Powder Technology 196 (2009) 89–101

Contents lists available at ScienceDirect

Powder Technology
j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / p ow t e c

Review

A review on development of nanofluid preparation and characterization

Yanjiao Li a,b,⁎, Jing'en Zhou a, Simon Tung c, Eric Schneider c, Shengqi Xi a
a
State Key Laboratory for Mechanical Behavior of Materials, School of Materials Science and Engineering, Xi'an Jiaotong University, Xi'an, 710049, China
b
Xi'an Research Instu. of Hi-Tech, Hongqing Town, Xi'an, 710025, PR China
c
GM R&D Center, 480-106-160, 30500 Mound Road Warren, MI 48090-9055, USA

a r t i c l e i n f o a b s t r a c t

Article history: Nanofluids, as a kind of new engineering material consisting of nanometer-sized additives and base fluids,
Received 4 October 2008 have attracted great attention of investigators for its superior thermal properties and many potential
Received in revised form 23 July 2009 applications. Many investigations on nanofluids were reported and especially some interesting phenomena,
Accepted 25 July 2009
new experimental results and theoretical study on nanofluids, in which consistent and inconsistent even
Available online 11 August 2009
contrary conclusions were reported, have been presented in literature. The aim of this review is to
summarize recent development in research on synthesis and characterization of stationary nanofluids and
Keywords:
Nanofluids
try to find some challenging issues that need to be solved for future research.
Thermal conductivity © 2009 Elsevier B.V. All rights reserved.
Stability
Viscosity

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
2. System of nanofluids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
3. Preparation of nanofluids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
4. Experimental investigations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
4.1. Stability of nanofluids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
4.2. Thermal conductivity of nanofluids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
4.2.1. Influence of nanoparticles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
4.2.2. Influence of base fluids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
4.2.3. Influence of the liquid–solid interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
4.2.4. Controversy about thermal conductivity enhancement of nanofluids . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
4.3. Viscosity of nanofluids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
5. Theoretical investigations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
5.1. Mechanism of thermal conductivity enhancement of nanofluids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
5.2. Analytical study on thermal conductivity of nanofluids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
6. Conclusions and challenges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99

1. Introduction compactness and high surface-to-volume ratio compared with

With progress of thermal science and thermal engineering, it is of a
strong interest to develop micro-scale liquid flow devices, which have
⁎ Corresponding author. State Key Laboratory for Mechanical Behavior of Materials,
School of Materials Science and Engineering, Xi'an Jiaotong University, Xi'an, 710049,
China. Tel.: +86 29 82663190; fax: +86 29 82660764.
E-mail addresses: lyj.xjtu@yahoo.com.cn (Y. Li), jezhou@mail.xjtu.edu.cn (J.' Zhou),
simon.c.tung@gm.com (S. Tung), eric.w.schneider@gm.com (E. Schneider),
xishq@mail.xjtu.edu.cn (S. Xi).
conventional flow system, for the requirement of enhancing heat
transfer in many diverse industries, including transportations, micro-
electronics, chemical engineering, aerospace and manufacturing.
Therefore, the management of thermal convection, which is the
most important heat transfer mode, has been widely investigated. It is
recognized that the main factor influencing heat transfer efficiency of
thermal convection is the thermal conductivity of heat transfer fluids,
and the low thermal conductivity of conventional fluids, such as water,
oil and ethylene glycol is a primary limitation in the development of
energy-efficient heat transfer medium. An effective way of improving

0032-5910/$ – see front matter © 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.powtec.2009.07.025
90 Y. Li et al. / Powder Technology 196 (2009) 89–101

thermal conductivity of fluids is to suspend small solid particles in the Table 1

fluids in order to change transport properties, flow and heat transfer Thermal conductivity of additives and base fluids used in nanofluids.

features of the liquids. Material Thermal conductivity (W/Mk)

It has been well known that micrometer- or millimeter-sized solid
particles added into a base liquid can improve the thermal
conductivity of the fluid. Although such solid additives may improve
heat transfer coefficient of the fluids, practical application is limited
because of the sedimentation of large particles, clogging flow
channels, erosion of pipelines and causing pressure drops.
Nanofluids, which are solid–liquid composite materials consisting
of nanometer sized solid particles, fibers, rods or tubes suspended in
different base fluids [1], provide a promising technical selection for
enhancing heat transfer because of its many advantages besides
anomalously high thermal conductivity. Nanofluids represent im-
proved stability compared with conventional fluids added with
micrometer- or millimeter-sized solid particles because of size effect
and Brownian motion of the nanoparticles in liquids. With such ultra-
fine nanoparticles, nanofluids can flow smoothly in a microchannel
without clogging and the size of the heat transfer system can be
reduced for the use of nanofluids with high heat transfer efficiency.
In the past several years, some review articles [2–8] involved in the
progress of nanofluid investigation were published. Although these
review papers have generally covered the current aspects of
experimental and theoretical studies of nanofluids, the state-of-the-
arts on nanofluids need to be re-surveyed due to a great number of
new papers on nanofluids published, in those some new phenomena
and new findings are reported. The purpose of this paper focuses on
the preparation and properties characterization of stationary nano-
fluids and the following aspects are involved:
• Selection of a nanofluid system
• Synthesis of nanofluids
• Stability of nano-liquid–solid suspension
• Thermal conductivity of nanofluids
• Viscosity of nanofluids
• Mechanism of enhanced thermal conductivity of nanofluids
• Thermal conductivity model for nanofluids
We also try to find some challenging issues that need to be solved
for future research based on the review on these aspects of nanofluids.
2. System of nanofluids
Theoretically, all solid nanoparticles with high thermal conduc-
tivity can be used as additives of nanofluids. These nanoparticles
that have been often used to prepare nanofluids reported in literature
are: (1) metallic particles (Cu, Al, Fe, Au, and Ag); (2) nonmetal
particles (Al2O3, CuO, Fe3O4, TiO2, and SiC); (3) carbon nanotube; and
(4) nanodroplet. The base fluids commonly used are water, oil,
acetone, decene and ethylene glycol. Generally, the thermal conduc-
tivity of solid is typically higher than that of liquids, seen from Table 1.
Some of the nanofluid systems reported in literature are showed in
Table 2.
Nanofluid dispersing solid nanoparticles such as metal, nonmetal
nanoparticles and carbon nanotubes have been widely investigated by
many researchers. Recently, a new kind of nanofluids dispersed with
nanodroplets was reported [23]. Such types of nanoemulsion fluids
possess long-term stability and can be easily mass produced. Although
the potential of nanodroplets in enhancing thermal conductivity of
nanofluids is doubtful, the development of nanoemulsion fluids may
open a direction for thermal fluid studies. Ma [24] also proposed a
new system which consisted of nanoparticles and liquid metal. Since
the development of nanofluid systems, the definition of nanofluids
needs to be modified. The nanofluids refer to a new kind of composite
materials consisting of nano-additives and base fluids. The additives
may be metal or nonmetal nanoparticles, nanofiber, nanorods,
nanotubes or nanodroplets and the base fluids are any fluids useful.
Metallic solids Cu 401
Al 237
Ag 428
Au 318
Fe 83.5
Nonmetallic solids Al2O3 40
CuO 76.5
Si 148
SiC 270
CNTs ~3000(MWCNTs) ~ 6000(SWCNTs)
BNNTs 260 ~ 600
Base fluids H2O 0.613
Ethylene glycol (EG) 0.253
Engine oil (EO) 0.145
At present, investigation on different nanofluid systems is in
experimental stage, and as for engineering application of a special
nanofluid system is rarely reported. For example, a system of
nanofluids applicable to an advanced vehicle with special properties
as following has not been proposed. The properties including:
• Electrically non-conductive for both safety and stack efficiency
• High thermal and heat transfer properties
• High density and specific heat to minimize the volume flow of
coolant that must be circulated to provide required cooling
• Viscosity needs to be favorable for pumping in a certain range of
temperature
• Material compatibility with the other materials in the fuel cell
system
• Ability to accept gaseous H2 in solution without degradation
• Ability to accept strong voltage fields
• No corrosion, erosion, and plugging of microchannels (0.30 mm)
• Coolant stability for several years
• No flammability and toxicity
In medical application aspect, nanofluids were expected to be
advanced drug delivery fluids that can be used for cancer therapy,
while there is no report on nanofluids with medical application at
present.
3. Preparation of nanofluids
Preparation of nanofluids is the key step in the use of nanoparticles
to improve the thermal conductivity of fluids. Two kinds of methods
have been employed in producing nanofluids. One is a single-step
method and the other is a two-step method.
The single-step method is a process combining the preparation of
nanoparticles with the synthesis of nanofluids, for which the nano-
particles are directly prepared by physical vapor deposition (PVD)
technique or liquid chemical method. In this method the processes of
drying, storage, transportation, and dispersion of nanoparticles are
avoided, so the agglomeration of nanoparticles is minimized and the
stability of fluids is increased. But a disadvantage of this method is
that only low vapor pressure fluids are compatible with the process.
This limits the application of the method.
Eastman et al. [9] has used a one-step physical method to prepare
nanofluids, in which Cu vapor was directly condensed into nanoparticles
by contact with a flowing low vapor pressure liquid (ethylene glycol).
Liu [10] synthesized nanofluids containing Cu nanoparticles in
water through chemical reduction method for the first time. Lo et al.
[25,26] prepared copper dioxide nanofluids by a single-step method
called SANSS. The established SANSS demonstrated to be effective in
avoiding particle aggregation and producing uniformly distributed and
well-controlled size of CuO nanoparticles dispersed in a deionized
water suspension. Zhu [27] presented a novel one-step chemical
 Y. Li et al. / Powder Technology 196 (2009) 89–101 91

Table 2
Nanofluids systems reported in literature.

System Synthesis process Particle loading Particle size Increase in thermal conductivity Ref
(vol%) (nm) (%)

Cu/EG Single-step 0.3 10 40 [9]

Cu/H2O Single-step 0.1 75 ~ 100 23.8 [10]
Cu/H2O Two-step 7.5 100 78 [11]
Fe/EG Single-step 0.55 10 18 [12]
Ag/toluene Two-step 0.001 60 ~ 80 16.5 (60 °C) [13]
Au/toluene Two-step 0.00026 10 ~ 20 21 (60 °C) [13]
Au/ethanol Two-step 0.6 4 1.3 ± 0.8 [14]
Fe3O4/H2O Single-step 4 10 38 [15]
TiO2/H2O Two-step 5 15 30–33 [16]
Al2O3/H2O Two-step 5 20 20 [17]
Al2O3/EG Two-step 0.05 60 29 [17]
CuO/H2O Two-step 5 33 11.5 [18]
SiC/H2O Two-step 4.2 25 15.9 [19]
NCTs/engine oil Two-step 2.0 20 ~ 50 nm 30 [20]
NCTs/poly oil Two-step 1.0 25 nm × 50 μm 160 [21]
NCTs/EG Two-step 1.0 15 × 30 μm 19.6 [22]
NCTs/H2O Two-step 1.0 15 × 30 μm 7.0 [22]
NCTs/decene Two-step 1.0 15 × 30 μm 12.7 [22]
H2O/FC-72 Two-step 12 9.8 nm 52 [23]

method for preparing copper nanofluids by reducing CuSO4·5H2O with
NaH2PO2·H2O in ethylene glycol under microwave irradiation. Non-
agglomerated and stably suspended Cu nanofluids were obtained.
The two-step method for preparing nanofluids is a process by
dispersing nanoparticles into base liquids. Nanoparticles, nanofibers
or nanotubes used in this method are first produced as a dry powder
by inert gas condensation, chemical vapor deposition, mechanical
alloying or other suitable techniques, and the nanosized powder is
then dispersed into a fluid in a second processing step. This step-by-
step method isolates the preparation of the nanofluids from the
preparation of nanoparticles. As a result, agglomeration of nanopar-
ticles may take place in both steps, especially in the process of drying,
storage, and transportation of nanoparticles. The agglomeration will
not only result in the settlement and clogging of microchannels, but
also decrease the thermal conductivity. Simple techniques such as
ultrasonic agitation or the addition of surfactants to the fluids are
often used to minimize particle aggregation and improve dispersion
behavior. Since nanopowder synthesis techniques have already been
scaled up to industrial production levels by several companies, there
are potential economic advantages in using two-step synthesis
methods that rely on the use of such powders. But an important
problem that needs to be solved is the stabilization of the suspension
prepared.
Hong [12,28] prepared Fe nanofluids by dispersing Fe nanocrystal-
line powder in ethylene glycol by a two-step procedure. The Fe
nanoparticles with a mean size of 10 nm were synthesized by a
chemical vapor condensation process. To avoid nanoparticles congre-
gating in nanofluids, they used an ultrasonic cell disrupter generating
ultrasonic pulses of 700 W at 20 kHz. Xuan [11,29,30] prepared Cu/
H2O, Cu/oil nanofluids by two-step method. In order to avoid
nanoparticle aggregation, surfactants and ultrasonic agitation were
employed. Murshed [16] reported TiO2 suspension in water prepared
by two-step method. Xie [17] prepared Al2O3/H2O, Al2O3/EG, Al2O3/
PO nanofluids by two-step method, and intensive ultrasonication and
magnetic force agitation were employed to avoid nanoparticle
aggregation. Liu [20] and Choi [21] produced carbon nanotube
suspensions by a two-step method.
4. Experimental investigations
4.1. Stability of nanofluids
Nanofluids are not a simple mixture of liquid and solid particles.
Nanoparticles tend to aggregate with the time elapsed for its high
surface-activity. The agglomeration of nanoparticles results not only
the settlement and clogging of microchannels but also the decreasing
of thermal conductivity of nanofluids. So the investigation on stability
is also a key issue that influenced the properties of nanofluids for
application.
Now many methods have been used to evaluate the stability of
nanofluids. The most simple and reliable method is sedimentation
method. In this method, the variation of concentration or particle size
of supernatant particle with sediment time was obtained by special
apparatus. The nanofluids were considered to be stable when the
concentration or particle size of supernatant particles keeps constant.
Sedimentation photograph of nanofluids in test tubes taken by a
camera was also a usual method for observing the stability of
nanofluids. Long period for observation is the defect of sedimentation
method compared with other methods. Zeta potential analysis is
another method for evaluating the stability of nanofluids, however,
this method has a limitation of the viscosity and concentration of
nanofluids. Hwang [31,32] studied the stability of nanofluids with a
UV–vis spectrophotometer. The variation of supernatant particle
concentration of nanofluids with sediment time can be obtained by
the measurement of absorption of nanofluids because there is a linear
relation between the supernatant nanoparticle concentration and the
absorbance of suspended particles. Li [33] studied the dispersion
behavior of Cu nanoparticles in water under different pH values,
different dispersant type and concentration by the method of Zeta
potential, absorbency and sedimentation photographs.
Peng et al. [34] studied the factors that influence the stability of
nanofluids. The test results showed that the most important factors
affecting the stability of suspensions were the nanoparticle's
concentration, dispersant, viscosity of base liquid and pH value. The
variety, diameter, density of nanoparticle and ultrasonic vibration also
influence the stability of nanofluids. Hwang [32] studied the stability
of nanofluids with the UV–vis spectrophotometer. He believed that
the stability of nanofluids was strongly affected by the characteristics
of the suspended particle and base fluids such as particle morphology,
chemical structure of particles and base fluids. Moreover, addition of a
surfactant can improve the stability of the suspensions. Wang [35,36]
also draw a conclusion that the equivalent diameter of nanoparticle
and dynamic viscosity of nanofluids were the most important factors
to affect the stability of nanoparticle suspension. Investigators [37,38]
also found that suitable pH value of nanofluids and mass fraction of
dispersant had an important effect on the stability of nanofluids.
The investigation of Li [33] showed that pH value, dispersant type
and concentration influenced the stability of Cu/H2O suspensions. The
92 Y. Li et al. / Powder Technology 196 (2009) 89–101
addition of optimizing concentrations of dispersants at pH 9.5 can
lead to the best stability of the Cu/H2O nanofluids. Xuan and Li [11]
used salt and oleic acid as dispersant to enhance the stability of Cu/oil
and Cu/water nanofluids, respectively. Oleic acid and cetyltrimethy-
lammonium bromide (CTAB) surfactants were used by Murshed et al.
[16]. Xie [22] proposed a method to produce stable and homogeneous
suspensions of CNTs in DW, EG, and DE. Entangled and undispersible
CNTs have been treated using concentrated acid to disentangle the
aggregates and to make dispersible CNTs. More hydrophilic surfaces
have been formed by introducing oxygen-containing functional
groups on CNT surfaces during the chemical treatment. Thus stable
CNT nanofluids in DW, EG, or DE were produced and these nanofluids
showed enhanced thermal conductivities. Yu [39] and co-workers
used plasma polymerization to modify the surface of diamond
nanoparticles. Through this treatment, a much lower water contact
angle, i.e., much stronger water affinity, was achieved with plasma
treated diamond nanoparticles as compared with the untreated
controls. As a result, a much more stable suspension of the treated
diamond nanoparticles in water was observed. Hong et al. [28]
reported Fe/ethylene glycol nanofluids. The ultrasonic pulses of
700 W at 20 KHz generated by an ultrasonic cell disrupter was used
to improve the dispersion of nanoparticles in the nanofluids. In
general, methods such as change of pH value, addition of dispersants,
treatment on surface of nanoparticles and ultrasonic agitation have
been used to improve stability of nanofluids. Although the methods
mentioned above were used to improve stability of nanofluids, only
several days or one month stable time of nanofluids was reported,
effective and simple methods that can keep long-term stability of
nanofluids are not available so far. More studies for changing the
surface properties of suspended particles and suppressing formation
of particle cluster to obtain stable and homogeneous suspension are
required.
It should be noted that at present no uniform standard methods for
examining the stability of nanofluids were proposed and thus it is
difficult to compare the stability of nanofluids reported by different
investigators.
4.2. Thermal conductivity of nanofluids
Thermal conductivity is an important parameter for enhancing
heat transfer performance of a heat transfer fluid. Many researchers
have reported experimental studies on the thermal conductivity of
nanofluids. The transient hot wire (THW) method, temperature
oscillation and the steady-state parallel plate method has been
employed to measure thermal conductivity of nanofluids. However,
the transient hot wire method h as been wi dely used
[11,12,16,25,28,40]. The transient hot wire technique works by
measuring the temperature/time response of the wire to an abrupt
electrical pulse. The wire is used as both heater and thermometer. A
derivation of Fourier's law and temperature data are used to calculate
the thermal conductivity. To avoid problems associated with the
measurement of electrically conducting fluids, the wire is coated with
a thin electrical insulation layer. For example, Zhang [18] reports on
accurate measurements of the effective thermal conductivity and
thermal diffusivity of various nanofluids by using a transient short-
hot-wire (SHW) technique. To remove the influences of the static
charge and electrical conductance of the nanoparticles on measure-
ment accuracy, the SHW probes are coated by a pure Al2O3 thin film
with a sputtering apparatus.
Since thermal conductivity of solid nanoparticles is much higher
than that of fluids, the suspended particles are expected to be able to
increase thermal conductivity and heat transfer performance. Fig. 1 is a
comparison of thermal conductivity increment of some nanofluids
reported in literature. In fact, many factors (such as size, volume
fraction, thermal conductivity of the nanoparticles, viscosity, temper-
ature and thermal conductivity of base fluids, as well as the membrane
Fig. 1. Comparison of experimental data on thermal conductivity of nanofluids.
of liquid molecules around the nanoparticles, etc.) may influence the
thermal conductivity of nanofluids. In the past few years, many
experimental investigations on the factors influencing thermal
conductivity of nanofluids have been reported.
4.2.1. Influence of nanoparticles
Nanoparticles, the additives of nanofluids, play an important role
in changing the thermal transport properties of nanofluids. The
investigations about nanoparticles mainly focus on the effects of their
volume fraction, thermal conductivity, morphology and Brownian
movement et al.
Most investigations revealed that the thermal conductivity of
nanofluids increases anomalously with increasing the volume fraction
of nanoparticles. The thermal conductivity of Fe/ethylene glycol
nanofluids increases nonlinearly with the volume fraction of Fe
nanoparticles [12]. The experimental results showed that the
enhanced thermal conductivity ratios increase with the increase of
volume fraction of Al2O3 nanoparticles in nano-suspensions [17]. Choi
[21] has produced oil-based nanofluids containing carbon nanotubes
and measured their effective thermal conductivity. The measured
thermal conductivity is anomalously greater than theoretical predic-
tions and is nonlinearly increasing with increasing of nanotube
loadings. Their argument for the enhanced thermal conductivity is
due to the nature of heat conduction in nanotube suspensions, and an
organized structure at the solid–liquid interface. The nonlinear
phenomenon can be ascribed to both the size and shape of nanotubes.
The results made by Zhang [18,40] and his co-workers showed that
the effective thermal conductivity and thermal diffusivity of Au/
toluene, Al2O3/water, and CNT/water nanofluids at different tem-
peratures increase with increasing of the particle concentration, the
particle thermal conductivity, and the ratio of the length to the
diameter of CNT. However, no anomalous enhancements in the
effective thermal conductivity of the nanofluids were observed. All of
the measured values at lower volume fractions agree well with those
predicted by the HC equation for the spherical particles and by the
unit-cell model equation of Yamada and Ota for the CNTs. Putnam and
his co-workers [14] studied the nanoparticle suspensions of Au/
ethanol and no significant enhancement of thermal conductivity was
found also.
Generally, nanoparticles with high thermal conductivities would
result in nanofluids with high enhancement of the thermal conduc-
tivity. The investigations conducted by Liu [10,20] and Xie [22]
compared the effects of CNT and CuO, CNT and Al2O3 nanoparticles on
the enhancement of thermal conductivity of nanofluids, respectively.
The results demonstrated that no matter what kind of base fluids was
used, the thermal conductivity enhancement of CNT suspensions was
 Y. Li et al. / Powder Technology 196 (2009) 89–101
much higher than those of oxide nanoparticle suspensions with the
same volume fraction of nanoparticles. The reason may be the
substantial difference in thermal conductivity and morphology
between oxide nanoparticles and carbon nanotubes. But some
investigators drew a completely different conclusion. For example,
although the bulk Fe3O4 crystal has a lower thermal conductivity than
Al2O3, CuO, and TiO2 bulk crystals the investigation [15] showed that
Fe3O4/aqueous nanofluids exhibited higher thermal conductivity than
the other oxide aqueous nanofluids at the same volume fraction of
nanoparticles. The author believed that the abnormal and nonlinear
increase in thermal conductivity of Fe3O4/aqueous nanofluids was
ascribed to the observed nanoparticles clustering and alignment. Just
like this case, Hong et al. [12,28] also observed the same phenomenon.
They prepared nanofluids with ethylene glycol containing Fe nano-
particles and compared the thermal conductivity of the nanofluids
with that of nanofluids containing Cu nanoparticles. Although Cu
nanoparticles have a higher thermal conductivity than Fe nanoparti-
cles, they found that the nanofluids containing Fe nanoparticles
showed a more effective thermal transport property than nanofluids
containing Cu nanoparticles dispersed with little agglomeration.
The morphology of nanoparticles, including SSA (specific surface
area), shape and size of the nanoparticles, may strongly affect the
thermal conductivity of nanofluids. Xie [17] investigated the thermal
conductivity behavior of nanosized Al2O3/PO and Al2O3/EG suspen-
sion systems and found that the thermal conductivity ratio increases
first, and then decreases with the increase of SSA of the nanoparticles
and the high thermal conductivity at a particle SSA of 25 m2 g− 1 was
observed. He argued that for the suspensions using the same base
fluid, the enhancement of the thermal conductivity depended on SSA
(specific surface area) of the nanoparticles and thus the highest
thermal conductivity can be expected with an optimal SSA. The
similar results were observed by Shou [41].
The shape of nanoparticles can also influence the enhancement of
effective thermal conductivity of nanofluids. Zhou and Gao [42,43]
presented differential effective medium theory to estimate the
effective thermal conductivity of nanofluids dispersed with non-
spherical solid nanoparticles by taking into account the interfacial
thermal resistance across the solid particles and the host liquids. They
found that high enhancement of effective thermal conductivity can be
achieved when the shape of nanoparticles is deviated much from the
spherical one. Yang et al. [44] investigated the temperature
dependence of thermal conductivity in nanorods-in-FC72 and
nanorods-in-oil nanofluids. The analysis of the experimental data
suggested that the dominant mechanism of thermal conductivity
enhancement in nanofluids strongly depends on the particle aspect
ratio, i.e. the ratio of the major axis to the minor axis of the particle.
The effects of the Brownian motion could be predominant for
spherical nanoparticles in nanofluids while the diffusive heat
conduction mechanism will gradually take over the dominance as
the aspect ratio increases.
Since contradictory results have been reported, more theoretical
and experimental work is needed to clarify the effect of nanoparticles
on the thermal conductivity of nanofluids.
4.2.2. Influence of base fluids
Besides the influence of nanoparticles, the thermal conductivity,
temperature and viscosity of base fluids also affect the enhancement
of thermal conductivity of nanofluids. However, research on the effect
of base fluids is relatively less reported.
Xie's investigation [17,22] indicated that suspensions using the
same nanoparticles, the enhanced thermal conductivity ratio is
reduced with increasing thermal conductivity of the base fluid.
Similar results were obtained by Hwang [31] for CuO-in-ethylene
glycol nanofluids and CuO-in-water nanofluids.
Li and Peterson [45] examined the effects of variation in the
temperature and volume fraction on the steady-state effective
93
thermal conductivity of CuO and Al2O3 nanoparticles suspensions.
The nanoparticles of CuO and Al2O3 with diameters of 29 and 36 nm
were blended with distilled water at 2%, 4%, 6%, and 10% volume
fractions and the resulting suspensions were evaluated at tempera-
tures ranging from 27.5 to 34.7 °C. The experimental results indicated
that the nanoparticle material, diameter, volume fraction, and bulk
temperature, all have a significant effect on effective thermal
conductivity of these suspensions. The thermal conductivity of
nanofluids increases significantly with increase of temperature. The
study on thermal conductivity of Al2O3/H2O, CuO/H2O, Al2O3/EG, CuO
/EG nanofluids conducted by Shou also showed that the thermal
conductivity of nanofluids increased linearly with the increase of
temperature [41]. Similar results were obtained by Patel et al. [13] and
Das et al. [46]. However, Yang et al. [44] studied the temperature
dependence of thermal conductivity enhancement in nanofluids
containing Bi2Te3 nanorods. The thermal conductivity has been
measured in FC72 with and without Bi2Te3 nanorods over a
temperature range from 3 to 50 °C and in hexadecane oil from 20 to
50 °C. The measured thermal conductivity enhancement decreased
with increasing temperature in both nanofluids systems. They argued
that the different trends observed in those nanofluids could be due to
the difference in the particle aspect ratio, which would have
important effects on both the particle Brownian motion and diffusive
heat conduction.
There are few reports about the direct relationship between
viscosity and thermal conductivity of nanofluids. However, the
viscosity of base fluids has significant effects on stability of nanofluids
[34].
4.2.3. Influence of the liquid–solid interface
Another important factor that may influence the thermal conduc-
tivity of nanofluids is the liquid–solid interface. Liquid molecules close
to the solid surface are known to form a layered structure. A liquid in
contact with a solid interface is more ordered than a bulk liquid. Since
crystalline solids (which are obviously ordered) display much better
thermal transport than liquids, such liquid layering is expected to lead
to a higher thermal conductivity. Fig. 2 is a sketch of a single particle
with an interfacial layer in a fluid medium proposed by Leong et al.
[47].
Many studies suggested that the nano-layer acts as a thermal
bridge between a solid nanoparticle and a bulk fluid and plays a key
role in enhancing the thermal conductivity of nanofluids. Yu and Choi
[48,49] argued that this ordered nano-layer has a major effect on
nanofluids thermal conductivity when the particle diameter was less
than 10 nm, and they modified the Maxwell model for the effective
thermal conductivity of liquid–solid suspensions including the effect
of this ordered nano-layer. Xie [50] suggested that the effective
thermal conductivity increased with increasing nano-layer thickness.
Especially in a small particle size range, the effects of particle size and
Fig. 2. Single spherical particle with interfacial layer in a fluid medium.
94 Y. Li et al. / Powder Technology 196 (2009) 89–101
nano-layer thickness become much more obvious, which implies that
manipulating nano-layer structure might be an effective method to
produce highly thermally conductive nanofluids. The calculated
values fit quite well with some currently available experiments.
Leong et al. [47] concluded that the interfacial layer was significant to
enhancing the thermal conductivity of nanofluids. But Xue et al.
[51] demonstrated that the layering of the liquid atoms at the liquid–
solid interface did not have any significant effect on thermal transport
properties.
4.2.4. Controversy about thermal conductivity enhancement of nanofluids
Over the past 10 years, many researchers reported enormous
increases of thermal conductivity with small nanoparticle loading,
while others reported a normal increase which can be reported by
effective medium theories. Whether the enhancement of thermal
conductivity is abnormal or not, the problem is a hotly debated issue
discussed by some investigators.
Evans et al. [52] presented a kinetic theory based analysis and
molecular dynamics simulations and led to a conclusion that the
thermal conductivity of nanofluids with well dispersed nanoparticles
was well described by the effective medium theory and did not show
any significant enhancements due to effects associated with Brownian
motion induced hydrodynamic effects. Keblinski and Prasher [53]
presented a critical analysis of the experimental data and found that
the well established effective medium theories for composite
materials were capable of explaining the vast majority of the reported
data without resorting to novel mechanisms such as Brownian motion
induced nanoconvection, liquid layering at the interface, or near-field
radiation. Nie [54] also proposed that the high thermal conductivity of
nanofluids observed was questionable based on his theoretical
analysis.
Although different experimental phenomenon and theoretical
analysis result were reported, there is no accepted conclusion on
thermal conductivity enhancement of nanofluids. Thus, more theo-
retical and experimental works are needed in the future.
4.3. Viscosity of nanofluids
Most published studies have focused on the heat transfer behavior
and stability of nanofluids in the past 10 years. In fact, viscosity is one
of the key properties of nanofluids. It is believed that viscosity is as
critical as thermal conductivity in engineering systems because the
nanofluid was expected to show an increase in thermal conductivity
without an increase in pressure drop, which in turn is related to fluid
viscosity. Among the few studies about viscosity of nanofluids, the
problems concerned by investigators are mainly focused on the
factors influencing the viscosity of nanofluids. These factors include
concentration and size of nanoparticles, temperature of nanofluids
and shear rate etc. The model for predicting the viscosity of nanofluids
was also an important issue interested in by many investigators.
Prasher et al. [55] investigated the viscosity of alumina-based
nanofluids. The experimental results showed that the relative
viscosity of alumina-based nanofluids was not a strong function of
temperature and nanoparticle diameter, but is a strong function of
nanoparticle volume fraction. Li et al. [56] measured the viscosity of
water with CuO nanoparticle suspensions using a capillary viscom-
eter. Results showed that the apparent viscosity of nanofluids
decreased with increasing temperature and increased somewhat
with increasing concentration of the nanoparticles, the effect of con-
centration was not as obvious as the effect of temperature. However,
as they pointed out, the capillary tube diameter may influence the
apparent viscosity for higher nanoparticle mass fractions, especially at
lower temperatures. Chen et al. [57] investigated ethylene glycol
based titania nanofluids and found that the nanofluids showed
Newtonian behavior over a shear rate range of 0.5–104 s− 1 at 293–
333 K. The shear viscosity of the nanofluids depends strongly on
temperature and the particle's concentration. Guo [58] measured the
viscosity of copper oxide dispersed in ethylene glycol and water
mixture and got the similar results. Nguyen et al. [59] investigated
experimentally the influence of both the particle size and the
temperature on dynamic viscosities of water-based nanofluids,
namely Al2O3–water and CuO–water mixtures. The additives used
were 36 nm alumina particles, 47 nm alumina particles and 29 nm
oxide copper particles. For all the nanofluids tested, their viscosities
were found to be strongly dependent on both temperature and
particle volume fraction. In general, the dynamic viscosity of
nanofluids increases with an augmentation of particle volume fraction
(for a given temperature), but decreases with increasing temperature
(for a given particle concentration). For Al2O3–water nanofluid, it was
observed that particle size effects were more important for high
particle concentrations. Viscosities of 47 nm particle fluids are clearly
higher than those of 36 nm size when the particle volume fractions
were higher than 4%. Viscosities corresponding to oxide copper–water
were the highest among the three kinds of nanofluids tested
especially for high concentration of nanoparticles. An important
discovery of this paper is hysteresis phenomenon, namely the
existence of a critical temperature beyond which nanofluids viscous
behavior becomes drastically altered. The author also pointed out that
the hysteresis phenomenon raised serious concerns regarding the use
of nanofluids for heat transfer enhancement purposes.
Das et al. [60] measured the viscosity of Al2O3/water nanofluids
against shear rate. Their results showed an increase of viscosity with
increased particle concentrations. There is strong possibility that
nanofluids may be non-Newtonian fluids, even viscoelastic in some
cases. Further experimental studies are needed to define the viscosity
models of nanofluids so they can be used in simulation studies. Ding et
al. [61] measured the viscosity of CNT/water nanofluids as a function
of shear rate. They observed that the viscosity of nanofluids increased
with increasing CNT concentration and decreasing temperature. Also,
the shear thinning behavior was found by the authors. That means the
nanofluids can provide better fluid flow performance due to the
higher shear rate at the wall, which results in low viscosity.
There exist few theoretical formulas that can be used to estimate
particle suspension viscosities. Almost all such formulas have been
derived from the pioneering work of Einstein (1906) [62], which are
based on the assumption of a linearly viscous fluid containing dilute,
suspended, spherical particles. The formula Einstein obtained was:
μnf = μbf ð1 + 2:5ϕÞ:
Here, µnf is the viscosity of suspension, µbf is the viscosity of base
fluid and ϕ is the volume fraction of particles in base fluid. Einstein's
formula is found to be valid for relatively low particle volume
fractions ϕ ≤ 0.02. Beyond this value, it underestimates the effective
viscosity of the mixture. Since the publication of Einstein's work,
many articles have been devoted to the “correction” of his formula.
Brinkman (1952) [63] has extended Einstein's formula for use with
moderate particle concentrations, as follows:
1
μnf = μbf :
ð1−ϕÞ2:5
In 1972, Lundgren [64] proposed the following equation under the
form of a Taylor series in ϕ:


25 2 3
μnf = μbf 1 + 2:5ϕ + ϕ + Oðϕ Þ :
4
It is obvious that if the terms of the second or higher order of ϕ are
neglected, the above formula reduces to that of Einstein.
In 1977, Batchelor [65] considered the effect due to the Brownian
motion of particles on the bulk stress of an approximately isotropic
 Y. Li et al. / Powder Technology 196 (2009) 89–101
suspension of rigid and spherical particles. He proposed the following
formula:
2
μnf = μbf ð1 + 2:5ϕ + 6:5ϕ Þ:
In fact, nearly no model mentioned above can predict the viscosity
of nanofluids precisely in a wide range of nanoparticle volume
fraction. Nguyen et al. [66] find that the conditional formula, including
the Einstein's formula and the ones proposed by Brinkman, Lundgren
and Batchelor have all underestimated the nanofluid's viscosity even
for a relatively low particle fraction. They also proposed correlations
for nanofluids consist of 47 nm and 36 nm Al2O3 nanoparticles and
water, respectively as following:
0:483ϕ
μnf = 0:904e μbf ð47nm Al2 O3 Þ
2
μnf = ð1 + 0:025ϕ + 0:015ϕ Þμbf ð36nm Al2 O3 Þ:
Lee [67] synthesized aqueous nanofluids containing low volume
concentrations of Al2O3 nanoparticles in the 0.01–0.3 vol% range.
Viscosity measurements showed that the viscosity of the Al2O3–water
nanofluids significantly decreased with increasing temperature. Further-
more, the measured viscosities of the Al2O3–water nanofluids showed a
nonlinear relation with the concentration even in the low volume
concentration (0.01%–0.3%) range, while the Einstein viscosity model
clearly predicts a linear relation and severely underestimate the viscosity.
Now most formulas were developed to relate viscosity as a function of
volume fraction of nanoparticles. In fact, temperature is also an important
factor influencing the viscosity of nanofluids. Some correlations between
viscosity and temperature were proposed by investigators.
Kulkarni et al. [68] proposed a correlation that relates viscosity of
copper oxide nanoparticles suspended in water and in the temper-
ature range of 5–50 °C


1
ln μs = A −B:
T
Here A and B are the functions of volume percentage ϕ. As this is
an aqueous solution, this correlation is not applicable for nanofluids in
the subzero temperature range.
Praveen [69] derived an exponential model for copper oxide
nanoparticles suspended in 60:40 (by weight) ethylene glycol and
water mixture as the following:
−BT
logðμs Þ = Ae
where µs is the copper oxide nanofluid viscosity in centipoise (cP), T is the
temperature in K and A, B are functions of particle volume percentage ϕ.
Fig. 3. Schematic diagrams of several possible mechanisms. (a) Enhancement of k due to formation of highly conductive layer-liquid structure at the liquid/particle interface;
(b) Ballistic and diffusive phonon transport in a solid particle; (c) Enhancement of k due to increased effective φ of highly conducting clusters.
95
It should be noted that there is some inconsistency among the few
studies on the rheology of nanofluids. For example, some studies
[55,70] showed the Newtonian behavior of nanofluids while others
observed obvious non-Newtonian behavior [71] The investigation of
Prasher [55] showed that the normalized shear viscosity depended
linearly on nanoparticle volume fraction, while Nguyen et al. [66]
found an exponential relationship between viscosity of nanofluids
and volume fraction of nanoparticles. The reasons for these phenom-
ena may lie in the difference of nanoparticle volume fraction, type of
nanoparticles, and chemical properties of base fluids.
5. Theoretical investigations
5.1. Mechanism of thermal conductivity enhancement of nanofluids
At present, mechanism investigations of nanofluids mainly focus
on the experimentally observed enhancement of the thermal
conductivity.
Kelblinski [72] and Eastman [73] discussed four possible explana-
tions for this anomalous increase: Brownian motion of the nanopar-
ticles, molecular-level layering of the liquid at the liquid/particle
interface, the nature of heat transport in the nanoparticles, and the
effects of nanoparticle clustering, which are schematically shown in
Fig. 3. They postulated that the effect of Brownian motion could be
ignored since contribution of thermal diffusion was much greater
than Brownian diffusion. However, they only examined the cases of
stationary nanofluids.
Xuan and Li [11] also discussed four possible reasons for the
improved effective thermal conductivity of nanofluids: the increased
surface area due to suspended nanoparticles, the increased thermal
conductivity of the fluid, the interaction and collision among particles,
the intensified mixing fluctuation and turbulence of the fluid, and the
dispersion of nanoparticles. Later, Li and Xuan [74] proposed another
factor that may contribute to the enhancement of thermal conduc-
tivity. They analyzed the mechanism of an intensified thermal
conductivity of nanofluids from the following two aspects, namely, a
changed liquid structure due to the addition of nanoparticles and the
micro-motion of nanofluids. Through the measurement of the thermal
conductivity of nanofluids under various temperatures it has been
verified that the micro-motion of the nanoparticles is a major factor
contributing to the enhancement of thermal conductivity of the
nanofluids.
Xie et al. [75] studied the mechanism of heat conduction in
nanofluids and they proposed that the effects of nonlinear heat
transfer in nanoparticles, micro convection caused by the Brownian
movement of nanoparticles, congregation of nanoparticles and orderly
96 Y. Li et al. / Powder Technology 196 (2009) 89–101
array of liquid molecules at the interface between the nanoparticles
surface and the base fluid are four main influencing factors.
Recently, the effect of nanoparticle aggregation has been proposed
as an important mechanism of thermal conductivity enhancement. By
using the three-level homogenization theory validated by MC
simulations of heat conduction on model fractal aggregates, Prasher
[76] showed that thermal conductivity of nanofluids based purely on
conduction phenomenon can be significantly enhanced as a result of
aggregation of nanoparticles. The conductivity enhancement due to
aggregation is a strong function of the chemical dimension of the
aggregates and the radius of gyration of the aggregates. Later, Evans
[77] revised the conclusion by considering interfacial thermal
resistance and proposed that the value of the thermal conductivity
enhancement was determined by the cluster morphology, filler
conductivity and interfacial thermal resistance. They also pointed
out that the degradation of thermal conductivity by interfacial
thermal resistance can be limited by large aggregate sizes. According
to this mechanism, there exists an optimized aggregation structure to
achieve maximum thermal conductivity, which is far beyond the
prediction from homogeneous dispersions. Zhu [15] studied the
thermal conductivities of Fe3O4 aqueous nanofluids and got the
same conclusion. They believed that the abnormal thermal conduc-
tivities of Fe3O4 nanofluids were attributed to the observed nano-
particle clustering and alignment.
The liquid–solid interfacial layer is an important structure in
nanofluids. There are two different viewpoints about the role of this
structure. Some investigators [47–49,72] considered that the liquid–
solid interfacial layers play an important role in the enhanced thermal
conductivity of nanofluids. They believed that the solid-like layer acts
as a thermal bridge between a solid nanoparticle and the base liquid
and so it is an important mechanism for explaining the anomalously
increased thermal conductivity of nanofluids. Some investigators
proposed an opposite viewpoint. Based on effective medium theory,
for large interfacial thermal resistance, the addition of particles will
decrease thermal conductivity of the composite. Thus it seems that
the nano-layer can only give a decrease in thermal conductivity rather
than enhancement due to the Kapitza resistance [6]. Putnam [78] used
a 3ω method to study the thermal conductivity of composites of
nanoscale alumina particles in polymethylmethacrylate (PMMA)
matrices and found that interface effects were relatively unimportant
for nanoparticles of radius N30 nm. The investigation of Evans [77]
and Nan [79] believed that for the majority of nanoparticles the
interfacial resistance will not significantly affect the composite
thermal conductivity. However, when small radius (or diameter)
particles coincide with large interfacial resistance, such as the case for
single wall carbon nanotubes, the composite thermal conductivity
may be significantly diminished.
Chon et al. [81] showed that the Brownian motion of the nano-
particles at the molecular and nanoscale levels is a main mechanism
controlling the thermal conductivity of nanofluids. Jang and Choi [81]
and Prasher [82] argued that convection due to the Brownian
movement of the nanoparticles is the main reason for the observed
enhancement in thermal conductivity of nanofluids. Koo et al. [83]
made a theoretical study on Brownian, thermo-phoretic, and osmo-
phoretic motion effects on the effective thermal conductivities of
nanofluids. They got a conclusion that the effect of nanoparticles on
Brownian motion is much more significant than the thermo-phoretic
and osmo-phoretic motion effects. On the other hand, some other
investigators believed that the Brownian motion played a small role in
the thermal conductivity enhancement. Kelblinski [72] performed
simulation of a single crystalline nanoparticle suspended in fluid and
found that the role of Brownian motion has little effect on the total heat
flux autocorrelation function. Evans et al. [52] suggested that the
contribution of Brownian motion to the thermal conductivity of the
nanofluids is very small and cannot be responsible for the extraordi-
nary thermal transport properties of nanofluids. They also supported
their argument using the molecular dynamics simulations and the
effective medium theory. However, they just limited their discussion
to stationary fluids, which weakens their results. Nie [54] conducted an
analysis for the enhancement of thermal conductivity due to the
Brownian motion of nanoparticles and got a conclusion that Brownian
motion of nanoparticles makes little contribution to the effective
thermal conductivity of nanofluids.
From the aforementioned discussions, we can see that so far no
accepted conclusions were obtained to explain the thermal behavior
of nanofluids although various mechanisms have been proposed
considering different factors. Controversy still exists. More theoretical
and experimental work need to be done.
5.2. Analytical study on thermal conductivity of nanofluids
From the experimental work of many researchers, it is evident that
the thermal conductivity of nanofluids depends on the thermal
conductivity of both base fluids and nanoparticle material, the volume
fraction, the surface area, the shape and the Brownian motion of
nanoparticles suspended in the liquid, the temperature of the base
fluid and the interfacial layer of liquid–solid suspension.
In the past decade many efforts have been made to predict the
enhancement of thermal conductivity of nanofluids based on
experiments and theoretical analysis and a number of models are
proposed. However, so far no model can be generally used to predict
the enhancing effects in thermal conductivity of different nanofluids.
The Maxwell model [84], a classical model for thermal conductivity,
was proposed to predict the thermal conductivity of homogeneous
suspensions with relatively large and spherical particles in the form as
follows:
kp + 2kb + 2ðkp −kb Þϕ
keff = k ð1Þ
kp + 2kb −ðkp −kb Þϕ b
where kp represents the thermal conductivity of the solid particles
added, kb is the thermal conductivity of the bulk liquid, and φ is the
particle volume fraction of the suspension, vol%. Maxwell's model
shows that the effective thermal conductivity of nanofluids relies on
the thermal conductivity of the spherical particle, the base fluid and
the volume fraction of the solid particles.
Since the Maxwell equation considers only spherical particles of
even distribution, some other possible factors influencing thermal
conductivity, such as diameter, surface area, shape of the added
particles, Brownian motion of suspended nanoparticles, interface
between the solid particles and the base fluid etc., are neglected.
Consequently, comparing with experimental data, the increase in
thermal conductivity predicted by the equation is underestimated. So,
considering more possible factors, various modified models were
proposed based on the traditional Maxwell model.
The liquid–solid interface is an important part of particle-in-liquid
suspensions. The nano-layer between the nanoparticles and base
fluids may be a dominant factor influencing thermal conductivity of
nanofluids. The nano-layer has not been taken into account for any
classical models. Recently, the research on nanofluids indicates that
the enhancement of thermal conductivity is due to the ordered
layering of liquid molecules near the solid particles, and some models
taking the effect of the nano-layer into account were developed.
Yu and Choi [48] proposed a modified Maxwell model considering
the effect of the nano-layer by replacing the thermal conductivity of
solid particles kp in Eq. (1) with the modified thermal conductivity of
particles kpe, which is based on the so called effective medium theory
[85]:
½2ð1−γÞ + ð1 + βÞ3 ð1 + 2γÞγ
kpe = kp ð2Þ
−ð1−γÞ + ð1 + βÞ3 ð1 + 2γÞ
 Y. Li et al. / Powder Technology 196 (2009) 89–101 97

where γ = klayer/kp is the ratio of nano-layer thermal conductivity to of the so called “complex nanoparticles”, which included the inter-
particle thermal conductivity and β = h/r is the ratio of the nano-layer facial shells between the nanoparticles and the base fluids.
thickness to the original particle radius. Hence, the Maxwell equation

(Eq. (1)) can be modified into: ϕ keff −kb ϕ ðkeff −k2 Þð2k2 + k1 Þ−αðk1 −k2 Þð2k2 + keff Þ
1− + =0
α 2keff + kb α ð2keff + k2 Þð2k2 + k1 Þ + 2αðk1 −k2 Þðk2 −keff Þ
ð6Þ
kpe + 2kb + 2ðkpe −kb Þð1 + βÞ3 ϕ
keff = kb : ð3Þ
kpe + 2kb −ðkpe −kb Þð1 + βÞ3 ϕ where ϕ and ϕ/α are the volume fraction of the nanoparticles and the
complex nanoparticles, respectively. α = (R/(R + t))3, R is the radius
of the nanoparticles and t is the thickness of interfacial shell. k1 and k2
This modified Maxwell model can predict the presence of very thin
are the thermal conductivity of the nanoparticle and interfacial shell,
nano-layers, particularly when the particle diameter is b10 nm. It can
respectively. The model can interpret the size dependence of the
be concluded that the smaller the particle, the higher the thermal
thermal conductivity of nanofluids. The theoretical results on the
conductivity increase. In addition, the very small size of nanoparticles
effective thermal conductivity of CuO/water and CuO/EG nanofluids
should markedly improve the stability of the suspension.
with interfacial shells are in good agreement with experimental data
Yu and Choi [49] proposed a modified Hamilton–Crosser model to
[90].
include the particle–liquid interfacial layer for non-spherical particles.
Xie et al. [50] considered the interfacial nano-layer with linear
The effective thermal conductivity is expressed as
thermal conductivity distribution and proposed an effective thermal

 conductivity model to account for the effects of nano-layer thickness,
nϕeff A
keff = 1+ k ð4Þ nanoparticles size, volume fraction, and thermal conductivities of
1−ϕeff A b
fluid, nanoparticles, and nano-layer. Their formula is
where A is defined by !
3Θ2 ϕ2T
keff = 1 + 3ΘϕT + k ð7Þ
1 ðkpj −kb Þ 1−ΘϕT b
A= ∑
3 j = a;b;c kpj + ðn−1Þkb
with
and
β1b ½ð1 + γÞ3 −βp1 = βb1 
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi Θ=
ða2 + tÞðb2 + tÞðc2 + tÞ ð1 + γÞ3 + 2β1b βp1
ϕeff = pffiffiffiffiffiffiffiffi
abc
where
is the equivalent volume concentration of complex ellipsoids, which is k1 −kb
an imaged structure of elliptical particles with surrounding nano- β1b =
k1 + 2kb
layers, (a, b, c are the semiaxes of monosized ellipsoidal particles and
a ≥ b ≥ c). With a general empirical shape factor n (n = 3Ψ−α, here α is kp −k1
an empirical parameter and Ψ is the particle sphericity), this modified βp1 =
kp + 2k1
HC model can predict the thermal conductivity of carbon nanotube-
in-oil nanofluids reasonably well. However, it is not able to predict the
kb −k1
nonlinear behavior of the effective thermal conductivity of nanofluids. βb1 =
kb + 2k1
Xue [86] presented a model for the effective thermal conductivity
of nanofluids considering the effect of the interface between the solid
γ = δ/rp is the thickness ratio of nano-layer and nanoparticle, δ is the
particles and the base fluid based on Maxwell model and the average
thickness of nano-layer and rp is the radius of nanoparticle. ϕT is the
polarization theory. His formula of effective thermal conductivity is
modified total volume fraction of the original nanoparticle and nano-

 layer, ϕT = ϕ(1 + γ)3. Here, k1 is the average thermal conductivity of
9 1−
ϕ keff −kb
λ 2keff + kb
+
ϕ keff −kc;x
λ keff + B2;x ðkc;x −keff Þ½ ð5Þ the nano-layer. They claimed that the calculated values could agree
well with some available experimental data.
As mentioned above, temperature is an important factor influenc-

+4
keff −kc;y

2keff + ð1−B2;x Þðkc;y −keff Þ
=0
where λ = abc/[(a + t)(b + t)(c + t)]. Here a, b, c is half-radii of the
assumed elliptical nanoparticles and t is the thickness of elliptical
shell. k c , j is the effective dielectric constant and B2,x is the
depolarization factor along the x-symmetrical axis which is derived
from the average polarization theory. Xue claims that his model
predictions are in good agreement with the experimental data for
nanotube-in-oil nanofluids at high thermal conductivity and nonlin-
earity with nanotube loadings. But through the comparison with Xue's
model and careful check, Yu and Choi [49] found that the predicted
values from Xue's model are inaccurate for the using of two incorrect
parameters. Kim's [87] paper also reveals this conclusion.
Xue and Xu [88] obtained an equation for the effective thermal
conductivity according to Bruggeman model [89]. Their model takes
account of the effect of interfacial shells by replacing the thermal
conductivity of nanoparticles with the assumed thermal conductivity
ing the thermal conductivity of nanofluids [41,44–46]. One key reason
for this phenomenon may be the Brownian motion of nanoparticles at
the molecular and nanoscale levels. Considering this factor, some new
models based on Maxwell model were developed.
Xuan et al. [91] considered the random motion of suspended
nanoparticles (Brownian motion) based on Maxwell model and
proposed a modified formula for the effective thermal conductivity
as follows:
sffiffiffiffiffiffiffiffiffiffiffiffiffi
kp + 2kb −2ðkb −kp Þϕ ρp ϕcp kB T
keff = k + ð8Þ
kp + 2kb + 2ðkb −kp Þϕ b 2 3πrc μ
where the Boltzmann constant kB = 1.381 × 10− 23 J/K, rc is the
apparent radius of clusters and depends on the fractal dimension of
the cluster structure. The predictions obtained from this model are in
satisfactory agreement with experiment results, especially when the
effect of nanoparticle aggregation is taken into account. But the model
may be wrong for the second term of the equation has wrong units.
98 Y. Li et al. / Powder Technology 196 (2009) 89–101
Wang et al. [92] proposed a fractal model for predicting the
thermal conductivity of nanofluids based on the effective medium
approximation and the fractal theory. The fractal theory was proposed
firstly by Mandelbrot [94], a French mathematician. It can well
describe the disorder and stochastic process of clustering and
polarization of nanoparticles within the mesoscale limit. The thermal
conductivity can be calculated as:
∞
ð1−ϕÞ + 3ϕ∫0 kc1 ðrÞnðrÞ = ½kc1 ðrÞ + 2kb dr
keff = ∞
kb
ð1−ϕÞ + 3ϕ∫0 kb ðrÞnðrÞ = ðkc1 ðrÞ + 2kb Þdr
where kc1(r) is the thermal conductivity of particle clusters and n(r) is
the radius distribution function. Considering the effect of size and
surface adsorption of nanoparticles, the proposed model compared
well with their experimental data for 50 nm CuO particles suspended
in deionized water with particle concentrations lower than 0.5 vol%.
A comprehensive model considering the large enhancement of
thermal conductivity in nanofluids and its strong temperature
dependence, which was deduced from the Stokes–Einstein formula,
was proposed by Kumar et al. [95]. The thermal conductivity
enhancement taking into account the Brownian motion of the
particles can be expressed as:
2kb T ϕrb
keff = kb + c kb
πvd2p kb ð1−ϕÞrp
where c is a constant, ν is the dynamic viscosity of the base fluid, and
dp is the diameter of the particles. However, the validity of the model
in the molecular size regime has to be explored and it may not be
suitable for a large concentration of the particles where interactions of
particles become important. Besides, just like Eq. (8), the second term
of the equation does not have a unit of thermal conductivity.
Bhattacharya et al. [95] developed a technique to compute the
effective thermal conductivity of a nanofluid using Brownian motion
simulation. They combined the liquid conductivity and particle
conductivity as follows:
keff = ϕkp + ð1−ϕÞkb
where kp is replaced by the effective contribution of the particles
towards the overall thermal conductivity of the system, kp = k T12 V
B
∑nj = 0 〈Q ð0ÞQ ðjΔtÞ〉Δt, kB is the Boltzmann's constant, T is the tem-
perature, V is the volume of the domain, n is the number of time steps
used in the simulation, Δt is the time step, and hQ(0)Q(jΔt)〉 is the time
autocorrelation function of Q(t), Q(t) is the time autocorrelation
function of the heat flux operator. But the unit for the right hand side
of the above equation is not W/m K.
Jang and Choi [81] devised a new theoretical model for predicting
thermal conductivity of nanofluids. Four modes of energy transport in the
nanofluids such as collision between base fluid molecules (kb(1 −ϕ)),
thermal diffusion of nanoparticles in fluids (kpϕ), collision between
nanoparticles due to Brownian motion (neglected), and thermal
interaction of dynamic or “dancing” nanoparticles with the base fluid
molecules (fhδT) were considered. The resulting expression for the
effective thermal conductivity of nanofluids is:
db 2 nanotubes.
keff = kb ð1−ϕÞ + kp ϕ + 3C k Re Prϕ
dp b dp
where h ∼ (kb/dp)Redp2Pr2 and δT ∼ δ/Pr presents the heat transfer
coefficient for the flow past nanoparticles and the thickness of the
thermal boundary layer, respectively. Here δ ∼ 3db and db and dp are
the diameter of the base fluid molecule and nanoparticle respectively.
Redp is the Reynolds number defined by Redp = ðC̄ RM dp Þ = v, C is a
proportional constant, C̄ RM is the random motion velocity of the
nanoparticles, ν is the dynamic viscosity of base fluid, Pr is the Prandtl
number. The model considered the effects of concentration, temper-
ature, and particle size. The predictions from this model were in
agreement with experimental data of Lee et al. [90] and Eastman et al.
[9]. However, it may not be suitable in the high temperature for the
Brownian motion effect was neglected.
Prasher [82] considered that convection caused by Brownian motion
of nanoparticles is primarily responsible for the enhancement in the
effective thermal conductivity of nanofluids. By introducing the general
correlation for heat transfer coefficient h, the modified Maxwell model
ð9Þ including the convection of the liquid near the particles due to Brownian
movement was proposed as follows:
" #
m 0:333 kp + 2kb + 2ðkp −kb Þϕ
keff = ð1 + ARe Pr ϕÞ kb ð13Þ
kp + 2kb −ðkp −kb Þϕ
where h = kb/a(1 + ARemPr0.333ϕ), a represents the radius of the
nanoparticles, A and m are constants. The Reynolds number can be
written as:
sffiffiffiffiffiffiffiffiffiffiffiffiffi
1 18kb T
Re = :
ν πρp dp
The test result showed that the model matches very well with the
ð10Þ
experimental data for different fluid temperature by assuming
m = 2.6 and 2.4 for ϕ = 4% and ϕ = 1%, respectively, but there is an
error that exists in the unit for the equation of Re.
A new model for nanofluids, which takes account the effects of
particle size, particle volume fraction and temperature dependence as
well as properties of the base fluid and the particle subject to
Brownian motion, was recently developed by Koo and Kleinstreuer
[96]. The resulting formula is
sffiffiffiffiffiffiffiffiffi
kp + 2kb + 2ðkp −kb Þϕ 4 kB T
keff = k + 5 × 10 βϕρp cp f ðT; ϕÞ: ð14Þ
kp + 2kb −ðkp −kb Þϕ b ρp D
ð11Þ Note that the first part of Eq. (14) is obtained directly from the
Maxwell model while the second part accounts for Brownian motion,
which causes the temperature dependence of the effective thermal
conductivity. ρp is the particle density, cp is the specific heat of the
particle, kB is the Boltzmann constant, T is the fluid temperature, and D
is its diameter. f (T,φ) can be assumed to vary continuously with the
particle volume fraction while β is related to particle motion. f (T, φ) =
(−6.04φ + 0.4705)T + (1722.3φ − 134.63). However, there are two
small mistakes in this model. First, the unit of the second term at the
right hand side of the above f (T, φ) function is dimensionless while the
first term is K. Second, the thermal conductivity unit for the second
part of the Eq. (14) can't be got.
What should be noted is that most of the models were developed
for spherical or elongated nanoparticles. Nanotubes usually possess a
large aspect ratio and different from nanoparticles which are usually
spherical or elongated. For the different shape of nanotubes and
nanoparticles, the thermal conductivity of nanofluids containing
nanotubes is difficult to predict by the models suitable for nanopar-
ticles. So it is necessary to develop models for nanofluid containing
ð12Þ
Nan et al. [97] presented a quite simple formula for the thermal
conductivity enhancement in carbon nanotube composites based on a
conventional model. This simple formula predicts much higher
thermal conductivity enhancement even in the dilute case of the
carbon nanotubes, due to ultrahigh thermal conductivity and aspect
ratio of the carbon nanotubes. Nan showed a comparison between his
model (keff = 1 + ϕkp/3 kb) and a recent experiment on nanotube
suspensions [98] in his paper. The result indicated that the Nan's
model [97] agreed well with the experimental observations.
 Y. Li et al. / Powder Technology 196 (2009) 89–101
However, this model does not considering the thermal resistance
across the carbon nanotube/fluid interface, the aggregation and twist
of the nanotubes in the composites. Later, Nan et al. [98] modified
their model and tried to describe the effect of the interface thermal
resistance. However, the model still cannot explain the nonlinear
phenomena of the effective thermal conductivity of nanotube
suspensions with nanotube volume fractions.
Recently, considering the very large axial ratio and the space
distribution of the CNTs, Xue [99] proposed a model of effective
thermal conductivity of CNTs-nanofluids based on Maxwell
theory. With the assumed distribution function P(Bj) = 2, the
corresponding expression of the effective thermal conductivity of
CNT-based nanofluids is
kp kp + kb
1−ϕ + 2ϕk ln
p −kb 2kb
keff = kb kp + kb
: ð15Þ
kb
1−ϕ + 2ϕ k ln
p −kb 2kb
Xue [100] also presented a model of the effective thermal con-
ductivity for carbon nanotube composites by incorporating the
interface thermal resistance with an average polarization theory.
The proposed model includes the effects of carbon nanotube length,
diameter, concentration, interface thermal resistance, and the thermal
conductivities of nanotubes and base fluid on the thermal conductiv-
ity of the nanofluids simultaneously.
" #
c c
keff −kb keff −k33 keff −k11
9ð1−ϕÞ +ϕ +4
2keff + kb keff + 0:14 L ðk33 −keff Þ
d c
2keff + 2 ðk11 −keff Þ
1 c
ð16Þ
where
kc kc
k11c = 2RK kc
k33c = 2RK kc
: heat transfer applications, Particuology 7 (2009) 141–150.
1+ d
1+ L
The model predicts that the large length of the carbon nanotubes
embedded plays a key role in thermal conductivity enhancement,
while the large interface thermal resistance across the nanotube–
matrix interface causes a significant degradation. Predictions from the
novel model are in excellent agreement with the experimentally
observed values of the effective thermal conductivity of carbon
nanotube nanofluids from Xie et al. [22].
Although many models have been proposed, there are still many
phenomena about heat transfer of nanofluids that can't be inter-
preted. No theoretical models available in open literature for
predicting the thermal conductivity of nanofluids can be satisfactory.
An accurate and versatile model is required to account for the
reported observations on nanofluids. To achieve this object, more
experimental data are needed and much more work need to be done.
6. Conclusions and challenges
This report presents recent developments in research on station-
ary nanofluids. In general, nanofluids show many excellent properties
promising for engineering application. Many interesting phenomena
and valuable conclusions have been got. But there are still several
important issues that need to be solved for applications of nanofluids
in engineering.
(1) Although many nanofluid systems have been prepared, nanofluid
systems with special properties that can meet practical engi-
neering requirement have not been developed. For example,
nanofluids with high thermal conductivity, long-term stability
and dielectric property, which are potentially applicable to
advanced vehicles, is not available at present and need to be
exploited.
99
(2) The long-term stability of nanofluids is a key issue for both
scientific and practical application. To date, the long-term
stability of most studied nanofluids is not confirmed and more
basic theoretical and experimental work is required for
improving the stability of nanofluids.
(3) The factors influencing enhancement of thermal conductivity
of nanofluids need to be investigated systematically. The
mechanism influencing the thermal conductivity of nanofluids
is still unknown although many models have been proposed
to deal with the abnormal increase in thermal conductivity of
nanofluids.
(4) Whether the enhancement of thermal conductivity is abnormal
or not need to be investigated thoroughly.
(5) No uniform standard was presented for experimental research
on nanofluids, including the preparation of nanofluids, the
thermal conductivity measurement of nanofluids and the
stability evaluation of nanofluids.
References
[1] S.U.S. Choi, Enhancing thermal conductivity of fluids with nanoparticles,
Developments and Applications of Non-Newtonian Flows, FED-vol.231/MD-vol.
66, 1995, pp. 99–105.
[2] Pawel Keblinski1, JeffreyA. Eastman, DavidG. Cahill, Nanofluids for thermal
transport, Mater. Today 06 (2005) 36–44.
[3] Xiang-Qi Wang, ArunS. Mujumdar, Heat transfer characteristics of nanofluids: a
review, Int. J. Therm. Sci. 46 (2007) 1–19.
[4] S.M.S. Murshed, K.C. Leong, C. Yang, Thermophysical and electrokinetic properties
of nanofluids—a critical review, Appl. Therm. Eng. 28 (2008) 2109–2125.
[5] Sarit Kumar Das, Stephen U.S. Choi, Hrishikesh E. Patel, Heat transfer in nanofluids—
=0
a review, Heat Transf. Eng. 27 (2006) 3–19.
[6] J.A. Eastman, S.R. Phillpot, S.U.S. Choi, P. Keblinski, Thermal transport in nanofluids,
Annu. Rev. Mater. Res. 34 (2004) 219–246.
[7] Yulong Ding, Haisheng Chen, Liang Wang, Chane-Yuan Yang, Yurong He, Wei
Yang, WaiPeng Lee, Lingling Zhang, Ran Huo, Heat transfer intensification using
nanofluids, KONA 25 (2007).
[8] Dongsheng Wen, Guiping Lin, Saeid Vafaei, Kai Zhang, Review of nanofluids for
[9] J.A. Eastman, S.U.S. Choi, Anomalously increased effective thermal conductivities
of ethylene glycol-based nanofluids containing copper nanoparticles, Appl. Phys.
Lett. 78 (2001) 718–720.
[10] Min-Sheng Liu, MarkChing-Cheng Lin, C.Y. Tsai, Chi-Chuan Wang, Enhancement
of thermal conductivity with cu for nanofluids using chemical reduction method,
Int. J. Heat Mass Transfer 49 (2006) 3028–3033.
[11] Y. Xuan, Q. Li, Heat transfer enhancement of nanofluids, Int. J. Heat Mass Transfer
21 (2000) 58–64.
[12] T.K. Hong, H.S. Yang, C.J. Choi, Study of the enhanced thermal conductivity of Fe
nanofluids, J. Appl. Phys. 97 (064311) (2005) 1–4.
[13] H.E. Patel, S.K. Das, T. Sundararagan, A.S. Nair, B. Geoge, T. Pradeep, Thermal
conductivities of naked and monolayer protected metal nanoparticle based
nanofluids: manifestation of anomalous enhancement and chemical effects,
Appl. Phys. Lett. 83 (2003) 2931–2933.
[14] Shawn A. Putnam, David G. Cahill, Paul V. Braun, Thermal conductivity of
nanoparticle suspensions, J. Appl. Phys. 99 (084308) (2006) 1–6.
[15] H. Zhu, C. Zhang, S. Liu, Effects of nanoparticle clustering and alignment on
thermal conductivities of Fe3O4 aqueous nanofluids, Appl. Phys. Lett. 89 (23123)
(2006) 1–3.
[16] S.M.S. Murshed, K.C. Leong, C. Yang, Enhanced thermal conductivity of TiO2–
water based nanofluid, Int. J. Therm. Sci. 44 (2005) 367–373.
[17] H. Xie, J. Wang, T. Xi, Y. Liu, F. Ai, Thermal conductivity enhancement of suspensions
containing nanosized alumina particle, J. Appl. Phys. 91 (2002) 4568–4572.
[18] X. Zhang, H. Gu, M. Fujii, Experimental study on the effective thermal conductivity
and thermal diffusivity of Nanofluid, Int. J. Thermophys. 27 (2006) 569–580.
[19] H.Q. Xie, et al., Study on the thermal conductivity of SiC nanofluids, J. Chin. Ceram.
Soc. 29 (2001) 361–364 (in Chinese).
[20] M.S. Liu, M.C. Lin, I.Te. Huang, C.C. Wang, Enhancement of thermal conductivity
with carbon nanotube for nanofluids, Int. Commun. Heat Mass Transf. 32 (2005)
1202–1210.
[21] S.U.S. Choi, Z.G. Zhang, W. Yu, F.E. Lockwood, E.A. Grulke, Anomalous thermal
conductivity enhancement in nanotube suspensions, Appl. Phys. Lett. 79 (2001)
2252–2254.
[22] Huaqing Xie, Hohyun Lee, Wonjin Youn, Mansoo Choi, Nanofluids containing
multiwalled carbon nanotubes and their enhanced thermal conductivities, J. Appl.
Phys. 94 (2003) 4967–4971.
[23] B. Yang, Z.H. Han, Thermal conductivity enhancement in water-in-FC72
nanoemulsion fluids, Appl. Phys. Lett. 88 (261914) (2006) 1–3.
[24] Kun-Quan Ma, Jing Liu, Nano liquid–metal fluid as ultimate coolant, Phys. Lett., A
361 (2007) 252–256.
[25] C.H. Lo, T.T. Tsung, L.C. Chen, C.H. Su, H.M. Lin, Fabrication of copper oxide nanofluid
using submerged arc nanoparticle synthesis system (SANSS), J. Nanopart. Res. 7
(2005) 313–320.
100 Y. Li et al. / Powder Technology 196 (2009) 89–101
[26] C.H. Lo, T.T. Tsung, L.C. Chen, Shape-controlled synthesis of Cu-based Nanofluid
using submerged arc nanoparticle synthesis system (SANSS), J. Cryst. Growth
277 (2005) 636–642.
[27] H. Zhu, Y. Lin, Y. Yin, A novel one-step chemical method for preparation of copper
nanofluids, J. Colloid Interface Sci. 227 (2004) 100–103.
[28] K.S. Hong, T.K. Hong, H.S. Yang, Thermal conductivity of Fe nanofluids depending
on the cluster size of nanoparticles, Appl. Phys. Lett. 88 (31901) (2006) 1–3.
[29] Q. Li, Y. Xuan, The heat conduction properties of fluxion and convective of Cu/
H2O nanofluids, Sci. China, Ser. E 32 (2002) 331–337 (in chinese).
[30] Y. Xuan, Q. Li, Heat transfer enhancement of nanofluids, J. Eng. Thermophys. 21
(2000) 466–470 (in Chinese).
[31] Y. Hwang, H.S. Park, J.K. Lee, W.H. Jung, Thermal conductivity and lubrication
characteristics of nanofluids, Curr. Appl. Phys. 6S1 (2006) e67–e71.
[32] Y. Hwang, J.K. Lee, C.H. Lee, Y.M. Jung, S.I. Cheong, C.G. Lee, B.C. Ku, S.P. Jang,
Stability and thermal conductivity characteristics of nanofluids, Thermochim.
Acta 455 (2007) 70–74.
[33] Xinfang Li, Dongsheng Zhu, Xianju Wang, Evaluation on dispersion behavior of
the aqueous copper nano-suspensions, J. Colloid Interface Sci. 3 (10) (2007)
456–463.
[34] X. Peng, X. Yu, Influence factors on suspension stability of nanofluids, J. Zhejiang
Univ.: Eng. Sci. 41 (2007) 577–580 (in Chinese).
[35] Buxuan Wang, Chunhui Li, Xiaofeng Peng, Research on stability of nano-particle
suspension, J. Univ. Shanghai Sci. Technol. 25 (2003) 209–212 (in Chinese).
[36] Buxuan Wang, Chunhui Li, Xiaofeng Peng, Stability of nano-particle suspensions,
J. Basic Sci. Eng. 11 (2003) 169–173 (in Chinese).
[37] Dean-Mo Liu, Influence of dispersant on powders dispersion and properties of
zirconia green compacts [J], Ceram. Int. 26 (3) (2000) 279–287.
[38] Dongsheng Zhu, Xinfang Li, Nan Wang, Xianju Wang, Jinwei Gao, Hua Li,
Dispersion behavior and thermal conductivity characteristics of Al2O3–H2O
nanofluids, Curr. Appl. Phys. 9 (2009) 131–139.
[39] Q. Yu, Y.J. Kim, H. Ma, Plasma treatment of diamond nanoparticles for dispersion
improvement in water, Appl. Phys. Lett. 88 (231503) (2006) 1–3.
[40] X. Zhang, H. Gu, M. Fujii, Effective thermal conductivity and thermal diffusivity of
nanofluids containing spherical and cylindrical nanoparticles, J. Appl. Phys. 100
(044325) (2006) 1–5.
[41] Q. Shou, R. Chen, Research on thermal conductivity of metal-oxide nanofluids,
Mater. Rev. 20 (2006) 117–119 (in chinese).
[42] X.F. Zhou, L. Gao, Effective thermal conductivity in nanofluids of nonspherical
particles with interfacial thermal resistance: differential effective medium
theory, J. Appl. Phys. 100 (024913) (2006) 1–6.
[43] L. Gao, X. Zhou, Differential effective medium theory for thermal conductivity in
nanofluids, Phys. Lett., A 348 (2006) 355–360.
[44] B. Yang, Z.H. Han, Temperature-dependent thermal conductivity of nanorod-
based nanofluids, Appl. Phys. Lett. 89 (083111) (2006) 1–3.
[45] CalvinH. Li, G.P. Peterson, Experimental investigation of temperature and volume
fraction variations on the effective thermal conductivity of nanoparticle
suspensions (nanofluids), J. Appl. Phys. 99 (084314) (2006) 1–8.
[46] S.K. Das, N. Putta, P. Thiesen, W. Roetzel, Temperature dependence of thermal
conductivity enhancement for nanofluids, J. Heat Transfer 125 (2003)
567–574.
[47] K.C. Leong, C. Yang, S.M.S. Murshed, A model for the thermal conductivity
of nanofluids—the effect of interfacial layer, J. Nanopart. Res. 8 (2006)
245–254.
[48] W. Yu, S.U.S. Choi, The role of interfacial layers in the enhanced thermal of
nanofluids: a renovated maxwell model, J. Nanopart. Res. 5 (2003) 167–171.
[49] W. Yu, S.U.S. Choi, The role of interfacial layers in the enhanced thermal
conductivity of nnanofluids: a renovated Hamilton–Crosser model, J. Nanopart.
Res. 6 (2004) 355–361.
[50] H. Xie, M. Fujii, X. Zhang, Effect of interfacial nanolayer on the effective thermal
conductivity of nanoparticle-fluid mixture, Int. J. Heat Mass Transfer 48 (2005)
2926–2932.
[51] L. Xue, P. Keblinski, S.R. Phillpot, S.U.-S. Choi, J.A. Eastman, Effect of liquid layering
at the liquid–solid interface on thermal transport, Int. J. Heat Mass Transfer 47
(2004) 4277–4284.
[52] W. Evans, J. Fish, P. Keblinski, Role of Brownian motion hydrodynamics on
nanofluid thermal conductivity, Appl. Phys. Lett. 88 (093116) (2006) 1–3.
[53] P. Keblinski, R. Prasher, J. Eapen, Thermal conductance of nanofluids: is the
conversy over? J. Nanopart. Res. 10 (2008) 1089–1097.
[54] C. Nie, W.H. Marlow, Y.A. Hassan, Discussion of proposed mechanisms of thermal
conductivity enhancement in nanofluids, Int. J. Heat Mass Transfer 51 (2008)
1342–1348.
[55] R. Prasher, D. Song, J. Wang, Measurements of nanofluid viscosity and its
implications for thermal applications, Appl. Phys. Lett. 89 (133108) (2006) 1–3.
[56] J.M. Li, Z.L. Li, B.X. Wang, Experimental viscosity measurements for copper oxide
nanoparticle suspensions, Tsinghua Sci. Technol. 7 (2002) 198–201.
[57] Haisheng Chen, Yulong Ding, Yurong He, Chunqing Tan, Rheological behaviour of
ethylene glycol based titania nanofluids, Chem. Phys. Lett. 444 (2007) 333–337.
[58] Shunsong Guo, Zhongyang Luo, Tao Wang, Jiafei Zhao, Kefa Cen, Viscosity of
monodisperse silica nanofluids, Bull. Chin. Ceram. Soc. 25 (2006) 52–55 (in Chinese).
[59] C.T. Nguyen, F. Desgranges, G. Roy, N. Galanis, et al., Temperature and particle-
size dependent viscosity data for water-based nanofluid-hysteresis phenome-
non, International Journal of Heat and Fluid Flow 28 (2007) 1492–1506.
[60] S.K. Das, N. Putra, W. Roetzel, Pool boiling characteristics of nano-fluids, Int. J. Heat
Mass Transfer 46 (2003) 851–862.
[61] Y. Ding, H. Alias, D. Wen, R.A. Williams, Heat transfer of aqueous suspensions of
carbon nanotubes (CNT nanofluids), Int. J. Heat Mass Transfer 49 (2006) 240–250.
[62] A. Einstein, Eine neue Bestimmung der Molekul-dimension (a new determina-
tion of the molecular dimensions), Annalen der Physik 19 (2) (1906) 289–306.
[63] H.C. Brinkman, The viscosity of concentrated suspensions and solution, J. Chem.
Phys. 20 (1952) 571–581.
[64] T.S. Lundgren, Slow flow through stationary random beds and suspensions of
spheres, J. Fluid Mech. 51 (1972) 273–299.
[65] G.K. Batchelor, The effect of Brownian motion on the bulk stress in a suspension
of spherical particles, J. Fluid Mech. 83 (1) (1977) 97–117.
[66] C.T. Nguyen, F. Desgranges, N. Galanis, G. Roy, T. Maré, S. Boucher, H. Angue
Mintsa, Viscosity data for Al2O3–water nanofluid–hysteresis: is heat transfer
enhancement using nanofluids reliable? Int. J. Therm. Sci. 47 (2008) 103–111.
[67] Ji-Hwan Lee, KyoSik Hwang, SeokPil Janga, ByeongHo Lee, JunHo Kim,
Stephen U.S. Choi, ChulJin Choi, Effective viscosities and thermal conductivities
of aqueous nanofluids containing low volume concentrations of Al2O3 nano-
particles, Int. J. Heat Mass Transfer 51 (2008) 2651–2656.
[68] D.P. Kulkarni, D.K. Das, G.A. Chukwu, Temperature dependent rheological
property of copper oxide nanoparticles suspension, J. Nanosci. Nanotechnol. 6
(2006) 1150–1154.
[69] PraveenK. Namburu, DevdattaP. Kulkarni, Debasmita Misra, DebendraK. Das,
Viscosity of copper oxide nanoparticles dispersed in ethylene glycol and water
mixture, Exp. Therm. Fluid Sci. 32 (2007) 397–402.
[70] Ying Yang, Z.George Zhang, EricA. Grulke, WilliamB. Anderson, Gefei Wu, Heat
transfer properties of nanoparticle-in-fluid dispersions (nanofluids) in laminar
flow, Int. J. Heat Mass Transfer 48 (2005) 1107–1116.
[71] Yurong He, Yi Jin, Haisheng Chen, Yulong Ding, Daqiang Cang, Huilin Lu, Heat
transfer and flow behaviour of aqueous suspensions of TiO2 nanoparticles
(nanofluids) flowing upward through a vertical pipe, Int. J. Heat Mass Transfer 50
(2007) 2272–2281.
[72] P. Keblinski, Mechanisms of heat flow in suspensions of nano-sized particles
(nanofluids), Int. J. Heat Mass Transfer 45 (2002) 855–863.
[73] J.A. Eastman, S.R. Phillpot, S.U.S. Choi, P. Keblinski, Thermal transport in
nanofluids, Annu. Rev. Mater. Res. 34 (2004) 219–246.
[74] Q. Li, Y. Xuan, A preliminary analysis of the intensified thermal conductivity
mechanism of nanofluids, J. Eng. Therm. Energy Power 17 (584) (2002) 568–571.
[75] H.Q. Xie, T.G. Xi, J.C. Wang, Study on the mechanism of heat conduction in
nanofluid medium, Acta Phys. Sin. 52 (2003) 1444–1449.
[76] R. Prasher, P.E. Phelan, P. Bhattacharya, Effect of aggregation kinetics on the
thermal conductivity of nanoscale colloidal solutions (nanofluid), Nano Lett. 6
(2006) 1529–1534.
[77] W. Evans, R. Prasher, J. Fish, P. Meakin, P. Phelan, P. Keblinski, Effect of aggregation
and interfacial thermal resistance on thermal conductivity of nanocomposite and
colloidal nanofluids, Int. J. Heat Mass Transfer 51 (2008) 1431–1438.
[78] ShawnA. Putnam, DavidG. Cahill, BenjaminJ. Ash, LindaS. Schadler, High-precision
thermal conductivity measurements as a probe of polymer/nanoparticle interfaces,
J. Appl. Phys. 94 (2003) 6785–6788.
[79] Ce-Wen Nan, R. Birringer, D.R. Clarke, H. Gleiter, Effective thermal conductivity of
particulate composites with interfacial thermal resistance, J. Appl. Phys. 81 (1997)
6692–6699.
[80] Chan Hee Chon, Kenneth D. Kihm, Shin Pyo Lee, Stephen U.S. Choi, Empirical
correlation finding the role of temperature and particle size for nanofluid (Al2O3)
thermal conductivity enhancement, Appl. Phys. Lett. 87 (153107) (2005) 1–3.
[81] S.P. Jang, S.U.S. Choi, Role of Brownian motion in the enhanced thermal
conductivity of nanofluids, Appl. Phys. Lett. 84 (2004) 4316–4318.
[82] R. Prasher, P. Bhattacharya, P.E. Phelan, Thermal conductivity of nanoscale
colloidal solutions (nanofluids), Phys. Rev. Lett. 94 (025901) (2005) 1–4.
[83] J. Koo, C. Kleinstreuer, Impact analysis of nanoparticle motion mechanisms on the
thermal conductivity of nanofluids, Int. Commun. Heat Mass Transf. 32 (2005)
1111–1118.
[84] J.C. Maxwell, A Treatise on Electricity and Magnetism, Second ed.Clarendon Press,
Oxford, UK, 1881.
[85] L. Schwartz, E. Garboczi, D. Bentz, Interfacial transport in porous media: application
to dc electrical conductivity of mortars, J. Appl. Phys. 78 (1995) 5898–5908.
[86] Q.-Z. Xue, Model for effective thermal conductivity of nanofluids, Phys. Lett., A
307 (2003) 313–317.
[87] J. Kim, Y.T. Kang, C.K. Choi, Analysis of convective instability and heat transfer
characteristics of nanofluids, Phys. Fluids 16 (2004) 2395–2401.
[88] Q. Xue, W.-M. Xu, A model of thermal conductivity of nanofluids with interfacial
shells, Mater. Chem. Phys. 90 (2005) 298–301.
[89] D.A.G. Bruggeman, Berechnung verschiedener physikalischer Konstanten
von heterogenen Substanzen, I. Dielektrizitatskonstanten und Leitfahigkeiten
der Mischkorper aus isotropen Substanzen, Ann. Phys. (Leipz.) 24 (1935)
636–679.
[90] S. Lee, S.U.S. Choi, S. Li, J.A. Eastman, Measuring thermal conductivity of fluids
containing oxide nnanoparticles, J. Heat Transfer 121 (1999) 280–289.
[91] Y. Xuan, Q. Li, W. Hu, Aggregation structure and thermal conductivity of nanofluids,
AIChE J. 49 (2003) 1038–1043.
[92] B.X. Wang, L.P. Zhou, X.F. Peng, A fractal model for predicting the eeffective
thermal conductivity of liquid with suspension of nnanoparticles, Int. J. Heat Mass
Transfer 46 (2003) 2665–2672.
[93] B.B. Mandelbrot, The Fractal Geometry of Nature, W.H. Freeman Press, San Francisco,
1982.
[94] D.H. Kumar, H.E. Patel, V.R.R. Kumar, T. Sundararajan, T. Pradeep, S.K. Das, Model
for heat conduction in nanofluids, Phys. Rev. Lett. 93 (144301) (2004) 1–4.
[95] P. Bhattacharya, S.K. Saha, A. Yadav, P.E. Phelan, R.S. Prasher, Brownian dynamics
simulation to determine the effective thermal conductivity of nanofluids, J. Appl.
Phys. 95 (2004) 6492–6494.
 Y. Li et al. / Powder Technology 196 (2009) 89–101 101

[96] J. Koo, C. Kleinstreuer, A new thermal conductivity model for nanofluids, J. Nanopart.
Res. 6 (2004) 577–588.
[97] C.-W. Nan, Z. Shi, Y. Lin, A simple model for thermal conductivity of carbon
nanotube-based composites, Chem. Phys. Lett. 375 (2003) 666–669.
[98] C.-W. Nan, G. Liu, Y. Lin, M. Li, Interface effect on thermal conductivity of carbon
nanotube composites, Appl. Phys. Lett. 85 (2004) 3549–3551.
[99] Q.Z. Xue, Model for thermal conductivity of carbon nanotube-based composites,
Phys., B Condens. Matter 368 (2005) 302–307.
[100] Q. Xue, Model for the effective thermal conductivity of carbon nanotube composites,
Nanotechnology 17 (2006) 1655–1660.