Light-stimulated cargo release from a core-shell structured nanocomposite for site-specific delivery

Article abstract of DOI:10.1016/j.jssc.2015.02.019

This paper reported a core-shell structured site-specific delivery system with a light switch triggered by low energy light (λ=510 nm). Its core was composed of supermagnetic Fe₃O₄ nanoparticles for magnetic guiding and targeting. Its outer shell consisted of mesoporous silica molecular sieve MCM-41 which offered highly ordered hexagonal tunnels for cargo capacity. A light switch N1-(4aH-cyclopenta[1,2-b:5,4-b′]dipyridin-5(5aH)-ylidene)benzene-1,4-diamine (CBD) was covalently grafted into these hexagonal tunnels, serving as light stimuli acceptor with loading content of 1.1 μM/g. This composite was fully characterized and confirmed by SEM, TEM, XRD patterns, N₂ adsorption/desorption, thermogravimetric analysis, IR, UV-vis absorption and emission spectra. Experimental data suggested that this composite had a core as wide as 150 nm and could be magnetically guided to specific sites. Its hexagonal tunnels were as long as 180 nm. Upon light stimuli of on and off states, controllable release was observed with short release time of ~900 s (90% capacity).

Full text of DOI:10.1016/j.jssc.2015.02.019

Journal of Solid State Chemistry 226 (2015) 179–185
Contents lists available at ScienceDirect
Journal of Solid State Chemistry
journal homepage: www.elsevier.com/locate/jssc
Light-stimulated cargo release from a core–shell structured
nanocomposite for site-specific delivery
a
a
a
b
a,
n
Yun Cai , Li Ling , Xiaofang Li , Meng Chen , Likai Su
a
Department of Neurology, Affiliated Hospital of Hebei University, Baoding 071000, PR China
Department of Rheumatology, Affiliated Hospital of Hebei University, Baoding 071000, PR China
b
a r t i c l e i n f o
a b s t r a c t
Article history:
This paper reported a core–shell structured site-specific delivery system with a light switch triggered by
Received 14 November 2014
Received in revised form
low energy light (λ¼510 nm). Its core was composed of supermagnetic Fe O nanoparticles for magnetic
3
4
guiding and targeting. Its outer shell consisted of mesoporous silica molecular sieve MCM-41 which
offered highly ordered hexagonal tunnels for cargo capacity. A light switch N1-(4aH-cyclopenta[1,2-
30 January 2015
Accepted 24 February 2015
Available online 5 March 2015
0
b:5,4-b ]dipyridin-5(5aH)-ylidene)benzene-1,4-diamine (CBD) was covalently grafted into these hex-
agonal tunnels, serving as light stimuli acceptor with loading content of 1.1
μ
M/g. This composite was
Keywords:
fully characterized and confirmed by SEM, TEM, XRD patterns, N adsorption/desorption, thermogravi-
2
Site-specific delivery
Core–shell structure
Magnetic feature
Light switch
metric analysis, IR, UV–vis absorption and emission spectra. Experimental data suggested that this
composite had a core as wide as 150 nm and could be magnetically guided to specific sites. Its hexagonal
tunnels were as long as 180 nm. Upon light stimuli of “on” and “off” states, controllable release was
observed with short release time of ꢀ900 s (90% capacity).
&
2015 Elsevier Inc. All rights reserved.
1
. Introduction
Core–shell structured nanocomposites are then proposed to
meet above requirements since they can combine advantages
of each component without compromising individual character
[4–7]. A representative core–shell structured site-specific delivery
system usually has a magnetic core which is responsible for
magnetic guiding [8,9]. Its outer shell is usually composed of
mesoporous or amorphous silicates which are here used to
support cargo molecules and other functional components.
Aiming at controllable and active release, stimulus-sensitive
switches should be embedded into such core–shell structured
nanocomposites, such as light, pH, competitive binding, enzymes
and chemical reactions [10–12]. Of all these candidates, light
switch has been considered promising owing to its virtue of
noninvasive therapy, which makes it free of biological interfer-
ences, giving instant and controllable release. Some precursive
efforts have tried azobenzene and its derivatives as light switch
owing to their trans/cis iso-merization and high compatibility
with supporting hosts [13,14]. However, these light switches need
high energy light peaking at 450 nm which causes server light
damage to live body is nearly impenetrable for live tissues and is
nearly impenetrable for live tissues. Thus, core–shell structured
site-specific delivery system with light switch triggered by low
energy light should be explored and developed.
For their potential application in drug transportation field, site-
specific delivery systems have been insensitively explored and
reported [1,2]. An ideal delivery system is supposed to release
therapeutic drug level at specific tissues, organs or even cellular
structures. Before reaching target organs or tissues, no cargo
molecules should be released since some of them are generally
toxic and harmful to normal organs as well as to abnormal ones. In
addition, a perfect delivery system should be highly stable to
experience various physiological conditions, holding cargo mole-
cules tightly and safely. So far, some candidates have been
proposed. However, few of them can meet all above requirements
well [3,4]. For example, biodegradable polymers have been
reported as a delayed release supporting host for drug molecules.
In this case, their release process is controlled by their hydrolysis-
induced erosion [3]. Their drawback localizes at the fact that such
release process happens immediately upon dispersion in water,
which makes them unsuitable for toxic drug delivery such as
antitumor medicament. What’s more, it is nearly impossible to
request a site-specific release from them. Thus, it is needed to
explore and develop smart delivery systems with both controllable
release and site-specific character.
Guided by above consideration, in this effort, we decide to
construct a core–shell structured site-specific delivery system with
light switch triggered by low energy light. To achieve this goal,
magnetic nanoparticles were used as core, mesoporous silicate
n
Corresponding author. Tel.: þ86 312 5981036; fax: þ86 312 5981036.
E-mail address: zhangdong19992003@163.com (L. Su).
http://dx.doi.org/10.1016/j.jssc.2015.02.019
0
022-4596/& 2015 Elsevier Inc. All rights reserved.

180
Y. Cai et al. / Journal of Solid State Chemistry 226 (2015) 179–185
3 4
Scheme 1. Synthetic route for CDU and construction strategy for CDU-MCM-41@Fe O .
MCM-41 was grew onto this core, serving as supporting host. As
shown in Scheme 1, a light switch derived from 1,10-phenanthro-
line was grafted into MCM-41 shell so that it could be triggered by
low energy light. The resulting nanocomposite was characterized
in detail, along with its controllable release character.
were obtained from Aldrich Chemicals Co. and used as received with
no purifications. Other chemicals, including NH O, HCl, NH AC,
N,N-dimethylformamide (DMF), cetyltrimethylammonium bromide
3
ꢁ H
2
4
(denoted as CTAB, AR grade), tetraethoxysilane (denoted as TEOS),
FeCl
3
ꢁ 6H
2
O, anhydrous sodium acetate, ethylene glycol, p-toluene
sulfuric acid, n-hexane, chloroform, toluene and ethanol, were
bought from Tianjin Chemical Company. All organic solvents were
purified through standard procedures. Solvent water used in this
work was deionized.
2
. Experimental details
2
.1. General information
Equipments used for sample characterization are listed as follows.
NMR, mass and IR spectra were obtained from a Varian INOVA 300
spectrometer, a Agilent 1100MS series/AXIMA CFR MALDI/TOF
Construction route and operating principle of our nanocomposite
(
denoted as CDU-MCM-41@Fe
3
O
4
) are shown in Scheme 1. Starting
(matrix assisted laser desorption ionization/time-of-flight) MS (COM-
ꢂ1
compounds, 1,10-phenanthroline, benzene-1,4-diamine, 3-(triethox-
ysilyl)propyl isocyanate (denoted as TESPIC) and rhodamine 6 G,
PACT) and a Bruker Vertex 70 FTIR spectrometer (400–4000 cm
,
KBr pellet), respectively. Elemental analysis was finished on a Carlo

Y. Cai et al. / Journal of Solid State Chemistry 226 (2015) 179–185
181
Erba 1106 elemental analyzer. Photophysical parameters were mea-
sured from a Hitachi F-4500 fluorescence spectrophotometer and a
HP 8453 UV–vis–NIR diode array spectrophotometer, respectively. X-
ray diffraction (XRD) data were collected by a Rigaku D/Max-Ra X-ray
30 min under N
dropwise added. The mixture was mechanically stirred at room
temperature for 10 h under N protection. Finally, crude product
was collected, dispersed in mixed solution of HCl (5 mL) and
ethanol (100 mL) and stirred for 24 h to remove template reagent
CTAB. The resulting sample was dried in vacuum at 50 1C to give
2
protection. During this time, TEOS (0.8 mL) was
2
diffractometer (
λ¼1.5418 Ǻ). Sample morphology was measured by
scanning electron microscopy (SEM), transmission electron micro-
scopy (TEM) and energy-dispersive analysis of X-ray (EDAX) on a
Hitachi S-4800 microscope and a JEM-2010 transmission electron
microscope (Japanese JEOL Company). Thermogravimetric analysis
3 4
MCM-41@Fe O .
2
.4. Construction of CDU-MCM-41@Fe
3 4
O
(
TGA) data were collected through a Perkin-Elmer thermal analyzer.
Magnetic feature was explored on a vibrating sample magnetometer
VSM) (Lake Shore Company) at room temperature. Mesoporous
structure was confirmed by N adsorption and desorption measure-
The final core–shell structured site-specific delivery system
(
CDU-MCM-41@Fe was constructed by the following procedure
3
O
4
2
[
16]. MCM-41@Fe O (0.2 g) and CDU (0.2 g) were dispersed in
3 4
ment with a Nova l000 analyzer at 77 K. Pore size and volume were
calculated by Barrett–Joyner–Halenda (BJH) model. Surface area was
determined with Brunauer–Emmett–Teller (BET) model. Above ope-
rations were finished in the air at room temperature without
specifications.
2
anhydrous toluene (20 mL) and heated to reflux under N protec-
tion for 12 h. After natural cooling, solid product was collected and
washed with anhydrous toluene and ethanol to give CDU-MCM-
3 4
41@Fe O .
2.2. Synthesis of CBD and CDU
2.5. Dye loading and release monitoring
0
N1-(4aH-cyclopenta[1,2-b:5,4-b ]dipyridin-5(5aH)-ylidene)benzene-
3 4
Rhodamine 6 G was first loaded into CDU-MCM-41@Fe O by
0
1
,4-diamine (CBD) was synthesized with 4aH-cyclopenta[1,2-b:5,4-b ]
3 4
the following procedure. CDU-MCM-41@Fe O (0.2 g) was soaked
dipyridin-5(5aH)-one (CDO) and benzene-1,4-diamine as starting com-
pounds. CDO was first prepared following a literature procedure [15].
CDO (10 mmol), benzene-1,4-diamine (12 mmol) and p-toluene sulfuric
acid (0.1 g) were dissolved in distilled toluene (20 mL) and heated to
in rhodamine 6 G solution (ethanol, 5 M/L) and stirred for 30 h
μ
under 510 nm radiation to achieve dye loading equilibrium. Then
this rhodamine 6 G-saturated nanocomposite was collected and
washed with ethanol until solution became clear. Dye loading
reflux under N
and recrystallized in MeOH/toluene to give pure CBD as red powder. H
NMR (300 MHz, CDCl ): 8.79 (s, 1H, pyridine Ar–H), 8.65 (s, 1H,
pyridine Ar–H), 8.25 (s, 2H, pyridine Ar–H), 7.39 (s, 2H, aniline Ar–H),
.05–6.99 (m, 2H, pyridine Ar–H), 6.80–6.78 (m, 2H, aniline Ar–H), 4.12
s, 2H, –NH ). Anal. Calcd for C17 12:C, 74.98; H, 4.44; N, 20.58.
Found: C, 74.83; H, 4.52; N, 20.64. EI–MS m/e: calc. for C17 12, 272.1;
2
protection. The resulting crude product was collected
content was determined as 1.1
MCM-41@Fe ) by concentration difference before and after
CDU-MCM-41@Fe loading with UV–vis absorption spectropho-
tometry. As for controllable release measurement, rhodamine 6 G-
saturated CDU-MCM-41@Fe suspension (10 mg in 5 mL etha-
nol) was exposed to periodical “on” and “off” light stimuli
¼510 nm). Its corresponding emission spectra were recorded
meanwhile. To avoid emission interference from rhodamine 6 G in
composite tunnels, CDU-MCM-41@Fe was concentrated at the
μ
M/g (against unloaded CDU-
1
3 4
O
δ
3 4
O
3
7
(
3 4
O
2
4
N H
N
4
H
(λ
þ
found, 272.1 [m] .
-(4-((4aH-cyclopenta[1,2-b:5,4-b ]dipyridin-5(5aH)-ylidene)
0
1
3 4
O
amino)phenyl)-3-(3-(triethoxysilyl)propyl)urea (CDU) was synthe-
sized with CBD and TESPIC as starting reagents. CBD (5 mmol) was
dissolved in TESPIC (25 mL) and treated with ultrasonic bath for
bottom of sample pool by a magnet so that it could not be excited
by excitation light.
3
0 min. The solution was then heated to reflux under N
After cooling, this solution was dispersed in cold n-hexane (100 mL).
Crude product was recrystallized from mixed solvent of CHCl :n-
12.12 (b, 1H, aniline–
NH)–CO), 11.47 (b, 1H, C–(NH)–CO), 8.78 (s, 1H, pyridine Ar–H), 8.64
s, 1H, pyridine Ar–H), 8.22 (s, 2H, pyridine Ar–H), 7.37 (s, 2H, aniline
Ar–H), 7.06–6.98 (m, 2H, pyridine Ar–H), 6.81–6.76 (m, 2H, aniline
Ar–H), 3.80–3.73 (q, 6H, O–(CH )–), 3.13–3.10 (m, 2H, N–(CH )–C),
.57–1.53 (m, 2H, C–(CH )–C), 1.20–1.15 (t, 9H, –CH ), 0.68–0.65 (t,
H, Si–(CH )–). Anal. Calcd for C27 Si: C, 62.40; H, 6.40; N,
3.48. Found: C, 62.36; H, 6.52; N, 13.54. EI–MS m/e: calc. for
2
protection.
3. Results and discussion
3
1
hexane (V:V¼2:8). H NMR (300 MHz, CDCl
3
):
δ
3.1. Characterization on CDU-MCM-41@Fe
3 4
O
(
(
3.1.1. Design strategy
Design strategy of CDU-MCM-41@Fe
aiming at a clear understanding on each functional component. As
a typical core–shell structure, Fe core is obviously applied for
3 4
O is explained as follows
2
2
1
2
1
C
2
3
3 4
O
2
N
5
H
33
O
4
magnetic guiding purpose. This magnetic core is coated by
amorphous silica so that its surface can be smoothed and covered,
minimizing its negative effect to cargo molecules. In addition, this
þ
27 5 33 4
N H O Si, 519.6; found, 519.2 [m] .
3 4
silica shell may decrease magnetic aggregation between Fe O
2
.3. Preparation of Fe nanoparticles, SiO
3
O
4
2
@Fe
3
O
4
and MCM-
nanoparticles, favoring the following MCM-41 growth procedure.
In this effort, MCM-41, a representative mesoporous silica mole-
cular sieve, serves as the supporting host for cargo molecules
owing to its big surface-area-to-volume ratio and highly ordered
hexagonal tunnels which are positive for cargo adsorption and
transportation [17]. Aiming at controllable release, a light switch
CDU derived from 1,10-phenanthroline was covalently embedded
into MCM-41 channels, avoiding light switch leakage. A C¼N bond
and a large conjugation plane exist in CDU so that its trans/cis iso-
merization could be triggered by low energy light. An organic dye
rhodamine 6 G was selected as cargo owing to its spectral
character. By monitoring rhodamine 6 G spectral character, con-
41@Fe
3
O
4
Core–shell structured supporting host was prepared using
nanoparticles as core, amorphous silica as middle smooth-
nanoparticles were first
prepared following a classic literature method [16]. Fresh Fe
nanoparticles (0.1 g) were then dispersed in mixed solvent of
deionized water (10 mL), NH O (1 mL) and TEOS (0.1 mL).
ꢁ H
Then this mixture was stirred under N protection at room
temperature for 30 min. The resulting sample (SiO @Fe ) was
collected and carefully washed with ethanol and then dispersed in
mixed solution of CTAB (0.3 g), NH O (1.2 mL) and deioinized
ꢁ H
water (40 mL). This mixture was stirred by a mechanical stirrer for
Fe
3 4
O
3 4
ing layer and MCM-41as outer shell. Fe O
3 4
O
3
2
2
2
3 4
O
3
2
trollable release character of CDU-MCM-41@Fe
evaluated.
3
O
4
can be

182
Y. Cai et al. / Journal of Solid State Chemistry 226 (2015) 179–185
3
.1.2. Morphology and magnetic feature
CDU-MCM-41@Fe morphology can be first evaluated from its
SEM image shown in Fig. 1. Those for Fe nanoparticles, SiO @Fe
and MCM-41@Fe are presented in Fig. 1 for comparison. Fe
3 4
into MCM-41 tunnels instead of MCM-41@Fe O surface, which will
be later proved.
3
O
4
3
O
4
2
3
O
4
Hysteresis loops of Fe
41@Fe O and CDU-MCM-41@Fe O
3 4 3 4
3
O
4
nanoparticles, SiO
are presented in Fig. 2. All
these samples are supermagnetic ones without any coercivity
values. Saturation value for Fe nanoparticles is determined as
2 3 4
@Fe O , MCM-
3
O
4
3
O
4
nanoparticles are generally spherical with mean diameter of 150 nm.
They tend to aggregate together owing to their magnetic nature. There
are multiple humps on their surface, showing coarse surface. After
3 4
O
82.6 emu/g, which is comparable to literature values [9,14,16].
silica coating procedure, SiO
2
@Fe
3
O
4
increases its diameter to 200 nm
After a series of modification procedures, saturation values are
with its surface greatly smoothed. From its TEM image shown in Fig. 1,
this core–shell structure can be confirmed. Magnetic aggregation
determined as 75.7 emu/g for SiO
41@Fe and 57.7 emu/g for CDU-MCM-41@Fe
These decreased values compared to saturation value of Fe
nanoparticles can be attributed to their decreased Fe weight
ratios. Owing to its hydrophilic surface and supermagnetic nature,
CDU-MCM-41@Fe can be easily dispersed in water, forming a
stable suspension, as shown by the inset of Fig. 2. In the presence
of external magnetic field, CDU-MCM-41@Fe can be magneti-
2
@Fe
3
O
4
, 60.5 emu/g for MCM-
3
O
4
3 4
O
, respectively.
between them, however, is still obvious. As for MCM-41@Fe
3
O
4
, silica
3 4
O
molecular sieve growth further increases its mean diameter to 550 nm
with smooth and porous surface. MCM-41 shell thickness can be
calculated as 180 nm, as suggested by its TEM image. These hexagonal
tunnels are longer than literature values and can offer a large capacity
for cargo molecules [14,16]. After light switch loading procedure, CDU-
3 4
O
3 4
O
3 4
O
MCM-41@Fe
3
O
4
mean diameter is nearly identical to that of MCM-
cally guided to specific places, meeting site-specific delivery
4
1@Fe , suggesting that these light switch molecules may be grafted
3
O
4
requirements.
3 4 2 3 4 3 4 3 4 2 3 4
Fig. 1. SEM images of Fe O nanoparticles (a), SiO @Fe O (b) and MCM-41@Fe O (c) and CDU-MCM-41@Fe O (d), along with TEM image of SiO @Fe O (e) and MCM-
4
1@Fe (f).
3 4
O

Y. Cai et al. / Journal of Solid State Chemistry 226 (2015) 179–185
183
procedure which partially compromised this highly ordered
structure.
N
2
adsorption/desorption measurement is performed on MCM-
4
1@Fe and CDU-MCM-41@Fe so that the mesostructure on
3
O
4
3 4
O
their surface can be analyzed in detail. Their adsorption/deso-
rption isotherms are typical IV-type ones (see Fig. S1, Supporting
information), confirming that there is a mesostructure in MCM-
4
1@Fe
has been well preserved after light switch encapsulation. As for
MCM-41@Fe , its pore size, pore volume and BET surface area
3 4 3 4
O and CDU-MCM-41@Fe O [16,17]. This mesostructure
3 4
O
3
2
are 2.7 nm, 0.462 cm /g and 676.4 m /g, respectively, which are
comparable to literature values [16,17]. After light switch loading
procedure, these values decrease to 2.3 nm, 0.259 cm /g and
3
2
4
57.4 m /g respectively, suggesting that these light switch mole-
Fig. 2. Hysteresis loops of Fe
CDU-MCM-41@Fe Inset (left): CDU-MCM-41@Fe
0 mL of ethanol) without external magnetic field. Inset (right): CDU-MCM-
1@Fe suspension (10 mg in 50 mL of ethanol) with external magnetic field.
3
O
4
nanoparticles, SiO
2
@Fe
3
O
4
, MCM-41@Fe
3 4
O and
cules have been successfully loaded into MCM-41 tunnels. This
large porous structure can readily load cargo molecules, such as
antitumor medicament.
3
O
4
.
3 4
O suspension (10 mg in
5
4
3 4
O
3.1.4. IR spectra and TGA data
Covalent grafting between light switch and supporting host can be
further confirmed by IR spectral analysis. IR spectra of CDU, MCM-
1@Fe and CDU-MCM-41@Fe are shown in Fig. 4. As for IR
spectra of MCM-41@Fe and CDU-MCM-41@Fe , a series of bands
and 860 cm , respectively, can be
4
3
O
4
3 4
O
3
O
4
3 4
O
ꢂ1
ꢂ1
ꢂ1
peaking at 516 cm , 753 cm
attributed to Si–O asymmetric stretching vibration, Si–O symmetric
stretching vibration and Si–O–Si plane bending vibration, respectively
[17], suggesting the successful formation of silica shell and MCM-41
tunnels in these samples. As for light switch CDU, a series of bands
ꢂ1
ranging from 2870 to 2980 cm can be attributed to vibration of –
(
CH
2
)
3
– group [17]. There is no N¼C¼O stretch vibration ranging
ꢂ1
from 2250 to 2275 cm , suggesting that CBD had completely reacted
with TESPIC. The characteristic band peaking at 1645 cm
ꢂ1
can be
attributed to stretching vibrations of C¼N bond. The intense absorp-
ꢂ1
ꢂ1
ꢂ1
tion bands peaking at 957 cm , 1086 cm and 1558 cm , respec-
tively, can be attributed to stretching vibrations of Si–O and Si–O
bonds. The appearance of CDU –(CH
2
)
3
– characteristic bands (2850 to
IR spectrum indicates that light
switch CDU has been immobilized in MCM-41@Fe tunnels. In
ꢂ1
3 4
2968 cm ) in CDU-MCM-41@Fe O
3 4
O
addition, its red shift suggests that the covalent grafting decreases
vibration frequencies. Some other characteristic bands, including C¼N
Fig. 3. (A) Wide-angle XRD patterns of Fe
1@Fe . (B) SAXRD patterns of MCM-41@Fe
3
O
4
nanoparticles and CDU-MCM-
and CDU-MCM-41@Fe
4
3
O
4
3
O
4
3 4
O .
3
.1.3. XRD patterns and N
Wide-angle XRD patterns of Fe
MCM-41@Fe are presented in Fig. 3A for comparison. Despite
2
adsorption/desorption
3 4
O
nanoparticles and CDU-
3 4
O
their different intensity values, the two curves own nearly iden-
tical diffraction peaks indexed as (2 2 0), (3 1 1), (4 0 0), (4 2 2),
(
5 5 1) and (4 4 0), respectively, which is consistent with literature
reports on Fe
there exists Fe
3
O
4
nanoparticles [14,16]. It is thus confirmed that
magnetic core in CDU-MCM-41@Fe which
3
O
4
3 4
O
has been well preserved after a series of modification procedures.
Small-angle XRD (SAXRD) patterns of MCM-41@Fe and CDU-
MCM-41@Fe are shown in Fig. 3B for comparison. A strong
Bragg refection peak and two weak ones which can be indexed as
1 0 0), (1 1 0) and (2 0 0) can be observed from both samples,
3 4
O
3 4
O
(
suggesting that there is mesostructure on their surface [14,16,17].
Their intensity difference can be attributed to probe loading
3 4 3 4
Fig. 4. IR spectra of CDU, MCM-41@Fe O and CDU-MCM-41@Fe O .

184
Y. Cai et al. / Journal of Solid State Chemistry 226 (2015) 179–185
3 4
Fig. 5. TGA and DTG curves of CDU-MCM-41@Fe O .
ꢂ1
ꢂ1
ꢂ1
bond (1645 cm ), Si–O and Si–O bonds (957 cm , 1086 cm and
1
ꢂ1
3 4
558 cm ), can be traced in CDU-MCM-41@Fe O IR spectrum.
However, more evidences are needed to say that CDU has been
covalently grafted into MCM-41 tunnels. Their covalent grafting will be
further proved below.
3 4
Light switch loading content in CDU-MCM-41@Fe O is then
determined by thermogravimetric analysis (TGA). Its TGA curve
shown in Fig. 5 is composed of four major parts. The first one
ranging from 25 to 188 1C is responsible for 6% weight loss and can
be attributed to thermal release of organic solvents and physi-
sorbed water molecules. The second one ranging from 210 to
3
80 1C and the third one ranging from 380 to 420 1C are totally
responsible for 19.8% weight loss. Considering their sharp
endothermic peaking at 404 1C, these two regions can be attrib-
uted to thermal breakdown and decomposition of light switch
Fig. 6. (A) UV–vis absorption spectra of CDU (1 μM in ethanol), CDU-MCM-
41@Fe O4 (10 mg in 50 mL of ethanol) and MCM-41@Fe O4 (10 mg in 50 mL of
3
3
CDU. In other words, CDU loading content in CDU-MCM-41@Fe
is as high as 19.8% which is slightly higher than literature values
14–17]. We attribute this improved loading content to the long
MCM-41 tunnels in CDU-MCM-41@Fe , as mentioned in Section
.1.2. As for the last one ranging from 420 to 575 1C, it is
3
O
4
ethanol). (B) Low temperature (77 K) emission spectra of CDU (1 μM in ethanol),
CDU-MCM-41@Fe (10 mg in 50 mL of ethanol) and MCM-41@Fe (10 mg in
0 mL of ethanol).
3
O
4
3 4
O
5
[
3 4
O
3
possible [14]. After being grafted into MCM-41 tunnels, these
absorption peaks have been well preserved. As for the high energy
absorption band, a slight blue shift tendency can be observed,
which should be attributed to CDU interaction with silica frame-
work. The low energy absorption band is widened, covering nearly
whole visible region, suggesting that the light switch in CDU-
responsible for 3.3% weight loss and can be assigned to thermal
degradation and deconstruction of organosilicate framework in
3 4
CDU-MCM-41@Fe O , such as Si–C, C–C and C–N bonds cleavage.
For comparison, TGA and DTG curves of CDU and MCM-41@Fe
3 4
O
are compared (see Figs. S2 and S3, Supporting information). As for
pure CDU, its thermal decomposition temperature is much lower
3 4
MCM-41@Fe O may be activated over a wide range from 400 to
than that of CDU-MCM-41@Fe
peaks (236 1C, 264 1C and 443 1C). MCM-41@Fe
3
O
4
, with three major endothermic
is thermally
600 nm. Upon various excitation wavelengths of 260 nm, 350 nm,
400 nm and 500 nm, no obvious emission signals can be observed
3 4
O
stable with slim weight loss, no obvious endothermic peak is
observed. This result confirms that CDU has been covalently
grafted into MCM-41 tunnels and our composite is not the mixture
of reactants.
3 4
from CDU and CDU-MCM-41@Fe O at room temperature (see Fig.
S4, Supporting information). Even when they are cooled to 77 K,
only quite weak emission signals can be detected, as shown in
Fig. 6. This fact suggests that all absorbed energy has been totally
exhausted by trans/cis iso-merization of light-switch. An efficient
trans/cis iso-merization motion can thus be expected for light
stimulated release.
3 4
3.2. Photophysical measurement on CDU-MCM-41@Fe O
We have above mentioned that light switch CDU is anticipated
to be triggered by low energy light. Photophysical characters of
CDU and CDU-MCM-41@Fe thus should be understood. Fig. 6
shows UV–vis absorption (abs.) and PL emission (em.) of CDU
3
O
4
3.3. Controllable release performance of CDU-MCM-41@Fe
3
O
4
(
1
μ
M in ethanol) and CDU-MCM-41@Fe
3
O
4
(10 mg in 50 mL of
(10 mg in 50 mL of
Rhodamine 6 G was loaded into CDU-MCM-41@Fe
3
O
4
so that
ethanol), along with those of MCM-41@Fe
ethanol) for comparison. CDU absorption spectrum is composed of
a high energy absorption part and a low energy absorption part.
The former one ranging from 220 to 370 nm can be attributed to
π
3
O
4
its controllable release performance can be determined. Here,
rhodamine 6 G was selected as cargo owing to its characteristic
spectrum. As shown in Fig. 7, rhodamine 6 G emission peaking at
540 nm obviously increases upon light stimulus “on” state. Given
light stimulus “off” state, rhodamine 6 G emission intensity is
greatly maintained or even constant. There can be find a positive
correlation between rhodamine 6 G emission intensity and light
“on” state during above periodically varying light stimuli. It can
n
-
π
absorption of aromatic rings in CDU, with multiple peaks of
2
41 nm, 262 nm and 313 nm [15]. The latter one ranging from 390
to 620 nm is a broad one peaking at 510 nm and can be assigned to
n
π
-
π
absorption of C¼N bond [15]. This C¼N absorption is red
shifted compared to literature reports based on N¼N bond
abs¼450 nm), which makes low energy excitation of light switch
thus be confirmed that CDU-MCM-41@Fe
3 4
O cargo release can be
(λ
controlled by light stimuli. In other words, controllable cargo

Y. Cai et al. / Journal of Solid State Chemistry 226 (2015) 179–185
185
release, while light switch localized outside of MCM-41 tunnels is
powerless to do so.
4. Conclusion
To sum up, a core–shell structured site-specific delivery system
CDU-MCM-41@Fe O was constructed with supermagnetic Fe O
3 4 3 4
as the core and mesoporous silica molecular sieve MCM-41 as the
shell. A light switch CDU which could be triggered by low energy
light was covalently grafted into MCM-41 hexagonal tunnels,
serving as light stimuli acceptor. CDU-MCM-41@Fe
fied and confirmed by SEM, TEM, XRD patterns, N
3
O
4
was identi-
adsorption/
2
desorption, TGA, IR, UV–vis absorption and emission spectra.
Experimental data suggested that CDU-MCM-41@Fe owned
Fig. 7. Emission intensity monitoring (λem¼540 nm) of CDU-MCM-41@Fe
10 mg in 50 mL of ethanol) loaded with rhodamine 6 G upon periodically changed
light stimuli (λex¼510 nm), on and off. Inset: corresponding emission spectra.
3 4
O
(
3 4
O
supermagnetic feature, which allowed its homogenous dispersion
in aqueous solutions under no external magnetic field and site
specific guiding under external magnetic field. Its hexagonal
tunnels were as long as 180 nm, favoring cargo capacity improve-
ment. Our light switch CDU was superior to literature ones by the
fact that it could be excited by low energy light (
Controllable release measurement on CDU-MCM-41@Fe
λ
¼510 nm).
3 4
O
sug-
gested that CDU could effectively release its cargo molecules upon
light stimuli with short release time of 900 s (90% capacity). These
factors suggested that CDU-MCM-41@Fe O could be further
3 4
developed for magnetism-guiding site-specific delivery. For
further improvements, real drug molecules should be tested. In
addition, its size should be decreased for biological applications.
Acknowledgments
Fig. 8. Cargo holding and release performance of CDU-MCM-41@Fe
3 4
O under
continuous light “on” and “off” states.
The authors thank our universities for financial supports.
3 4
release from CDU-MCM-41@Fe O has been realized with the help
from light switch CDU.
Appendix A. Supporting information
Cargo holding and release performance of CDU-MCM-
1@Fe under continuous light “on” and “off” states is analyzed
4
3
O
4
Supplementary data associated with this article can be found in
the online version at http://dx.doi.org/10.1016/j.jssc.2015.02.019.
as follows. As shown in Fig. 8, under darkness environment,
rhodamine 6 G emission intensity remains constant at a low level,
suggesting that few rhodamine 6 G molecules have been released.
When exposed to continuous light excitation, rhodamine 6 G
emission intensity increases sharply in the first 900 s, suggesting
that most cargo molecules have been released during this time.
Then the release procedure is greatly slowed, showing smooth
release process. The whole release process can be finished in
References
[1] M.E. Caldorera-Moore, W.B. Liechty, N.A. Peppas, Acc. Chem. Res. 44 (2011)
1061–1070.
[
[
2] E.A. Rozhkova, Adv. Mater. 23 (2011) H136–H150.
3] I.I. Slowing, B.G. Trewyn, S. Giri, V.S.Y. Lin, Adv. Funct. Mater. 17 (2007)
1225–1236.
2
200 s. Compared to literature reports on similar core–shell
[4] Q. Yuan, Y.F. Zhang, T. Chen, D.Q. Lu, Z.L. Zhao, X.B. Zhang, Z.X. Li, C.H. Yan
W.H. Tan, ACS Nano 6 (2012) 6337–6344.
structured site-specific delivery systems and biodegradable poly-
mers, this release time seems to be long [14,18]. However, CDU-
MCM-41@Fe
[
[
5] Y. Deng, Y. Cai, Z. Sun, D. Zhao, Chem. Phys. Lett. 510 (2011) 1–13.
6] S.L. Gai, P.P. Yang, C.X. Li, W.X. Wang, Y.L. Dai, N. Niu, J. Lin, Adv. Funct. Mater.
20 (2010) 1166–1172.
3 4
O can finish nearly 90% of its total capacity within
[
[
[
7] J. Liu, S.Z. Qiao, Q.H. Hu, G.Q. Lu, Small 7 (2011) 425–443.
8] X. Liang, X. Li, L. Jing, X. Yue, Z. Dai, Biomaterials 35 (2014) 6379–6388.
9] Y. Zhu, T. Ikoma, N. Hanagata, S. Kaskel, Small 6 (2010) 471–478.
9
00 s, which makes it promising for further application. Here, this
site-specific delivery system with magnetism-guiding and con-
trollable release has been fully confirmed and characterized.
In order to further confirm this controllable release perfor-
mance, emission monitoring of rhodamine 6 G, MCM-41@Fe
and the mixture of CDU and MCM-41@Fe is recorded under
continuous radiation and continuous darkness (see Supporting
information, Figs. S5–S8). None of these components alone could
respond to light stimuli. This fact confirms that CDU-MCM-
[10] R. de la Rica, D. Aili, M.M. Stevens, Adv. Drug Delivery Rev. 64 (2012) 967–978.
[11] J. Croissant, J.I. Zink, J. Am. Chem. Soc. 134 (2012) 7628–7631.
12] P. Yang, S. Gai, J. Lin, Chem. Soc. Rev. 41 (2012) 3679–3698.
[
[
3 4
O
13] Y. Wen, L. Xu, W. Wang, D. Wang, H. Du, X. Zhang, Nanoscale 4 (2012)
3
O
4
4
473–4476.
[14] Y. Wang, B. Li, L. Zhang, H. Song, L. Zhang, ACS Appl. Mater. Interfaces 5 (2013)
1–15.
15] L. Zhang, S. Yue, B. Li, D. Fan, Inorg. Chim. Acta 384 (2012) 225–232.
16] Y.Y. Wang, B. Li, L.M. Zhang, P. Li, L.L. Wang, J. Zhang, Langmuir 28 (2012)
1657–1662.
1
[
[
4
1@Fe
nation of CDU and MCM-41@Fe
bilized in MCM-41 tunnels could control and accelerate cargo
3
O
4
responds to light stimuli owing to the covalent combi-
[
17] B. Lei, B. Li, H. Zhang, S. Lu, Z. Zheng, W. Li, Y. Wang, Adv. Funct. Mater. 16
3 4
O : only the light switch immo-
(2006) 1883–1891.
[
18] M. Chen, D. Zhang, M.H. Xu, S.Q. Zhao, C.X. Sun, Opt. Mater. 36 (2013) 321–327.