Introduction

Color is the visual impression created by the electromagnetic radiation emitted or reflected from an object, which forms the beautiful and colorful world. In principle, colors can be classified into two categories (Sakai et al. 2018; Kim et al. 2019). The one is pigment color, which is due to chemical hues or dyes. It is presumably the most common existence form of color in the nature (Kumar and Sinha 2004; Wrolstad and Culver 2012). However, the pigments may fade due to photochemical degradation, and partly are toxic even cancerigenic, which limit the applications in the fields of food and cosmetic (Samain et al. 2013, 2011; Auger and McLoughlin 2016; Vermeulen et al. 2018). Compared with pigment color, the other color formation, structural color, has some distinctive advantages, such as stability, durability, safety, and metallic luster (Li et al. 2018, 2017; Wang et al. 2021a, 2021b; Takeoka 2018), because the structural color is caused by reflecting certain wavelengths of light with special microstructure (Kohri 2020).

Photonic crystals (PCs) are special photonic materials in possession of photonic band gap owing to the alternating arrangement of two kinds of matrix with different refractive indexes, which results in the reflection of certain wavelength light. When the wavelength locates in the range of the visible spectrum, it exhibits brilliant color, i.e., structural color (Sharp et al. 2002; Subramania et al. 2010; Iwayama et al. 2013). The self-assembly of colloidal nanoparticles is the most common method to prepare PCs materials because it only requests general instruments and makes scale production possible (Debord et al. 2002; Shim et al. 2012; Yamamoto et al. 2014, 2017; Yuan et al. 2016). Instead, a kind of force, such as gravity (Yu et al. 2020), centrifugal force (Chen et al. 2015), surface tension (Mashkour et al. 2019), oversaturation (Fu et al. 2017), and electric field (Chen et al. 2017), needs to be introduced to drive the colloidal nanoparticles assembled orderly. However, it always takes long time for these methods, which is adverse to the production and applications of PCs. The introduction of superparamagnetic nanoparticles solves the problem well. Magnetic responsive photonic crystals (MRPCs) are a unique kind of responsive PCs, their reversible assembly and structural color can be controlled momently with a magnetic field (Fu et al. 2012; Ge et al. 2011). They have received more and more attention for their applications for the detections of temperature (Wang et al. 2015), air humidity (Xuan et al. 2011), mechanical force (Jia et al. 2015), electric field (Liu et al. 2013), solvent (Wang et al. 2011), as well as chemical (Fenzl et al. 2014), and biological molecules (Peterson et al. 2014). As a rule, superparamagnetic nanoparticles for constructing MRPCs are composed of a magnetic core and a protective shell, and they should be controlled at sub-micrometer level. When the magnetic field is applied, superparamagnetic nanoparticles are driven to assemble into one dimensional chain along the magnetic field (Hu et al. 2011). The structural color is controlled by magnetic field intensity, because it can change the balance between the magnetic attraction from the magnetic cores and repulsion force provided with the coatings among the superparamagnetic nanoparticles, and result the interparticle space changing. Up to now, Fe3O4 nanoparticles are the most widely reported superparamagnetic cores in the literatures (Zhang et al. 2012; Yang et al. 2016; He et al. 2012; Ge and Yin 2008). The high saturation magnetization of Fe3O4 colloidal nanocrystal clusters (CNCs) is important. He et al. reported MRPCs based on solvation force from Fe3O4@SiO2 with core–shell structure, the preparation contains two steps: hydrothermal and a Stöber method (He et al. 2012). Nevertheless, these processes need harsh experiment conditions with high temperature and high pressure even long preparation time, which resulting in limited practical applications. As a consequence, it is necessary to develop a simple, fast, gentle, and controllable strategy for the preparation of Fe3O4@SiO2 as building blocks of MRPCs.

In this work, Fe3O4 CNCs are easily obtained by a miniemulsion and evaporation process (as shown in Fig. 1). Firstly, sodium dodecyl sulfate (SDS) as an emulsifier and oleic acid (OA)-modified Fe3O4 (OA-Fe3O4) nanoparticles dispersed in chloroform as an oil phase were miniemulsified to form O/W miniemulsion with a strong ultrasonic treatment. A rotary evaporation was employed to remove the chloroform. Meanwhile, OA-Fe3O4 nanoparticles assemble into Fe3O4 CNCs gradually, and Fe3O4 CNCs were obtained within 30 min. Compared with previous reports, it took much short time. Then, the Fe3O4 CNCs were collected by a magnet and dispersed in a polyvinylpyrrolidone (PVP) aqueous solution to replace the SDS molecules. One reason was a strong binding between C = O group of PVP and Fe–O of Fe3O4 CNCs, and the excellent hydrophilicity of PVP made the Fe3O4 CNCs dispersed in water stably for long. The other reason was that uncharged PVP replaces of negatively charged SDS to get rid of electrostatic repulsion with SiO32− in Stöber system. Finally, SiO2 shells coat on Fe3O4 CNCs with a modified Stöber method. Herein, the size of Fe3O4 CNCs and thickness of SiO2 shells could be tuned easily by changing the amount of emulsifier SDS, miniemulsion parameters, and the amount of tetraethyl orthosilicate (TEOS), which made the preparation of Fe3O4@SiO2 more controllable and smart.

Fig. 1
figure 1

Schematic diagram of the rapid preparation of Fe3O4@SiO2 nanoparticles

Full size image

Experimental section

Chemicals and materials

Ferric chloride hexahydrate (FeCl3·6H2O), ferrous sulfate heptahydrate (FeSO4·7H2O), anhydrous ethanol (CH3CH2OH), ammonium hydroxide (NH3·H2O), OA, lauryl sodium sulfate (SDS), TEOS, chloroform (CHCl3), and PVP were purchased from National Medicines Corporation Ltd., China. All chemicals were of analytic grade and used as received without any further purifications.

Fabrication of the Fe3O4 CNCs

According to literature (Wang et al. 2006), 450 mg OA-modified Fe3O4 nanoparticles (about 10 nm) was dispersed in chloroform (2.5 mL) completely via a 10-min ultrasonic treatment (F-030S, Suzhou Maihong Electric Appliance Co. Ltd., China). Then, it was mixed with 40 mL 1.5 mg/mL SDS aqueous solution in a 50 mL beaker, and miniemulsified with ultrasonication (240 W) (JP-1200Y, Wuxi Jiuping Instrument Co. Ltd., China) in ice-water bath for 4 min. After that, the prepared O/W miniemulsion was transferred into a 250 mL single-port flask, and rotatory evaporation was conducted at 40 °C to remove the solvent within 15 min. Then, the Fe3O4 CNCs was obtained through magnetic separation.

Fabrication of the Fe3O4@SiO2 MCNPs

A total of 0.5 g Fe3O4 CNCs was added into 20 mL 2.67% PVP aqueous solution and the mixture was treated with ultrasonic for 10 min. Eighty-milliliter anhydrous ethanol and 4-mL ammonium hydroxide were mixed in a 250 mL flask and stirred for 10 min. Next, 0.9 mL TEOS was added dropwise into the mixture and stirred for 30 min. Then, above-mentioned 20 mL PVP-modified Fe3O4 CNCs suspension was added into the mixture and stirred for 40 min. All the stirring speed set at 350 rpm. Finally, Fe3O4@SiO2 nanoparticles were magnetically separated by a magnet (0.3 T) and washed with deionized water (50 mL) for 3 times. The magnetic colloidal nanoparticles (MCNPs) were dispersed in deionized water for further analysis.

Characterization

The morphology of the Fe3O4@SiO2 was characterized by a JEM-2100 (JEOL) transmission electron microscopy (TEM) at a 200 kV accelerating voltage. The lattice parameter was characterized using a D8 (Bruker AXS) X-ray diffraction (XRD) with Cu Kα radiation at a scan rate of 5°/min. A WQF-600 N (Beijing Rayleigh Analytical Instrument Co. Ltd.) Fourier transform infrared (FT-IR) spectrometer was employed to get the FT-IR spectrum. The magnetic properties were recorded by a MPMS XL-7 (QUANTUM) vibrating sample magnetometer (VSM). The reflectance spectrum was measured via using a FLA5000 + optic spectrometer (Flight Technology Co. Ltd., China). A Nano Brook Zeta PALS (Brookhaven) nanoparticle sizer was employed to measure the particle size. The photographs were taken by a Xiaomi mobile phone (Mix2s).

Results and discussions

Characteristic of the MCNPs

TEM was the most intuitive tool to investigate the morphology of the nanoparticles. Figure 2a showed the TEM image of a single Fe3O4@SiO2 nanoparticle, and it had regularly spherical core–shell structure. During miniemulsification process, oil droplets containing OA-Fe3O4 nanoparticles dispersed in chloroform were stabilized in aqueous solution by SDS, where hydrophobic alkyl tail inserted in the oil phase while outward hydrophilic head got away from oil phase. With chloroform evaporation, Fe3O4 nanoparticles became closer to each other, Fe3O4 CNCs occurred eventually. The coating of SiO2 protected the Fe3O4 CNCs from the aggregation with each other. Therefore, the MCNPs were in random order without any coagulation as shown in Fig. 2b. However, ordered one-dimension chains were observed in the TEM image of the sample prepared under a magnetic field (Fig. 2c). It revealed that the Fe3O4@SiO2 nanoparticles were superparamagnetic. When a magnetic field was applied, MCNPs produced magnetic attraction, which is balanced with the repulsive force brought by the silica shell, and the nanochains were aligned in the direction of the magnetic field.

Fig. 2
figure 2

TEM images of a single Fe3O4@SiO2 particle, b Fe3O4@SiO2 sample dried without a magnetic field, and c Fe3O4@SiO2 sample dried under a magnetic field

Full size image

The XRD patterns of OA-Fe3O4 (black line) and Fe3O4@SiO2 (red line) were shown in Fig. 3. The main characteristic peaks located at 30.4°, 35.8°, 43.5°, 53.8°, and 57.3° correspond to (220), (311), (400), (422), and (511) planes, respectively, which demonstrated OA-Fe3O4 with cubic spinel structure. In comparison, there was no significant difference between XRD patterns of OA-Fe3O4 and that of Fe3O4@SiO2. It implied that the miniemulsification and rotary evaporation process have not change the crystal form of Fe3O4.

Fig. 3
figure 3

XRD patterns of OA-Fe3O4 (black line) and Fe3O4@SiO2 (red line)

Full size image

The FT-IR spectra also proved SiO2 shells on the surface of MCNPs. Figure 4a showed the FT-IR spectra of OA-Fe3O4, SiO2, and Fe3O4@SiO2 nanoparticles. The characteristic peaks at 475 cm−1 and 1106 cm−1 were corresponding to the stretching vibration of Si–O groups and Si–O-Si groups, respectively. Moreover, the characteristic peaks at 601 cm−1 attributed to the Fe–O groups. Thus, it was testified that the as-obtained nanoparticles consisted of Fe3O4 and SiO2. The magnetic properties of Fe3O4@SiO2 were characterized by a VSM magnetometer as shown in Fig. 4b. The hysteresis loop was measured in the range from − 3 to 3 T. The Fe3O4@SiO2 exhibited a high saturation magnetization (Ms) as 47.01 emu·g−1. Ms is a key parameter for MCNPs to magnetically assemble MRPCs and regulate the structural color. In general, MCNPs with Fe3O4 CNCs had enough Ms to control the structural color covering entire visible spectrum. The residual magnetizations (Mr) was calculated as 2.04 emu·g−1 and the coercivity was as low as 33.4 Oe. It was consistent with TEM characterization that the Fe3O4@SiO2 exhibited superparamagnetism.

Fig. 4
figure 4

a FT-IR spectra of OA-Fe3O4 (blue line), SiO2 (red line), and Fe3O4@SiO2 (black line); b hysteresis loops of Fe3O4@SiO2 nanoparticles

Full size image

Photonic characteristics

A total of 70 mg Fe3O4@SiO2 nanoparticles were dispersed in 10 mL deionized water to obtain the colloidal dispersion, followed by transferring to a glass vial. When an appropriate magnetic field was applied to the dispersion, a specific wavelength of light was reflected owing to the photonic band gap built by Fe3O4@SiO2 particles. The reflection spectra of the MRPCs built with Fe3O4@SiO2 aqueous dispersion under the different magnetic fields were recorded with a reflection spectrometer. The magnetic field intensity altered via changing the distance between the sample bottle and the magnet. The probe integrated with the light source and signal receiver was put into the magnetic responsive colloidal dispersion, and a magnet was placed below the vial, which provided a bottom-up magnetic field. The 100% baseline was collected in the Fe3O4@SiO2 aqueous dispersion without magnetic field. The magnetic field intensity changed from 0 to 200 mT measured by a magnetometer. As was observed in Fig. 5b, the reflection wavelength shifted from 623 to 478 nm along with the increase of magnetic intensity, and the color changed from red to blue, which covered the whole visible spectrum. Figure 5a showed the digital photograph of MRPCs under the magnetic field with different magnetic intensities. Because the increase of the magnetic intensity resulted in the enhancement of the magnetic attraction among the MCNPs, the distance between the particles became shorter and the reflective wavelength blue-shifted. The MRPCs possessed the excellent optical properties such as rapid and reversible response, brilliant color, and wide controllable color range. It was testified that miniemulsification process could be used to prepare the magnetic cores of MCNPs. The size of Fe3O4 CNCs could be tuned by regulating the experimental parameters during miniemulsification. It exhibited potential applications in the designs and constructions of MRPCs.

Fig. 5
figure 5

a Digital photographs and b reflectance spectra of Fe3O4@SiO2 MRPCs in the magnetic field at different intensities

Full size image

The effect of miniemulsion conditions on Fe3O4 CNCs

A list of single factor experiments was designed to investigate the effect of SDS concentrations, oil/water volume ratio, and the amount of added OA-Fe3O4 on the particle sizes of Fe3O4 CNCs. The SDS concentrations changed from 0.25 to 2.0 mg/mL while other conditions were the same as mentioned in the “Experimental section.” SDS concentrations mainly affected the emulsifying efficiency. It was well known that the size of oil droplet in O/W emulsion was influenced significantly by the emulsifier concentration. In general, the higher the SDS concentration, the smaller the emulsion droplets. After the solvent was evaporated from the oil droplets, the particles self-assembled and became smaller. As expectedly, the particle sizes exhibited a descending trend as the SDS concentrations increase, shown in Fig. 6a. Figure 6b showed the effect of oil/water volume ratio in miniemulsion system. The oil/water volume ratio was tuned by changing the volume of oil phase and remaining the volume of water phase unchanged. As the oil/water volume ratio increased from 1:20 to 1:5, the particle sizes showed an increasing trend gradually. Because the relative amount of emulsifier decreased with the increase of oil phase, the emulsion droplets became smaller. However, as the oil/water volume ratio was 1:40, the oil phase was so small that it was not enough to disperse all of OA-Fe3O4 particles. So, the size of particles was large unusually. Figure 6c showed the effect of the amount of OA-Fe3O4. The amount of OA-Fe3O4 changed from 0.05 to 0.45 g while other conditions were the same as mentioned in the “Experimental section.” It was obviously that the particle sizes of Fe3O4 CNCs increase as the amount of OA-Fe3O4 added into the oil phase, which was on account of more OA-Fe3O4 contained in the emulsion droplets. Therefore, the particle sizes of Fe3O4 CNCs could be tuned easily by changing the emulsion conditions such as the concentrations of emulsifier, oil/water ratio, and the amount of OA-Fe3O4 in the oil phase. It was beneficial to the controllable preparation of Fe3O4 CNCs.

Fig. 6
figure 6

Hydration diameters of Fe3O4 CNCs nanoparticles prepared by using different SDS concentrations (a), oil/water ratio (b), and amount of OA-Fe3O4 in the oil phase (c)

Full size image

The effect of amount of TEOS on Fe3O4@SiO2

The thickness of SiO2 shell on the surface of Fe3O4@SiO2 nanoparticles was controlled by changing the amount of TEOS in the sol–gel method. Fig. S1 showed the TEM images of Fe3O4@SiO2 nanoparticles prepared with 0.3 mL, 0.9 mL, and 1.5 mL TEOS, and the thickness of SiO2 of 14.72 nm, 20.64 nm, and 22.31 nm, respectively. Moreover, the FT-IR spectra of Fe3O4@SiO2 nanoparticles prepared with different amounts of TEOS are shown in Fig. S2. The characteristic peaks located at 606 cm−1 and 471 cm−1 are corresponding to Fe–O and Si–O groups, respectively. Compared with Fe–O group, the relative absorbance degree of Si–O group became stronger with TEOS increase from 0.3 to 1.5 mL. It indicated that the amount of SiO2 and the thickness of SiO2 shells were increasing. Figure 7a showed the hysteresis loops of Fe3O4@SiO2 nanoparticles prepared with different amounts of TEOS. The Ms decreased as the amount of TEOS increased because of the increase of thickness of the SiO2 shells (Table S1). The digital photograph of Fe3O4@SiO2 nanoparticles with different amounts of TEOS was recorded under a same magnet (Fig. 7b). It was observed obviously that when the amount of TEOS added in Stöber system was 0.3 mL, i.e., the thickness of SiO2 shell was 14.72 nm, the variance of structural color responding to the magnetic field could cover the whole visible light. Figure 7c showed the reflection wavelength of above-mentioned MRPCs under different magnetic fields. When the amount of TEOS added increased from 0.3 to 0.9 mL, the range of reflection wavelength obviously red-shifted, owing to the decrease of Ms and the increase of SiO2 shells thickness. But when the SiO2 continued to increase, Ms was not enough to magnetically assemble Fe3O4@SiO2 nanoparticles well under the weaker magnetic field. Therefore, the range of structural color was narrowed as the amount of TEOS at 1.2 mL and 1.5 mL. The amount of TEOS added in the sol–gel system significantly affected the tuning width of structural color of MRPCs.

Fig. 7
figure 7

a Hysteresis loops, b digital photograph under a same magnetic field, and c reflection wavelength of Fe3O4@SiO2 nanoparticles prepared with different amounts of TEOS under different magnetic fields

Full size image

Because this work was aimed at a rapid preparation of size-tunable nanoparticles, a comparison with these of previous reports in aspects of particles sizes, magnetic properties, and fabrication time is necessary. Table 1 showed several kinds of nanoparticles. Firstly, the fabrication process we reported costs the least time. Secondly, the saturation magnetism enhances normally by increasing the particle size in view of the amount of magnetic substance. But, the magnetic particles of this work have a high saturation magnetism with a small particle size, which make it possible to assemble into photonic crystals under a small magnetic field. The Fe3O4/PAA particle mentioned below not only has a high saturation magnetism, but it also has a big size. Also, the fabrication needs a solvothermal method, which requires special equipment and high temperature. This work provided a rapid strategy to prepared size-tunable nanoparticles to construct magnetically responsive photonic crystals. The magnetically responsive photonic crystals can be used to prepare photonic crystals polymer film because the Fe3O4@SiO2 with high saturation magnetism can easily assemble into photonic crystals in thick solution.

Table 1 The particle sizes, saturation magnetism, and fabrication time of the Fe3O4@SiO2 nanoparticles and several other nanoparticles of previous literatures
Full size table

Conclusions

In summary, we fabricated the Fe3O4 CNCs via miniemulsification and evaporation processes successfully, and then Fe3O4 CNCs were further capsulated to form Fe3O4@SiO2 nanoparticles. The fabrication process was fast and gentle only in two steps: miniemulsification and rotary evaporation within 30 min. The particle sizes of Fe3O4 CNCs could be tuned easily in the miniemulsion process by regulating SDS concentration, OA-Fe3O4 amount, and oil/water phase ratio, respectively. Moreover, the thickness of SiO2 shells on the surface of Fe3O4@SiO2 could be directly adjusted by TEOS amount in the sol–gel system. When the amount of TEOS increased from 0.3 to 1.5 mL, the Ms of corresponding Fe3O4@SiO2 decreased from 47.01 to 35.2 emu·g−1 and the thickness increased from 14.72 to 22.31 nm, respectively. A wide adjustable reflection wavelength shifted from 623 to 478 nm (Δλ = 145 nm) when the thickness of SiO2 shells was 14.74 nm, which covered the whole visible spectrum. Therefore, it provided a simple, rapid, facile procedure to fabricate Fe3O4@SiO2 with controllable size, Mr, and shell thickness, as well as MRPCs with controllable structural color.