Materials Transactions, Vol. 45, No. 4 (2004) pp. 1375 to 1378 #2004 The Japan Institute of Metals EXPRESS REGULAR ARTICLE

A Study of Magnetic Field Effect on Nanofluid Stability of CuO

Ho Chang1;*, Tsing-Tshih Tsung1, Chii-Ruey Lin2, Hong-Ming Lin3, Chung-Kwei Lin4, Chih-Hung Lo1 and Hung-Ting Su1

1Department of Mechanical Engineering, National Taipei University of Technology, Taipei, 10608, R.O. China 2Graduate Institute of Mechatronic Engineering, National Taipei University of Technology, Taipei, 10608, R.O. China 3Department of Materials Engineering, Tatung University, Taipei,10452, R.O. China 4Department of Material Science, Feng Chia University, Taizhong,40724, R.O. China

This study investigates the effect of additional magnetic field on the stability of CuO nanofluid. Experiments are conducted by imposing an additional magnetic field to the CuO nanofluid prepared by the self-developed Arc-Submerged Nanoparticle Synthesis System (ASNSS), so as to investigate the aggregation phenomenon and the stability of the nanoparticle suspension. It is subsequently known that the permeance strength, time and frequency of the additional magnetic field can affect the CuO nanofluid. Under the influence a strong magnetic field, the longer the permeance time, the more apparent the sedimentation phenomenon will be owing to the aggregation of the nanoparticles. However, the permeance frequency has a relatively slight effect on the CuO nanofluid.

(Received January 20, 2004; Accepted February 27, 2004) Keywords: arc-submerged nanoparticle synthesis system (ASNSS), nanofluid, magnetic field effect, surface potential

1. Introduction Table 1 Process variables of preparing nanofluid by means of ASNSS. Working condition Description In recent years, application and technological development Peak current (A) 2.5 of nanomaterials have been growing rapidly worldwide in Breakdown voltage (V) 150 engineering industries and academic fields. Nanomaterials Pulse duration (ms) 25 are usually defined as the materials with their sizes ranged Off time (ms) 25 from 1 to 100 nanometers. It has the characteristics of size, Temperature of surface, quantum and quantum tunneling effects. The 5 dielectric fluid (C) physical properties of nanomaterials are reflected in the Tool polarity positive areas of heat transfer, electricity, magnetism and mechanics, Dielectric fluid deionized water as well as its chemical properties which are obviously different from those of bulks. These phenomena motivate researchers to further investigate the physical and chemical properties of the surface structure of nanomaterials. More- Nanoparticle Synthesis System (ASNSS)5) under the influ- over, the thermal conductive efficiency of the nanofluid is ence of both weak and strong magnetic fields. inversely proportional to the size of the particles.1) A maximum increase in thermal conductivity of approximately 2. Experimental 20% was observed in the study for 4 vol% of CuO nano- particles with average diameter of 35 nm dispersed in The nanofluid used in the experiments is prepared by the ethylene glycol.2) Furthermore, the effective thermal con- ASNSS, and the process variables are shown in Table 1. By ductivity has shown to increase up to 40% for the nanofluid remaining in static state, the prepared nanofluid formed consisting of ethylene glycol with approximately 0.3 vol% stable suspension particles. Figure 1 is the TEM image of the Cu nanoparticles of mean diameter <10 nm.3) In addition, prepared CuO nanoparticle of acicular structure, and having when the nanofluids are in high electric potential, their an average length of 60 nm and width of 25 nm. thermal conductivity would be increased.4) Carrying an Figure 2 is a schematic diagram of the experimental setup excellent thermal conductivity, CuO nanofluid can be used in for the magnetic field. Two pairs of magnets, which are machine tools as a highly effective circulation fluid. When capable of creating the same magnetic field, are installed on the machine tool is in motion, the magnetic field created by the platform. They are then set at weak (600–1000 Gauss) the power source and dynamic systems would affect the and strong magnetic fields (1850–3000 Gauss) for compara- material properties of the circulation fluid. Thus, it is very tive studies. The magnetization frequency is controlled by the important to investigate the effect of magnetic field on the rotative speed that is pre-set by the motor. Moreover, the nanoparticle suspension. installation platform has to be made of diamagnetic material, This study develops a magnetic environment system to which has lesser influence on the magnetic field, while its simulate an environment having additional magnetic field base has to be made of shock-absorbent material to avoid the imposed onto the fluid. It further investigates the stability of interference of vibration. Then, 15 cm3 of CuO nanofluid, the CuO nanofluid that is prepared by Arc-Submerged which is prepared by the process variables as shown in Table 1, are extracted and poured into a cylindrical test tube *Corresponding author, E-mial: [email protected] made of glass with a diameter of 10 mm and length of 1376 H. Chang et al.

The total magnetic flux density B produced by additional magnetic field can be deduced by eqs. (1) and (2):6)

B ¼ 0H þ 0M or B ¼ H ð1Þ

0 in eq. (1) denotes the permeability in a vacuum, which is 4 107; H is the magnetic field strength, M is the magnetization, is the permeability of the material and M can be represented by the following equation:

M ¼ n m H ð2Þ

where n is the magnetic dipoles of the unit volume, and m is the magnetization coefficient of the material. The nanofluid employed in this experiment is CuO, and its permeability and magnetization coefficients are both 0:9999 ; 1. When mea- suring the additional weak and strong magnetic fields by gauss meter, the range of average magnetic field on top of the magnet and center of test tube lies within 650–1000 Gauss and 1850–3000 Gauss, respectively. Therefore, the range of magnetic flux density produced towards the nanofluid lies within the range of 0.065–0.1 T and 0.185–0.3 T, res- pectively. Fig. 1 TEM image of CuO nanoparticles. 3. Results and Discussion

N S The experimental results indicate that under an additional magnetic field action, the CuO nanoparticle suspension would aggregate after being permeated for 10 minutes, M causing their mean particle size to increase as shown in Figs. 3 and 4 according to the measurement of HORIBA LB- 500 particle size distribution analyzer. Also revealed in the figures is the fact that particles created by strong magnetic Magnet Motor Nanofluid field are apparently coarser than those by weak magnetic field. It is further seen by comparing Figs. 3 and 4 that, whether the magnetic field action is strong or weak, different Fig. 2 Schematic diagram of the experimental setup. permeance frequencies would not impose an apparent change to the particle size. The aggregation of nanoparticle suspen- sion can be attributed to the activation of the surface atoms of 100 mm. The test tube is then placed on a fixed position at the the nanoparticles by the additional magnetic field. These center of a turntable. After the tests, a particle size atoms, which make up a relatively large proportion of the distribution analyzer and TEM are used to perform inspection nanoparticles, begin to spin when activated. At the same of particle size distribution and morphological observation. time, additional magnetic field causes the electron energy The following tests are conducted. (1) Use permeance time as a variable: To investigate the effect of additional magnetic field on the changes of 180 nanoparticles inside the nanofluid under different time periods, namely 10, 20, 30, 40, 50 and 60 minutes. 160 /nm 140

Moreover, in order to learn the influence of magnetic D field on CuO nanofluid under a long time period, an 120 1.933Hz additional test of 8-hour permeance time is conducted. 100 (2) Use intermittent permeance frequency as a variable: 5.133Hz 80 Power supply is used to control the rotative speed of the 7.133Hz variable speed motor so as to acquire different 60 permeance frequencies. When the voltage is set at 40

10 V, 15 V and 20 V, the rotative speed of motor would Mean Particle Size, 20 be 58 rpm, 154 rpm and 214 rpm, respectively, and the 0 permeance frequency would be 1.933 Hz, 5.133 Hz and 0 102030405060 7.133 Hz, respectively. Time, t /min (3) Use the strength of magnet flux density as a variable: The magnetic flux density is set at strong and weak Fig. 3 Relationship between permeance time and particle size (weak magnetic fields for comparison. magnetic field). A Study of Magnetic Field Effect on Nanofluid Stability of CuO 1377

250

/nm 200 D 1.933Hz 150 5.133Hz 7.133Hz 100

50 Mean particle size, 0 0102030405060 Time, t /min

Fig. 4 Relationship between permeance time and particle size (strong magnetic field).

Fig. 6 TEM images of nanofluid after weak magnetic field action.

Table 2 Time needed for nanoparticles to fully precipitate because of aggregation.

Permeance Time needed for the time (min) sedimentation of nanofluid (h) 10 48 20 45 30 36 40 24 50 12 60 4 488 2

aggregate and result in a total sedimentation phenomenon. As indicated in Table 2, the longer the permeance time, the Fig. 5 TEM images of nanofluid after strong magnetic field action. shorter the time required for the particles to precipitate under static state. After the nanoparticle suspension has been permeated, the levels to change, thus creating surface and interface effects.7) nanoparticle suspension shows a drop in electric potential These changes would further lead to the difference between because the additional magnetic field can reduce the the crystal field environment and bonding energy of the repulsion force of the static electricity between the particles, surface atoms of particles and internal atoms. Plenty of thus enabling the nanoparticles to aggregate. The Zeta dangling bonds would then occur on the particle surface. Potential Analyzer (Zeta Plus) of Brookhaven Instruments These bonds possess nonsaturated property and enable the Corporation is used to measure surface potentials in the particles to integrate with other particles, so as to form larger experiment. Figures 7 and 8 illustrate the Zeta potential particles.8) Figures 5 and 6 are the TEM images of nanofluid changes of nanoparticle suspension under two different under strong and weak magnetic field action, respectively. As additional magnetic fields. Comparing these two figures seen in Fig. 5, the particles appear in the shape of bamboo shows that after being permeated for 10 minutes, the Zeta leaves, and their length and width are larger than those not potential of nanofluid would drop immediately. Moreover, under magnetic field action, which are in the shape of under strong magnetic field, Zeta potential of nanofluid is acicular as shown in Fig. 1. In addition, comparing Figs. 5 lower than that influenced by weak magnetic field action. and 6 reveals that under strong magnetic field, the particles However, whether the magnetic field action is strong or are also larger in size than those under weak magnetic field. weak, different permeance frequencies does not impose an Furthermore, under strong magnetic field action, nanofluid of apparent effect on the Zeta potential. different permeance time is extracted and maintained in static A relative movement has occurred on the intersecting state to observe the time required for nanoparticles to surface of the suspension particles and the fluid, which 1378 H. Chang et al.

25 30 /mV 20 E 25 /mV 15 1.933 Hz E 20 1.933Hz 10 5.133 Hz 7.133 Hz 15 5.133Hz 5 7.133Hz 10

Zeta Potential, 0 0102030405060 5 Zeta Potential,

Time, t /min 0

Fig. 7 Relationship between surface potential and permeance time (weak 0 102030405060 magnetic field). Time, t /min

Fig. 8 Relationship between surface potential and permeance time (strong magnetic field). further led to the interface electromotive effect, thus chang- ing the surface potential of the suspension. The surface molecules in deionized water and particles inside the inside the CuO nanofluid would become unstable within suspension formed a fixed layer, and when the particles and a short period of time. The repulsion force of the static fluid start to move relatively, this fixed layer would move electric charge between the suspension particles would together with the particles, and the variation in electric be reduced, and thereby enabling them to aggregate. It potential of the particle movement equals that of the fixed is also known from the experiments that strong layer and the inner part of fluid. The relationship between the magnetic field contributes to coarser particles than repulsion force of the static electricity and the surface those formed under weak magnetic field. potential is indicated in eq. (3):9) (2) Under magnetic field action, the longer the permeance time, the shorter the time required for the nanoparticles V ¼ 2 " " r 2 ekH ð3Þ rep 0 d to aggregate and completely precipitate. However, the From eq. (3), Vrep is the repulsion force of static electricity, " level of permeance frequency has no apparent effect on and "0 are the permittivity of a vacuum and the dielectric nanoparticle suspension. fluid, respectively, r is the radius of particles, d is the surface (3) Affected by additional magnetic field, the repulsion 2 2 1=2 potential, and k is ðzi CiF =""0RTÞ . From eq. (3), it is force between the particles decreases, causing the known that the strength of magnetic field H can lower Vrep, electric potential to drop and allowing the nanoparticles and according to the Principles of Schulze-Hardy, the lower to aggregate. the electric potential, the larger the aggregation force between the particles will be. REFERENCES The experimental results also reveal the average particle size upon 8 hours of continuous permeance is similar to that 1) P. Keblinski, S. R. Phillpot, S. U. S. Choi and J. A. Eastman: Int. J. Heat upon 60 minutes. However, it is known from Fig. 2 that after Mass Transfer. 145 (2002) 855–863. 2) H. X, J. Wang, T. Xi, Y. Liu, F. Ai and Q. Wu: J. Appl. Phys. 91 (2002) 8 hours of permeance, nanofluid would start to precipitate 4568–4572. within 2 hours. This indicates that after long hours of 3) J. A. Eastman, S. U. S. Choi, S. Li, W. Yu and L. J. Thompson: Appl. permeance, the particles will remain in an unstable state, and Phys. Lett. 78 (2001) 718–720. can easily be precipitated within short hours owing to the 4) Y. Xuan and W. Roetzel: Int. J. Heat Mass Transfer. 43 (2000) 3701– aggregation of particles. 3707. 5) T. T. Tsung, H. Chang, L. C. Chen, L. L. Han, C. H. Lo and M. K. Liu: Mater. Trans. 44 (2003) 1138–1142. 4. Conclusions 6) W. D. Kinger, H. K. Bowen and D. R. Uhlmann: Introduction to Ceramics, 2nd edn, (New York: John Wiley & Sons, 1994) pp. 67–74. This study investigates the effect of additional magnetic 7) R. H. Kodama: J. Magn. Magn. Mater. 200 (1999) 359–372. field on the nanofluid applied on machine tool as circulative 8) M. F. Hansen and S. Morup: J. Magn. Magn. Mater. 184 (1998) 262– 274. fluid. From the experimental results and discussion, the 9) W. F. Twan: Fundamentals and application of Nanoparticles, (Taipei, following conclusions are made. Taiwan: Fu-Han publications, 1995) pp. 34–30. (1) Affected by the additional magnetic field, the particles