Temperature and magnetic dual responsive restricted access material for the extraction of malachite green from crucian and shrimp samples

Article abstract of DOI:10.1039/d0nj00230e

A new restricted access material (RAM) based on magnetic graphene oxide (MGO) and a temperature-response polymer brush was developed and used as an adsorbent for the magnetic solid-phase extraction (MSPE) of malachite green (MG). The adsorption and desorption of MG could be achieved on the obtained smart material only through changing the temperature of the solution. At 20 °C, lower than the lower critical solution temperature (LCST) of poly (N-isopropylacrylamide) (PNIPAAM), the temperature-response polymer chains are swollen and are in a hydrophilic state, which can improve mass transfer and binding capacity, and further enhance recognition capability and extraction efficiency for MG. When the temperature is above the LCST, the adsorbent shows hydrophilic to hydrophobic surface property alterations, leading to a gradually weakening interaction between MG and the adsorbent material, so the captured MG molecules were released. The smart material and aqueous solution achieve separation using a magnet within 2 min. The main parameters affecting the extraction efficiency were optimized. An effective analytical method for MG was developed by combining with a UV-Vis spectrophotometer. In real sample analysis, the average recoveries with their respective relative standard deviation (RSD%) for crucian and green shrimp were 79.7-90.1 (3.8-5.3%) and 73.6-81.2 (3.7-4.3%), respectively. It indicated that the new method can be applied in the determination of MG in complex real samples. This journal is

Full text of DOI:10.1039/d0nj00230e

NJC
View Article Online
View Journal
PAPER
Temperature and magnetic dual responsive
restricted access material for the extraction
of malachite green from crucian and shrimp
samples†
Cite this:DOI: 10.1039/d0nj00230e
ab
ab
ab
Yuanyuan Li,
*
Xiangzhi Feng,
Yulong Ma,
Tong Chen,c Wenxin Ji,ab
Xiaoxia Ma,ab Yang Chend and Hong Xue
A new restricted access material (RAM) based on magnetic graphene oxide (MGO) and a temperature-
response polymer brush was developed and used as an adsorbent for the magnetic solid-phase
extraction (MSPE) of malachite green (MG). The adsorption and desorption of MG could be achieved on
the obtained smart material only through changing the temperature of the solution. At 20 1C, lower than
the lower critical solution temperature (LCST) of poly (N-isopropylacrylamide) (PNIPAAM), the temperature-
response polymer chains are swollen and are in a hydrophilic state, which can improve mass transfer and
binding capacity, and further enhance recognition capability and extraction efficiency for MG. When the
temperature is above the LCST, the adsorbent shows hydrophilic to hydrophobic surface property
alterations, leading to a gradually weakening interaction between MG and the adsorbent material, so
the captured MG molecules were released. The smart material and aqueous solution achieve
separation using a magnet within 2 min. The main parameters affecting the extraction efficiency
were optimized. An effective analytical method for MG was developed by combining with
a
Received 18th January 2020,
Accepted 18th March 2020
UV-Vis spectrophotometer. In real sample analysis, the average recoveries with their respective
relative standard deviation (RSD%) for crucian and green shrimp were 79.7–90.1 (3.8–5.3%) and
73.6–81.2 (3.7–4.3%), respectively. It indicated that the new method can be applied in the determination
of MG in complex real samples.
DOI: 10.1039/d0nj00230e
rsc.li/njc
organisms and wastewater. Although many physical and chemical
treatment technologies have been developed, such as ozonation,4
Introduction
Malachite green (MG) is a N-methylated diamino triphenyl- oxidation,5 membrane filtration,6 flocculation,7 bio-sorption
methane dye, which can be used to treat bacterial infections in methods8 and electrochemical methods,9 and exhibit some inter-
marine animals as an effective antibacterial and antifungal esting characteristics, some disadvantages such as slow removal
agent. It is also a dye for wool, silk, cotton, leather, etc.1–3 As process, high dissolved oxygen demand, and secondary pollution
a cytotoxic chemical substance, MG has potential carcinogenic, to the environment still exist.
teratogenic and mutagenic hazards for mammalian cells. MG
The adsorption method based on nanomaterials is considered
could seriously threaten human health through channels such as a superior method for the extraction of MG, due to low cost,
as the food chain and water. Therefore, researchers have paid simplicity of operation, easy implementation and insensitivity to
attention to the extraction and determination of MG from living toxic pollutants.10 Recently, graphene oxide (GO) as a nanosized
carbon material, has been actively explored for use in a broad
range of applications. GO has a large surface area and many
a State Key Laboratory of High-efficiency Coal Utilization and Green Chemical
hydrophilic groups such as hydroxyl, epoxide and carboxyl,11,12 so
Engineering, Ningxia University, Yinchuan 750021, China.
it can be dispersed well in solution as an ideal adsorbent for the
separation and enrichment of various compounds. However, the
recovery is tedious and time-consuming for GO,13,14 so magnetic
nanoparticles were introduced to simplify the recycling process.15
Luo et al. prepared a Fe3O4@SiO2@G adsorbent for the magnetic
solid-phase extraction (MSPE) of sulfonamide antibiotics in
water.16 Lu and coworkers developed the Fe3O4@G@TiO2-based
E-mail: liyuanyuan_abc@sohu.com; Fax: +86 951 2062323; Tel: +86 951 2062330
b College of Chemistry and Chemical Engineering, Ningxia University,
Yinchuan 750021, China
c Comprehensive Technology Centre, Zhenjiang Customs District,
Zhenjiang, 212000, China
d Shanghe New Materials Company, Zhenjiang, 212000, China
e Zhenjiang Centre for Disease Control and Prevention, Zhenjiang 212000, China
† Electronic supplementary information (ESI) available. See DOI: 10.1039/d0nj00230e
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2020

View Article Online
Paper
NJC
MSPE method for the extraction and enrichment of phospho- Experimental
peptides from biological samples.17 Therefore, magnetic nano-
particles were loaded on graphene which not only exposes more
Material and reagents
GO was purchased from Hunan Fenghua Material Development
Co., Ltd (Hunan, China). DA hydrochloride (98%), 2,20-bipyridine
(bpy), 2-bromoisobutyrylbromide (BiBB, 98%), N-isopropyl-
acrylamide (NIPAAM, 98%), MG (99%), iron(II) chloride tetra-
hydrate (FeCl2ꢀ4H2O) and tris(hydroxymethyl) aminoethane (Tris)
were supplied by Macklin Biochemical Co., Ltd. Iron(III) chloride
hexahydrate (FeCl3ꢀ6H2O), copper(I) bromide (CuBr), ammonia
solution (NH3ꢀH2O, 99.95%), N,N-dimethylformamide (DMF),
and cis-oleic acid were obtained from Sinopharm Chemical
Reagent Co., Ltd (Shanghai, China). Triethylamine (TEA) and
N-hexane were purchased from Tianjin Damao Chemical
Reagent Factory (Tianjin, China). NIPAAM was used after recrys-
tallization with hexane. CuBr was purified in glacial acetic acid and
then washed with ethanol. All other chemicals are of analytical
grade. Deionized water (18.25 MO) was obtained from a purification
system (ULUPURE) throughout the experiments.
active sites by reducing the aggregation of graphene, but also has
magnetic responsiveness. It can be easily gathered using a mag-
netic field, and filtration, centrifugation and precipitation proce-
dures can be omitted.18 Besides, undesirable mass-transfer and
binding capacity also seriously impact the accuracy of quantitative
analysis in the progress of adsorption. Thus, some adsorbent
materials that can improve mass transfer and binding capacity
need to be prepared based on a green synthesis strategy.19–21 Smart
materials, particularly stimuli-responsive polymers, could reversibly
change their structures and functions due to external physical,
chemical or biological stimuli. It provides an alternative for
improving mass transfer and binding capacity. Li et al. reported
that a novel phycocyanin-imprinted ratiometric fluorescence nano-
sensor (CdTe QDs) was developed for temperature-regulated
sensing and detection of a phycocyanin target.22 Xiong and
coworkers developed Zipper-like thermo-responsive molecularly
imprinted polymers (E2-MIPs) for the selective recognition and
extraction of estradiol (E2) by temperature regulation.23 They
can regulate the binding affinity between the target analyte and
the sorbent, achieving a capture and controllable release target
analyte from abundant impurities.
In our work, a poly(N-isopropylacrylamide) (PNIPAAM) thermo-
sensitive polymer was grafted on magnetic GO (MGO) to prepare a
temperature and magnetic dual responsive smart material using
surface-induced ATRP technology,24 which not only improves
mass-transfer and binding capacity between the target analyte
and the substrate, but also effectively avoids using organic solvents
which can cause secondary pollution during the extraction pro-
cess. PNIPAAM is a widely studied thermo-sensitive polymer. Its
lower critical solution temperature (LCST) is about 32 1C.25–29 It
can undergo a sharp volume phase transition in aqueous
solution by temperature variation around LCST.30–32 For surface
modification using functional polymer brushes, a ‘‘grafting-from’’
approach is better than a ‘‘grafting-to’’ approach.33,34 Meanwhile,
the majority of synthetic routes of ATRP use chemical coupling
agents to fix the polymerization initiator onto substrates, which
is environmentally unfriendly and causes secondly pollution.
Inspired by mussels,35,36 dopamine (DA) was applied as a
bionic anchor for the polymerization of surface modification
chemistry because it is non-toxic and stable. DA is a hydrophilic
polymer, which can exhibit excellent dispersion in water and
make effective contact between composite materials and
analytes.34 In addition, the p electrons of PDA can help absorb
aromatic compounds through p–p stacking interactions.37
Therefore, we used PDA as a platform to anchor halogen-
containing macromolecular initiators.
Instruments and measurements
X-ray diffraction (XRD) patterns were recorded using a D8
Advance X-ray diffractometer. Transmission electron micro-
scopy (TEM) images were characterized using a H-7500 TEM
spectrometer (Hitachi, Japan). Thermo-gravimetric analysis (TGA)
curves were conducted using a SETSYS16 Integrated Thermal
Analyzer (Setaram, France). Magnetic hysteresis loops were con-
ducted using a vibrating sample magnetometer (VSM) on a
measurement system (Quantum Design, USA). X-ray photoelectron
spectroscopy (XPS) spectra were obtained using an AXIS ULTRA
DLD (Kratos, UK). The absorbances of MG solutions were obtained
using a UV-Vis spectrophotometer (Beijing General Analysis
TU-1901) at a maximum wavelength (lmax = 617 nm).
Preparation of MGO composite
MGO composite was prepared using a one-step co-precipitation
method.39 Briefly, GO (0.3 g) was dispersed into deionized
water (50 mL) with ultrasonic treatment for 4 hours to form a
uniform dispersion. Next, FeCl3ꢀ6H2O (4.0 g) and FeCl2ꢀ4H2O
(2.5 g) were dissolved in deionized water and added to the
above dispersion. Meanwhile, 25 mL ammonia solution was
quickly added to the reaction mixture, and 0.75 mL oleic acid
was injected into the mixture after stirring for 10 minutes.
Then, the reaction mixture was heated using a thermostatic
magnetic stirrer at 90 1C for 3 h. After that, MGO was washed
several times with deionized water and ethanol. The product
was collected by magnetic separation and dried under vacuum
at 70 1C overnight.
Preparation of polydopamine-coated MGO based on ATRP
initiator
Compared with the existing technology, this work proposed
a novel green synthesis strategy including environmental
friendliness, excellent thermosensitivity, convenient separability, ATRP initiator (DA-BiBB) was prepared using Edmondson’s
good mass transfer and binding capacity. Moreover, the tem- method.40 BiBB and DA could copolymerize into a PDA film
perature regulation system is simple, convenient and environ- containing ATRP-initiating groups. Briefly, DA (400 mg,
mentally friendly.38 The smart material shows a certain application 2.10 mmol) was dissolved in DMF (10 mL) and was degassed
potential in MSPE.
by purging with dry N2 for 10 min. BiBB (2.10 mmol) and TEA
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2020

View Article Online
NJC
Paper
(2.10 mmol) were added dropwise to the solution and reacted
In a batch process, 5 mg of the adsorbent was added to a
for 30 minutes at 0 1C. The ice-water bath was removed, and the 50 mL conical flask, and 10 mL of MG solution was accurately
reaction occurred for 3 h under a N2 atmosphere at room measured with a pipette and added into the conical flask. All
temperature. Then, this mixture was transferred to a glass beaker the adsorption experiments were performed in a thermo-static
(500 mL). Tris (480 mg, 4.0 mmol), MGO (0.15 g) and deionized oscillator at 20 1C with 150 rpm. After the equilibrium point,
water (100 mL) were added to the mixture, and stirred for 24 h at the conical flasks were taken out and separation was achieved
room temperature. The obtained product (MGO@PDA-Br) was between the MG solution and the absorbent material within
washed several times with deionized water and ethanol and 2 minutes by a magnet, and the solution was filtered with a 0.22 mm
separated by magnetic separation, then dried under vacuum at PTFE syringe filter. The absorbances of MG (lmax = 617 nm) before
60 1C overnight for further use.
and after adsorption were measured using a UV-Vis spectro-
photometer in each adsorption experiment. Then, the MG
concentration was calculated according to the standard working
curve (Table S1 and Fig. S1, ESI†).
Synthesis of PNIPAAM grafted MGO (MGO@PDA@PNIPAAM)
100 mg MGO@PDA-Br was dispersed in methanol/water (v/v,
10/10) by ultrasonication in a dry Schlenk flask. NIPAAM (2.0 g),
CuBr (0.12 g), and bpy (0.36 g) were added to the mixture under
a N2 atmosphere, followed by further bubbling with N2 for
30 minutes. Then, polymerization was performed for 24 h at
60 1C under a N2 atmosphere. The final product MGO@PDA@
PNIPAAM was collected with a magnet, intensively washed with
ethanol and water respectively, and then dried under vacuum at
40 1C for 12 h. The synthesis process is shown in Scheme 1.
Optimization of variables
Various parameters were optimized, such as MG concentration
(0.1–35 mg mLꢁ1), contact time (10–180 min), pH (3–10) and ionic
strength (0.02–0.1 M). By changing the initial concentration of MG,
the effect on adsorption capacity was investigated under constant
conditions of pH, dosage of adsorbent and ambient temperature
(20 1C). Then, the effect of contact time was investigated. In each
series of experiments, the samples were tested at specified intervals
(10–180 min). Subsequently, the effect of pH (3–10) value was studied
at the optimal concentration and time. Finally, the effect of ionic
strength was researched at an NaCl concentration of 0.02–0.1 M. The
adsorption capacity was evaluated by the following equation:
Batch adsorption experiments
Equilibrium data is important to obtain more insight into
adsorption and bonding properties between the target molecules
and the adsorbent.41 It also helps to understand the mechanisms
in the adsorption process. In this work, some parameters were
optimized. Besides, Langmuir, Freundlich isotherms, pseudo-first-
ðCo ꢁ CeÞm
qe ¼
V
order and pseudo-second-order kinetics were used to investigate where, qe (mg gꢁ1) is the adsorption quantity; Co (mg mLꢁ1) and Ce
the adsorption mechanism of MG on the adsorbent material.
(mg mLꢁ1) are the initial and equilibrium concentration of MG;
Scheme 1 The synthesis and adsorption process of MGO@PDA@PNIPAAM.
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2020

View Article Online
Paper
NJC
V (mL) is the volume of MG solution; m (mg) is the amount of
MGO@PDA@PNIPAAM. All the adsorption experiments were per-
formed at least in triplicate for data analysis.
Desorption and regeneration
After the extraction procedure was performed under optimal
conditions, the sorbent and MG solution were quickly separated
by an external magnetic field force and the solution was then
discarded. Then, the target analyte was eluted with water (with-
out additional chemicals) at 45 1C. Reasons for the selection of
temperature can be attributed to the temperature-responsive
behavior of MGO@PDA@PNIPAAM by a temperature variation
around 32 1C. The adsorbent material was reused for 5 cycles to
assess the recyclability.
Fig. 1 TGA curves of MGO (a), MGO@PDA-Br (b), MGO@PDA@PNIPAAM (c).
Sample preparation
(Fig. 1b) demonstrated a weight loss of approximately 39.65%
with the same temperature range due to the loss of PDA-Br.
When NIPAAM was grafted on to MGO@PDA-Br via ATRP
technology, the weight loss was around 45.38%, owing to the
decomposition of PNIPAAM (Fig. 1c).
A stock MG solution was prepared at a concentration of
1.0 mg mLꢁ1, using fresh ultrapure water in a tight and light-
resistant volumetric flask and stored in a freezer at 4 1C. MG is
stable in the temperature range studied. MG standard solution
at a concentration of 100 mg mLꢁ1 was prepared by diluting
stock standard solution with ultrapure water. Then, it was
diluted with different volumes of ultrapure water to prepare the
working solutions with various concentrations of 0.1–35 mg mLꢁ1
for the subsequent experiments. In all experiments, the stock
solution can be stored in a freezer at 4 1C for one month. Other
standard working solutions need to be reconstituted at each use.
Crucian and green shrimp samples were obtained from a
local market. Before extraction, the crucian tissue was separated
from bones and skin, and the shrimp shell and line were
removed from the shrimp. They were ground into a homogenate
using a mortar. 25 g of two samples and 20 mL of acetonitrile
were transferred to a 50 mL centrifuge tube, and were sonicated
for 15 minutes, respectively. They were centrifuged at 8000 rpm
for 5 minutes and the supernatants were collected. This process
was repeated three times. Then, all the extraction solutions were
merged, and concentrated via a rotary evaporator at 50 1C. The
residues on the wall of the bottle were washed with 1 mL of
acetonitrile 3 times and were then concentrated further. They
were filtered through a 0.22 mm PTFE syringe filter and trans-
ferred to volumetric flasks, which were diluted to a final volume
of 50 mL with ultrapure water. They were stored at 4 1C and used
in the MSPE procedure.
The magnetic properties of the functional GO were studied
by VSM. Fig. 2 shows the magnetic hysteresis loops of MGO,
MGO@PDA-Br, and MGO@PDA@PNIPAAM. They indicate that
all functionalized GO have superparamagnetic properties.
Their magnetization saturation (Ms) values are 52.97, 38.87
and 32.66 emu gꢁ1, respectively. MGO was covered by organic
substances including PDA-Br and PNIPAAM, which caused the
Ms value reduction. However, it still exhibits excellent magnetic
responsiveness. For example, MGO@PDA@PNIPAAM can be
completely separated from MG solution by an external magnetic
field within 2 minutes (Fig. 2, inset). Furthermore, MGO@PDA-Br
and MGO@PDA@PNIPAAM were investigated via incubating
dispersions using a constant-temperature water bath at 20 1C
and 45 1C. When the temperature of the solution is below the
LCST, the PNIPAAM chains are swollen in water, which dispersed
homogeneously in water (Fig. 3A-b). When the temperature is
higher than the LCST, they become shrunken and are precipitated
Results and discussion
Characterization of MGO@PDA@PNIPAAM
The synthesized composite was characterized using different
techniques. A TGA technique was used to determine the degree
of organic functionalization. Fig. 1 shows the TGA curves of
MGO, MGO@PDA-Br, and MGO@PDA@PNIPAAM. The weight
loss of MGO (Fig. 1a) was about 20.15% from 25 1C to 800 1C,
which was attributed to the evaporation of water molecules,
removal of oxygen-containing functional groups and the decom-
position of carbon skeleton.42 Compared with it, MGO@PDA-Br
Fig. 2 The magnetic hysteresis loops of MGO (a), MGO@PDA-Br (b),
and MGO@PDA@PNIPAAM (c). The optical image of MG solution with
MGO@PDA@PNIPAAM: left inset (without an external magnetic field), right
inset (with an external magnetic field).
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2020

View Article Online
NJC
Paper
Fig. 3 The optical images of MGO@PDA-Br (a) and MGO@PDA@
PNIPAAM (b) dispersion in water at 20 1C (A) and 45 1C (B).
out from the water (Fig. 3B-b). It is indicated that MGO@PDA@
PNIPAAM has excellent thermosensitivity in water around LCST.
However, MGO@PDA-Br was evenly dispersed at 20 1C or 45 1C,
which suggested that it remained without thermosensitivity.
Therefore, both typical temperatures of 20 1C and 45 1C were
studied. Meanwhile, it also provided evidence that PNIPAAM
chains were successfully immobilized on the surfaces of MGO.
The TEM images are shown in Fig. 4. Firstly, Fig. 4a presents
the TEM image of GO, which has a large number of adsorption
sites. As can be seen, Fig. 4b reveals a different surface with
many particles (size around 10 nm) due to the presence of
Fe3O4 nanoparticles on the GO surface, which demonstrated
that GO had been successfully modified by Fe3O4 nanoparticles
and endowed the GO nanocomposite with excellent magnetic
separability. As shown in Fig. 4c, a thin coating can be clearly
observed compared with MGO, which was attributed to the fact
that a PDA-Br layer was formed when BiBB-DA was anchored
onto the MGO.43 After PNIPAAM was grafted onto MGO@PDA-
Br, its TEM image (Fig. 4d) shows some furry structures
compared with the former. It could be observed that there were
a lot of curled and tangled brushes evenly decorated on the
MGO sheet, which can imply the successful preparation of
MGO@PDA@PNIPAAM.
Fig. 5 XRD pattern of GO (a), MGO (b), MGO@PDA-Br (c), and MGO@
PDA@PNIPAAM (d).
XRD was used to characterize the crystalline structures
of GO (a), MGO (b), MGO@PDA-Br (c), and MGO@PDA@
PNIPAAM (d). The results are shown in Fig. 5. The peak of
GO located at 2y E 111, which was consistent with the reported
results.44 The spectrum of MGO includes six characteristic
diffraction peaks at 30.321, 36.561, 43.361, 53.761, 57.201 and
63.961. These peaks are characteristic of Fe3O4. Therefore, it can
be inferred that there was a considerable connection between
GO and Fe3O4, retaining the crystal structure of Fe3O4 and
ensuring the magnetic property of the composite material. As
shown in Fig. 5c, the result revealed that the PDA film introduced
on the surface of MGO did not change the crystal structure, and
MGO@PDA-Br still had a typical cubic spinel structure. As shown
in Fig. 5d, there was a specific broad peak at 2y E 151 after
PNIPAAM was bonded, which indicated that the polymer was
successfully grafted on MGO@PDA-Br.
Fig. 6 displays the XPS spectra of MGO (a), MGO@PDA-Br (b)
and MGO@PDA@PNIPAAM (c). They confirm that the initiator
was immobilized and PNIPAAM was successfully grafted on GO
by analyzing the surface chemical compositions of those nano-
composites. As shown in Fig. 6a, the appearance of the Fe 2p
Fig. 4 TEM images of GO (a), MGO (b), MGO@PDA-Br (c), and MGO@
Fig. 6 XPS spectra of MGO (a), MGO@PDA-Br (b), and MGO@PDA@
PDA@PNIPAAM (d).
PNIPAAM (c).
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2020

View Article Online
Paper
NJC
peak component at the binding energy around 709.96 eV in the where qe (mg gꢁ1) is the adsorption capacity, Ce (mg mLꢁ1) is the
wide scan spectrum of MGO, indicated that Fe3O4 particles equilibrium concentration in water, qmax (mg gꢁ1) is the theoretical
were successful immobilized on GO. In the wide scan spectrum saturate adsorption capacity in the Langmuir model, and KL
of MGO@PDA-Br (Fig. 6b), the N 1s peak at 399.64 eV, the Br 3d (mL mgꢁ1) is the Langmuir isotherm constant demonstrating
peak, Br 3p peak and Br 3s peak at 69.99 eV, 182.25 eV and the tendency of adsorption, KF is the Freundlich isotherm
255.35 eV appeared, respectively. It confirmed that MGO was constant and n is the heterogeneity factor.
modified with PDA layers, bearing the ATRP initiator. In addi-
As shown in Fig. 8, the two isotherm adsorption models
tion, the increase of C 1s, N 1s and O1s and the disappearance (20 1C and 45 1C) were linearly fitted, and the calculated model
of the Br peak in Fig. 6c were observed. They indicated that constants and statistical parameters are summarized in Table 1.
PNIPAAM was successfully grafted onto the MGO@PDA-Br.
R2 values show a strong linear relationship between the predicted
and experimental Langmuir model values, which indicate a
monolayer adsorption of MG onto the MGO@PDA@PNIPAAM.
Lower R2 values of the Freundlich isotherm model confirm that
experimental data does not agree well with the Freundlich
isotherm.
Optimization of the enrichment conditions for MG
Adsorption isotherms. The MGO@PDA@PNIPAAM smart
material shows a temperature responsive adsorption performance
in low/high temperature processes. To investigate the maximum
adsorption capacity of the smart material, a constant dosage (5 mg)
of MGO@PDA@PNIPAAM and different concentrations of MG
(0.1–35 mg mLꢁ1) were used for MSPE. After adsorption equilibrium,
the absorbances of MG solution were measured using a UV-Vis
spectrophotometer. As shown in Fig. 7, the maximum adsorption
capacity was obtained when the concentration was 30 mg mLꢁ1 at
20 1C and 20 mg mLꢁ1 at 45 1C. This is because the material shows a
hydrophilic property at 20 1C, so that more hydrophilic MG
molecules can be adsorbed. As the temperature increased from
20 1C to 45 1C, the PNIPAAM chains changed from a swollen state to
a shrunken and hydrophobic state, which resulted in the hydro-
phobic property of the material. Therefore, the interaction weakens
between the smart material and MG, and causes desorption of the
captured MG molecules from the adsorbent.
Adsorption kinetics. To understand the adsorption mechanisms
and the influence of adsorption time, kinetic experiments were
performed at 20 1C and 45 1C. 5 mg adsorbent and concentrations
(30 mg mLꢁ1 at 20 1C and 20 mg mLꢁ1 at 45 1C) of MG were constant,
and the contact time was investigated from 10 to 180 minutes
during the MSPE process. As shown in Fig. 9, the adsorption
processes were very quick before 30 minutes, which might be
attributed to a large number of binding sites on the surfaces of
MGO@PDA@PNIPAAM and collisional probability between MG
and the sorbent. As time increased to 90 minutes, the adsorption
equilibrium could be achieved, which could be attributed to the
saturation of active sites on surfaces of the smart material. They
need 90 minutes to reach equilibrium at different temperatures
(20 1C or 45 1C), which were mainly ascribed to the fact that MG
may need more time through the layer of collapsed PNIPAAM
chains to reach the surface of the MGO sheet.
Two parameter isotherm models of Langmuir and Freundlich
were used to obtain the best of the equilibrium curves. Langmuir
isotherm (eqn (1)) assumes homogeneous adsorption and
adsorption only occurring at specific homogeneous sites. Freundlich
isotherm (eqn (2)) is an empirical model which describes multi-
layer adsorption.
The pseudo-first order model (eqn (3)) assumes that the
sorption rate decreases linearly with the increase of adsorption
capacity. The pseudo-second order kinetics model (eqn (4))
assumes that the rate-limiting step is the interaction between
the two reagent particles, which is used to describe a chemical
adsorption.
ꢀ
ꢁ
Ce
qe
1
1
¼
þ
Ce
(1)
(2)
q
maxKL
qmax
ln(qe ꢁ qt) = ln qe ꢁ k1t
(3)
(4)
In Ce
In qe ¼ In KF þ
n
t
1
t
¼
þ
qt k2qe2 qe
Where qe and qt (mg gꢁ1) are the amounts of analyte on the
adsorbents at equilibrium and time t (min). k1 (minꢁ1) is the
rate constant in the pseudo-first order adsorption model and
k2 g (mg min)ꢁ1 is the rate constant in the pseudo-second order
adsorption model.
As shown in Fig. 10, the obtained experimental data were
fitted to different kinds of kinetic models about pseudo-first
order and pseudo-second order. The fitting results of kinetic
data and statistical parameters are listed in Table 2. According
to the results shown, R2 of the pseudo-second-order kinetic
model was higher than the pseudo-first-order kinetic model and
the value of qe,cal was also close to the experimental value of
qe,exp. It indicated that the adsorption could be well described by
Fig. 7 Adsorption isotherms of MG on the smart material.
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2020

View Article Online
NJC
Paper
Fig. 8 The linear fit of the Langmuir isotherm and Freundlich isotherm at 20 1C and 45 1C.
Table 1 Isothermal parameters for the adsorption of MG onto the smart material
Langmuir model
Freundlich model
Contaminant
MG
Temperature (1C)
qmax (mg gꢁ1
)
KL (L mgꢁ1
)
R2
KF (mg1ꢁn Ln gꢁ1
)
n
R2
20
46.2961
0.7224
0.9957
12.4460
1.8797
0.8664
45
21.2314
0.4507
0.9977
3.9204
1.4624
0.9340
property of the adsorbent changed with pH. The green-blue
color of MG remains unaffected at acidic and neutral pH, but
under the condition of alkaline pH, could cause a change of the
structure and stability. In this work, the effect of solution pH on
adsorption efficiency was investigated by adjusting the pH from
3 to 10 at two different temperatures (20 1C and 45 1C). Mean-
while, the characteristics of the smart material were further
explained by comparing with MGO.
As shown in Fig. 11, a lower pH (pH o 7.0) was not
conducive to the progress of adsorption, a possible reason for
this could be attributed to the fact that excess H+ ions compete
with the positively charged MG molecules to reach the adsorbent
surfaces when pH o 7, so the adsorption capacity was greatly
reduced. When pH = 7, the qe always reached the largest value
(qe,max) for each adsorbent. The concentration of H+ could be
neglected, so the competition between H+ and MG molecules
was also negligible. It is worth noting that the color of the
MG solution became gradually lighter or disappeared due to
the change of the structure of MG molecule when pH 4 9.0.
Fig. 9 Adsorption kinetics of MG on the smart material.
the pseudo-second-order mechanism and the experimental
process was mainly based on the chemical adsorption.
Effect of solution pH on adsorption efficiency
The pH value of the solution has an important influence during Thus, it is beneficial for practical applications without pH
the MSPE process, because the structure of MG and surface adjustment. As shown in Fig. 11, the adsorption capacity of
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2020

View Article Online
Paper
NJC
Fig. 10 The linear fit of pseudo-first-order kinetic model and pseudo-second-order kinetic model at 20 1C and 45 1C.
Table 2 Kinetic parameters for the adsorption of MG onto the smart material
Pseudo-first order
Pseudo-second order
Contaminant
MG
Temperature (1C)
qe1 (mg gꢁ1
)
K1 (minꢁ1
)
R2
qe2 (mg gꢁ1
)
K2 (g (mgꢁ1 minꢁ1))
R2
20
43.6062
0.0434
0.9563
42.9185
0.0019
0.0037
0.9967
45
15.2732
0.0349
0.9442
19.2308
0.9962
indicated that MGO@PDA@PNIPAAM has a temperature-
responsive adsorption behavior. The adsorption amount of
MGO@PDA@PNIPAAM attain Qe,max = 41 mg gꢁ1 at 20 1C.
Although PNIPAAM grafted onto the MGO sheet may cover
partial surfaces of MGO@PDA@PNIPAAM, the p electrons of
PDA can help it to absorb MG via p–p stacking interactions.31
Moreover, the hydrophilic smart material shows excellent dis-
persibility in solution and promotes an effective contact
between the composite and the target analyte. In the case of
45 1C, the adsorption capacity of MGO@PDA@PNIPAAM was
slumped to 19 mg gꢁ1. The reason for this might be that the
whole PNIPAAM chains were changed from a swollen state to a
shrunken one, leading to the interaction gradually weakening
between MG and the adsorbent material, so the captured MG
molecules were released. In the meantime, it is also directly
proved that the grafted PNIPAAM brushes play a key role in
MSPE. Besides, another hint was obtained from the temperature-
Fig. 11 Effect of pH on the adsorption of MG onto MGO@PDA@PNIPAAM
at 20 1C (a) and 45 1C (d), and onto MGO at 45 1C (b) and 20 1C (c).
MGO slightly increased when the temperature changed from responsive adsorption behavior that a change of temperature
20 1C to 45 1C. Compared with MGO, the smart material could promote the regeneration of the hydrophilic adsorbent
adsorption behavior was completely opposite. Therefore, it is when it was higher than the LCST of NIPAAM.
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2020

View Article Online
NJC
Paper
Effect of ionic strength on adsorption efficiency
selected as the eluent volume accumulated. After desorption,
the collected MGO@PDA@PNIPAAM material was washed alter-
nately with hot/cold water several times, and then dried in
vacuum at 60 1C overnight for the next cycle.
During the reusing process, the number of cycles is an
important parameter for obtaining good recoveries. The results
showed that MGO@PDA@PNIPAAM could be reused at least
5 times without a significant loss of adsorbent capacity as
shown in Fig. 13(b). The adsorption capacity of MG was still
above 36 mg gꢁ1 after five cycles of use. Therefore, MGO@PDA@
PNIPAAM has good reusability.
The effect of ionic strength on adsorption efficiency was
studied in the coexistence of NaCl. The salt ionic concentration
has a significant effect on the adsorption capacity when the
temperature is at 20 1C and 45 1C. As the concentration of NaCl
increased from 0.02 to 0.1 M in Fig. 12, the adsorption capacity
of the smart material decreased. A possible reason for this
could be attributed to the fact that Na+ competed with MG
molecules in the solution, and occupied the active sites on the
surfaces of the adsorbent during the adsorption process, so the
adsorption capacity of the smart material was weakened for
MG. It was also indirectly proved that there exist an electrostatic
interaction between MG and the adsorbent during the MSPE
process. Additionally, the hydrogen bond and p–p stacking
interaction might also influence the adsorption in the extraction
process. Possible mechanisms for this are shown in Scheme 1.
Method validation
The proposed method was evaluated by several parameters
such as the linearity, limits of detection (LOD), recovery, reuse
and precision. The concentration of calibration curve ranged
from 0.1–35 mg mLꢁ1, the linearity was A = 0.1907C ꢁ 0.1121
and the correlation coefficient (R2) was 0.9987. Good linearity
was obtained in the studied concentration range. The enrich-
Desorption and regeneration
The regeneration performance of MGO@PDA@PNIPAAM was ment factor (EF) was 83 and the limit of detection (LOD) was
assessed. The desorption experiment was carried out using a 0.08 mg mLꢁ1. The accuracy of the method was evaluated by
thermostatic oscillator at 45 1C, and the material was regenerated recovery tests. The precision of the method was evaluated by
by changing the water temperature without other additional repeatability (intraday) and reproducibility (interday) studies,
chemicals. During the desorption process, the smart material performed by spiking known quantities of the standard
was eluted with 3 mL hot water at each time point, and repeatedly solution (different concentrations of 10, 20 and 30 mg mLꢁ1
)
until maximum recovery was obtained. The effect of eluent to known amounts of real samples. The resultant samples were
volume accumulated was evaluated using 3, 6, 9, 12 and 15 mL analyzed using this method, and the results are shown in
hot water. The results are shown in Fig. 13(a) and 12 mL was Table 3. Accordingly, a good correlation exhibits the effective
Fig. 12 Effect of ionic strength.
Fig. 13 The volume of eluent (a), and the adsorption capacity after five cycles of reuse (b).
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2020

View Article Online
Paper
NJC
Table 3 The recovery results of MG at three different spiking levels in real
samples and intra-day and inter-day precisions
Comparison with other methods
The analytical results obtained using the proposed method are
compared with those used in the literature for analyzing MG.
The methods were compared by evaluating the parameters
including analysis technique, dosage of sorbent, linear range
and regeneration method. The results are presented in Table 4.
The proposed method mainly exhibits lower amounts of the
sorbent and a wide linear range. The material was eluted and
regenerated in alternatively hot/cold water without an organic
solvent. It suggested that the proposed method was much
simpler and environmentally friendly because the consumption
of solvent decreased.
Spike
level
Intra-day
precision
Inter-day
precision
Real
Recovery RSD
sample (mg mLꢁ1) (%)
(%, n = 3) (RSD, %, n = 6) (RSD, %, n = 3)
Green
0
ND
—
—
—
shrimp 10
76.5
81.2
73.6
ND
3.7
4.3
3.9
—
4.2
5.6
3.5
—
3.7
4.3
5.7
5.3
4.7
6.1
—
4.6
3.9
6.5
20
30
0
10
20
30
Crucian
82.4
79.7
90.1
5.3
3.8
4.7
performance of the proposed method for the extraction of MG
in complex environmental samples.
Conclusions
Analysis of real samples
A temperature and magnetic dual responsive restricted access
To validate the applicability and reliability of the smart mate- material MGO@PDA@PNIPAAM was prepared, characterized
rial, it was used as a sorbent for pre-concentration and deter- and evaluated. The material as an adsorbent could accomplish
mination of MG in real samples. The preprocessed samples of adsorption and desorption of MG only by changing the
crucian and green shrimp were spiked with MG concentrations temperature of the aqueous solution without other additional
of 10 mg mLꢁ1, 20 mg mLꢁ1 and 30 mg mLꢁ1, respectively. The chemicals. Compared with the previous reported methods45–48
recoveries of MG were evaluated in real samples under optimized using an organic solvent as an eluent, the new method does not
conditions using the developed method. The results are listed in cause secondary environmental pollution. At the same time, the
Table 3. It showed that MG was not detected in un-spiked real smart material could be quickly and completely separated from
samples. In the spiked samples, the recoveries ranged from the aqueous solution by magnets within 2 minutes in the
79.7 to 90.1% for crucian and 73.6 to 81.2% for green shrimp, recycling process, which greatly saves time and simplifies
which indicated that the developed method is suitable for the the cumbersome separation process. An effective enrichment
pre-concentration and determination of MG in real samples.
method for MG was developed and successfully applied to the
Fig. 14 shows the spectrum of the unspiked real sample (a), analysis of MG in real samples with satisfactory recoveries. The
the spiked real sample with 3 mg mLꢁ1 (b), and the eluted material and the developed method will be convenient, reliable
spectrum (c). It indicates that the smart material could con- and promising for the determination of MG in complex
centrate and enrich MG.
samples.
Fig. 14 Typical spectrum of the unspiked real sample (a), the spiked real sample (b), and the eluted spectrum (c).
Table 4 Comparison of the proposed method with other methods applied for MG
Adsorbent
Analysis technique
Dosage of sorbent (mg)
Linear range
Regeneration method
Ref.
MBCNF/GOPA
Fe3O4@Mel
MMIPs
Fe3O4@PEI-MOF-5
MGO@PDA@PNIPAAM
UV-Vis
UV-Vis
HPLC-UV
UHPLC-MS/MS
UV-Vis
5
300
10
10
5
5–50 g mLꢁ1
Acetic acid/methanol
Acetic acid
Methanol/acetic acid
Methanol/formic acid
Hot/cold water
45
46
47
48
0.1–8 g mLꢁ1
2–1000 ng mLꢁ1
1–500 ng mLꢁ1
0.1–35 g mLꢁ1
This work
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2020

View Article Online
NJC
Paper
21 A. J. F. B. Jinhua Li and L. Chen, Analyst, 2018, 143,
3570–3578.
22 X. Wu, X. Wang, W. Lu, X. Wang, J. Li, H. You, H. Xiong and
L. Chen, J. Chromatogr. A, 2016, 1435, 30–38.
23 H. Xiong, X. Wu, W. Lu, J. Fu, H. Peng, J. Li, X. Wang,
H. Xiong and L. Chen, Talanta, 2018, 176, 187–194.
Conflicts of interest
There are no conflicts to declare.
Acknowledgements
This work was supported by the Research Project for Science
and Technology of Higher Education of Ningxia Provience of
China (NGY 2018014), the Ningxia Natural Science Foundation
of China (2019AAC03023), the East-West Cooperation Project of
Ningxia Key R&D Plan (2017BY064,2019BFH02014), the Zhenjiang
Social Development Project (SH2018017) and the National First-rate
Discipline Construction Project of Ningxia (NXYLXK2017A04).
´
´
24 A. Gruszkiewicz, M. Słowikowska, G. Grzes, A. Wojcik,
J. Rokita, A. Fiocco, M. Wytrwal-Sarna, M. Marzec, B. Trzebicka,
M. Kopec, K. Wolski and S. Zapotoczny, Eur. Polym. J., 2019, 112,
´
817–821.
25 R. Dong, J. Li, H. Xiong, W. Lu, H. Peng and L. Chen,
Talanta, 2014, 130, 182–191.
26 A. X. W. A. Lingxin Chen, Chem. Soc. Rev., 2016, 45, 2137–2211.
27 C. Boutris, E. G. Chatzi and C. Kiparissides, Polymer, 1997,
38, 2567–2570.
References
˘
28 S. Çakmak, A. S. Çakmak and M. Gu¨mu¨s-derelioglu, Mater.
1 A. Mohamed, M. M. Ghobara, M. K. Abdelmaksoud and
G. G. Mohamed, Sep. Purif. Technol., 2019, 210, 935–942.
2 C. Pradeep Sekhar, S. Kalidhasan, N. Rajesh and V. Rajesh,
Chem, 2009, 77, 842–847.
3 W. Qu, T. Yuan, G. Yin, S. Xu, Q. Zhang and H. Su, Fuel,
2019, 249, 45–53.
Sci. Eng., C, 2013, 33, 3033–3040.
29 K. Nagase, T. Okano and H. Kanazawa, Nano-Struct. Nano-
Objects, 2018, 16, 9–23.
30 Q. Yu, Y. Zhang, H. Chen, Z. Wu, H. Huang and C. Cheng,
Colloids Surf., B, 2010, 76, 468–474.
31 S. Burkert, E. Bittrich, M. Kuntzsch, M. Mu¨ller, K. Eichhorn,
C. Bellmann, P. Uhlmann and M. Stamm, Langmuir, 2010,
26, 1786–1795.
¨
4 E. Kusvuran, O. Gulnaz, A. Samil and O. Yildirim, J. Hazard.
Mater., 2011, 186, 133–143.
5 M. S. Sajab, C. H. Chia, C. H. Chan, S. Zakaria, H. Kaco, 32 J. Dong, J. Weng and L. Dai, Carbon, 2013, 52, 326–336.
S. W. Chook, S. X. Chin and A. A. M. Noor, RSC Adv., 2016, 6, 33 C. Chen, M. Liu, C. Gao, S. Lu¨, J. Chen, X. Yu, E. Ding, C. Yu,
19819–19825.
J. Guo and G. Cui, Carbohydrate, 2013, 92, 621–628.
34 Y. Lin, H. Lin, J. Lin, B. Hsieh and Y. Sheng, J. Colloid
Interface Sci., 2012, 387, 106–114.
35 M. Kohri, Y. Shinoda, H. Kohma, Y. Nannichi, M. Yamauchi,
S. Yagai, T. Kojima, T. Taniguchi and K. Kishikawa, Macro-
mol. Rapid Commun., 2013, 34, 1220–1224.
36 S. Hong, Y. S. Na, S. Choi, I. T. Song, W. Y. Kim and H. Lee,
Adv. Funct. Mater., 2012, 22, 4711–4717.
37 X. Wang, G. Song and C. Deng, Talanta, 2015, 132, 753–759.
6 S. M. Doke and G. D. Yadav, Chem, 2014, 117, 760–765.
7 H. Wu, R. Yang, R. Li, C. Long, H. Yang and A. Li, Environ.
Sci. Pollut. Res., 2015, 22, 13038–13048.
8 W. Tsai and H. Chen, J. Hazard. Mater., 2010, 175, 844–849.
9 F. Guenfoud, M. Mokhtari and H. Akrout, DiamRelater,
2014, 46, 8–14.
10 Y. Dai, N. Zhang, C. Xing, Q. Cui and Q. Sun, Chem, 2019,
223, 12–27.
11 H. He and C. Gao, ACS Appl. Mater. Interfaces, 2010, 2, 3201–3210. 38 Y. Song, X. Song, H. Yuan and C. Cheng, New J. Chem., 2016,
12 P. Sharma and M. R. Das, J. Chem. Eng. Data, 2012, 58,
151–158.
13 H. C. Schniepp, J. Li, M. J. McAllister, H. Sai, M. Herrera-
40, 3194–3207.
39 C. H. Hong, M. W. Kim, W. L. Zhang, I. J. Moon and
H. J. Choi, J. Colloid Interface Sci., 2016, 481, 194–200.
Alonso, D. H. Adamson, R. K. Prud’Homme, R. Car, D. A. Saville 40 B. Zhu and S. Edmondson, Polymer, 2011, 52, 2141–2149.
and I. A. Aksay, J. Phys. Chem. B, 2006, 110, 8535–8539.
14 Z. Wang, J. Zhang, P. Chen, X. Zhou, Y. Yang, S. Wu, L. Niu,
Y. Han, L. Wang, P. Chen, F. Boey, Q. Zhang, B. Liedberg
and H. Zhang, Biosens. Bioelectron., 2011, 26, 3881–3886.
15 Y. Ma, Y. Li, G. Zhao, L. Yang, J. Wang, X. Shan and X. Yan,
Carbon, 2012, 50, 2976–2986.
41 T. P. Krishna Murthy, B. S. Gowrishankar, M. N. Chandra
Prabha, M. Kruthi and R. Hari Krishna, Microchem. J., 2019,
146, 192–201.
42 Y. Yao, S. Miao, S. Liu, L. P. Ma, H. Sun and S. Wang, Chem.
Eng. J., 2012, 184, 326–332.
43 M. Kohri, H. Kohma, Y. Shinoda, M. Yamauchi, S. Yagai,
T. Kojima, T. Taniguchi and K. Kishikawa, Polym. Chem.,
2013, 4, 2696.
16 Y. Luo, Z. Shi, Q. Gao and Y. Feng, J. Chromatogr. A, 2011,
1218, 1353–1358.
17 J. Lu, C. Deng, X. Zhang and P. Yang, ACS Appl. Mater. 44 Y. Yao, S. Miao, S. Liu, L. P. Ma, H. Sun and S. Wang, Chem.
Interfaces, 2013, 5, 7330–7334. Eng. J., 2012, 184, 326–332.
18 X. Li, G. Zhu, Y. Luo, B. Yuan and Y. Feng, TrAC, Trends 45 P. Arabkhani and A. Asfaram, J. Hazard. Mater., 2020, 384, 121394.
Anal. Chem., 2013, 45, 233–247.
46 S. A. R. Mirzajani, J. Ind. Eng. Chem., 2015, 23, 171–178.
19 A. Ostovan, M. Ghaedi, M. Arabi, Q. Yang, J. Li and L. Chen, 47 Z. Lin, H. Zhang, L. Li and Z. Huang, React. Funct. Polym.,
ACS Appl. Mater. Interfaces, 2018, 10, 4140–4150. 2016, 98, 24–30.
20 A. R. Bagheri, M. Arabi, M. Ghaedi, A. Ostovan, X. Wang, 48 Z. Zhou, Y. Fu, Q. Qin, X. Lu, X. Shi, C. Zhao and G. Xu,
J. Li and L. Chen, Talanta, 2019, 195, 390–400. J. Chromatogr. A, 2018, 1560, 19–25.
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2020