Analytica Chimica Acta 877 (2015) 90–99

Contents lists available at ScienceDirect

Analytica Chimica Acta

journal homepage: www.elsevier.com/locate/aca

Magnetic graphene oxide modified with choline chloride-based deep

eutectic solvent for the solid-phase extraction of protein
Yanhua Huang, Yuzhi Wang * , Qi Pan, Ying Wang, Xueqin Ding, Kaijia Xu, Na Li,
Qian Wen
State Key Laboratory of Chemo/Biosensing and Chemometrics, College of Chemistry and Chemical Engineering, Hunan University, Changsha 410082, PR China

H I G H L I G H T S G R A P H I C A L A B S T R A C T

A strategy for extraction of protein

based on DES-coated magnetic
graphene oxide.

The deep eutectic solvents were
based on choline chloride.

Bovine serum albumin was used as
the analyte.

The material prepared works for the
acidic but not the basic or the neutral
proteins.

A R T I C L E I N F O A B S T R A C T

Article history: Four kinds of green deep eutectic solvents (DESs) based on choline chloride (ChCl) have been synthesized
Received 4 January 2015 and coated on the surface of magnetic graphene oxide (Fe3O4@GO) to form Fe3O4@GO-DES for the
Received in revised form 27 March 2015 magnetic solid-phase extraction of protein. X-ray diffraction (XRD), vibrating sample magnetometer
Accepted 28 March 2015
(VSM), Fourier transform infrared spectrometry (FTIR), field emission scanning electron microscopy
Available online 31 March 2015
(FESEM) and thermal gravimetric analysis (TGA) were employed to characterize Fe3O4@GO-DES, and the
results indicated the successful preparation of Fe3O4@GO-DES. The UV–vis spectrophotometer was used
Keywords:
to measure the concentration of protein after extraction. Single factor experiments proved that the
Choline chloride-based deep eutectic
solvent
extraction amount was influenced by the types of DESs, solution temperature, solution ionic strength,
Magnetic graphene oxide extraction time, protein concentration and the amount of Fe3O4@GO-DES. Comparison of Fe3O4@GO and
Protein Fe3O4@GO-DES was carried out by extracting bovine serum albumin, ovalbumin, bovine hemoglobin and
Magnetic solid-phase extraction lysozyme. The experimental results showed that the proposed Fe3O4@GO-DES performs better than
Fe3O4@GO in the extraction of acidic protein. Desorption of protein was carried out by eluting the solid
extractant with 0.005 mol L1 Na2HPO4 contained 1 mol L1 NaCl. The obtained elution efficiency was
about 90.9%. Attributed to the convenient magnetic separation, the solid extractant could be easily
recycled.
ã 2015 Elsevier B.V. All rights reserved.

* Corresponding author. Tel.: +86 731 88821903; fax: +86 731 88821848.
E-mail address: wyzss@hnu.edu.cn (Y. Wang).

http://dx.doi.org/10.1016/j.aca.2015.03.048
0003-2670/ ã 2015 Elsevier B.V. All rights reserved.
 Y. Huang et al. / Analytica Chimica Acta 877 (2015) 90–99
1. Introduction
Protein is an essential material of organisms. It plays an
important role in life activities [1]. It is extremely necessary to
obtain the protein that contains no harmful levels of contaminants.
Whereas, the target protein always mix with other impurities,
along with the self-owned particularity and complexity of protein,
which make the degeneration of protein being extremely easy to
occur during the separation and purification process. Consequent-
ly, the separation and purification of proteins become the
bottleneck of biotechnology industry development.
Separation methods based on magnetic materials have
attracted widespread attention in recent years [2,3]. The magnetic
materials used in solid phase extraction (SPE) do not need to pack
into the cartridge, and the problems of high pressure in SPE column
can be prevented. Furthermore, the centrifugation or filtration
step, which is time-consuming and requires additional apparatus-
es, is not needed [4]. The use of external magnetic field in magnetic
SPE (MSPE) can save processing time [5]. Obviously, MSPE can
improve the extraction efficiency and simplify the process of
preprocessing [6,7]. Additionally, the interfacial area between
sorbents and analytes increases due to the use of dispersive
extraction mode, and the vast majority of magnetic sorbents can be
easily recycled and reused subsequently [8,9]. MSPE shows a
comprehensive advantage of simplicity and saving time and labor,
which makes it a technique with bright application prospect for
the sample preparation [10,11].
In MSPE, it is particularly important to select appropriate
sorbent in order to obtain high recovery. In recent years, intense
interest has grown in graphene oxide (GO) as novel sorbents in
analytical chemistry [12]. GO possesses a large number of
oxygenated groups both on the basal plane and at the edges of
GO sheets, such as carboxyl and hydroxyl [13]. The existence
of these functional groups makes GO more hydrophilic.
Moreover, these functional groups can form hydrogen bonding
or electrostatic interaction with adsorbates containing oxygen-
and nitrogen-functional groups [12]. Furthermore, GO possesses
huge surface area. As a consequence, GO seems to be ideal sorbents
in SPE and it is potential for the efficient extraction of proteins.
However, GO is hard to be separated from aqueous solution. To
overcome this drawback, the combination of magnetic materials
with GO to form composites have become a research hotspot [14].
Due to the large surface area of GO, there is a platform available for
loading magnetic nanoparticles (MNPs). Among the widely used
MNPs, Fe3O4 nanoparticles are well received because of their low
cost and low toxicity [15]. Combined with Fe3O4, magnetic GO
(Fe3O4@GO) will get new and/or enhanced functionalities which
cannot be achieved by either component alone [14] and show great
promise for MSPE applications. Recently, magnetic GO has been
reported to be used as adsorbent toward heavy metal ions (such as
cobalt [16], copper [17], arsenic [18] and mercury [19]) and dye
[20,21]. In addition, there is research focused on magnetic GO as
adsorbent toward simultaneous removal of heavy metal ions and
dyes [22].
Ionic liquids (ILs) as novel green solvents have become one of
the hotspots of green chemistry. That is because ionic liquids have
unique properties and special characters, which are different from
the traditional solvent. A new kind of IL analogues, which is defined
as deep eutectic solvent (DES), has emerged to address the
shortcomings of ILs that include high cost and toxicity. It has been
found that DESs have similar characteristics and properties to ILs
[23]. In general, two or three cheap and safe components, which
can be associated with each other through hydrogen bond
interactions, are capable of forming a DES [24]. One of the most
common components used in DES is a quaternary ammonium salt,
which is called choline chloride (ChCl). That is because of the fact
91
that ChCl is very cheap, biodegradable and non-toxic. In addition,
ChCl can be obtained either by means of extracting from biomass
or by synthesizing from fossil reserves through a very high atom
economy process [25]. Combined with another component that
includes hydrogen bond containing functional groups (called
hydrogen bond donors), such as urea, acid amides, carboxylic acids
and polyols, ChCl is easy to form a DES. DESs based on ChCl present
many advantages as compared with traditional ILs [26], such as it is
relatively simple to synthesis, further purification is not needed,
the majority of them are biodegradable [27], biocompatible [28]
and non-toxic [29]. What is more, the atom utilization rate in the
synthesis process can reach 100%, absolutely conforming to the
requirements of the atom economy of green chemistry. Therefore,
it is worthwhile to study DESs, making it serve various fields better.
Materials functionalized with ionic liquids (ILs) are receiving
considerable attention. During the last five years, a few studies
have been focused on ionic liquid-coated magnetic GO as
adsorbents for extraction of heavy metal ions [30], nitrobenzene
compounds [31] and proteins [32]. Although ILs lose liquid state
when immobilized on the surface of GO, the other unique
properties such as polarity and low volatility are maintained
[33]. Moreover, the introduction of ionic liquid moieties
could improve the performance of the resulting materials and
attract a broad range of applications [34]. For this reason, the
introduction of DES into magnetic GO is expected to increase the
water-solubility of magnetic GO and improve the extraction
efficiency of proteins.
In this paper, four kinds of deep eutectic solvents, which are
based on ChCl, have been prepared and coated on the surface of
magnetic GO to form Fe3O4@GO-DES. To our knowledge, it is the
first time that DESs derived from ChCl are applied to the
modification of magnetic GO. The resulting DES-coated Fe3O4@GO
has been used for magnetic solid-phase extraction of proteins.
Bovine serum albumin (BSA) was chosen as the model protein due
to its relatively high structural stability [35]. After extraction, the
concentration of the protein in the supernatant was measured by
the UV–vis spectrophotometer at 278 nm. The presented
Fe3O4@GO-DES-MSPE technique has also been used to extract
ovalbumin (OVA), lysozyme (Lys) and bovine hemoglobin (BHb).
2. Experimental
2.1. Apparatus
The main instruments: UV-2450 UV–vis spectrophotometer
(Shimadzu, Japan); FT-IR spectrometer (PerkinElmer, USA);
JSM-6700F field emission scanning electron microscopy (JEOL,
Japan); STA 409 thermal gravimetric analyzer (Netzsch, Germany);
EV11 Vibrating Sample Magnetometer (MicroSense, USA); D/Max
2500 X-ray diffraction (Rigaku, Japan); Mos-500 circular dichroism
(CD) spectrometer (Biologic, France); incubator shaker (QYC 200;
FuMa Experimental Equipment Co., Ltd., Shanghai, China);
ultrapure water instrument (RM 220; LiDe Experimental
Equipment Co., Ltd., Shanghai, China).
2.2. Chemicals and reagents
All reagents used were of at least analytical grade and needed
no further purification. Graphite powder, KMnO4, BaCl2, NaCl,
Na2HPO4, H2O2 (30%), FeCl36H2O, FeSO47H2O, ammonium
hydroxide, methanol, D-(+)-glucose, urea, glycerol and bovine
serum albumin (BSA) were purchased from Sinopharm Chemical
Reagent Co., Ltd. (Shanghai, China). NaNO3 was supplied by Taishan
Chemical Co., Ltd. (Guangdong, China). Concentrated sulfuric acid
and hydrochloric acid were obtained from Zhuzhou Star Glass Co.,
Ltd. (Hunan, China). Hydrazine hydrate (80%) was purchased from
92 Y. Huang et al. / Analytica Chimica Acta 877 (2015) 90–99

Table 1
Deep eutectic solvents based on choline chloride and four different hydrogen bond donors.

Abbreviation ChCl Hydrogen bond donor Mole ratio of ChCl to HBD

Compound Structure
DES1 Urea O 1:2
-
Cl
N+

H 2N NH2
OH

DES2 D-(+)-Glucose 1:1

HO

O
HO
HO

OH OH

DES3 Ethylene glycol 1:2

HO
OH

DES4 Glycerol OH 1:1

HO OH

Shanghai Shanpu Chemical Co., Ltd. (Shanghai, China). Ethylene
glycol was procured from Hengyang Gaixin Chemical Reagent Co.,
Ltd. (Hunan, China). Choline chloride (ChCl), ovalbumin (OVA),
lysozyme (Lys), and bovine hemoglobin (BHb) were obtained from
Shanghai Yuanye Biological Technology Co., Ltd. (Shanghai, China).
Ultrapure water was used as the working medium.
2.3. Preparation of graphene oxide (GO)
GO was prepared by oxidizing graphite powder based on a
modified Hummers method [36]. 23 mL of concentrated H2SO4
solution was mechanically stirred 30 min in ice-water bath. Then,
0.5 g of graphitic powder and 0.5 g of NaNO3 were added in turn
and kept on stirring 30 min. 3.0 g of KMnO4 was gradually added
into the above mixture under continuous vigorous stirring and the
temperature was kept below 15  C. The stirring was kept on for 2 h
in ice-water bath, then a water bath replaced the ice-water bath
and the mixture was stirred 30 min at 35  C. In order to dilute the
mixture, 100 mL of ultrapure water was slowly added, and then
continued stirring 30 min at 95  C. The reaction was terminated by
slowly added 30% H2O2. After cooling down, the solid product was
collected by centrifugation and washed with 1 mol L1 HCl until
SO42 ions were no longer detectable with BaCl2. Finally, the solid
product was washed with ultrapure water until the washing
solution was neutral. And then freeze dried. At the end, a loose
brown power was obtained.
2.4. Preparation of magnetic graphene oxide (Fe3O4@GO)
The Fe3O4@GO particles were prepared by the modified
co-precipitation method [37,38]. First, 0.2 g of GO in 50 mL of
water was ultra-sonicated 30 min. 0.006 mol of FeCl36H2O and
0.003 mol of FeSO47H2O were dissolved in 25 mL water in
sequence and the resulting aqueous solution was slowly added
into the above mixture at room temperature with violent stirring.
The violent stirring was maintained 2 h for ion exchange. Then,
450 mL of 80% hydrazine hydrate was added into the mixture
quickly, ammonium hydroxide was dropwise added until pH 9. The
reaction was kept at room temperature for 30 min, followed by 2 h
at 60  C in water bath. The solid product was collected by magnetic
separation and washed with ultrapure water until the washing
solution was neutral. Finally, the Fe3O4@GO particles were freeze
dried.
2.5. Preparation of magnetic graphene oxide-deep eutectic solvent
(Fe3O4@GO-DES)
Four kinds of deep eutectic solvents were prepared by heating
two eutectic mixtures at 100  C with constant stirring until a
homogeneous liquid was formed [39]. The deep eutectic solvents
(as shown in Table 1 and Fig. 1) were based on choline chloride
(ChCl) and four different hydrogen bond donors (HBDs) with
desired molar ratio.

Fig. 1. Preparation of deep eutectic solvents (DES2 as representative).

 Y. Huang et al. / Analytica Chimica Acta 877 (2015) 90–99 93

The preparation of Fe3O4@GO-DES was carried out
according to the previous report with some modification
[30]. 1 mL of DES was dissolved in 3 mL of methanol. 0.2 g
of Fe3O4@GO was taken in a round bottomed flask and
2 mL of the dissolved DES was added and sonicated for 1 h.
After an interval of 30 min, the residual dissolved DES was
added and kept sonicating for further 1 h. Let the resulting
product stand overnight and then washed twice with methanol.
Finally, the resulting DES impregnated Fe3O4@GO was dried at
80  C and then used for further studies. Scheme 1 illustrates
the preparation of Fe3O4@GO-DES and its application in MSPE
of protein.
2.6. Magnetic solid-phase extraction of proteins
In the MSPE experiments, Fe3O4@GO-DES was used as extractant
and BSA was chosen as representative to research the extraction
capacity. 10 mg of Fe3O4@GO-DES and 1 mL of the protein solution
with the concentration of 1 mg mL1 were displaced into a 2 mL
centrifuge tube and then shaken at room temperature with a
shaking speed of 200 rpm for 1 h. After extraction, Fe3O4@GO-DES
was recovered from the suspension using an external magnetic field
and the concentration of the protein in the supernatant was
measured by the UV–vis spectrophotometer at the maximum
absorption wavelength (278 nm).

Scheme 1. Preparation of Fe3O4@GO-DES and its application for the MSPE of protein.
94 Y. Huang et al. / Analytica Chimica Acta 877 (2015) 90–99

The extraction amount (Q) was calculated according to the

following equation:
ðC 0  CÞV
Q¼ (1)
m
where Q (mg g1) is the mass of protein adsorbed onto a unit
amount of Fe3O4@GO-DES, C0 and C (mg mL1) are the initial and
final concentration of proteins in the solution, V (mL) is the volume
of the initial solution and m (mg) is the mass of Fe3O4@GO-DES.

3. Results and discussion

3.1. Characterization

3.1.1. X-ray diffraction

The XRD patterns of GO, Fe3O4@GO and Fe3O4@GO-DES are
presented in Fig. 2 (Fe3O4@GO-DES2 was selected as the model to
show their similar physical and chemical properties). Obviously,
the significant structural changes occurred while GO turned into
Fe3O4@GO-DES. In Fig. 2a, a considerably sharp diffraction peak,
which was resulted from the (0 0 2) plane of GO, appeared at 10.12 .
This was along with a wide diffraction peak at 20.98 , which was
attributed to the short-range order in stacked graphene sheets
[40]. From Fig. 2b and c, it can be observed that the diffraction peak
at 21.02 that originated from GO indicates the existence of GO.
Moreover, six primary diffraction peaks for Fe3O4, appearing at
30.1, 35.5 , 43.1, 53.5 , 57.3 and 62.7, can be seen clearly in
Fe3O4@GO and Fe3O4@GO-DES. These peaks belonged to the (2 2 0),
(3 11), (4 0 0), (4 2 2), (5 11) and (4 4 0) planes, respectively (JCPDS
card: 019-0629). The XRD patterns indicated that Fe3O4@GO-DES
possesses magnetic properties, making it possible for the magnetic
separation. Fig. 3. (A) Magnetic hysteresis loops of Fe3O4@GO (a) and Fe3O4@GO-DES (b), and
(B) the magnetic response of Fe3O4@GO-DES to external magnetic field.

3.1.2. Vibrating sample magnetometer

It is completely important for the investigated materials to have
sufficient magnetic property in order to be applied to magnetic
solid-phase extraction. Hence, vibrating sample magnetometer
(VSM) was used to study the magnetic properties of the materials
at ambient temperature. From Fig. 3A (Fe3O4@GO-DES2 as
representative), it can be seen that the magnetic hysteresis loop
of Fe3O4@GO and Fe3O4@GO-DES is S-like curve. The saturation
magnetization of Fe3O4@GO and Fe3O4@GO-DES are 34.65 emu g1
and 27.73 emu g1, respectively. Compared with Fe3O4@GO, the
saturation magnetization of Fe3O4@GO-DES is decreased. This is to
be expected, because the DESs exist on the surface of Fe3O4@GO
[41]. Luckily, the saturation magnetization of Fe3O4@GO-DES was
adequate to ensure the convenient separation of the materials
from solutions. Fig. 3B shows that an external magnet was able to
collect the Fe3O4@GO-DES from aqueous solution.
3.1.3. Fourier transform infrared spectrometry
The FTIR spectra were used for identifying the successful
modification of magnetic particles with deep eutectic solvent. The
results were shown in Fig. 4 (Fe3O4@GO-DES2 as representative).
For the FTIR spectrum of DES2, the appearance of a broad band

Fig. 2. X-ray diffraction patterns of GO (a), Fe3O4@GO (b) and Fe3O4@GO-DES (c). Fig. 4. FTIR spectra of DES (a), GO (b), Fe3O4@GO (c) and Fe3O4@GO-DES (d).
 Y. Huang et al. / Analytica Chimica Acta 877 (2015) 90–99
Fig. 5. FESEM images of GO (a), Fe3O4@GO (b) and Fe3O4@GO-DES (c).
between 3600 and 3000 cm1 ( OH) and the peaks around
2939 cm1 (CH3), 2883 cm1 (CH2) and 1041 cm1 (C N) indicat-
ed the successful synthesis of DES2. The FTIR spectrum of GO
showed the peaks at 1071 cm1 (COC), 1226 cm1 (C OH),
1739 cm1 (C¼O stretching vibrations of COOH) and 3419 cm1
(OH), revealing the existence of the oxygen-containing
functional groups. The peak appeared at 1645 cm1 was assigned
to the stretching vibrations of C¼C in the carbon skeletal
network. The FTIR spectrum of Fe3O4@GO and Fe3O4@GO-DES2
differed from that of GO as demonstrated by the weakening of
the peaks that located at 1739 cm1 (C¼O) and 3419 cm1 (OH).
Furthermore, they both presented the peak that originated
from Fe O at 582 cm1, proving the presence of Fe3O4. For
Fe3O4@GO-DES2, the new peak appeared at 1033 cm1, which
N, confirmed the
corresponded to the stretching vibration of C
successful modification of Fe3O4@GO with DES2.
3.1.4. Field emission scanning electron microscopy
FESEM images of GO, Fe3O4@GO and Fe3O4@GO-DES are
demonstrated in Fig. 5 (Fe3O4@GO-DES2 as representative). From
Fig. 5a, it can be obviously seen that GO appear to be sheet-like
structure, which has smooth surface and wrinkled edge.
After forming magnetic Fe3O4@GO, it can be found that Fe3O4
particles are distributed on the surface of GO, making the
surface of Fe3O4@GO rough. Fig. 5c shows the surface morphology
of the obtained Fe3O4@GO-DES. We can observe that there
was a morphological difference between Fe3O4@GO and
Fig. 6. TGA curves of Fe3O4@GO (a) and Fe3O4@GO-DES (b).
95
Fe3O4@GO-DES. The Fe3O4@GO-DES is porous with rougher
surface, indicating Fe3O4@GO-DES possesses large surface area.
3.1.5. Thermal gravimetric analysis
The experimental results of TGA are provided in Fig. 6
(Fe3O4@GO-DES2 as representative). Under experimental condi-
tions, there is a weight loss of about 2.7% for Fe3O4@GO, which was
attributed to the removal of solvent and water and the
decomposition of labile oxygen. For Fe3O4@GO-DES, the weight
loss, which was originated from the removal of solvent and water
and the decomposition of labile oxygen, was around 3.0% when the
temperature was up to 150  C. In the range of 150–800  C, the
decomposition of deep eutectic solvent occurred, and the weight
loss was approximately 11.5%. That is, the amount of deep eutectic
solvent deposited on the Fe3O4@GO particles was about 11.5%. In
comparison, Fe3O4@GO-DES exhibited more weight loss than
Fe3O4@GO as the temperature rose to 800  C, confirming the
existence of deep eutectic solvents on the Fe3O4@GO particles.
3.2. Magnetic solid-phase extraction of proteins
BSA was selected as representative protein to investigate the
extraction capacity of the proposed magnetic solid-phase
extraction based on deep eutectic solvent-coated Fe3O4@GO.
Several parameters, including the types of DESs, the solution
temperature, the solution ionic strength, the extraction time, the
protein concentration and the amount of Fe3O4@GO-DES, were
examined.
3.2.1. Effect of the types of DESs
Four kinds of deep eutectic solvents with different hydrogen
bond donors have been selected to coat the surfaces of magnetic
particles, which have been investigated for the extraction of BSA.
The experimental results indicated that the extraction amount of
the investigated deep eutectic solvents were different (DES1:30.0
mg g1, DES2:48.4 mg g1, DES3:38.0 mg g1, DES4:22.8 mg g1). It
can be seen that the extraction amount was not only affected by the
types of the hydrogen bond donors, but also affected by the
proportion of functional groups. For example, DES1 and DES3 have
the same proportion of functional groups, but the types of the
functional groups are different ( NH2 and OH, respectively).
Hydrogen bonding interactions serve as the important driving
force in protein extraction [39].  OH is much easier to form
hydrogen bond with protein, because the electronegativity of
oxygen is greater than that of nitrogen. Therefore, the extraction
amount of Fe3O4@GO-DES3 is higher than that of Fe3O4@GO-DES1.
For DES2, DES3 and DES4, they have the same types of functional
groups, but the proportion of functional groups is different. The
proportion of functional groups in DES2 is the highest, so it has the
96 Y. Huang et al. / Analytica Chimica Acta 877 (2015) 90–99
most sites to form hydrogen bond with protein. Consequently, the
extraction amount of Fe3O4@GO-DES2 is the highest. Finally,
Fe3O4@GO-DES2 was chosen in the following studies.
3.2.2. Effect of solution temperature
In order to study the influence of temperature on the
extraction capacity of Fe3O4@GO-DES2, the experiment was
carried out at different temperature. The experimental results
were shown in Fig. 7a, which suggested that solution temperature
obviously affected the extraction of BSA. As the solution
temperature rising, the extraction capacity increased. This
observation indicated that the extraction is an endothermic
process. When the temperatures go from 30 to 40  C, the
increasing of extraction capacity turns slow. It is likely that the
hydrophobic interactions enhance as the temperature increases,
but the hydrogen bonding interactions between the deep eutectic
solvents coated on the surfaces of magnetic particles and the
amino acid residue could be destroyed when the temperature is
enough high. It is worth mentioning that higher temperature can
denature the protein. Therefore, to prevent protein to be
denatured, we have tried to process the extraction at lower
temperature. From experimental results, although the extraction
amount has not reached the best value when the extraction
performed at room temperature, it has gained a better value. In
addition, we have taken full account of the simplicity of
operation. Ultimately, we have chosen room temperature.
3.2.3. Effect of solution ionic strength
Fig. 7b depicted extraction tendency of BSA as the concentra-
tion of NaCl in the range of 0.0–0.5 mol L1. It is so clear that the
increasing of ionic strength within a certain range obviously
decreases the extraction capacity. It could be speculated that the
Fig. 7. Effects of (a) solution temperature, (b) solution ionic strength, (c) extraction time, (d) protein concentration and (e) amount of Fe3O4@GO-DES2.
electrostatic interaction is one of the main driving forces during
the extraction, and an ion-exchange process between sodium ions
and deep eutectic solvents may exist. When the concentration of
NaCl was over 0.15 mol L1, the downward trend was not as obvious
as in the concentration of less than 0.15 mol L1. Then, the
extraction capacity which was close to the minimum value, but
not a negligible value, was achieved. Further increasing of ionic
strength would not obviously change the extraction capacity. The
reason may be that the main driving forces of the extraction
included hydrophobic interaction and hydrogen bonding interac-
tion, not just electrostatic interaction. Since the addition of NaCl
would not help to improve the extraction capacity, therefore there
is no need to add NaCl in the subsequent experiments.
3.2.4. Effect of extraction time
The extraction tendency of BSA as time passed was presented in
Fig. 7c. If the protein was totally extracted during the experiment, it
is likely to cause the illusion of balance. In order to avoid the
illusion of balance happening, 2 mg L1 of protein was used. From
the relation curve between extraction capacity and time, decrease
in the slope of the curve was observed, meaning that the decline in
extraction rates happened. In the first hour, the extraction capacity
increased rapidly, and then the upward trend became slower
gradually in the following 11 h. After 12 h, the extraction reached
equilibrium and the extraction capacity nearly reached a constant
value. This phenomenon can be explained by the fact that there are
many activated sites available for the extraction of protein at the
beginning, making the extraction easy to proceed. As time went on,
more action sites had been occupied, so that the extraction rates
became slower and slower. Ultimately, the extraction reached
equilibrium. Therefore, 12 h was chosen for proceeding the
extraction.
 Y. Huang et al. / Analytica Chimica Acta 877 (2015) 90–99
3.2.5. Effect of protein concentration
Fig. 7b shows the extraction capacities of Fe3O4@GO-DES2
at different initial BSA concentrations, with the values in the
range of 0.1–2.0 mg mL1. For obtaining equilibrium, the solutions
were shaken for 24 h. In the beginning, as the increment of the
beginning concentrations of BSA, the BSA uptake increased
linearly. Additionally, BSA was nearly totally extracted by
Fe3O4@GO-DES2 when the beginning concentrations were less
than 1.0 mg mL1, meaning that the extraction efficiency was
almost reached 100%. This result demonstrated that there were
sufficient action sites on Fe3O4@GO-DES2 and that the extraction
capacity completely depended on the amount of BSA transferred
from the aqueous solutions to the surfaces of Fe3O4@GO-DES2 at
lower beginning concentrations. Moreover, when the beginning
concentration was larger than 1.0 mg mL1, the extraction
capacities almost remained unchanged whereas the extraction
efficiencies decreased linearly, because the action sites on
Fe3O4@GO-DES2 were saturated.
3.2.6. Effect of the amount of Fe3O4@GO-DES2
Considering that the amount of Fe3O4@GO-DES2 was a vital
factor related to the extraction capacity, different amount of
Fe3O4@GO-DES2 ranging from 2 to 22 mg were studied. The
solutions were shaken for 24 h to achieve balance. As illustrated in
Fig. 7e, the extraction capacity of Fe3O4@GO-DES2 decreased with
the increasing of amount of Fe3O4@GO-DES2. However, the
extraction efficiency presented different changing tendency. From
the inset, it can be seen that the extraction efficiency increased
with the amount of Fe3O4@GO-DES2 increasing. This phenomenon
may be attributed to the fact that more number of action sites
would be available for protein molecules at higher amount of
Fe3O4@GO-DES2 [42]. When the amount of Fe3O4@GO-DES2 was
more than 14 mg, most of the protein had been extracted and the
extraction efficiency remained approximately constant.
3.3. Comparison of Fe3O4@GO and Fe3O4@GO-DES2
Fe3O4@GO and Fe3O4@GO-DES2 were used to extract proteins
under the same conditions with the purpose of evaluating the
advantage of the proposed magnetic solid-phase extraction based
on deep eutectic solvent-coated Fe3O4@GO. From the results
Fig. 8. (A) Comparison of Fe3O4@GO and Fe3O4@GO-DES2 and (B) zeta potentials of Fe3O4@GO and Fe3O4@GO-DES2.
97
showed in Fig. 8A, it is obvious that both materials had the ability
to extract proteins, while the extraction capacity showed different
value. GO can be used to extract protein because of the existence of
many oxygen-containing functional groups, which can form
hydrogen bonding and electrostatic interaction with proteins
[12]. Besides, GO can interact with proteins through hydrophobic
interactions [43]. Both the materials (without and with DES)
combined the advantages of Fe3O4 and GO, so they can be used for
magnetic solid-phase extraction of proteins. During the process of
extraction, these three forces (hydrogen bonding, hydrophobic
interaction and electrostatic interaction) are considered to affect in
different extent for both materials. For Fe3O4@GO-DES2, the
introduction of DES2 makes the hydrogen bonding stronger,
because there are many hydroxyl groups in DES2. Since there
are no long alkyl chains, the change of hydrophobic interaction
may be little. According to Fig. 8B, after introducing DES2, the
surface charge of the materials changed, negatively charged
(Fe3O4@GO, zeta potential = 10.6 mV) to positively charged
(Fe3O4@GO-DES2, zeta potential = 19.4 mV), so the electrostatic
interaction changed. It can be seen from Fig. 8A that DES improves
the extraction capacity only for acidic proteins but not neutral and
basic proteins. These phenomena may be attributed to the
electrostatic interaction. The pH value of ultrapure water was
determined and the obtained result was 6.32. When using
ultrapure water as the working medium, the surfaces of Fe3O4@GO
(zeta potential = 10.6 mV), BSA (pI = 4.8) and OVA (pI = 4.7) were
negatively charged, while the surfaces of Fe3O4@GO-DES2
(zeta potential = 19.4 mV), BHb (pI = 6.9) and Lys (pI = 11.0) were
positively charged. Therefore, the existence of DES can help to
improve the extraction capacity only for acidic proteins.
3.4. Desorption studies
Desorption of analytes from the solid extractants is an
important part in the entire process. Therefore, a series of
desorption experiments were proceeded. It is well known that
hydrogen bonds can be broken under alkaline condition.
Accordingly, alkaline environment can destroy the hydrogen
bonds between Fe3O4@GO-DES2 and BSA. Besides, the extraction
amounts of BSA decrease at high ionic strength. Based on the above
reasons, 0.005 mol L1 Na2HPO4 contained 1 mol L1 NaCl was
98 Y. Huang et al. / Analytica Chimica Acta 877 (2015) 90–99

Fig. 9. (A) The CD spectra of BSA in different media and (B) extraction amount of protein in different cycles.

selected as the eluent to release BSA from the solid extractants.
During the desorption process, the alkaline condition provided by
Na2HPO4 broke the hydrogen bonds and the eluent also weakened
the electrostatic interaction. The elution efficiency was calculated.
The experimental result showed that the elution efficiency was
about 90.9%. Circular dichroism spectra (CD spectra) have been
used to determine the secondary structure of BSA eluted from the
solid extractants. In CD spectra, the double negative peaks
appeared at 210 nm and 220 nm are characteristic of a-helical
structure of protein [44,45]. From Fig. 9A, it can be seen that the
spectral shape of BSA eluted from the solid extractants is similar to
that of BSA in aqueous solution, and both of them show the double
negative peaks at the same wavelengths. The obtained result
suggests that the secondary structures of BSA eluted from the solid
extractants are not changed.
After desorption, the magnetic materials were recovered by an
external magnetic field, then were used to extract protein again.
Three consecutive extraction–desorption cycles were carried out.
The results are shown in Fig. 9B. It can be seen a significant loss of
the extraction capacity of protein after three consecutive cycles. It
may be due to the loss of DES on the surface of Fe3O4@GO and the
loss of magnetic materials (both materials are weighed once and
washed multiple times). When compared with Fe3O4@GO, it can
be found that the extraction capacity of the recycled magnetic
materials and Fe3O4@GO are similar. It may be proved that the
loss of DES occurred during the process of elution.
To further understand whether DES still existed on the surface
of Fe3O4@GO, the FTIR spectra of Fe3O4@GO-DES2 (before
extraction and after elution) have been measured (shown in
Fig. 10A). Compared the FTIR spectra, it can be found that the
disappearance of C N at 1033 cm1 in Fe3O4@GO-DES2 (after
elution), indicating that DES has transferred from the
magnetic materials to the eluent in the process of elution.
Therefore, after elution, DES existed in the protein solution. To
find out DES existed alone or combined with protein, the
FTIR spectra of DES, BSA and the BSA solution (after elution)
have been measured (shown in Fig. 10B). Amide is a critical
part of the protein: the IR absorption band of amide I appeared at
1700–1600 cm1, while the amide II appeared at 1600–1500 cm1
[46]. As shown in Fig. 10B, after elution, the adsorption bands
of BSA (the amide I band at 1658 cm1 and the amide II band
at 1541 cm1) have changed to be 1639 cm1 and 1477 cm1. The
adsorption bands were red shifted. It may be attributed to
the hydrogen bonding interaction between BSA and DES. That
is, the desorbed proteins which combined with DES have
existed in the BSA solution (after elution). The new peak
appeared at 989 cm1 (which corresponded to the stretching
vibration of CN, have changed from 1041 cm1 to 989 cm1 due
to the hydrogen bonding interaction between BSA and DES)
indicating the existence of DES. However, much to our regret, we
are unable to judge the presence of free DES in the protein
solution from the FTIR spectra.

Fig. 10. (A) The FTIR spectra of Fe3O4@GO-DES2 before extraction (a) and after elution (b), and (B) the FTIR spectra of DES (a), pure BSA (b) and the BSA solution (c, after
elution). (For interpretation of the references to color in text, the reader is referred to the web version of this article.)
 Y. Huang et al. / Analytica Chimica Acta 877 (2015) 90–99 99

Table 2 Acknowledgements
The results of the precision, repeatability and stability experiments.

Precision measurement results (n = 5) The authors greatly appreciate the financial supports by the
Repeats 1 2 3 4 5
National Natural Science Foundation of China (No. 21175040;
Absorbance 0.636 0.630 0.635 0.632 0.634 No. 21375035; No. J1210040) and the Foundation for Innovative
Extraction amounts (mg g1) 99.3 100.2 99.4 99.9 99.6 Research Groups of NSFC (Grant No. 21221003).
RSD (%) 0.39

References
Repeatability measurement results (n = 5)

Sample number 1 2 3 4
Absorbance 0.636 0.637 0.633 0.636
Extraction amounts (mg g1) 99.3 99.0 99.8 99.3
RSD (%) 0.47
Stability measurement results (n = 5)
Day number 1 2 3 4
Absorbance 0.636 0.639 0.634 0.639
Extraction amounts (mg g1) 99.3 98.8 99.6 98.8
RSD (%) 0.51
3.5. Validation of the proposed method
The stability, repeatability and precision experiments have
been used to validate the proposed method. In this part,
10 mg of Fe3O4@GO-DES2 and 1 mL of BSA aqueous solution
(2.0258 mg mL1) were displaced into a 2 mL centrifuge tube and
then shaken at room temperature with a shaking speed of
200 rpm for 24 h to reach equilibrium. Apparatus precision was
evaluated by measuring the absorbance value of the protein
in the supernatant for five times under the same conditions. The
relative standard deviation (RSD) of extraction amount is 0.39%
(n = 5), indicating that the UV–vis spectrophotometer has
satisfactory precision. Repeatability experiments were carried
out by measuring five copies of the same sample at the same
conditions. The results show that the RSD is 0.47% (n = 5), proving
the good repeatability of the method. Stability experiments were
performed by detecting a sample continuously within five days
under identical conditions. The value of RSD is 0.51% (n = 5),
demonstrating the favorable stability of the proposed method.
The results of the experiments are given in Table 2.
4. Conclusion
It is the first time to report the magnetic solid-phase
extraction of protein based on deep eutectic solvent-coated
magnetic GO (Fe3O4@GO-DES). The deep eutectic solvents used in
this study are environmentally friendly and have an atom
utilization rate of 100% in the synthesis process. Additionally,
the synthetic materials of DESs are abundant and cheap. Thus,
DESs are expected to be used in various fields. Accordingly, the
proposed method has used DESs to functionalize the magnetic
materials for extraction of proteins. The research result shows
that the proposed Fe3O4@GO-DES is more superior to Fe3O4@GO
in the extraction of acidic proteins. Single factor experiments
illustrated that the extraction efficiency was influenced
by the solution temperature, the solution ionic strength, the
extraction time, the protein concentration and the amount of
Fe3O4@GO-DES. The magnetic materials can be easily recycled,
but the extraction capacity shows significant loss when reusing
it, making us feel regret. Even so, the proposed Fe3O4@GO-DES
has been successfully applied in magnetic solid-phase extraction
of protein and exhibits potential applications in the extraction of
other biomolecules.
5 [1] M. Kumar, J. Lawler, Sep. Purif. Technol. 130 (2014) 112–123.
0.630 [2] K. Aguilar-Arteaga, J.A. Rodriguez, E. Barrado, Anal. Chim. Acta 674 (2010)
100.2 157–165.
[3] L.G. Chen, T. Wang, J. Tong, TrAC, Trends Anal. Chem. 30 (2011) 1095–1108.
[4] T. Wei, S. Kaewtathip, K. Shing, J. Phys. Chem. C 113 (2009) 2053–2062.
[5] A.K. Gupta, M. Gupta, Biomaterials 26 (2005) 3995–4021.
[6] Y.B. Luo, Q.W. Yu, B.F. Yuan, Y.Q. Feng, Talanta 90 (2012) 123–131.
5 [7] A.A. Karamani, A.P. Douvalis, C.D. Stalikas, J. Chromatogr. A 1271 (2013) 1–9.
0.632 [8] M.Y. Hou, X.H. Zang, C. Wang, Z. Wang, J. Sep. Sci. 36 (2013) 3242–3248.
99.9 [9] Y.N. Jiao, S.L. Fu, L. Ding, Q. Gong, S.H. Zhu, L.B. Wang, H. Li, Anal. Methods 4
(2012) 2729–2734.
[10] D.N. Huang, C.F. Fu, Z.B. Li, C.H. Deng, J. Sep. Sci. 35 (2012) 1667–1674.
[11] Y. Zou, Y.Z. Chen, Z.H. Yan, C.Y. Chen, J.P. Wang, S.H. Yao, Analyst 138 (2013)
5904–5912.
[12] R. Sitko, B. Zawisza, E. Malicka, TrAC, Trends Anal. Chem. 51 (2013) 33–43.
[13] W. Gao, L.B. Alemany, L.J. Ci, P.M. Ajayan, Nat. Chem. 1 (2009) 403–408.
[14] S.W. Li, Z.Y. Niu, X. Zhong, H.L. Yang, Y.J. Lei, F.W. Zhang, W.Q. Hu, Z.P. Dong, J.
Jin, J.T. Ma, J. Hazard. Mater. 229 (2012) 42–47.
[15] W.N. Wang, R.Y. Ma, Q.H. Wu, C. Wang, Z. Wang, J. Chromatogr. A 1293 (2013)
20–27.
[16] M.C. Liu, C.L. Chen, J. Hu, X.L. Wu, X.K. Wang, J. Phys. Chem. C 115 (2011)
25234–25240.
[17] X.J. Hu, Y.G. Liu, H. Wang, A.W. Chen, G.M. Zeng, S.M. Liu, Y.M. Guo, X. Hu, T.T. Li,
Y.Q. Wang, L. Zhou, S.H. Liu, Sep. Purif. Technol. 108 (2013) 189–195.
[18] V. Chandra, J. Park, Y. Chun, J.W. Lee, I.C. Hwang, K.S. Kim, ACS Nano 4 (2010)
3979–3986.
[19] V. Chandra, K.S. Kim, Chem. Commun. 47 (2011) 3942–3944.
[20] H.M. Sun, L.Y. Cao, L.H. Lu, Nano Res. 4 (2011) 550–562.
[21] Y.J. Yao, S.D. Miao, S.Z. Liu, L.P. Ma, H.Q. Sun, S.B. Wang, Chem. Eng. J. 184 (2012)
326–332.
[22] J.H. Deng, X.R. Zhang, G.M. Zeng, J.L. Gong, Q.Y. Niu, J. Liang, Chem. Eng. J. 226
(2013) 189–200.
[23] A.P. Abbott, D. Boothby, G. Capper, D.L. Davies, R.K. Rasheed, J. Am. Chem. Soc.
126 (2004) 9142–9147.
[24] F. Pena-Pereira, J. Namiesnik, ChemSusChem 7 (2014) 1784–1800.
[25] Q.H. Zhang, K.D.O. Vigier, S. Royer, F. Jérôme, Chem. Soc. Rev. 41 (2012)
7108–7146.
[26] M. Hayyan, M.A. Hashi, A. Hayyan, M.A. Al-Saadi, I.M. AlNashef, M.E.S.
Mirghani, O.K. Saheed, Chemosphere 90 (2013) 2193–2195.
[27] Y.H. Yu, X.M. Lu, Q. Zhou, K. Dong, H.W. Yao, S.J. Zhang, Chem. Eur. J. 14 (2008)
11174–11182.
[28] K.D. Weaver, H.J. Kim, J.Z. Sun, D.R. MacFarlane, G.D. Elliott, Green Chem. 12
(2010) 507–513.
[29] F. Ilgen, D. Ott, D. Kralisch, C. Reil, A. Palmberger, B. König, Green Chem. 11
(2009) 1948–1954.
[30] L.L. Li, C.N. Luo, X.J. Li, H.M. Duan, X.J. Wang, Int. J. Biol. Macromol. 66 (2014)
172–178.
[31] X.J. Cao, L.X. Shen, X.M. Ye, F.F. Zhang, J.Y. Chen, W.M. Mo, Analyst 139 (2014)
1938–1944.
[32] X.Q. Ding, Y.Z. Wang, Y. Wang, Q. Pan, J. Chen, Y.H. Huang, K.J. Xu, Anal. Chim.
Acta 861 (2015) 36–46.
[33] L. Vidal, M.L. Riekkola, A. Canals, Anal. Chim. Acta 715 (2012) 19–41.
[34] S.J. Guo, S.J. Dong, E.K. Wang, Adv. Mater. 22 (2010) 1269–1272.
[35] P. Du, J. Zhao, H. Mashayekhi, B.S. Xing, J. Phys. Chem. C 118 (2014)
22249–22257.
[36] W.S. Hummers Jr., R.E. Offeman, J. Am. Chem. Soc. 80 (1958) 1339–1339.
[37] M.Z. Kassaee, E. Motamedi, M. Majdi, Chem. Eng. J. 172 (2011) 540–549.
[38] M. Zhang, Y.Z. Wang, X.P. Jia, M.Z. He, M.L. Xu, S. Yang, C.J. Zhang, Talanta 120
(2014) 376–385.
[39] Q. Zeng, Y.Z. Wang, Y.H. Huang, X.Q. Ding, J. Chen, K.J. Xu, Analyst 139 (2014)
2565–2573.
[40] C. Nethravathi, M. Rajamathi, Carbon 46 (2008) 1994–1998.
[41] X.P. Jia, M.L. Xu, Y.Z. Wang, D. Ran, S. Yang, M. Zhang, Analyst 138 (2013)
651–658.
[42] B.M. Borah, B. Saha, S.K. Dey, G. Das, J. Colloid Interface Sci. 349 (2010) 114–121.
[43] J.B. Liu, S.H. Fu, B. Yuan, Y.L. Li, Z.X. Deng, J. Am. Chem. Soc. 132 (2010)
7279–7281.
[44] Y. Yue, H.W. Li, T.Y. Liu, Y.Q. Wu, Vib. Spectrosc. 74 (2014) 137–141.
[45] C.B. Xu, J.L. Gu, X.P. Ma, T. Dong, X.L. Meng, Spectrochim. Acta, Part A 125 (2014)
391–395.
[46] M.L. Sheng, Y. Gao, J.Y. Sun, F. Gao, Biosens. Bioelectron. 58 (2014) 351–358.