2011_Optimization of adsorption conditions of BSA on thermosensitive magnetic

Prévia do material em texto

O
c
M
a
b
c
a
A
R
R
1
A
A
K
A
N
R
M
T
1
i
T
n
t
p
h
i
f
o
n
l
w
[
p
a
i
i
t
i
0
d
Colloids and Surfaces B: Biointerfaces 84 (2011) 477–483
Contents lists available at ScienceDirect
Colloids and Surfaces B: Biointerfaces
journa l homepage: www.e lsev ier .com/ locate /co lsur fb
ptimization of adsorption conditions of BSA on thermosensitive magnetic
omposite particles using response surface methodology
ing-Min Songa, Christopher Branford-Whiteb, Hua-Li Niea,c,∗, Li-Min Zhua,∗
College of Chemistry, Chemical Engineering and Biotechnology, Donghua University, 2999 North Remin Road, Songjiang University City, Shanghai 201620, China
Institute for Health Research & Policy, London Metropolitan University, 166-220 Holloway Road, London, UK
Key Laboratory of Science & Technology of Eco-Textile (Donghua University), Ministry of Education, P.R. China
r t i c l e i n f o
rticle history:
eceived 30 August 2010
eceived in revised form
1 December 2010
ccepted 1 February 2011
a b s t r a c t
Thermosensitive core–shell magnetic composite particles with a magnetic silica core and a rich poly
(N-vinylcaprolactam) (PNVCL) shell layer were developed for studying the adsorption of bovine serum
albumin (BSA) in a batch system. Various analytical and spectroscopic techniques including SEM, FT-IR,
VSM and DSC were used to characterize the adsorbents prepared in this study. The combined effects of
operating parameters such as initial temperature, pH and initial BSA concentration on the adsorption
vailable online 26 February 2011
eywords:
dsorption
-vinylcaprolactam
esponse surface methodology
were analyzed using response surface methodology. The optimum conditions were 40 ◦C, pH 4.68, and
initial BSA concentration 2.0 mg/mL. Desorption experiments were conducted by altering the system
temperature where a high recovery rate of protein was obtained. The separation process developed here
indicates that the dual-responsive smart adsorbent could be an ideal candidate for the separation of
protein.
agnetic separation
hermosensitive
. Introduction
Nowadays, magnetic nanoparticles with tailored surface chem-
stry have been widely used in diverse areas of biotechnology.
hanks to their unique size, biocompatibility, and superparamag-
etic properties, magnetic nanoparticles are emerging as promising
ools for isolating, separating, or concentrating biological sam-
les from liquid suspensions [1,2]. Various magnetic nanoparticles
ave been used with different surface-functionalizations that allow
mmobilized affinity ligands to capture target biomaterials. Sur-
ace modifications are usually achieved through the attachment
f inorganic shells or/and organic molecules that stabilize the
anoparticles and protect them from being oxidised. Recently, a
ot of investigations have been devoted to designing nanoparticles
ith surface coating layers sensitive to variations of temperature
3,4], pH value [5,6], and specific analytes [7]. These nanocom-
osites make a great contribution to the extension of the area for
pplying new functional materials.
Thermosensitive polymers are attracting increasing research
nterests due to their unique property and potential application
n biomedical fields. They can respond to the external stimuli
hrough an on–off switch mechanism [8]. The most widely stud-
ed temperature sensitive polymer is poly (N-isopropylacrylamide)
∗ Corresponding authors. Tel.: +86 21 67792659; fax: +86 21 67792655.
E-mail addresses: niehuali@dhu.edu.cn (H.-L. Nie), lzhu@dhu.edu.cn (L.-M. Zhu).
927-7765/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
oi:10.1016/j.colsurfb.2011.02.002
© 2011 Elsevier B.V. All rights reserved.
(PNIPAM). Recently, another thermo-sensitive polymer, poly (N-
vinylcaprolactam) (PNVCL) which presents higher biocompatibility
compared with PNIPAM, is being extensively studied [9–11]. PNVCL
is water-soluble and hydrophilic below the lower critical solution
temperature (LCST). Above this temperature, PNVCL undergoes a
sharp coil-to-globule transition to form inter- and intra-chain asso-
ciation resulting in loss of solubility and hydrophobic aggregate
[12]. It has been demonstrated that this swelling/deswelling behav-
ior in response to external temperature changes is favorable for
protein adsorption, drug release and biomolecular conjugates.
Having the double response characteristics of magnetism and
temperature, thermosensitive magnetic composite particles play
an important role in biomedicine and bioseparation. Several stud-
ies have been reported about protein extraction and protein
concentration using the magnetic thermosensitive adsorbents.
Elaissari and Bourrel [13] utilized thermosensitive core–shell
magnetic latex with a magnetic polystyrene core and poly (N-
isopropylacrylamide) shell to adsorb the human serum albumin.
Shamim et al. [14] employed poly (N-isopropylacrylamide) coated
nanomagnetic particles as adsorption material for the separation
of bio-molecules. These processes involved the variation of physic-
ochemical parameters such as pH, ionic strength, temperature,
extract time, and initial protein concentration. However few stud-
ies have applied systematically statistical methods to investigate
the combination of parameters that provides optimum protein sep-
aration. Response surface methodology (RSM) is proven to be an
effective means for the above-mentioned purpose. It is a collection
dx.doi.org/10.1016/j.colsurfb.2011.02.002
http://www.sciencedirect.com/science/journal/09277765
http://www.elsevier.com/locate/colsurfb
mailto:niehuali@dhu.edu.cn
mailto:lzhu@dhu.edu.cn
dx.doi.org/10.1016/j.colsurfb.2011.02.002
478 M.-M. Song et al. / Colloids and Surfaces B: Biointerfaces 84 (2011) 477–483
Fe2+, Fe3+ γ-Fe2O3
NaOH
Si OCH2CH3
OCH2CH3
OCH2CH3
CH3CH2O
γ-Fe2O3
Si
Si
Si
OH
OH
OH
Si
Cl
Cl
Cl
NO
CH CH2
γ-Fe2O3
Si
Si
Si
O
SiO
CH2CH
O
NO
CHCH2 *
Si SiSi
O
Si
O
O
* n
γ-Fe2O3
CH2CH
hesis
o
c
[
t
a
T
c
(
(
s
(
i
u
a
d
o
f
2
2
t
b
w
6
i
f
c
a
2
m
o
d
w
a
n
Fig. 1. Schematic illustration of the synt
f statistical and mathematical techniques which has been suc-
essfully used for developing, improving, and optimizing processes
15]. In this methodology, multivariate experiments are designed
o reduce the number of assays necessary to optimize the process,
s well as to collect data more precisely than univariate strategies.
herefore, it is less time-consuming and more effective than other
onventional methods [16].
In this paper, we prepared a smart material—thermoresponsive
PNVCL) polymer-conjugated magnetic silica composite particles
PNVCL–SiO2–MNPs) and the obtained multifunctional micro-
pheres were used as a carrier for adsorbing bovine serum albumin
BSA), which is chosen as a target protein. The objective of this
nvestigation is to develop an approach that will provide a better
nderstanding of the combined effects of the key processing vari-
bles (temperature, initial pH and initial BSA concentration) on the
esired response (adsorption capacity), as well as to determine the
ptimum conditions for the next phase of purifying BSA directly
rom the crude bovine serum.
. Materials and methods
.1. Materials
Bovine serum albumin (BSA), N-vinylcaprolactam, N,N,N,N-
etramethyl ethylenediamine (TEMED), N,N-methylene
is-acrylamide (MBA) and potassium persulfate (99%) (KPS)
ere purchased from Sigma Chem. Corp., China. Ferric chloride
-hydrate, ferrous chloride, sodium hydroxide, tetraethylorthosil-
cate (TEOS, 98%) and vinytriethoxysilane (VTES) were obtained
rom Sinopharm Chemical Reagent Co., Ltd., Shanghai. All other
hemicals were the analytic grade reagents commercially available,
nd used without further purification.
.2. Preparation of maghemite nanoparticles
A maghemite nanoparticles dispersion was prepared using the
ethod already described [17]. This involves the co-precipitation
f FeCl2 and FeCl3 by adding a concentrated solution of 5 M NaOH
ropwise into the mixture of iron salts. The molar ratio of Fe3+/Fe2+
as fixedto 2:1. The alkaline solution was stirred for 15 min at 40 ◦C
nd then heated at 80 ◦C for 30 min. The obtained ultrafine magnetic
anoparticles were washed three times with doubly distilled water
of magnetic PNVCL composite particles.
and collected via an external magnetic field, and finally they were
redispersed in water for further use.
2.3. Synthesis of silica coated maghemite nanoparticles
A silica coating of maghemite nanoparticles was obtained using
alkaline hydrolysis of TEOS. First a suspension of the previously syn-
thesized magnetic nanoparticle colloid (5 mL, 80 wt.%) was diluted
with 80 mL ethanol and 20 mL water. Then, under continuous
mechanical stirring, 2 mL ammonia solution (30 wt.%) was added
and the pH of the reaction solution was adjusted to 9.0. The pre-
cursor of tetraethylorthosilicate (TEOS, 1.0 mL) was consecutively
dropped into the reaction mixture. The reaction was kept at 40 ◦C
for 24 h.
2.4. Modification of silica coated maghemite nanoparticles
The surface of silica-coated magnetic nanoparticles was
endowed with reactive C C bonds through modification with
vinyltriethoxysiliane (VTES) using the method reported by
Bourgeat-Lami and Lang [18]. This procedure was achieved by
adding VTES in excess amount to the reaction mixture and stirring
them for 12 h at 40 ◦C. The silica-coated magnetic nanoparticles
grafted with VTES were separated with a magnet and washed
repeatedly with ethanol and water to remove superfluous VTES.
Then water was added to give a colloid suspension (20 wt.%).
2.5. Synthesis of thermoresponsive magnetic composite particles
The Fe2O3/SiO2/PNVCL composite microspheres were prepared
using precipitation polymerization. Typically, 5 mL VTES-modified
Fe2O3/SiO2 dispersion (20 wt.%) was mixed with 10 mL aque-
ous NVCL (monomer, 5 wt.%), 10 mL aqueous MBA (cross-linker,
0.5 wt.%), 100 �L TEMED which was used as an accelera-
tor in the reaction and 40 mL doubly distilled water by
mechanical stirring. After degassing under nitrogen for 30 min,
the solution was heated to 70 ◦C, and 10 mL aqueous APS
(initiator, 0.2 wt.%) was injected to initiate the polymeriza-
tion. 5 h later, the final products were enriched by magnetic
separation, washed with doubly distilled water, and finally
stored in deionized water before characterization and appli-
cation. The binding procedures and schematic illustration of
faces B: Biointerfaces 84 (2011) 477–483 479
P
F
2
t
(
t
F
m
m
F
M
S
m
c
fi
n
s
o
1
2
m
(
B
p
i
t
q
w
b
i
i
(
2
f
m
b
o
T
Table 1
Experimental design of the adsorption of BSA on magnetic PNVCL composite
particles.
Symbols Factors Coded levels
−2 −1 0 +1 +2
M.-M. Song et al. / Colloids and Sur
NVCL with maghemite nanoparticles are shown in detail in
ig. 1.
.6. Characterization
The morphology and size of the magnetic PNVCL composite par-
icles were determined by scanning electron microscopy (SEM)
JEOL, JSM-5600LV, Japan). Prior to scanning samples were sput-
er coated for 90 s with gold using a JEOL JFC-1200 fine coater.
ourier transform infrared spectroscopy (FT-IR) spectra of the
aghemite nanoparticles, silica-coated maghemite nanoparticles,
agnetic PNVCL composite particles and NVCL were obtained by
T-IR spectrophotometer (NEXUS-670, Thermo Nicolet Corp., USA).
agnetization measurements were preformed with a Vibrating-
ampling Magnetometer (VSM) (Princeton Applied Research,
odel-155) at room temperature. Hysteresis measurements were
onducted at 300 K on freeze dried samples in an applied magnetic
eld up to 1 T. The volume–phase transition behavior of the mag-
etic PNVCL composite particles was investigated by differential
canning calorimetry (DSC). The calorimetric analysis was carried
ut on a PYRIS Diamond DSC. The samples were scanned from 0 to
00 ◦C at a scan rate of 5 ◦C/min.
.7. Batch adsorption studies
Batch adsorption experiments were performed in 50 mL Erlen-
eyer flasks containing 10 mL adsorption solution and 50 mg
60 wt.%) wet magnetic PNVCL composite particles. Adsorption of
SA from aqueous solution to the adsorbents was studied at various
H values in a 0.05 M Na2HPO4–NaH2PO4 buffer. Adsorption exper-
ments were conducted in a shaker at 100 rpm for 3 h at different
emperatures (ranging from 20 ◦C to 40 ◦C).
The amount of adsorbed BSA was calculated using Eq. (1):
= (Ci − Cf )V
m
(1)
here q is the amount of BSA adsorbed onto unit mass of the adsor-
ents (mg/g); Ci and Cf are the concentrations of the BSA in the
nitial and final solutions after adsorption, respectively (mg/mL); V
s the volume of BSA solution (mL); m is the mass of the adsorbents
g).
.8. Experimental design
The adsorption conditions were optimized using response sur-
ace methodology (RSM). RSM is essentially a particular set of
athematical and statistical methods for designing experiments,
uilding models, evaluating the effects of variables, and searching
ptimum conditions of variables to predict targeted responses [19].
he main advantage of this methodology is an effective reduction
Fig. 2. SEM images of PNVCL–SiO2–MNPs with diffe
A Temperature (◦C) 20 25 30 35 40
B Initial pH 4 5 6 7 8
C Initial BSA concentration (mg/mL) 0.4 0.8 1.2 1.6 2.0
in experimentation and a quantification of factors as independent
variables that lead to optimum conditions [20].
To assess the effect of variables on the response in the region
of investigation, a central composite design (CCD) is chosen for
RSM to design the experiment. This model is well suited to fitting a
quadratic surface and usually works well for process optimization.
The CCD system is an effective design that is ideal for sequential
experimentation and allows a reasonable amount of information
for testing lack of fit without needing a large number of design
points [21].
Therefore, Box–Benhnken central composite design with three
factors was applied using Design-Expert 7.1.6 without any blocking.
The main factors considered to influence adsorption are tempera-
ture, initial pH and initial BSA concentration. The levels and ranges
chosen for the factors are shown in Table 1.
2.9. Desorption studies
Desorption was studied as a function of temperature and
was carried out under alkaline conditions using a 0.05 M
Na2HPO4–NaH2PO4 buffer (pH 8.0) and 1 M NaCl. After equilib-
rium was achieved for adsorption in the Na2HPO4–NaH2PO4 buffer
(10 mL, 0.05 M, pH 5.0, containing 1.5 mg/mL of BSA) at different
temperatures, the supernatant was decanted from the latex par-
ticles using a magnet and the particles were washed with water
to remove the unadsorbed protein. BSA-saturated-magnetic PNVCL
composite particles were then mixed with 10 mL of elution solution
at 20 ◦C which is far below the LCST (31 ◦C) of PNVCL. 3 h later, the
supernatant was withdrawn and analyzed in a UV spectrometer.
3. Results and discussion
3.1. Characterization of thermosensitive magnetic composite
particles
The size distribution and morphology of magnetic PNVCL com-
posite particles were observed by SEM microscopy. The typical SEM
micrographs for the PNVCL coated magnetic composite particles
are shown in Fig. 2(a) and (b). It exhibited that the chemically modi-
fied magnetic particles were multi-dispersed. The size of the PNVCL
rent magnifications (a) 5.0k× and (b) 10.0k×.
Rubênia Silveira
Highlight
Rubênia Silveira
Highlight
Rubênia Silveira
Highlight
Rubênia Silveira
Highlight
480 M.-M. Song et al. / Colloids and Surfaces B: Biointerfaces 84 (2011) 477–483
F
N
e
t
d
c
a
f
s
c
o
i
s
c
a
c
a
a
c
t
S
i
s
t
(
g
w
u
m
z
c
n
i
m
u
P
r
a
t
F
i
m
L
y = 37.02 + 13.42x1 − 4.23x2 + 10.99x3 − 2.79x1x2 + 7.32x1x3
− 0.88x2x3 + 0.65x21 − 3.79x22 − 0.28x23 (2)
where y is predicted response, xi is variables.
ig. 3. The FT-IR spectra: (a) MNPs; (b) SiO2–MNPs; (c) PNVCL–SiO2–MNPs; (d)
VCL.
ncapsulated magnetic composite particles distributed from 2 �m
o 8 �m, with an average diameter of about 5 �m (Fig. 2(a)). Fig. 2(b)
emonstrated that the prepared magnetic PNVCL composite parti-
le was in a standard spherical form which provided alarge surface
rea.
FT-IR technique was applied to ensure the success of the sur-
ace modification. Fig. 3 shows the FT-IR spectra of �-Fe2O3 (a),
ilica coated magnetic particles (b), PNVCL modified magnetic silica
omposite particles (c) and NVCL (d), respectively. By comparison
f the FT-IR patterns (Fig. 3a and b), the presence of the silica coat-
ng changed the FT-IR pattern of �-Fe2O3 (Fig. 3a) significantly. As
een in Fig. 3b, the presence of the siloxane groups Si–O–Si was
onfirmed by the band at 1089 cm−1. Besides, the adsorption band
t 811 cm−1 indicates the existence of silanol groups (Si–OH). The
haracteristic adsorption peak of magnetic fluid also appears at
round 586 cm−1. This peak demonstrates the presence of �-Fe2O3
s a result of the successful coating procedure. Moreover, when
omparing the two FT-IR spectra (Fig. 3b and c), it is apparent that
he chemical modification significantly altered the FT-IR pattern of
iO2-coated MNPs. In the spectrum of NVCL (Fig. 3d), the character-
stic peak of NVCL is located at 1487 cm−1 corresponding to the C–N
tretching vibration [22], which is also observed in Fig. 3c, while
he adsorption peaks 3109 cm−1 (CH and CH2 ) and 1660 cm−1
C C) disappeared in Fig. 3c. These results provide evidences of the
rafting of PNVCL onto SiO2 coated maghemite nanoparticles.
The magnetic properties of MNPs and PNVCL–SiO2–MNPs
ere examined by VSM (Fig. 4). The magnetization curve of
ntreated �-Fe2O3 revealed a typical saturation magnetization of
aghemite nanoparticles of 13.61 emu/g whereas the magneti-
ation of PNVCL–SiO2–MNPs was found to be 9.87 emu/g. This
ould be attributed to the surface modification of maghemite
anoparticles. Moreover, there are no remanence and coerciv-
ty observed, which suggests that PNVCL–SiO2–MNPs acting as
agnetic adsorbents are superparamagnetic. Thus the large sat-
ration magnetization and superparamagnetic properties make
NVCL–SiO2–MNPs very susceptible to the external magnetic field,
aising the possibility of reuse without aggregation after removing
n external magnetic field.
DSC measurement was used to determine the thermosensi-
ive properties of the synthesized PNVCL–SiO2–MNPs. As shown in
ig. 5, a clear endothermic point could be seen in the curve and this
s due to the phase transition of PNVCL and the maximum endother-
ic peaks of the sample at 33.4 ◦C, which is little higher than the
CST of pure PNVCL (31 ◦C) [23]. This phenomenon may be ascribed
Fig. 4. VSM curves of the MNPs and PNVCL–SiO2–MNPs.
to the chemical bonding between PNVCL and silica suppressing the
thermodynamic behavior of PNVCL [24].
3.2. Development of regression model equation
Central composite design was used to develop a correlation
between the adsorption conditions and adsorption capacity of
thermosensitive magnetic composite particles. Runs 15–20 at the
center point were used to determine the experimental error.
According to the sequential model sum of squares, the models were
selected based on the highest order polynomials where the addi-
tional terms were significant and the models were appropriate.
To analyze the adsorption capacity of magnetic PNVCL compos-
ite particles, the quadratic model was selected as suggested by the
software. According to the experimental design, the result was ana-
lyzed and approximating functions of adsorption was obtained in
Eq. (2).
Fig. 5. DSC scan plot of magnetic PNVCL composite particles.
M.-M. Song et al. / Colloids and Surfaces B: Biointerfaces 84 (2011) 477–483 481
Table 2
ANOVA results of response surface quadratic model according to adsorption capacity
of the magnetic PNVCL composite particle for BSA.
Source Sum of squares Degree
of free-
dom
Mean square F-value Prob > F
Model 6008.47 9 667.61 86.28 <0.0001
A 2882.62 1 2882.62 372.54 <0.0001
B 285.78 1 285.78 36.93 0.0002
C 1932.04 1 1932.04 249.69 <0.0001
AB 62.38 1 62.38 8.06 0.0194
AC 428.66 1 428.66 55.40 <0.0001
B2 339.44 1 339.44 43.87 <0.0001
Residual 69.64 9 7.74
R
w
3
t
a
i
a
s
m
a
c
i
a
t
a
B
m
r
t
r
p
d
r
d
F
[27]. At isoelectric points electrostatic repulsion between adsorbed
Lack of fit 69.35 5 13.87 194.62 <0.0001
Pure error 0.29 4 0.071
2 = 0.9885; R2
adj
= 0.9771; adequate precision = 29.96.
A positive sign in front of the terms indicates synergistic effect,
hereas a negative sign indicates antagonistic effect.
.3. Analysis of variance (ANOVA)
The adequacy of the response surface quadratic model was fur-
her justified through analysis of variance (ANOVA) and the results
re presented in Table 2. As noted the model F-value of 86.28
mplies that the model is significant. Moreover, values of “prob > F”
re less than 0.0500 suggest that model terms are significant. In this
tudy, A, B, C, B2, AB and AC are significant model terms. Insignificant
odel terms, which have limited influence, such as A2, C2 and BC
re excluded from the study to improve the model. Adequate pre-
ision measures the signal to noise ratio and a ratio greater than 4
s desirable. In this case, a ratio of 29.96 was achieved indicating an
dequate signal, thus the model is regarded as being fit to navigate
he design space.
The normal probability and studentized residuals plot for the
dsorption capacity of magnetic PNVCL composite particles for
SA are shown (Fig. 6). Residual values demonstrated that the
odel satisfied the assumptions of ANOVA where the studentized
esiduals measured the number of standard deviations separating
he actual and predicted values. It also demonstrated that neither
esponse transformation was needed nor there was an apparent
roblem with normality [25]. Fig. 7 shows the actual and the pre-
icted adsorption capacity plots. Actual values were the measured
esponse data for a particular run and the predicted values were
etermined by approximating functions employed for the models.
ig. 6. The studentized residuals and normal % probability plot for BSA adsorption.
Fig. 7. The actual and predicted plots for BSA adsorption.
The determination coefficient (R2) was evaluated as 0.9885, and
the value of adjusted R-square (R2
adj
) was 0.9771. This demonstrates
that there is a good degree of correlation between actual and the
predicted data [26].
3.4. The optimization of adsorption conditions
To investigate the effects of the three factors on the adsorp-
tion of BSA, the response surface methodology was used and
three-dimensional plots were applied. Based on the ANOVA results
obtained, temperature and initial BSA concentration were found to
have significant effects on the adsorption of the protein. Initial pH
on the other hand imposed the least effect on the response. Fig. 8
shows a 3-dimensional representation of the combined effects of
temperature and pH on adsorption capacity for BSA at the fixed
initial BSA concentration of 1.2 mg/mL. An optimum point of max-
imum adsorption capacity was obtained as 72.74 mg/g at 40 ◦C,
pH 4.68, which is close to the isoelectric point of BSA (pI = 4.7)
molecules is at a minimum and molecules have a higher structural
stability, so there is a smaller tendency to spread at the interface.
This results in maximum adsorption at the isoelectric point [28].
Fig. 8. The combined effects of temperature and initial pH on BSA adsorption (pro-
tein concentration was fixed at 1.2 mg/mL).
Rubênia Silveira
Highlight
482 M.-M. Song et al. / Colloids and Surfaces B
F
a
s
(
t
o
t
c
c
a
h
a
r
e
c
s
a
a
t
a
t
r
ig. 9. The combined effects of initial BSA concentration and temperature on BSA
dsorption (pH was fixed at 4.68).
The three-dimensional response surfaces which were con-
tructed to show the effects of two most important variables
temperature and initial BSA concentration) on the adsorption of
he protein at pH 4.68 are given in Fig. 9. The adsorption capacity
f BSA increased from 72.74 to 125.09 mg/g when BSA concentra-
ion increased from 1.2 mg/mL to 2.0 mg/mL at 40 ◦C. The uptake
apacity increased with the rise in temperature and initial BSA
oncentration. A higher ratio of the surface binding site on the
dsorbentto the BSA could be obtained at higher temperature and
igher initial BSA concentration. Similar observations have been
lso revealed in studies on cadmium biosorption using Saccha-
omyces cerevisiae [29] and in the case of carotene recovery from
xtracted oil of POME by adsorption chromatography [30]. In this
ase the main factor explaining this characteristic is that adsorption
ites remain unsaturated during the adsorption reaction. Addition-
lly, the hydrophobic interaction between the adsorbent’s surface
nd BSA becomes more intense at a higher temperature due to
he thermosensitive property of the adsorbent. However increasing
dsorption temperature beyond 40 ◦C would damage the struc-
ures and lower the activity of protease and thus reduce the protein
ecovery.
Fig. 10. Effect of temperature on adsorption and desorption. [
: Biointerfaces 84 (2011) 477–483
3.5. Desorption
The desorption reaction performed at 20 ◦C which is conduc-
tive to the rapid dissociation of protein from the thermosensitive
magnetic composite microspheres. Fig. 10 shows the effect of tem-
perature on adsorption and desorption. It is observed that more
than 85% of protein could be desorbed from the latex particles
when the adsorption was done near the LCST (30 ◦C) of modified
PNVCL. On the other hand, about 74% of BSA was desorbed from the
latex particles when adsorption process happened at 40 ◦C which
is much higher than the LCST of modified PNVCL. Although the
amount of protein adsorbed increased with greater temperature,
a smaller amount of protein was desorbed. This is owing to the
easy deformation of protein molecules at high temperature through
the interaction between polymer chains and the molecules, which
results in the mechanical entrapment of BSA by PNVCL chains dur-
ing the shell shrinkage.
4. Conclusions
In this study, a novel thermosensitive magnetic composite
microsphere with a �-Fe2O3/SiO2 core and thermosensitive PNVCL
shell was successfully prepared for the adsorption of BSA. The influ-
ences of operating parameters such as temperature, initial pH, and
initial BSA concentration for the adsorption of BSA on the ther-
mosensitive magnetic composite particles were evaluated using
RSM. The relationship between the response and the independent
variables was examined via the quadratic approximating func-
tions of adsorption capacity at the end of 180 min adsorption. The
RSM analysis demonstrates that optimum adsorption conditions
were pH 4.68, 40 ◦C, and initial BSA concentration of 2.0 mg/mL. At
optimum adsorption conditions, the predicted uptake capacity of
the thermosensitive magnetic composite particle for BSA reached
125.09 mg/g. Desorption experiments show that temperature plays
a significant role in the separation process. In conclusion the out-
comes from this investigation support the view that the magnetic
PNVCL composite particle is an effective adsorbent providing an
alternative strategy for the separation of protein with possible com-
mercial and industrial applications.
Acknowledgements
This work was supported by the National Natural Science
Foundation of China (21006010) and the Fundamental Research
Funds for the Central Universities. This work was also supported
in part by UK–CHINA Joint Laboratory for Therapeutic Textiles
based in Donghua University Biomedical Textile Materials “111
Project” Ministry of Education of P.R. China (No. B07024) and the
Research Fund for the Doctoral Program of Higher Education (No.
2009007512001).
References
[1] T. Kappler, M. Cerff, K. Ottow, T. Hobley, C. Posten, Biotechnol. Bioeng. 102
(2009) 535.
[2] T.E. Kappler, B. Hickstein, U.A. Peuker, C. Posten, J. Biosci. Bioeng. 105 (2008)
579.
[3] J.E. Wong, A.K. Gaharwar, D. Muller-Schulte, D. Bahadur, W. Richtering, J. Colloid
Interface Sci. 324 (2008) 47.
[4] S. Chen, Y. Li, C. Guo, J. Wang, J.H. Ma, X.F. Liang, L.R. Yang, H.Z. Liu, Langmuir
23 (2007) 12669.
[5] F.Q. Wang, P. Li, J.P. Zhang, A.Q. Wang, Q. Wei, Drug Dev. Ind. Pharm. 36 (2010)
867.
[6] K.S. Sivudu, K.Y. Rhee, Colloids Surf., A 349 (2009) 29.
[7] H.Y. Tsai, J.R. Chan, Y.C. Li, F.C. Cheng, C.B. Fuh, Biosens. Bioelectron. 25 (2010)
2701.
[8] W.F. Lee, W.F. Yuan, J. Appl. Polym. Sci. 84 (2002) 2523.
[9] C.K. Cheee, S. Rimmer, I. Soutar, L. Swanson, React. Funct. Polym. 66 (2006) 1.
10] A. Imaz, J. Forcada, J. Polym. Sci., Part A: Polym. Chem. 48 (2010) 1173.
Rubênia Silveira
Realce
faces B
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
M.-M. Song et al. / Colloids and Sur
11] S. Shah, A. Pal, R. Gude, S. Devi, Eur. Polym. J. 46 (2010) 958.
12] A. Imaz, J. Forcada, Eur. Polym. J. 45 (2009) 3164.
13] A. Elaissari, V. Bourrel, J. Magn. Magn. Mater. 225 (2001) 151.
14] N. Shamim, L. Hong, K. Hidajat, M.S. Uddin, J. Colloid Interface Sci. 304 (2006)
1.
15] S.J. Wang, F. Chen, J.H. Wu, Z.F. Wang, X.J. Liao, X.S. Hu, J. Food Eng. 78 (2007)
693.
16] M.A.M. Mune, S.R. Minka, I.L. Mbome, Food Chem. 110 (2008) 735.
17] Y.S. Kang, S. Risbud, J.F. Rabolt, P. Stroeve, Chem. Mater. 10 (1998) 1733.
18] E. Bourgeat-Lami, J. Lang, J. Colloid Interface Sci. 197 (1998) 293.
19] P.N. Sampaio, C.R.C. Calado, L. Sousa, D.C. Bressler, M.S. Paris,.L.P. Fonseca, Eur.
Food Res. Technol. 231 (2010) 339.
20] T. Sahan, H. Ceylan, N. Sahiner, N. Aktas, Bioresour. Technol. 101 (2010) 4520.
[
[
[
[
: Biointerfaces 84 (2011) 477–483 483
21] S. Sharma, A. Malik, S. Satya, J. Hazard. Mater. 164 (2009) 1198.
22] M. Prabaharan, J.J. Grailer, D.A. Steeber, S.Q. Gong, Macromol. Biosci. 8 (2008)
843.
23] R.X. Liu, M. Fraylich, B.R. Saunders, Colloid Polym. Sci. 287 (2009) 627.
24] T. Serizawa, K. Wakita, M. Akashi, Macromolecules 35 (2002) 10.
25] A. Özer, G. Gürbüz, A. C¸alimli, B.K. Körbahti, J. Hazard. Mater. 152 (2008) 778.
26] A.L. Ahmad, C.J.C. Derek, M.M.D. Zulkali, Sep. Purif. Technol. 62 (2008) 702.
27] K.Y. Chun, P. Stroeve, Langmuir 18 (2002) 4653.
28] E.B. Altintas, A. Denizli, J. Appl. Polym. Sci. 103 (2007) 975.
29] F. Ghorbani, H. Younesi, S.M. Ghasempouri, A.A. Zinatizadeh, M. Amini, A.
Daneshi, Chem. Eng. J. 145 (2008) 267.
30] A.L. Ahmad, C.Y. Chan, S.R. Abd Shukor, M.D. Mashitah, Sep. Purif. Technol. 73
(2010) 279.
	Optimization of adsorption conditions of BSA on thermosensitive magnetic composite particles using response surface method...
	Introduction
	Materials and methods
	Materials
	Preparation of maghemite nanoparticles
	Synthesis of silica coated maghemite nanoparticles
	Modification of silica coated maghemite nanoparticles
	Synthesis of thermoresponsive magnetic composite particles
	Characterization
	Batch adsorption studies
	Experimental design
	Desorption studies
	Results and discussion
	Characterization of thermosensitive magnetic composite particles
	Development of regression model equation
	Analysis of variance (ANOVA)
	The optimization of adsorption conditions
	Desorption
	Conclusions
	Acknowledgements
	References