Xiaomei Qin1, Shihao Yang1, Yapei Chen1,2, Xiaoyun Qin1, Jianbo Zhao1, Wenjun Fang3, Dan Luo2. 1. School of Material and Chemical Engineering, Zhengzhou University of Light Industry, Zhengzhou, Henan Province 450000, China. 2. State Key Laboratory of Heavy Oil Processing, College of New Energy and Materials, Beijing Key Laboratory of Biogas Upgrading Utilization, China University of Petroleum Beijing, Beijing 102249, China. 3. Department of Chemistry, Zhejiang University, Hangzhou, Zhejiang Province 310027, China.
Abstract
Palladium nanoparticles, which were prepared by modified hyperbranched polyglycerol (mHPG) as stabilizers, can be dispersed well in nonpolar organic solvents and form highly stable nanofluids. The influences of three mHPG products modified with cyclohexanethiol (CSHPG), dodecanethiol (DSHPG), and octadecanethiol (OSHPG) on the preparation and stability of the palladium nanoparticles were investigated. The stability and thermal conductivity enhancement of the hydrocarbon-based nanofluids with Pd@mHPG (Pd@CSHPG, Pd@DSHPG, and Pd@OSHPG) compared to the corresponding base fluid were investigated at different temperatures. The average diameters of nanoparticles stabilized by CSHPG, DSHPG, and OSHPG are within 2.7-3.6 nm. The palladium nanoparticles could be dispersed well in the nonpolar base fluid such as decalin. The nanofluids with Pd@DSHPG and Pd@OSHPG could remain stable for up to 330 days at room temperature. The nanofluid with Pd@DSHPG or Pd@OSHPG could be stable for more than 24 h at 110 °C. The thermal conductivity of the nanofluids improved with increasing temperature and the mass fraction of nanoparticles compared to the corresponding base fluid. The long alkyl chain-modified HPG can give better protection for nanoparticles from agglomeration and assist metal nanoparticles in enhancing the thermal conductivity of nanofluids.
Palladium nanoparticles, which were prepared by modified hyperbranched polyglycerol (mHPG) as stabilizers, can be dispersed well in nonpolar organic solvents and form highly stable nanofluids. The influences of three mHPG products modified with cyclohexanethiol (CSHPG), dodecanethiol (DSHPG), and octadecanethiol (OSHPG) on the preparation and stability of the palladium nanoparticles were investigated. The stability and thermal conductivity enhancement of the hydrocarbon-based nanofluids with Pd@mHPG (Pd@CSHPG, Pd@DSHPG, and Pd@OSHPG) compared to the corresponding base fluid were investigated at different temperatures. The average diameters of nanoparticles stabilized by CSHPG, DSHPG, and OSHPG are within 2.7-3.6 nm. The palladium nanoparticles could be dispersed well in the nonpolar base fluid such as decalin. The nanofluids with Pd@DSHPG and Pd@OSHPG could remain stable for up to 330 days at room temperature. The nanofluid with Pd@DSHPG or Pd@OSHPG could be stable for more than 24 h at 110 °C. The thermal conductivity of the nanofluids improved with increasing temperature and the mass fraction of nanoparticles compared to the corresponding base fluid. The long alkyl chain-modified HPG can give better protection for nanoparticles from agglomeration and assist metal nanoparticles in enhancing the thermal conductivity of nanofluids.
As
an innovative kind of thermal fluid based on nanotechnology,
nanofluids can significantly improve the efficiency of the energy
conversion process.[1−3] A nanofluid refers to a stable colloidal mixture
of nanoparticles in a base fluid.[4] Since
the suspended nanoparticles in nanofluids could improve the thermal
conductivity of a common fluid, they have tremendous potential for
different industrial cooling applications such as heat exchangers,
cooling devices, solar collectors, and so on.[5,6] Moreover,
the nanofluid could be regarded as promising coolants and fuels to
be utilized in an aircraft power system.[7] A growing number of studies concentrate on nanofluids suspending
metal nanoparticles used in various thermal fields.[8−10] However, the
preparation and use of nanofluids with metal nanoparticles still face
some limitations, especially the acquisition of a solid–liquid
suspension with good stability, which is still a major challenge.[11,12]The small size, high surface activity, and van der Waals interactions
among metal nanoparticles make them difficult to be dispersed well
in the base fluid.[13,14] A nanofluid with long-term stability
is what we expected. The agglomeration and sedimentation of nanoparticles
are hindrances to acquire nanofluids with high stability, which may
even cause pipe blockage and finally influence the performance of
fluids in usage.[15] Preparation of metal
nanoparticles with good dispersing ability is the key to obtain an
oil-based nanofluid with good stability and high thermal conductivity
enhancement. The conventional method is adding surfactants or modification
agents, while excess surfactants may lead to uneven phonon scattering
and affect important properties, including viscosity and heat transfer
ability of nanofluids.[16−18]The metal nanoparticles could be dispersed
by the hyperbranched
polymers. As a class of polymer with highly branched three-dimensional
spherical structures, the hyperbranched molecules have unique physical
and chemical properties, such as good solubility, low viscosity, and
so on.[19−21] Furthermore, the hyperbranched polymer, which has
a large number of functional groups, can be used as a kind of effective
surface modifier to stabilize metal nanoparticles.[22] The nanoparticles could be suspended in fluids without
increasing the pressure drop and pumping power caused by increasing
the density and viscosity of nanofluids.[23,24] Liu et al.[25] synthesized the hyperbranched
polymer-decorated Au nanoparticles in aqueous media with a remarkable
stability at 200 °C. Skaria et al.[26] proposed a convenient approach to prepare hyperbranched polymers
with a thioether shell for stabilization of coinage nanoparticles
in organic media. Hyperbranched polyglycerol (HPG) has high-flexible
aliphatic polyether backbones and possesses the ability to stabilize
metal nanoparticles close to that of dendrimers. The internal spatial
structure could limit the growth of particles and could endow nanoparticles
with small particle sizes and a large specific surface area. Through
modification of HPG with nonpolar alkyl groups, the highly organosoluble
amphiphilic structures with the hydrophobic shells could be obtained.[19] However, only a few applications of hyperbranched
polymers in oil-based or hydrocarbon-based nanofluids have been reported
to date.[27,28]In this work, the hydrophobic palladium
nanoparticles are prepared
with long-chain alkyl- or cycloalkyl group-modified hyperbranched
polyglycerol (mHPG) as dispersants and stabilizers. mHPG could provide
metal nanoparticles salient dispersibility and stability in oil base
fluids, and the process allowed us to prepare palladium nanofluids
with good stability. The influences of three mHPG products modified
with cyclohexanethiol (CSHPG), dodecanethiol (DSHPG), and octadecanethiol
(OSHPG) on the preparation and stability of the palladium nanoparticles
were investigated. The thermal stability of nanofluids in the temperature
range of 110–140 °C was investigated. The effects of the
palladium nanoparticles (Pd@OSHPG, Pd@DSHPG, and Pd@CSHPG) that enhance
the thermal conductivity of nanofluids were studied, and the viscosity
of nanofluids was measured.
Results and Discussion
Characterization of HPG and mHPG
The 1H
NMR spectra of HPG, HPG-MA, and mHPG (CSHPG, DSHPG,
and OSHPG) are shown in Figure a. The significant signals for ether bonds of HPG appear at
3.2–4.0 ppm. After the successful transesterification reaction,
vinyl group signals can be seen around 5.6 and 6.1 ppm, and methyl
signals of the methacryloyl group appear at 1.9 ppm. After the thiol-ene
click reaction of HPG-MA with alkylthiol, methyl and methylene signals
for alkyl groups of CSHPG, DSHPG, or OSHPG can be observed between
0.8 and 2.1 ppm. The disappearance of signals for C=C bonds
at 5.6 and 6.1 ppm also indicate that the three mHPG products have
been successfully prepared.
Figure 1
(a) 1H NMR spectra and (b) FTIR spectra
of HPG, HPG-MA,
and mHPG (CSHPG, DSHPG, and OSHPG).
(a) 1H NMR spectra and (b) FTIR spectra
of HPG, HPG-MA,
and mHPG (CSHPG, DSHPG, and OSHPG).The FTIR results of HPG, HPG-MA, and mHPG are shown in Figure b. The broad absorption
bands of the O–H and C–O vibration of HPG appear around
3400 and 1120 cm–1, respectively. After transesterification,
the absorption peak of C=C (around 1570 cm–1) is observed. The absorption bands around 3003 and 1720 cm–1 belong to the stretching vibrations of =C—H and C=O,
respectively. In the FTIR spectra of CSHPG, DSHPG, and OSHPG, the
absorption band of C=C decreases and the typical band of S–C
appears in a range of 600–700 cm–1. The signals
of the 1H NMR and FTIR spectra are consistent with previous
reports.[29,30]
Characterization of Pd@mHPG
The morphology
and size distribution of Pd@mHPG (Pd@CSHPG, Pd@DSHPG, and Pd@OSHPG)
are observed by transmission electron microscopy (TEM) (Figure ). Three kinds of Pd nanoparticles
were well dispersed. The average particle sizes of Pd@CSHPG, Pd@DSHPG,
and Pd@OSHPG are about 3.4, 2.7, and 3.6 nm, respectively. It can
be seen that the prepared Pd nanoparticles have small particle sizes
(2.7–3.6 nm) and narrow distribution due to the encapsulation
capabilities of mHPG. The outer shell of the hyperbranched polymer
can protect the nanoparticles from aggregation between multiple molecules
and provide good dispersiblity of the Pd nanoparticles in nonpolar
base fluids. The particle size contributes in dispersion stability
and predominantly influences the thermal conductivity of the nanofluids.
The small nanoparticle size could enhance the surface-to-volume ratio
and improve interfacial heat exchange resulting from the improved
particle specific surface area and mobility of the small nanoparticles.[31,32] In other words, the small particle size means that the metal nanoparticle
has a large specific surface area and good thermal conductivity. As
a contrast, the Pd nanoparticles stabilized by dodecanethiol (Pd@S)
were prepared, and the TEM image of Pd@S is shown in Figure S1 of the Supporting Information. The average particle size of Pd@S is about 3.0 nm with unobvious
aggregation.
Figure 2
TEM images and size distributions of (a) Pd@CSHPG, (b)
Pd@DSHPG,
and (c) Pd@OSHPG.
TEM images and size distributions of (a) Pd@CSHPG, (b)
Pd@DSHPG,
and (c) Pd@OSHPG.Figure shows the
X-ray photoelectron spectroscopy (XPS) spectrum of Pd@DSHPG. It is
clear that the Pd nanoparticles contain S, C, Pd, and O elements.
This suggests the formation of Pd@DSHPG. The Pd 3d spectrum shows
two peaks (336.8 and 335.8 eV) for 3d5/2 and 342.05 eV
for Pd 3d3/2. The binding energy of 335.8 eV (Pd 3d5/2) corresponds to the metallic Pd0. The peak around
336.8 eV corresponding to Pd2+ is attributed to anchoring
with the thioether structures.
Figure 3
(a) Survey XPS spectrum of Pd@DSHPG. (b)
High-resolution XPS spectrum
of Pd in Pd@DSHPG.
(a) Survey XPS spectrum of Pd@DSHPG. (b)
High-resolution XPS spectrum
of Pd in Pd@DSHPG.
Stability
of Nanofluids
Considering
that nanofluids need to be stored for a long time, the stability of
nanofluids of Pd@mHPG under different conditions was investigated.
The schematic diagram of the stability of nanofluids with 0.1 wt %
nanoparticles at room temperature is shown in Figure . It can be seen that the nanoparticles stabilized
by different mHPG products are well dispersed in the base liquid to
form a uniform and quasi-homogeneous system (photograph shown in Figure a). The stability
of nanofluids with Pd@CSHPG, Pd@DSHPG, or Pd@OSHPG was tested against
time at room temperature by a UV–vis spectrograph. The relative
content of metal nanoparticles in the nanofluid is calculated by the
ratio of absorbance at different storage times to the initial absorbance.
The relative content of Pd@DSHPG or Pd@OSHPG in the nanofluid remains
almost unchanged, and the nanofluids with Pd@DSHPG and Pd@OSHPG remain
stable for 330 days (Figure b). The relative content of Pd@CSHPG in the nanofluid decreased
significantly in the first few days, and little precipitation is observed
at the bottom of the liquid. The rate of deposition becomes relatively
slow in the following days. The CSHPG encapsulated nanoparticles are
less stable than the other two. It is probably because the peripheral
ligand of CSHPG is the cyclohexyl groups, and the interaction between
metal nanoparticles and CSHPG is limited due to the steric hindrance
of CSHPG. The metal nanoparticles gradually agglomerate with increasing
storage time, and more than 50% of the Pd@CSHPG nanoparticles are
stably dispersed for 330 days. The TEM images of Pd@CSHPG, Pd@DSHPG,
and Pd@OSHPG in nanofluids after a 330 day storage period are shown
in Figure S2. It can be seen that the Pd@CSHPG
nanoparticles are partly agglomerated, and Pd@DSHPG and Pd@OSHPG in
nanofluids remain highly dispersed.
Figure 4
(a) Photographs of nanofluids with Pd@CSHPG,
Pd@DSHPG, and Pd@OSHPG
with a mass fraction of 0.1%. (b) Relative content of the suspended
particle in nanofluids with storage time at room temperature.
(a) Photographs of nanofluids with Pd@CSHPG,
Pd@DSHPG, and Pd@OSHPG
with a mass fraction of 0.1%. (b) Relative content of the suspended
particle in nanofluids with storage time at room temperature.The nanoparticles could agglomerate at high temperatures.
The stability
of three nanofluids under different temperatures was evaluated by
a UV–vis spectrograph. The nanofluids were considered to be
unstable when there was precipitation, and the absorbance decreased
markedly. The heating process stopped when the absorbance dropped
to zero. The heating stability times for nanofluids with 0.1 wt %
Pd@CSHPG, Pd@DSHPG, or Pd@OSHPG at different temperatures (110–140
°C) are shown in Figure . Every experimental result for thermal stability is the average
of three measurements. It was shown that the thermal stability times
for three nanofluids decreased with increasing temperature. The nanofluid
with Pd@DSHPG and Pd@OSHPG could be stable for more than 24 h at 110
°C. The stability time of a nanofluid with Pd@CSHPG is less than
that of Pd@DSHPG or Pd@OSHPG at the same heating temperature. The
heating stability for the nanofluids with 0.1 wt % Pd@S is shown in Figure S3. The thermal stability times for the
nanofluids with Pd@S decreased with increasing temperature, and the
thermal stability time is less than 24 h at 110 °C. DSHPG or
OSHPG could utilize its backbone to encapsulate the metal nanoparticles,
while the steric hindrance of CSHPG limits the interaction with metal
nanoparticles as the peripheral ligand of CSHPG is the cyclohexyl
groups. The progressive breaking of the skeleton of polymer molecules
with increasing temperature causes the coated metal nanoparticles
to be exposed, and the thermal stability of the nanoparticles reduces
significantly. The hyperbranched molecules modified with alkyl chains
could effectively protect the metal nanoparticles in a certain temperature
range, which expands the applicable conditions for the storage and
application of nanofluids.
Figure 5
Thermal stability of nanofluids with Pd@CSHPG,
Pd@DSHPG, and Pd@OSHPG
at a mass fraction of 0.1%.
Thermal stability of nanofluids with Pd@CSHPG,
Pd@DSHPG, and Pd@OSHPG
at a mass fraction of 0.1%.
Thermal Conductivity of Nanofluids
Compared
with traditional heat transfer fluids, nanofluids have opened
up new directions in the field of thermal conductivity enhancement.[6,33] The thermal conductivity of nanofluids with Pd@CSHPG, Pd@DSHPG,
or Pd@OSHPG was measured for different mass fractions at different
temperatures. In this work, the improvement effect of nanoparticles
on the thermal conductivity for nanofluids is expressed as the ratio
of thermal conductivity (k/k0). k is the thermal conductivity of the nanofluid,
and k0 is that of the nonpolar base fluid
(decalin). As shown in Figure , the thermal conductivity of nanofluids with Pd@CSHPG, Pd@DSHPG,
or Pd@OSHPG increased with increasing temperature and/or the mass
fraction of nanoparticles. The experimental results showed obvious
improvements in thermal conductivity for three nanofluids compared
to the thermal conductivity for the base fluid. Figure a shows the variation of thermal conductivity
for nanofluids with Pd@CSHPG at different mass fractions with temperatures.
At 25 °C, the enhancement of thermal conductivity is about 1,
3, or 5% for 0.05, 0.1, or 0.3 wt % of Pd@CSHPG, respectively. As
the temperature increased to 50 °C, the improvements in thermal
conductivity of the nanofluids were observed to be approximately 16,
20, and 35%. For 0.1 wt % Pd@DSHPG, the enhancement of thermal conductivity
of the nanofluid is 4% at 20 °C and 30% at 50 °C. For the
nanofluids with Pd@OSHPG, the thermal conductivity increases significantly.
When the added nanoparticles of Pd@OSHPG increases from 0.05 to 5
wt %, the enhancement of thermal conductivity for nanofluids is from
15 to 74% at 50 °C. The increase in thermal conductivity is nonlinear
with increasing temperature and the mass fraction, which matches with
other reports.[25] This nonlinear behavior
depends on the nature of the nanoparticles and base fluids. The distance
between particles decreases with increasing particle concentration.
More particles cause the collision between particles in the fluid
to be more severe, and the energy transfer is more efficient. Moreover,
the palladium nanoparticles stabilized by long alkyl chain-modified
HPG have small particle sizes and uniform distribution and show better
stability and thermal conductivity.
Figure 6
Ratio of thermal conductivity (k/k0) of nanofluids with (a)
Pd@CSHPG, (b) Pd@DSHPG, and
(c) Pd@OSHPG. (d) Comparison of k/k0 for nanofluids with different Pd nanoparticles at a
mass fraction of 0.1%.
Ratio of thermal conductivity (k/k0) of nanofluids with (a)
Pd@CSHPG, (b) Pd@DSHPG, and
(c) Pd@OSHPG. (d) Comparison of k/k0 for nanofluids with different Pd nanoparticles at a
mass fraction of 0.1%.For better understanding, Figure d compares the effect
of thermal conductivity enhancement
for nanofluids with different Pd nanoparticles at the same mass fraction
(0.1 wt %). In the temperature range of 25–50 °C, the
nanofluid with Pd@OSHPG has higher thermal conductivity at the same
mass fraction. The order of nanoparticles in nanofluids for thermal
conductivity enhancement is Pd@OSHPG > Pd@DSHPG > Pd@CSHPG.
The thermal
conductivity enhancement of Pd@mHPG is larger than that of Pd@S for
the nanofluids with the same nanoparticle concentration. The longer
alkyl chain-modified HPG can give better protection for nanoparticles
and assist metal nanoparticles in enhancing the thermal conductivity
of nanofluids.Figure represents
the density and viscosity measurements of the base fluid and nanofluids
with 0.1 wt % nanoparticles at 25–50 °C. It can be seen
that the density and viscosity decrease with increasing temperature,
which is due to a weakening of intermolecular forces. At 50 °C,
the increases of density and viscosity for nanofluids compared to
those for the base fluid are less than 0.12 and 2%, respectively.
In the literature,[33] the viscosity and
density of distilled water and GNP-Ag were reported. The results shows
that the viscosity increases by about 30% for nanofluids with 0.1
wt % nanoparticles compared to the viscosity of the base fluid (40
°C). Meanwhile, the density increases by about 0.09% for 0.1
wt % nanoparticles compared to the density of the base fluid. In general,
the change of viscosity for the Pd nanofluids prepared in this paper
is very small and negligible compared to the literature. The results
could be useful for engineering application.
Figure 7
(a) Densities and (b)
viscosities of the base fluid and nanofluids
at different temperatures.
(a) Densities and (b)
viscosities of the base fluid and nanofluids
at different temperatures.
Conclusions
Pd@CSHPG, Pd@DSHPG, and Pd@OSHPG
were prepared with modified HPG.
The prepared Pd nanoparticles were quasi-spherical in shape and small
in particle size (3.4, 2.7, and 3.6 nm). Stability studies indicated
that the modified hyperbranched molecules enhanced the dispersion
and stability of nanoparticles in the nonpolar base fluid without
any additives. The nanofluids with Pd@DSHPG and Pd@OSHPG remain stable
for up to 330 days, and the CSHPG encapsulated nanoparticles are less
stable than the other two. The nanofluid with Pd@DSHPG or Pd@OSHPG
could be stable for 24 h at 110 °C. With increasing temperature,
the thermal stability of nanofluids decreased. The thermal conductivity
study showed that the thermal conductivity enhanced with increasing
temperature and the mass fraction of nanoparticles. The experimental
results showed significant improvements in thermal conductivity for
hydrocarbon-based nanofluids compared to the corresponding base fluid.
In general, the palladium nanoparticles stabilized by long-chain alkyl-modified
HPG show better stability and thermal conductivity. The results could
supply useful information for designing smart coolants for a variety
of thermal management systems.
Experimental Section
Materials and Reagents
Potassium
tetrachloropalladate (98%), cyclohexanethiol (98%), 1-dodecanethiol
(98%), 1-octadecanethiol (97%), glycidyl methacrylate (97%), decalin
(98%), toluene (99%), and 4-dimethylaminopyridin (99%) were all purchased
from Aladdin Reagent Co., Ltd. (Shanghai, China). Benzophenone (99%),
tetrabutylammonium bromide (TOAB; 98%), and dimethyl sulfoxide (DMSO)
were obtained from Macklin Biochemical Co., Ltd. (Shanghai, China).
HPG (Mn = 6659, Mw/Mn = 1.3) was prepared in our
laboratory, according to the literature.[34,35] The hyperbranched polyglycerols were synthesized by anionic ring-opening
polymerization, and the monomers in the reactions should be added
very slowly to ensure low polydispersity.
Preparation
of Palladium Nanoparticles and
Nanofluids
mHPG products (CSHPG, DSHPG, and OSHPG) were prepared
according to the following procedure, as shown in Scheme . mHPG with thioether structures
was synthesized by click chemistry of HPG-MA (transesterification
product of HPG with GMA[29]) and cyclohexanethiol,
1-dodecanethiol, or 1-octadecanethiol. The modification to terminal
alkyl groups imparted hydrophobicity to HPG, giving the modified product
mHPG (CSHPG, DSHPG, and OSHPG) good solubility in the oil phase.
Scheme 1
Preparation Process of mHPG (CSHPG, DSHPG, and OSHPG)
The oil-soluble palladium nanoparticles were prepared
by the phase-transfer
method. Typically, 20 mL of K2PdCl4 (0.61 mmol)
aqueous solution and 1.40 mmol of tetrabutylammonium bromide (TOAB)
were dissolved in 100 mL of toluene. The mixture was stirred for 1
h. After standing stratification, the organic phase was collected.
Then, 0.2 g of CSHPG, DSHPG, or OSHPG was added. After that, 20 mL
of NaBH4 (4.00 mmol) aqueous solution was added dropwise,
and the reaction was stirred vigorously overnight. The organic phase
was removed by rotary evaporation at 45 °C. Finally, the product
was obtained by vacuum drying at 25 °C. A series of oil-based
nanofluids were prepared by dispersing a certain amount of Pd@CSHPG,
Pd@DSHPG, and Pd@OSHPG in decalin with ultrasonication.To better
understand the role of Pd@mHPG, the dodecanethiol-coated
Pd nanoparticles (Pd@S) were synthesized by the phase-transfer method;
the details are shown in the Supporting Information.[36,37]
Characterization
Nuclear magnetic
resonance spectra (1H NMR) were carried out by a Bruker
Advance III 600 MHz spectrometer. HPG and mHPG were tested with CH3OD and CDCl3 as the solvents, respectively. Fourier
transform infrared spectra (FTIR) were acquired by a Nicolet 5700
spectrometer (Thermo Scientific). The morphology of Pd nanoparticles
was performed on a JEM-2100 transmission electron microscope (JEOL).
The elemental composition and chemical state of nanoparticles were
recorded by an EscaLab 250Xi X-ray photoelectron spectroscope (Thermo
Scientific). The stability of nanofluids was monitored by UV–vis
spectroscopy on an Evolution 201 spectrophotometer (Thermo Scientific).
The thermal conductivity of nanofluids was measured by using a KD2
Pro thermometer with a KS-1 probe sensor (Decagon Devices). The result
was reported as average after at least six measurements for each temperature
and concentration after equilibrium of nanofluids. The viscosities
and densities of the base fluid and nanofluids were measured by a
Lovis 2000 M automatic microviscometer and DMA 5000 M density meter
(Anton Paar), respectively.
Authors: Nelson K Y Wong; Ripen Misri; Rajesh A Shenoi; Irina Chafeeva; Jayachandran N Kizhakkedathu; Mohamed K Khan Journal: J Biomed Nanotechnol Date: 2016-05 Impact factor: 4.099
Authors: Virender K Sharma; Christie M Sayes; Binglin Guo; Suresh Pillai; Jason G Parsons; Chuanyi Wang; Bing Yan; Xingmao Ma Journal: Sci Total Environ Date: 2018-11-03 Impact factor: 7.963