Advanced
Dissipative Particle Dynamics Simulation on the Formation Process of CeO<sub>2</sub> Nanoparticles in Alcohol Aqueous Solutions
Dissipative Particle Dynamics Simulation on the Formation Process of CeO2 Nanoparticles in Alcohol Aqueous Solutions
Journal of the Korean Chemical Society. 2012. Aug, 56(4): 431-439
Copyright © 2012, The Korean Chemical Society
  • Received : June 04, 2012
  • Accepted : July 12, 2012
  • Published : August 20, 2012
Download
PDF
e-PUB
PubReader
PPT
Export by style
Article
Author
Metrics
Cited by
TagCloud
About the Authors
Qi Zhang
Jing Zhong
zjwyz@cczu.edu.cn
Bao-Zhu Yang
Wei-Qiu Huang
Ruo-Yu Chen
Jun-Min Liao
Department of Chemistry, National Sun Yat-sen University, Kaohsiung, 80424, Taiwan, R.O.C.
Chi-Ruei Gu
Department of Chemistry, National Sun Yat-sen University, Kaohsiung, 80424, Taiwan, R.O.C.
Cheng-Lung Chen
Department of Chemistry, National Sun Yat-sen University, Kaohsiung, 80424, Taiwan, R.O.C.
chen1@mail.nsysu.edu.tw

Abstract
Dissipative particle dynamics (DPD) was carried out to study the nucleation and crystal growth process of CeO 2 nanoparticles in different alcohol aqueous solutions. The results showed that the nucleation and crystal growth process of CeO 2 can be classified into three stages: nuclei growth, crystal stabilization and crystal aggregation except the initial induction stage, which could be reproduced by collecting simulation results after different simulation time. Properly selecting the sizes of CeO 2 and water bead was crucial in the simulation system. The influence of alcohol type and content in solutions, and precipitation temperature on the particle dimension were investigated in detail and compared with the experimental results. The consistency between simulation results and experimental data verify that the simulation can reproduce the macroscopic particle aggregation process. The effect of solvent on the nucleation and crystal growth of CeO 2 nanoparticles are different at three stages and can not be simply described by Derjaguin-Landau-Verwey-Overbeek (DLVO) theory or nucleation thermodynamics theory. Our work demonstrated that DPD methods can be applied to study nanoparticle forming process.
Keywords
INTRODUCTION
In recent years, nanometer-sized particle of cerium dioxide (CeO 2 ) has been widely used in many fields, such as gas sensor, 1 electrode materials for solid oxide fuel cells, 2 catalytic supports for automobile exhaust system, 3 UV absorbent and filters 4 and abrasives of the chemical mechanical polishing slurry, 5 since CeO 2 particle has high stability at high temperature, high hardness and reactivity. 6
Over the past decade, several techniques have been developed to produce the CeO 2 particles, such as: sol-gel process, 7 hydrothermal synthesis, 8 forced hydrolysis, 9 microemulsion, 10 electrochemical synthesis 11 and precipitation. 12 - 20 The liquid-phase precipitation method is more attractive among these techniques because the low cost salt precursors, simple operation and ease of mass production. 17
In general, the liquid-phase precipitation process includes three stages: chemical reaction, nucleation, and crystal growth. 13 These stages are fast in most cases, but difficult to control and observe in experiments. The dimension of CeO 2 particles formed in water by liquid-phase precipitation is in the range of 10-15 nm from XRD measurement. 17 Some researchers reported that adding low dielectric medium, such as alcohols, to the aqueous solution could alter the thermodynamics of reaction system and nucleation kinetics in nanoparticle forming process, which lead the formation of particle with smaller size. 16 , 18 , 19 , 21 Chen 16 and Li et al. 18 , 19 have reported that the particle size of CeO 2 decreased to 8-10 nm by introducing the monohydric alcohols - methanol (MeOH), ethanol (EtOH), iso-propanol ( i -PrOH), tert -butanol ( t -BuOH), into the aqueous solutions. The decrease of particle size varied with the type and quantity of alcohols adding into the aqueous solutions, and can be explained by the Derjaguin-Landau-Verwey-Overbeek (DLVO) theory and nucleation thermodynamics theory for crystal.
Chen et al. 16 found that the particle size of CeO 2 decreases with the increasing of alcohol content in solutions, and the size of particle formed in 67 vol.% alcohol aqueous solutions at 50 ℃ decrease as MeOH aq > EtOH aq > i -PrOH aq ≈ t -BuOH aq. The order is directly proportional to the dielectric constant, ε , of solvent. In Li’s work, 18 , 19 they found the particle size of samples obtained in 75 vol.% alcohol aqueous solutions at 75 ℃ decreases in the order: i -PrOH aq > EtOH aq > MeOH aq > t -BuOH aq. The order, however, is not proportional to the ε value of solvent. They also found there has a minimum particle size with the increasing of alcohol content in solutions. Their experimental results can not be explained simply by DLVO theory or nucleation thermodynamics theory for crystal. In addition to this, these two experiments are not consistent, which perhaps due to the different initial concentration of reactants and reaction time. It is difficult to measure the forming kinetics for CeO 2 nanoparticle because the formation process is too fast.
In addition to the experimental efforts, computer simulation is a possible way to study the phase separation and particle aggregation. Conesa investigated the structures and relative stabilities of several CeO 2 surfaces and of anion vacancy centers using theoretical molecular mechanics (MM) method. The simulations reproduce the experimental bulk modulus of CeO 2 within ~10% error. 22 Gotte et al. simulated the bulk and (011) surfaces of reduced and unreduced CeO 2 by MD method, and concluded that the main structural change occurring on ceria reduction are the new equilibrium positions for the oxygen ions. 23 Baudin et al. studied the dynamics, structure and energetics of the (111), (011) and (001) surface of CeO 2 by molecular dynamics (MD) simulation method within NPT ensemble. 24 They reported the structure and defects of ceria without considering the forming process of CeO 2 nanoparticle.
For the nanoparticle forming process with solvents and metal-oxide molecules, the DPD method is especially appropriate because it is capable of studying systems containing millions of atoms in the nano-second time domain, and can describe adequately the phase separation process. 25 - 28 The method was introduced by Hoogerbrugge and Koelman in 1992. 29 In our previous work, DPD have been used to simulate phase behaviors of EO 20 PO 70 EO 20 (P123) in water, ethanol/water solutions and silica sol. 30 The results indicated that DPD simulation method can be used to study the influence of template on the structure of mesoporous material and guide the preparation of mesoporous material. We also applied DPD to investigate the self-assembled aggregation problems of surfactants and a gold nanoparticle system, 30 , 31 and found that the major difficulty to reproduce macroscopic experimental phase behavior by the simulation lies on the choosing of bead size and interaction parameters. Although Groot and Warren discussed the DPD method in detail and gave suggestions on the selection of DPD parameters, 32 some fundamental problems still remained unclear. No consistent protocol has been reported to determine these two important parameters.
In this work, the growth of CeO 2 nanoparticle based on the precipitation process was simulated. The nucleation and crystal growth of the particles were investigated by using DPD simulation. The way to define DPD beads were discussed in the simulation section. To be consistent with the experimental works, various alcohol aqueous solutions: MeOH aq, EtOH aq, i -PrOH aq, t -BuOH aq, were used in the simulation. The effects of alcohol type, concentration, and precipitation temperature on the particle dimension were investigated. Our simulation provides a possible tool for the investigation of the nucleation and crystal growth of nanoparticles.
SIMULATION METHOD
- Crystal Growth Theory
The growth of crystal is always analyzed by DLVO theory and nucleation thermodynamics theory. Based on the DLVO theory, the interaction potential between particles can be described as, 19
PPT Slide
Lager Image
where, A is Hamker constant; κ is Debye-Hückel parameter; d is the diameter of the particle; ψ is surface potential; ε is dielectric constant of solvents; and V is interactive potential between particles respectively. The first and second term in the equation represent electro-attractive and repulsive interactions respectively. According to this equation, the electrostatic repulsion force between particles would decrease with the decreasing of the ε values of solvents, which leads the particles easy to aggregate and growth. Therefore, within certain concentration range and at certain precipitation temperature, the particle dimension is inversely proportional to the ε values of solvents if the electrostatic repulsion force between particles dominated in the nanoparticle forming process.
On the other hand, the forming of particles strongly depends on the supersaturation of solute according to the nucleation thermodynamic theory, which can be expressed by Kelvin equation, 33
PPT Slide
Lager Image
where S is supersaturation of solute; m is the weight of solute molecule; γ is the interfacial energy between solute and solution phase; ρ is the density of solute; k is the Boltzmann’s constant; and T is the Kelvin temperature respectively. Conventionally, S is defined as the ratio of solute concentration ( C ) to the equilibrium concentration of solute ( Cl ) in the saturated solution.
PPT Slide
Lager Image
According to Debye-Hückle theory, Cl can be expressed as: 16
PPT Slide
Lager Image
where ε 0 is the permittivity in vacuum and ε is the dielectric constant in a given solution; r + and r represent the radius of ions charged z + and z respectively; and e represents the elementary charge ( e = 1.602×10 -19 C).
In combination of equations (2), (3) and (4) yields,
PPT Slide
Lager Image
From this equation, the particle size was directly proportional to the ε value of solvents if the supersaturation of solute controlled the nanoparticle forming process.
Based on the above analysis, a change in ε value of solvent can alter the electrostatic repulsion forces between particles and the supersaturation of solute in the opposite direction.
- DPD Theory
The outline of the DPD model here is based on the work by Groot and Warren. 32 Molecules are divided into beads, and the motions of beads are governed by Newton’s equations:
PPT Slide
Lager Image
PPT Slide
Lager Image
where
PPT Slide
Lager Image
are the position vector, velocity and total force on the i th bead respectively. The total force exerted on bead i contains three parts.
PPT Slide
Lager Image
where FCij is the conservative force which describe soft repulsive central force between bead i and j ; FDij and FRij are the dissipative and random forces respectively. The significance of these terms is investigated in detail elsewhere. 29 , 32
To simulate a system, a set of interacting parameters αij ’s between different types of beads must be determined. The DPD simulation should reproduce correct phase behaviors as observed experimentally with these pre-determined parameters. There are several methods suggested in the literature to evaluate interaction parameters. 34 We adopted Monte Carlo method to evaluate the interaction parameters by compromising the accuracy and computation time. The derivation of αAB is based on Flory-Huggins mixing parameter χAB where the subscript AB denotes type A and B beads, 35 which is related to the mixing energy by
PPT Slide
Lager Image
The maximum repulsion αAB deduced from this χAB parameter at a given particle density ρ = N/V , as suggested by Groot and Warren. 32
PPT Slide
Lager Image
PPT Slide
Lager Image
The αAA term is derived from the compressibility of pure component A ( αAA = 75 kBT/ρ ). 32 All simulations were performed with Cerius2 and Material Studio software packages. 36
- Determination of Simulation Parameters
To carry out DPD simulation, molecular fragments are usually treated as beads. There is no definitive way to divide the chain molecule into fragments. In the first attempt, the molecules of MeOH, EtOH, i -PrOH, t -BuOH, EG, water and CeO 2 are directly used as beads for DPD simulation. The structures of these molecules were optimized with UNIVERSAL force field (UFF). 37 All atomic partial charges of molecules were obtained from quantum mechanical ZINDO method based on the optimized structures except CeO 2 molecule. The atomic partial charges of CeO 2 molecule was calculated by the Density Functional Theory (DFT) with Gaussian 03 software. 38 The hybrid and correlation functionals B3LYP and SDD potential form was used. The interacting parameter χAB was calculated by UFF. 37 These DPD simulations failed to reproduce aggregation behavior as observed experimentally. This was because the sizes of CeO 2 and water are too much smaller than those of alcohols.
To generate CeO 2 bead with its size comparable to the alcohols, a small CeO 2 cluster with fluorite structure containing two CeO 2 molecules was adopted to represent CeO 2 bead. Similarly, a water cluster with cubic ice crystal structure containing two water molecules was choosing to represent water bead. 1 (a) and 1 (b) shows the calculated interaction parameters αAB ’s of CeO 2 , water bead and various alcohols at 50 and 75 ℃.
PPT Slide
Lager Image
All systems were simulated at the particle density ρ = 5. A 3D box of size 30×30×30 DPD units with periodic boundary conditions was adopted in the simulation. The dissipative parameter γij was set to a value of 4.5 kBT , and the time step Δt was set as 0.05 according to Ref. 35. 100,000 time steps were carried out for each system, and the molar ratio of [CeO 2 ]:[solvent] was kept at 0.20. The initial conformations and velocities were random. After comparing many times of simulations, we found the repeatability was good. Therefore, the parameters adopted in our simulation are reasonable and the conclusions are reliable. In additon, this parameterization method has been also used in wide-range systems on DPD simulation, such as: self-assembled aggregation problems of surfactants, 39 , 40 gold nanoparticle system, 31 etc., and gives good results.
RESULTS AND DISCUSSION
The particle sizes of CeO 2 formed in different alcohol aqueous solutions and at different simulation time were evaluated and compared with experimental results. The effects of volume content of alcohol in solutions and precipitation temperature on the particle sizes were also investigated.
- Growth Process of CeO2in Alcohol Aqueous Solutions
Giving in . 1 are snapshots of the CeO 2 particles formed in MeOH aq solutions at different simulation time. The red bead represents the CeO 2 , and the solvent was neglected in all Figures for clarity. The simulation temperature was set at 50 ℃, and 50 vol.% MeOH aq solution was used. The results show that CeO 2 molecules can form spherical aggregated particles at different simulation time. The number of aggregated particles decreased with the increasing of simulation time, but the dimension of particles increased, which reproduced the aggregation process in the precipitation experiments in alcohol aqueous solutions. The relationships of particle dimension and aggregation number are similar in other alcohol aqueous solutions.
. 2 shows the dimensions of particles formed in different 50 vol.% alcohol aqueous solutions at 50 ℃. The particle size was calculated by averaging the radial length of particle in x, y and z directions. According to the dimension of molecules defined for DPD beads and the particle density ρ = 5 used in the simulation, a DPD unit is about 0.74 nm. Then the particle size can be transferred from DPD unit into “nm” unit.
The process of CeO 2 particles formed in monohydric alcohols (MeOH, EtOH, i -PrOH, t -BuOH) has been simulated with the increasing simulation time. In the first 10,000 time steps, nuclei had been already formed. Between 10,000 time steps and 30,000 time steps, the dimensions of CeO 2 particles grow rapidly. This is because the particles were near to each other. After the fast growth stage, the particles grow gently until 70,000 time steps; since grown particles were already separated in distance from each other. After 70,000 time steps, the dimensions of the particles increased rapidly. This result indicates that the aggregation happened between those large particles.
PPT Slide
Lager Image
Simulation results of CeO2 particles at different simulation time (50 ℃, MeOH/water = 50 vol.%, molar ratio of [CeO2]: [solvent] = 0.20) (a)10,000; (b) 30,000; (c) 50,000; (d) 70,000; (e) 90,000; (f) 100,000.
PPT Slide
Lager Image
Dimension of CeO2 particles formed at different simulation time (50, alcohol/water = 50 vol.%, molar ratio of [CeO2]:[solvent] = 0.20).
Comparing the simulated dimension of the CeO 2 particles with the experimental measured size from XRD measurement, 16 - 19 we found that the sizes of CeO 2 particles formed within 30,000 time steps are very close to the measured sizes of CeO 2 single crystal. 16 , 18 , 19 Therefore the particles formed within 30,000 time steps have not finished the nuclei growth. After this point, further nucleation occurred between particles, and this stage can be classified as nuclei growth stage. The size of particle increased gently when the simulation time extend from within 30,000 to 70,000 time steps, which indicated that the CeO 2 crystals formed in alcohol aqueous solutions were stable in a certain period. This stage can be classified as stabilization stage. The size of particles after 70,000 steps is much larger than that of single crystal from XRD results. This stage can be classified as aggregation stage. Our simulation shows that the growth of CeO 2 particle in the alcohol aqueous solutions undergoes different stages.
- Effect of Alcohol Type in Alcohol Aqueous Solutions
The dimension of particles formed at different simulation time also depends on the type of alcohol as shown in . 2 . The size of CeO 2 particle are in the order t -BuOH(aq) > i -PrOH(aq) > EtOH(aq) > or ≈ MeOH(aq) when the simulation time less than 30,000 time steps (in nuclei growth stage), which is almost inversely proportional to the ε values of alcohol solutions. The trend is quite consistent with the results reported by both Hu et al. 21 and Chen et al. 16 They studied the relationship between incubation time and average hydrodynamic diameter of zirconia particles in different alcohol(aq) solutions at equal concentrations of alcohol and water (50 vol.%), and found that the nucleation rate of zironia particles in different alcohol solutions were in the order i -PrOH(aq) > EtOH(aq) > MeOH(aq) after the initial induction stage. Therefore, we can deduce that the particle growth was controlled by electrostatic repulsion between particles in this stage because the particle dimension was inversely proportional to the ε values of alcohol solutions.
In the range of 30,000 to 70,000 time steps (crystal stabilization stage), we found that the dimension of CeO 2 particles grows gently independent of type of alcohol. This indicated that electrostatic repulsion between particles still dominant the forming of particles in the crystal stabilization stage.
When simulation time is more than 70,000 time steps (crystal aggregation stage), the order of particle size changed abruptly, which decreases as following sequence: MeOH(aq) > i -PrOH(aq)> EtOH(aq) > t -BuOH(aq). This indicates that the mechanism of particle formation changed in this stage. Within this range the order of particle size is almost directly proportional to the ε values of alcohol(aq) solutions except the EtOH(aq), which indicated that the crystal growth of CeO 2 was controlled by the supersaturation of CeO 2 in the solutions according to the nucleation thermodynamic theory. The order of particle size obtained in our simulation is quite consist with the experimental results reported by Chen et al. at the same experimental conditions, 16 but the simulated particle sizes are slightly larger than the reported values, which perhaps due to the vigorous stirring in the experiments inhibiting the aggregation of CeO 2 particles.
The consistency of simulation results with the experimental results in monohydric alcohol(aq) solutions indicates that DPD simulation can describe the sequential forming process of CeO 2 nanoparticle, including nuclei growth, crystal stabilization and crystal aggregation stages. We can also infer that the growth of particle at the nuclei growth and crystal stabilization stages was controlled by electronic repulsion between particles, which can be described by DLVO theory; and the growth of particle at the crystal aggregation stage was controlled by the supersaturation of CeO 2 in the solutions, which can be described by nucleation thermodynamics theory. In other words, the nuclei growth and crystal repairing is faster for molecules in solvents with lower ε values, and the crystal aggregation is slower for particle in solvents with lower ε value. The two mechanisms are effective in the whole particle growth process and compete with each other.
- Effect of Alcohol Content in Alcohol Solutions
The ε value of solvents can be adjusted by changing the content of alcohol in the alcohol(aq) solutions. The addition of alcohol into water would decrease the ε value and the dissolvent ability of solvents. The solution is therefore supersaturated and capable of producing finer particle. 21 On the other hand the electrostatic repulsion between particles would decreases with the decreasing of the ε value, which leads to form larger CeO 2 particles. Chen et al. 16 and Li et al. 18 reported the effect of alcohol/water ratio on the dimension of CeO 2 particle in separated works. Chen et al. found that the dimension of particle decreases with the increasing of alcohol/water ratio. 16 While Li et al. found a minimum value in particle dimension during the increasing of EtOH(aq) ratio. 18
The effect of alcohol/water ratio on the dimension of CeO 2 nanoparticle by DPD at 50 ℃ was studied, and the results are shown in . 3 - 5 . . 3 gives the size of particles formed after 30 000 simulation steps. We found that the effect of alcohol/water ratio on the dimension of CeO 2 particle was varied with the alcohol type used in solutions. The general trend, except EtOH(aq), is the size of CeO 2 particle decreases with the increasing of alcohol/water ratio, but the change details are different for different solvents.
PPT Slide
Lager Image
Dimension of CeO2 particles formed in different alcohol/water mixed solvents (50 ℃, molar ratio of CeO2/solvent = 0.20, 30,000 time steps).
PPT Slide
Lager Image
Dimension of CeO2 particles formed in different alcohol/water mixed solvents (50 ℃, molar ratio of CeO2/solvent = 0.20, 50,000 time steps).
PPT Slide
Lager Image
Dimension of CeO2 particles formed in different alcohol/water mixed solvents (50 ℃, molar ratio of CeO2/solvent = 0.20, 100,000 time steps).
For the particles formed in MeOH(aq) and t -BuOH(aq), the change trends of particle sizes are almost consistent when the alcohol content in solutions raised from 20 vol.% to 50 vol.%, then decreased when the alcohol content larger than 50 vol.% after 30,000 time steps simulation. This indicates that the effect of electrostatic repulsion and that of supersaturation of CeO 2 reach equilibrium when the alcohol content was in the range of 20 vol.% to 50 vol.%, but the effect of supersaturation of CeO 2 become dominant when the alcohol content in solutions larger than 50 vol.%. In i -PrOH(aq), the particle dimension decreases with the increasing of alcohol content in the range of 20% to 50%, and slightly increases in the range of 50 vol.% - 67 vol.%, then drop obviously in the range of 67-80 vol.% after 30,000 time steps simulation. This indicates that the supersaturation of CeO 2 almost dominate the whole process. The effect of electrostatic repulsion was only significant within the range of 50 vol.% to 67vol.% alcohol content. In the EtOH(aq), there existed a minimum point in the curve of particle dimension vs. the alcohol content. Therefore, the supersaturation of CeO 2 control the forming of particles in the initial particle forming stage, then the electrostatic repulsion become dominant when the alcohol content was larger than 67 vol.%. This trend is as same as reported by Li et al. 19 and confirmed that our simulation can reproduce the effect of alcohol content in alcohol(aq) solutions on the particle dimension.
When the simulation time extend to 50,000 time steps, the sizes of particles formed in MeOH(aq) and t -BuOH(aq) become larger when the alcohol content in alcohol(aq) solutions raised from 20 vol.% to 50 vol.% as shown in . 4 . This indicates that the equilibrium between electrostatic repulsion and supersaturation of CeO 2 at 30,000 time steps was shifted and the electrostatic repulsion became dominant. For other alcohols and alcohol content, the relationship between particle size and alcohol content in alcohol(aq) solutions is almost unchanged though the particle size increase when the simulation steps prolonged to 50,000 steps, This indicates that the influence of alcohol content on the particle size is similar for nuclei growth stage and crystal stabilization stage.
The change of particle dimension with alcohol content in alcohol(aq) solutions after 100,000 time steps is different with those after 30,000 time steps and 50,000 time steps as shown in . 5 . There exists the minimum or maximum point for almost all the alcohol(aq) solutions studied in our simulation, which indicates that electrostatic repulsion force and supersaturation of CeO 2 still compete with each other and one of these two factors controls the aggregation.
The effect of alcohol content in alcohol(aq) solutions on the particle dimension shows some irregular phenomena because electrostatic repulsion forces and supersaturation of CeO 2 in solution competed with each other in different alcohol solutions. The trend of nanoparticle in the crystal forming stage can not be predicted by any simple theory. Our DPD simulation, however, can reproduce the experimental results on the dimension of CeO 2 nanoparticle. The method is adequate to investigate such problems.
- Effect of Precipitation Temperature
The increase of temperature would decrease the ε value of alcohol, 21 and increase the supersaturation of CeO 2 . Therefore temperature increasing would decrease the size of the nanoparticle. So the simulation in nuclei growth stage (less than 30,000 time steps) was investigated in this part.
The dimension of CeO 2 particle decreases when the simulation temperature increased from 50 ℃ to 75 ℃ in alcohol(aq) solutions with 50 vol.% alcohol as shown in . 6 , which consisted with the rule deduced from the relationship between supersaturation and temperature. The other interesting phenomenon occurred with the increasing of temperature is that the relationship between particle size and different alcohol solutions. The order of particle size changed from t -BuOH(aq) > i -PrOH(aq) > EtOH(aq) > MeOH(aq) to i -PrOH(aq) > MeOH(aq) > EtOH(aq) ≈ t -BuOH(aq) with the increasing of temperature. This indicated that supersaturation of CeO 2 also became important for particle growth. This phenomenon become obviously at 75 ℃ and alcohol/water = 75 vol%. The dimension of particle shows a maximum value in i -PrOH(aq) at 75 ℃ and i -PrOH(aq) = 75 vol%, which consisted with the experimental results reported by Li et al. 19 at the same experimental conditions. This consistency confirmed that our DPD simulation can correctly reproduced experimental results at different temperature.
PPT Slide
Lager Image
Dimension of CeO2 particles formed at different temperature (molar ratio of CeO2/solvent = 0.20, 30,000 time steps).
The effect of alcohol content in the alcohol(aq) solutions on the particle size is also investigated at 75 ℃. After 30,000 time step, the dimension of particles formed in MeOH(aq) decreases when the alcohol content increased from 50 vol.% to 75 vol%; increases slightly in EtOH(aq), and increases markedly in i -PrOH(aq); but decreases in t -BuOH(aq). The irregular trends is similar with the phenomena observed at 50 ℃, and can be explained by the competition between electrostatic repulsion force and the supersaturation of CeO 2 in solutions which influence the particle dimension contrarily. The detail in mechanism need further study.
CONCLUSION
DPD simulation was carried out to investigate the forming process of CeO 2 nanoparticle in alcohol(aq) solutions. The formation can be divided into nuclei growth, crystal stabilization and crystal aggregation, which can be observed by collecting the simulation results at different simulation time. The nuclei growth stage within the range of less than 30,000 time simuation steps, and the particles grew rapidly in this stage; crystal stabilization stage is within the range of 30,000 and 50,000 simulation time steps, and the size of particle remains unchange in this stage; particle aggregation is within the range after 50,000 simulation time steps, the particles further aggregate and the dimension of particles was larger than that of single crystal.
The dimension of CeO 2 nanoparticle is inversely proportional to the ε value of alcohol(aq) solutions, and electrostatic repulsion force between particles controls the forming process when the CeO 2 nanoparticle is in the nuclei growth stage; and that would change to be directly proportional to the ε value of alcohol(aq) solutions and the supersaturation of CeO 2 in solutions controls the forming process when the CeO 2 nanoparticle has entered into the crystal aggregation stage. The precipitation kinetics is faster for molecules in solvents with lower ε values, and the crystal aggregation kinetics is slower for particle in solvents with lower ε value.
The particle dimension decreases with the increasing of alcohol content in alcohol(aq) solutions and temperature when the particles were in the nuclei growth stage. The effect of alcohol content in alcohol(aq) solutions on the CeO 2 particle dimension is irregular because the two actions − electrostatic repulsion force between particles and the supersaturation of CeO 2 in solutions would compete with each other during the change of alcohol content and influence the particle dimension contrarily.
The simulation results produce from different type and content of alcohol in alcohol(aq) solutions, and different temperatures consist with the experimental data obtained at same conditions. Our work has demonstrated that DPD methods can be applied to the study of nanoparticle forming process.
Acknowledgements
This work is supported by the Project Funded by the Priority Academic Program Development of Jiangsu Higher Education Institutions, the Natural Science Foundation of Jiangsu Province (BK2011240), Major Projects of Natural Science Research of Colleges and Universities in Jiangsu Province (11KJA610002), the Transformation of Scientific and Technological Achievements and Industrialization plan (Innovation Funds) of Changzhou (CC20110003), the Jiangsu College Graduate Research and Innovation Projects (CXZZ11_0374).
References
Izu N. , Shin W. , Murayama N. , Kanzaki S. 2002 Sensors and Actuator B: Chemical 87 95 -    DOI : 10.1016/S0925-4005(02)00224-1
Yahiro H. , Baba Y. , Eguchi K. , Arai H. 1988 J. Electrochem. Soc. 135 2077 -    DOI : 10.1149/1.2096212
Kašpar J. , Fornasiero P. , Hickey N. 2003 Catal. Today 77 419 -    DOI : 10.1016/S0920-5861(02)00384-X
Yamashita M. , Kameyama K. , Yoshida S. , FUjishiro Y. , Kawai T. , Sato T. 2002 J. Mater. Sci. 37 683 -    DOI : 10.1023/A:1013819310041
Jiang M. , Wood N. D. , Komanduri R. 1998 J. Eng. Mater. Technol. (Trans. ASME) 120 304 -    DOI : 10.1115/1.2807019
Zhou X. D. , Huebner W. , Anderson H. U. 2002 Appl. Phys. Lett. 80 3814 -    DOI : 10.1063/1.1481244
Chu X. , Chung W. , Schmidt L. D. 1993 J. Am. Ceram. Soc. 76 2115 -    DOI : 10.1111/j.1151-2916.1993.tb08344.x
Hirano M. , Kato E. 1999 J. Am. Ceram. Soc. 82 786 -    DOI : 10.1111/j.1151-2916.1999.tb01838.x
Hsu W. P. , Ronnquist L. , Matijevic E. 1988 Langumir 4 31 -    DOI : 10.1021/la00079a005
Masui T. , Fujiwara K. , Machida K. I. , Adachi G. Y. , Sakata T. , Mori H. 1997 Chem. Mater. 9 2197 -    DOI : 10.1021/cm970359v
Zhow Y. , Philips R. J. , Switzer J. A. 1995 J. Am. Ceram. Soc. 78 981 -    DOI : 10.1111/j.1151-2916.1995.tb08425.x
Djuričić B. , Pickering S. 1999 J. Eur. Ceram. Soc. 19 1925 -    DOI : 10.1016/S0955-2219(99)00006-0
Zhou X. D. , Huebner W. , Anderson H. U. 2003 Chem. Mater. 15 378 -    DOI : 10.1021/cm020597c
Li Q. , Han Z. H. , Shao M. W. , Liu X. M. , Qian Y. T. 2003 J. Phys. Chem. Solids 64 295 -    DOI : 10.1016/S0022-3697(02)00299-8
Chen J. Q. , Chen Z. G. , Li J. C. 2004 J. Mater. Sci. Technol. 20 438 -
Chen H. I. , Chang H. Y. 2004 Colloids Surf., A 242 61 -    DOI : 10.1016/j.colsurfa.2004.04.056
Chen H. I. , Chang H. Y. 2005 Ceram. Int. 31 792 -
Li X. Z. , Chen Z. G. , Chen J. Q. , Chen Y. , Ni C. Y. 2005 J. Rare Earths 23 321 -
Li X. Z. , Chen Y. , Chen Z. G. , Chen J. Q. 2006 Electronic Components & Materials 25 43 -
Chen P. L. , Chen I. W. 1993 J. Am. Ceram. Soc. 76 1577 -    DOI : 10.1111/j.1151-2916.1993.tb03942.x
Hu M. Z. C. , Payzant E. A. , Byers C. H. 2002 J. Colloid Interface Sci. 222 20 -    DOI : 10.1006/jcis.1999.6610
Conesa J. C. 1995 Surf. Sci. 339 337 -    DOI : 10.1016/0039-6028(95)00595-1
Gotte A. , Hermansson K. , Baudin M. 2004 Surf. Sci. 552 273 -    DOI : 10.1016/j.susc.2004.01.032
Baudin M. , Wójcik M. , Hermansson K. 2000 Surf. Sci. 468 51 -    DOI : 10.1016/S0039-6028(00)00766-4
Cao X. R. , Xu G. Y. , Li Y. M. , Zhang Z. 2005 J. Phys. Chem. A 109 10418 -    DOI : 10.1021/jp053636r
Liu H. , Qian H. J. , Zhao Y. , Lu Z. Y. 2007 J. Chem. Phys. 127 144903 -    DOI : 10.1063/1.2790005
Soto-Figueroa C. , Vicente L. , Martinez-Magadán J. , Rodriguex-Hidalgo M. 2007 J. Phys. Chem. B 111 11756 -    DOI : 10.1021/jp074122q
Lisal M. , Brennan J. K. , Smith W. R. 2006 J. Chem. Phys. 125 164905 -    DOI : 10.1063/1.2359441
Hoogerbrugge P. , Koolman J. 1992 Europhys. Lett. 19 155 -    DOI : 10.1209/0295-5075/19/3/001
Chen Y. , Zhong J. , Huang W. Q. , Chen R. Y. , Hu W. H. , Chen C.-L. 2010 Chem. J. Chinese Universities 31 1827 -
Juan S. C. C. , Hua C. Y. , Chen C. L. , Sun X. Q. , Xi H. T. 2005 Mol. Simul. 31 277 -    DOI : 10.1080/08927020500035747
Groot R. D. , Warren P. B. 1997 J. Chem. Phys. 107 4423 -    DOI : 10.1063/1.474784
Mullin J. W. 1993 Crystallization Butterworth-heinemann Boston, U.S.A.
Bicerano J. 1996 Prediction of Polymer Properties 2nd ed. Marcel Dekker New York
Flory P. 1953 Principles of Polymer Chemistry Cornell University Press Ithaca
2007 Materials Studio 4.2; Discover/Accelrys San Diego, CA
Rappé A. K. , Casewit C. J. , Colwell K. S. , Goddard III W. A. , Skiff W. M. 1992 J. Am. Chem. Soc. 114 10024 -    DOI : 10.1021/ja00051a040
Frisch M. J. , Trucks G. W. 2004 GAUSSIAN 03 revision D.02 Gaussian, Inc. Wallingford, CT
Ryjkitia E. , Kuhn H. , Rehage H. , Muller F. , Peggau J. 2002 Angew. Chem. Int. Ed. 41 983 -
Kuo M. Y. , Yang H. C. , Hua C. Y. , Mao S. Z. , Deng F. , Wang W. W. , Du Y. R. 2004 Chem. Phys. Chem. 5 575 -    DOI : 10.1002/cphc.200301036