbed species were removed.

III. Results A. Preparation of the Copper Oxide. Several experiments were conducted varying the starting materials, solvents, stirring time, and drying methods, and all were found to have an effect on the surface area of the resulting sample. The best results were obtained by using anhydrous copper chloride as the starting material, rather than copper oxide, copper metal, copper methoxide, copper ethoxide, or copper sulfate. These data support earlier findings that using water-based solvents leads to low surface areas. Time was also found to be a factor in the surface area: enough had to be allowed for hydrolysis, but too much resulted in lowering the surface areas. Four different methods were studied for removing the solvent from the copper hydroxide powder: oven drying in air; hypercritical drying using an autoclave; vacuum stripping; and filtration with washing. The autoclave approach was not successful since the hot alcohol solvents caused the reduction of Cu2+ to Cu0. The most successful method proved to be filtration with washing. B. Activation of the Copper Oxide. Copper oxide was activated (heat treated) under argon flow or under dynamic vacuum. The surface areas did not vary much with either method used. During activation the surface area increases, then goes through a maximum, and then decreases. This decrease at temperatures above 250 °C is due to sintering. C. Copper Oxide Characterization. By careful characterization of the CuO samples, it became clear that the NC-CuO samples had morphology different from that of the commercial (CM) CuO samples. (1) Brunauer-Emmet-Teller Method (BET). Commercial CuO is most commonly prepared by high-temperature methods, and CM-CuO typically had surface areas within the range of 0.40-0.80 m2/g. Our NC-CuO samples typically possessed surface areas within the range of 120-140 m2/g after heat treatment at 250 °C. When heated at higher temperatures, the crystallites began to

Carnes et al. Table 1. Surface Area, Pore Volume, and Diameter of CM-CuO, and NC-CuO Powder after Heat Treatment at 250 °C sample

surface area (m2/g)

av pore vol (cm/g)

av pore diameter (nm)

CM-CuO NC-CuO

0.53 136

0.0041 0.22

31 8.4

Table 2. Pressure and Resulting Pore Shape for NC-CuO pressure (PL)

pore shape

0 2000 5000 10 000 20 000

cylindrical pores open at both ends slit-shaped pores, the space between parallel plates slit-shaped pores, the space between parallel plates slit-shaped pores, the space between parallel plates slit-shaped pores, the space between parallel plates

Table 3. Elemental Analysis for NC-CuO Heat Treated at 250 °C element

% calcd

% exptl

copper oxygen carbon hydrogen chlorine

80 21 0 0 0

78 20