All reagents used were analytical grade and available commercially. Styrene, FeCl₃・6H₂O, FeCl₂・7H₂O, KNO₃, hydrogen peroxide (30 wt%), and aqueous ammonia solution (28 wt%) were purchased from Shanghai Lingfeng Chemical Reagent Co., Ltd., (China). 2,2’-Azobis (2-methylpropi-onamidine) dihydrochloride (AAPH), polyvinylpyrrolidone (PVP, Mw = 30000), 3-aminopropyl triethoxy-silane (APTES), Hexamethylenetetramine (HMT), and 3-mercaptopropyltriethoxysilane (MPTS) purchased from Sigma-Aldrich Co. The aforementioned chemicals were used as received.

3.0 g of styrene, 1.5 g of PVP and 0.39 g of AAPH were dissolved in 100 mL of deionized water, and then the mixture was maintained at 70˚C in nitrogen atmosphere. After stirring for 24 h, the polystyrene emulsion was filtered by micro-pore film filter andplaced at room temperature for 2 days. Then, 40 mL of ethanol, 5 mL of aqueous ammonia and 1.0 g of 3-mercaptopropyltriethoxysilane (MPTS) were added into polystyrene emulsion, this solution under vigorous stirring was heated up to the temperature of 50˚C for 4 h. Then the product was separated by centrifugal hydroextractor, following washed with deionized water three to four times and ethanol two times. The thiol-functionalized hollow silica spheres (THSS) were prepared.

The as prepared THSS (0.8 g) were dissolved into 30 wt% 32 mL hydrogen peroxide, and kept the oxidation reaction at 50˚C for 12 h under a constant stirring at 300 rpm. The obtained sulfonated hollow silica spheres were separated again by centrifugation, following washed with deionized water three to four times and ethanol two times. Finally, the SHSS were re-dispersed into deionized water.

First, 0.3 g of freeze-dried sulfonated SiO₂ template spheres were dispersed in a solution composed of 50 mL ethanediol and 50mldeionized water. Next, 2.0 g of Hexamethylenetetramine (HMT) and 0.2 g of potassium nitrate (KNO₃) were added with stirring for 30min, then 20 mL 0.05 M iron salts (Fe³⁺, Fe²⁺) was added to the reaction mixture under vigorous stirring and subsequtently heated at 80˚C for 5 h. The obtained magnetic hollow spheres were purified by centrifugation and dried under vacuum at ambient temperature. With all of the other parameters being kept constant, the iron salts concentration, and the molar ratio of [Fe²⁺]/[Fe³⁺] were varied to evaluate their effects on the morphologies of MHSS. Table 1 summarized the nomenclatures we used for products adopting iron salts concentrations and different molar ratio of [Fe²⁺]/[Fe³⁺].

The dried MHSS were re-suspended in the dried toluene (50 ml), followed by the

addition of 3-aminopropyl triethoxysilane (APTES, 1 ml); the mixture was then heated to reflux for 12 h. The amino-functionalized magnetic MHSS were separated by a magnet and washed with ethanol and water, and then dried for 12 h.

The was reduced using stannous chloride, typically, 8 mg of SnCl₂, 0.1 ml of 0.5 mol/L sodium gluconate solution, 0.4 ml of PBS (pH = 5.7), and 5 mg of unlabeled compound (MHSS-a) (amino-functionalized-MHSS) were mixed in the vial to form homogeneous solution, which were subjected to ultrasonic treatment for 15 min. Then 1ml of 5 mCi/ml eluate was added, the mixture was heated to 70˚C for 2 h, the labeled compound (MHSS-^99mTc) were separated by magnetic field and washed with deionized water, and the radiochemical purity of the labeled compound was checked with Capintec CRC-15R external dose calibrator.

To further investigate the biodistribution of nanoparticles, 0.5 mCi/1ml ⁹⁹Tc^m labelled MHSS solution were injected into the female mice via caudal vein. After a certain time in magnetically targeted therapy, the animals were sacrificed with an overdose of sodium pentobarbital, and the main organs (the blood, heart, lung, liver, spleen, pancreas, stomach, small intestine, large intestine, brain, muscle and bone) were detached and weighed. The main organs were rinsed with buffer and remaining cell-associated radioactivity was measured with a γ-counter respectively.

A Hitachi S-4800 field emission scanning electron microscope (SEM) and a PhilipsTecnai-12transmission electron microscope (TEM) were used to characterize the morphology of the products. The infrared spectra were recorded by a Nicolet Impact 410FT-IR using KBr disks. X-Ray Powder Diffraction (XRD) was used to characterize the crystalline phases of the materials. A vibrating-sample magnetomer (VSM JDM-13) was used to characterize the magnetic properties.

As shown in Figure 2(a), Figure 2(b), it can be seen that the THSS spheres have clear cavity, and uniform size. The particle sizes of the samples are approximately 100 nm diameters, and the shell thickness is about 15 nm and the core diameter is about 75 nm. The SEM image recorded in Figure 2(c) clearly exhibited a uniform spheres structure with diameters and the sizes are close to the TEM results.

FTIR measurements can be used to deduce the surface properties of materials. Figure 2(d) illustrates the FTIR spectra of THSS samples. The peak at 2555 cm⁻¹ is attributed to the stretching vibrations of S-H, the peaks at 698 cm⁻¹ and 753 cm⁻¹ are attributed to the stretching vibration of the monoreplacement in the phenyl ring group of samples, and the peak at 1035 cm⁻¹ is attributed to the vibration of Si-O-Si. Similarly, two peaks at 1641 cm⁻¹ and 3435 cm⁻¹ are assigned to the O-H bond. These findings indicate that the mercapto group is presented in THSS.

Nano SiO₂ can be successfully coated on the surface of polystyrene because of the mutual interaction between positive charged polystyrene and negative charged MPTS, which may be hydrolyzed into SiO₂ carried thiol group under catalysis by ammonia. Therefore, the formation process of hollow silica spheres was related to ammonia, which not only served as a reactant to form SiO₂, but also as an alkali media to “dissolve” PS template spheres with ethanol [18] [19] . This method is a simple and “green” method to remove any templates of polystyrene cores in synthesis process which need not any additional dissolution, calcination, and etching process.

The TEM images of as-synthesized SHSS are illustrated in Figure 3(a) and Figure 3(b). The TEM observation indicates that as synthesized SHSS exhibiting hollow structure. The average size is around 100 nm, with the hollow diameter of 75 nm and the silica shell of 15 nm. Some shells of SiO₂ were linked among the particles (Figure 3(b)),

because SiO₂ formed by MPTS hydrolysis were the reunion of Si-O-Si. The morphology of the obtained SHSS recorded in Figure 3(c) exhibits uniform spherical shape with the size of 100 nm. Figure 3(d) displays FTIR spectroscopy of as-synthesized SHSS structure, in which the structural change of hollow silica spheres can be observed. The absorption peaks around 698 and 753 cm⁻¹ are attributed to the phenyl group, and the peak at 1035cm⁻¹ is assigned to the vibration of Si-O-Si. The pair peaks at 1641 cm⁻¹ and 3435 cm⁻¹ are assigned to the O-H bond. In addition, the peak at 2555 cm⁻¹ has vanished in comparison to Figure 2(d), but appearance of a new peak at 1450 cm⁻¹ can be attributed to the absorption peaks of sulfonic group, indicating mercapto group (S-H) has been oxidized to sulfonic groups.

Figure 4 shows the TEM images of MHSS prepared by different ratio of Fe²⁺ to Fe³⁺. The TEM image of MHSS-a (Figure 4(a)) demonstrates that the well dispersed MHSS has a uniform spherical hollow structure with a diameter around 105 nm and wall thickness of approximately 6 - 6.5 nm. It is obvious that black Fe₃O₄ completely coated on silica layer when the ratio of [Fe²⁺]/[Fe³⁺] is 2:1 and the iron concentration is 0.10 M, because both the ratio of Fe²⁺ to Fe³⁺ and the total iron concentration are two key factors influences the size of magnetic particles on the surface of MHSS.

Generally speaking, under the perfect condition, co-precipitation occurs when the mole ratio of [Fe²⁺] to [Fe³⁺] is 1:2. But actually, the Fe²⁺ is easily oxidized when exposed in air, so the mole ratio of [Fe²⁺] to [Fe³⁺] is kept constant at 2:1 to avoid formation of

orange red FeOOH phase [20] . The completely coprecipitation of Fe³⁺ with Fe²⁺ ions generally occurs at above pH 9.2, which can be written in the following:

On the other hand, deposition speed of Fe³⁺ ions are faster than that of Fe²⁺ ions, due to their different between Fe³⁺ ion and Fe²⁺. The_, Fe(OH)₃ = 3.98 × 10⁻³⁸, and, Fe(OH)₂ = 7.94 × 10⁻¹⁶, so the precipitate of Fe³⁺ and Fe²⁺ occur at pH 3 - 4 and pH 8 - 9 based on the calculations from their respectively. When coprecipitation of Fe³⁺ with Fe²⁺ ions occurs at 2:1 ratio of [Fe²⁺] to [Fe³⁺], the precipitate of Fe³⁺ preferentially occurs, then co-precipitation of Fe²⁺ with Fe³⁺ promoted the Fe₃O₄ crystal deposition on SHSS surface. But when the mole ratio of [Fe²⁺] to [Fe³⁺] is 2:2, shortfall amount of Fe²⁺ ions are not enough to support whole Fe³⁺ ions approaches to Fe₃O₄ crystal deposition, due to Fe²⁺ liable to be oxidized by trace amount oxygen in the air. Thus, the excess Fe³⁺ ions can accelerate the electrostatic adsorption between new-formed Fe₃O₄ crystals, resulting in formation of MHSS-b with severe aggregation on SHSS surface, as shown in Figure 4(b).

Whether an absence of Fe³⁺ ions is favorable to Fe₃O₄ crystal deposition on SHSS surface, there is unsatisfied answer shown in Figure 4(c). It is because at 2:0 ratio of [Fe²⁺] to [Fe³⁺], the Fe³⁺ ions employed for Fe₃O₄ crystal deposition can only be obtained by oxidization of Fe²⁺ with trace hydrogen peroxide residued in synthesis process. Therefore few obtained Fe³⁺ ions can be applied to participate Fe₃O₄ crystal deposition on SHSS surface, which leads to generation of MHSS-c with the small amount of magnetic particles covered on SHSS surface (Figure 4(a)) [21] .

Second, the influence of the total iron concentration must be considered, when the mole ratio of [Fe²⁺] to [Fe³⁺] is controlled at 2:1, and the total iron concentration is 0.05 M, the image of MHSS-d recorded in Figure 4(d) still shows hollow silica/magnetic composite spheres, it can be seen Fe₃O₄ layer of 2 - 3 nm thickness are coated on SHSS surface, which is much thinner than the Fe₃O₄ coating layer prepared at the total iron concentration of 0.10 M. It demonstrates that total iron concentration can influence the thickness of Fe₃O₄ coating layer on the SHSS surface, appropriate iron concentration can improve Fe₃O₄ crystal deposition on the SHSS surface, less iron concentration can only form thin Fe₃O₄ coating layer.

To further confirm the crystalline structure, XRD measurement was used to analyze the sample of varying the concentration of iron salts and different ratio of [Fe²⁺]/[Fe³⁺]. The diffraction peaks of the particles synthesized in this study were measured to be 2θ = 29.70, 35.06, 42.71, 56.60, 62.20, corresponding to (2 2 0), (3 1 1), (4 0 0), (5 1 1) and (4 4 0) crystal planes of the Fe₃O₄ respectively (Figure 5). It can be found that well-resolved diffraction peaks appeared on the pattern, and the pattern can be indexed to Fe₃O₄ (JCPDS#79-0418) according to the reflection peak positions and relative intensities, which confirms the magnetite structure of this sample. The consistence of four curves in Figure 5 demonstrated that there was no crystal transition of the magnetic nanoparticles when the iron salts concentrations and the molar ratio of [Fe²⁺]/[Fe³⁺] change.

The results of indexing to the powder X-ray diffraction pattern of Figure 5 are listed in Table 2. It can be inferred from Table 2 that all MHSS samples have monoclinic

Note: D is calculated value, D_o is original value.

symmetry, the lattice parameters of MHSS-a, MHSS-b, MHSS-c and MHSS-d samples were (a = 2.2597, b = 1.8097, c = 2.6941, β = 100.512˚), (a = 2.2682, b = 1.8068, c = 2.6965, β = 100.25˚), (a = 2.2725 nm, b = 1.808 nm, c = 2.6788 nm, β = 100.242˚) and (a = 2.2680 nm, b = 1.8102 nm, c = 2.6.949 nm, β = 100.44˚) respectively, the minimum relative deviation between the calculated D and original D_o is 0.0008‰ for MHSS-a. The MHSS-a has a largest cell angle (β) of unit cell at 100.512˚, the MHSS-d has a second large cell angle (β), and the MHSS-c has a minimum cell angle (β). Because there is intimate relationship between spherical shell’s surface curvature and the cell angle of unit cell [22] , as shown in the model structure (Figure 6), the unit cell dimensions a, b and c(001) and cell angle β will vary from unit cell to unit cell along the surface curvature of the spherical shell. Thus the larger angle the unit cell shows, the more spherical shell curvature, and the sample has a well-defined spherical shell. In comparison with Figure 4, the image of MHSS-a recorded in Figure 4(a) has an obvious spherical structure, the image of MHSS-d has approximate spherical structure, indicating the findings are related to their cell angles (β) of unit cell.

The Fe₃O₄ content of the spheres surface can be controlled by changing the amount of the ratio of [Fe²⁺]/[Fe³⁺] or the concentration of iron salts. Magnetic testing results of MHSS were shown in Figure 7 and Table 3. The magnetization saturation (Ms) of MHSS spheres are shown in Table 3, it can be seen from the results that MHSS-ahasthe highest saturation magnetization at 34.38 emu/g, MHSS-b has the second at 24.38 emu/g, MHSS-d has the third at 18.77 emu/g, but MHSS-c has the minimum value at 11.80 emu/g. The coercivity are close to zero, indicating they all belong to super- paramagnetic system. It can be seen that the Ms value of the MHSS increases with the concentration ofsalt solution increasing, due to a difference of the magnetic layer thickness. And Ms value of the MHSS increases with an increase of Fe³⁺ content in the ratio of [Fe²⁺]/[Fe³⁺], due to appropriate Fe³⁺ amount being able to improve deposition of Fe₃O₄. Considering MHSS-a having the highest saturation magnetization at 34.38 emu/g, MHSS-a could be selected as carrier of targeted drugsin this paper, providing the empirical research basisfor highly efficient antitumor with low toxicity.

For the radiolabeling of MHSS, the unlabelled amino-functionalized-MHSS were mixed with radioactive Na⁹⁹Tc^m O₄ in the presence of SnCl₂, which could reduce ⁹⁹Tc^m (VII) into ⁹⁹Tc^m(III) ions. [23] , it was found that ⁹⁹Tc^m3+ ions could also be anchored on the inner cavity of MHSS by coordinating bond between low valent ⁹⁹Tc^m3+ ions and -NH₂ groups (Figure 8). The obtained labelled MHSS compound was washed with physiological saline, and following seperated by strong magnet for three times to remove excess free ⁹⁹Tc^m. The radiochemical purity of the labeled compound was checked with Capintec CRC-15R external dose calibrator. The purified MHSS-NH₂-^99mTc was dispersed in physiological saline to generate 20 mg/2.5 mL of labeled compound solution by supersonic treatment. The labeling yield of ^99mTc was then determined to be 81%. To determine the radiostability, MHSS-NH2-^99mTc were incubated in mouse serum for 40 h. Nanoparticles remained stable with>96% of the nanoparticles remaining ^99mTc- bound after 24 h. At the final time point (72 h), 90.2 ± 1.7% of nanoparticles were found to be stable, thus confirming the enhanced stability of MHSS-NH2-^99mTc in mouse serum (Figure 9).

⁹⁹Tc^m labelled MHSS particles not only enabled enhanced internal radioisotope therapy [24] , but also allowed us to quantitatively track the in vivo behaviors of those nanoparticle after they were administered into animals. The female normal mouse was intravenously injected with MHSS-NH₂-^99mTc and then scanned under single photoemission-computed tomography (SPECT) imaging. Due to the wide distribution of each labelled MHSS particles throughout the body after injection of MHSS-NH₂-^99mTc, high levels of radioactivity accumulation were observed in the liver, spleen, kidneys, and bladder after 2 h post injection, but a small amount of radioactive particles uptake in the lungs (Figure 10(a)). The results show the labelled MHSS particles can successfully break through the pulmonary circulation, and excreted through the urinary system. Instead the MHSS-NH₂-^99mTc particles were swallowed by reticuloendothelial system (lung, liver and spleen).

In order to explore the targeting properties of labelled MHSS particles, the VX2 tumor mice, a fast-growing tumor model were each injected 0.5 mCi/1 ml of MHSS-NH2-^99mTc via tail vein, and the tumor of nudemice was placed on the magnetic pole of the YMC-11 medical pulse magnetic field generator with the output voltage of pulsed magnetic fields being controlled at 700 V for 2 h. In Figure 10(b), the right forelimb (A) is normal, the left forelimb (B) is the tumor region with targeting, the left hind limb (C) is the tumor region with no targeting and the right hind limb (D) is a sense of inflammation.

Figure 10(b) shows nude mice SPECT imaging targeting to forelimb after injection, the intensive distribution of black dots in the image correspond to the large amount of ⁹⁹Tc^m labelled MHSS particles were accumulated into the tumor sites by a combination of passive magnetic targeting and active targeting mechanisms. So in comparison to the in vivo bio-distribution of ⁹⁹Tc^m labelled MHSS particles, the targeting SPECT imaging show the activity concentration of labelled MHSS particles in tumour were significantly higher than that without targeting.

Table 4 shows the results of magnetically targeted radiotherapy employed ⁹⁹Tc^m labelled MHSS, the liver uptake of ⁹⁹Tc^m labelled MHSS was 16.4979, 14.3941, 13.4595, 13.4550 and 14.2896 % ID/g at 15, 30, 60, 120 and 180 min respectively, the radioactivity in the spleen was 13.9908, 14.3385, 12.0734, 12.9458 and 13.8828 % ID/g at 15, 30, 60, 120 and180 min, respectively, the results show MHSS-a has the highest liver uptake of ⁹⁹Tc^m for 180 min after the injection, when the absolute organ uptake is compared, uptake by the spleen is the second highest, and uptake by the lung is the third highest, the findings are in a good agreement with SPECT imaging data with high levels of

⁹⁹Tc^m labelled MHSS uptake in the liver, spleen and lung. On the contrary, low radioactivity uptake was observed in pancreas, brain and bone, probably because ⁹⁹Tc^m labelled MHSS is rapidly cleared by pancreas, brain and bone, and excreted by mouse into the urine.