1 Introduction

Recent technological breakthroughs and the desire for new features have created a huge demand for new materials. Many of the usual materials, such as metals, ceramics or plastics, cannot fulfill all of the technology requirements for many novel and demanding applications. In recent years, the research for the preparation of core–shell nanoparticles with a well-defined structure has attracted considerable interest due to their excellent physical and chemical properties, such as optical, electrical, thermal, mechanical, magnetic and catalytic ones, compared to pure materials. [1, 2].

Nanomaterials consisting of a core and a shell are commonly referred as core–shell nanomaterials/nanoparticles (CSNs or CSNp) [3, 4]. Core–shell nanomaterials are extremely functional materials with modified properties [5]. The majority of the CSNs are manufactured in order to combine two materials—and therefore two or more properties in one structure. Thereby, the new structure is characterized by the properties of both core and shell, giving a variety of new potentials and countless combinations [6]. Their production can be realized either in a one-step process where the core particles are synthesized in situ, this being followed by the coating of the shell material, or in a two-step process. In the latter, the core is firstly synthesized and then the synthesis of the shell follows [5, 7] employing several synthetic procedures such as sol–gel [8], microemulsion [9], polymerization [10] or deposition–precipitation [11]. By this way, various types of materials can be combined such as dielectric materials, metals and semiconductors, where the core consists of one of the materials and the shell of another or the same material [5, 12].

There are manifold purposes of coating the core particle, for instance, surface modification, the ability to increase the functionality, stability and dispersibility, the controlled release of the core, the reduction in consumption of precious materials and so forth [6, 13]. Apart from the improved material properties, core/shell materials are also significant from a financial viewpoint. For example, a precious material can be coated over an inexpensive material in order to reduce its consumption as opposed to making the same size structure composed of the pure precious material [4, 5, 14, 15].

Cerium dioxide or ceria (CeO2) is considered as a very useful material as it can be used in various applications such as polishing material [16] or solid oxide fuel cell electrolyte [17]. Moreover, it is not toxic, a fact that makes it useful in optoelectronics and photocatalysis [18]. Owing to the high oxygen storage capacity and the ability to stabilize the dispersed metal cation species, ceria has attracted researchers’ attention finding many applications in environmental catalysis [19]. Environmental pollution is a major issue and has attracted the interest of researchers, primarily because of adverse effects on human health and all living organisms. Industry causes many environmental problems, releasing a wide range of toxic substances into the environment [20]. Heterogeneous catalysis, as a means of anti-pollution, has also attracted much interest and has become an emergent technology approach for a sustainable environment, such as the removal of organic pollutants from aqueous effluents [21].

It is well known that size, shape, surface state and crystallinity are crucial factors for the catalytic performance of nanomaterials. Due to their tendency for aggregation during the catalytic process, their catalytic activity and stability are declined [22].

Therefore, different approaches have been investigated to control these properties. Conventional production methods such as precipitation and sol–gel usually result in large particles (about 50 nm), whereas the thermal degradation of cerium salts leads to particles of high porosity and specific surface area, but the control of particle size and morphology is very limited [23]. Many researchers wishing to overcome these problems have developed various methods of preparing silica–ceria core–shell particles as silica can offer many advantages such as good dispersion, narrow size distribution and controlled particle size [24,25,26,27,28,29]. However, in previous studies [30], the final particles showed intense agglomeration due to the formation of hydrogen bonds because of the presence of water during the precipitation process. In addition, the resulting suspension contained both coated silica particles and pure ceria particles. Thus, the development of core–shell structures which leads to good dispersion, narrow size distribution and controlled particle size will favor the catalytic process.

In the present study, the main objective is to develop a core–shell material comprising a uniform nano-CeO2 shell onto silica cores produced by an environmentally friendly method, as well as to reduce the cost of the final material as cerium oxide is an expensive material. The experimental procedure took place at ambient temperature, while the neutralizing agent for the precipitation of cerium oxide on the silica surface was a polymer–water solution. Finally, the optimum core–shell structure and the pure CeO2 were studied for their catalytic properties by their ability to catalyze the reduction of 4-nitrophenol to 4-aminophenol in the presence of sodium borohydride. The results showed that the development of core–shell structure affects the catalytic performance of CeO2.

2 Materials and methods

2.1 Materials

Tetraethyl orthosilicate (TEOS, 98%) was used as a precursor for the silica nanoparticles, and hyperbranched poly(ethylene)imine was used to modify the surface of the silica with the amine groups. Cerium (III) nitrate hexahydrate (99%) was used as a precursor for the ceria shell layer. 4-Nitrophenol (Fluka) was used as the organic pollutant, and sodium borohydrate (NaBH4, 99%) was used as the reducing agent. All materials, except 4-nitrophenol, were purchased from Sigma-Aldrich and used without further purification.

2.2 Synthesis of the core–shell nanoparticles

Firstly, silica cores were prepared through the hydrolysis–condensation reaction of silicic acid in the presence of hyperbranched poly(ethylene)imine [31]. To create the core–shell structure, the silica cores were calcined at 800 °C for 3 h under airflow and then they were redispersed in water.

Then, silicas’ surface was modified with HBPEI (SiO2/ΗΒΡΕΙ) and the cores were redispersed in water again. Following that, cerium (III) nitrate hexahydrate (in a percentage of 25, 33 or 50 w/w on silica amount) was added and the sol was kept under stirring for approximately 30 min. Finally, an appropriate amount of 1:10 w/v HBPEI/water sol was added dropwise in order to increase the pH value up to 10.5. The resulting suspension was kept under stirring for 24 h in order for the reactions to be completed. This was followed by the centrifugation and washing with water steps, drying with the freeze drying (FD) method and finally calcination at 600° C for 2 h. The as-obtained samples are described in Table 1.

Table 1 Codes and w/w ratios of the precursors for the final synthesized materials

2.3 Characterization techniques

The morphology of the core–shell structures as well as of pure SiO2 and CeO2 was examined via transmission electron microscopy (TEM) with a high-resolution transmission electron microscope (HRTEM, JEOL 2100 HR) which is equipped with a JED-2300T energy-dispersive X-ray spectrometer and a digital scanning image observation device (STEM, JEOL EM 24511 SIOD). Scanning electron microscopy (SEM) images were recorded with JEOL 6380LV. X-ray diffraction (XRD) patterns were obtained on a diffractometer (Bruker D8 Focus) with nickel-filtered CuKa radiation (k = 1.5406 Å, 40 kV, 40 mA), in the range of 10–80°. FTIR spectroscopy was also conducted using a Nicolet Magna-IR spectrometer 550. N2 adsorption/desorption measurements were taken using a Micromeritics ASAP 2000 apparatus, and the surface area was calculated by the BET equation. A UV–Vis spectrophotometer (Hitachi U-3010) was used in order to obtain diffuse reflectance spectra.

2.3.1 4NP reduction catalysis

The catalytic properties of the material with the optimum core–shell structure and of pure CeO2 were examined through the reduction of 4-nitrophenol (Fluka) to 4-aminophenol in the presence of NaBH4 (sodium borohydrate 99%, Sigma-Aldrich) as the reductant under room temperature. In aqueous solutions of 4NP in a concentration of 4 ppm, 1.6 g/l NaBH4 and 0.2 g/l of catalyst were added. The progress of the reaction was monitored using UV–Vis spectroscopy (UV–Vis Spectrometer, Cary 100). The samples were filtered and measured in the range of 200–800 nm at different times.

3 Results and discussion

3.1 Silica–ceria core–shell nanoparticles

The preparation of core–shell nanoparticles is a very difficult process as various difficulties are encountered, namely the agglomeration of the cores, the tendency to create separate particles from the coating material instead of coating the cores, the incomplete coating and the difficult control of the reaction degree [5].

Firstly, an attempt was made for biomimetic synthesis, where the cerium precursor was added to the silica nanospheres without being calcined and chemically surface-modified. However, inhomogeneities in the final coating as well as the creation of independent ceria nanoparticles were observed (Fig. S1). A pH study was also conducted, using phosphate buffer at various pH values but without leading to the desired result (Fig. S2). For the above reasons, the nanospheres were calcined in order to remove any free amino groups from the surface and then they were subjected to surface chemical modification using HBPEI. In order to simplify the process as well as to minimize the use of harmful substances, the process was carried out in an aqueous medium and at ambient temperature. In similar studies [26, 27, 32, 33], part or the whole reaction was carried out at higher temperature (70–100 °C) as well as with the use of toxic agents such as nitric acid or organic solvents.

The development of the shell was carried out through precipitation on the surface of silica which bears free amino groups. Initially, the cerium ions are bound by the amino groups of HBPEI present on the silica surface. With the addition of the polymer solution, a gradual increase in pH was observed accompanied by a color change. It has been reported that the formation of CeO2 from nitrate baths proceeds through intermediate steps that involve at first the formation of unstable metal–polymer complexes followed at intermediate pH values by the precipitation of the white trivalent cerium hydroxide. With the further increase in pH, the Ce(III) ion starts being oxidized into Ce(IV) forming again stable complexes with the polymer amine groups and thus being redissolved. Finally, at the elevated pH of 10.5 hydration and dehydration reactions follow leading sequentially to the formation of intermediate Ce(IV) species up to the final cubic nano-CeO2 that finally forms after a 24-h stirring period (Fig. 1) [34,35,36,37].

Fig. 1
figure 1

The formation process for the ceria-coated silica spheres

3.2 Morphology of silica cores, cerium oxide and silica–ceria nanoparticles

In order to confirm the synthesis of ceria shell on the silica surface, HRTEM analysis was performed. Figure 2a, b presents the SEM and TEM images of silica cores. It can be seen that the bare silica particles have a smooth surface with a mean size at around 300 nm. Α HRTEM image of the microstructure of pure cerium dioxide is presented in Fig. 2d. The material consists of uniform, almost cubic-shaped grains. The average grain size, calculated using the Image Pro program, was estimated at around 10–20 nm. Further, the selected area electron diffraction pattern (Fig. 2e) in combination with the results of X-ray diffraction analysis proved that the final material is nanocrystalline.

Fig. 2
figure 2

SEM image (a), TEM image (b) and SAED pattern (c) of bare SiO2 cores. TEM images (d, f) and SAED pattern (e) of CeO2 after calcination at 600 °C

Figure 3 shows the HRTEM micrographs of the ceria-coated silica particles prepared using 25, 33 and 50% w/w cerium nitrate on the silica cores’ amount. From the TEM images, it is clear that a distinctive crystalline (as shown by SAED analysis, Fig. 3e) ceria shell was formed on silica surface, the quality of which, however, depends on ceria precursor quantity employed. More specifically, in the case of using the low precursor quantity of 25% w/w on silica (Fig. 3a), the coating is successful with all the ceria materials being formed onto silica surface. However, the surface coverage is incomplete presenting many gaps. Instead, when the cerium nitrate quantity was increased to 33% w/w (Fig. 3b, c), the silica particles are completely coated with a uniform and very successful ceria layer. Finally, in the case of using 50% w/w cerium nitrate/silica cores, an inhomogeneous and thick ceria coating along with the formation of separate ceria nanoparticles is observed (Fig. 3d).

Fig. 3
figure 3

TEM image (a) of ScP25%CeP, TEM images (b, c) and SAED pattern (e) of ScP33%CeP, TEM image (d) of ScP50%CeP

Comparing the pure ceria material to that formed on silica cores, it should be pointed out that the ceria of the shell material presents a significantly reduced particle size. In particular, after calcination at 600 °C, the size of cerium oxide in the core–shell material varies between 2 and 6 nm, whereas that of the pure cerium oxide was 10–20 nm. The significantly reduced size was achieved possibly due to two key reasons. Firstly, the presence of the SiO2/HBPEI cores facilitates the nucleation rate. Secondly, the good dispersion of the CeO2 particles on the silica cores leads to the reduction in the grain growth during the final calcination process.

In addition, the percentage of CeO2 in each core–shell structure was calculated using the EDS data and Eq. 1:

$$\% {\text{CeO}}_{2} = \frac{{{\text{gr}}_{{{\text{CeO}}_{ 2} }} }}{{{\text{gr}}_{{{\text{CeO}}_{2} }} + {\text{gr}}_{{{\text{SiO}}_{2} }} }}*100$$
(1)

With the use of STEM and EDS, a number of measurements for separate core–shell nanoparticles were taken. An average % estimation of the measurements is given in Table 2.

Table 2 % CeO2 content of each silica–ceria core shell structure

3.3 XRD measurements

The phase composition of pure ceria as well as that of silica–ceria core–shell nanoparticles was investigated by XRD. Figure 4 shows the XRD pattern of CeO2 powder after calcination at 600 °C where the peaks of the cubic structure of cerium dioxide are evident [(111), (200), (220), (311), (222), (400), (331) and (420)] (JCPDS No. 34-0394).

Fig. 4
figure 4

XRD pattern of CeO2 after calcination at 600 °C

The XRD patterns of the core–shell materials (Fig. 5), combined with the results of the selective area electron diffraction pattern and the EDS analysis, confirmed the shell creation since the main peaks [(111), (200), (220), (311), (222), (400), (331) and (420)] of the cubic structure of cerium dioxide (JCPDS No. 34-0394) are presented. As shown in Fig. 5, increasing the weight ratio of Ce(NO3)3·6H2O/SiO2/ΗΒΡΕΙ from 0.25:1 to 0.33:1 and finally to 0.5:1, the intensity of ceria peaks is also increased. Further, these XRD patterns present a broad peak at about 22 degrees, which corresponds to the amorphous structure of silica [38, 39].

Fig. 5
figure 5

XRD patterns of the core–shell structures: (a) ScP25%CeP, (b) ScP33%CeP and (c) ScP50%CeP

3.4 FTIR analysis

Further analysis was performed using Fourier transform (FTIR) infrared spectroscopy. As shown in Fig. 6, no change is detected by increasing the percentage of cerium nitrate in the final material. The peaks at 1066, 797 and 443 cm−1 are attributed to stretching and bending vibrations of Si–O–Si bonds [31, 40, 41], indicating the condensation of silanols to Si–O–Si bonds and the formation of the silica network. The presence of CeO2 is confirmed by the peaks at around 405 cm−1 which are attributed to Ce–O bonds [38, 42, 43].

Fig. 6
figure 6

FTIR spectra of the silica nanospheres (i), cerium oxide (iv) and core–shell structures ScP25%CeP (ii), ScP33%CeP (iii) and ScP50%CeP (v)

3.5 N2 adsorption analysis

Based on the physicochemical analysis results that have been described above, the material with the optimal coating, i.e., ScP33%CeP, was selected for further analysis. Measurements of the specific surface area of the core–shell material as well as of the pure materials that formed the core and the shell were taken. The N2 sorption–desorption isotherms of CeO2, SiO2 and ScP33%CeP are shown in Fig. 7. As it is evident from the IUPAC classification [44], cerium dioxide is a mesoporous material (type IV), and it has a hysteresis loop corresponding to H2 class [45]. The specific surface area of the cerium oxide was recorded at 48.7 m2/g. SiO2 cores and ScP33%CeP core–shell material, as shown in Fig. 7, are non-porous materials with a low specific surface area of 25.3 and 27.5 m2/g, respectively. Therefore, it can be concluded that by preparing a new core–shell structure, the specific surface area of the final material is smaller compared to that of pure CeO2. This is expected given the small amount and the low thickness of the material deposited on the surface of much larger silica cores having a surface area of ~ 25 m2/g.

Fig. 7
figure 7

Adsorption and desorption isotherms of ScP33%CeP, CeO2 and SiO2

3.6 UV–Vis measurements

Diffuse reflectance spectroscopy was performed in order to estimate the band gaps of the ScP33%CeP, CeO2 and SiO2. A plot of the absorption function [F(R) × E]1/2 versus E (eV) is depicted in Fig. 8b, where F(R) = (1 − R)2/2R is the Kubelka–Munk function expressed through the measured diffuse reflectance R of the materials and E is the electromagnetic wave energy. The band gaps were estimated by extrapolating the linear section of the absorption function and measuring the value at its intersection with the energy axis, e.g., at zero absorption [46,47,48].

Fig. 8
figure 8

a Reflectance spectra of CeO2 (i), ScP33%CeP (ii) and SiO2 (iii), b absorption function [f(R) × E]1/2 versus energy E for CeO2 (i) and ScP33%CeP (ii)

The absorption spectrum of cerium dioxide, as well as the value of the energy band gap Ebg = 2.53 eV, coincides with the literature data on doped CeO2 [49] as the usual Ebg value for ceria is 3.1–3.3 eV [50,51,52]. Therefore, it can be concluded that the experimental process described above leads to the preparation of CeO2 with a sufficiently reduced energy band gap. As it is evident and consistent with the leterature [53, 54], the reflection spectrum for pure SiO2, given in Fig. 8a, exhibits negligible absorption only in the ultraviolet region. Instead, the pure CeO2 as well as the core–shell material ScP33%CeP absorbs in both the visible and ultraviolent regions. The measurement of the energy gap of SiO2 material due to the very low absorption was not reliable, but in the literature it is given at around 9 eV [55]. Comparing the spectra and the energy gap values of pure CeO2 and the core–shell material ScP33%CeP, a change in absorbance after the creation of the new core–shell structure is shown. In particular, the absorption spectrum of the core–shell structure is shifted to smaller wavelengths, while it exhibits a higher reflectance (Fig. 8a). The core–shell material has a higher energy gap value of 3.01 eV than that of pure CeO2, which is could be attributed to the smaller particle size of CeO2 [56] in the core–shell structure, as it was observed by TEM analysis.

4 Catalytic activity of ScP33%CeP and CeO2 for 4NP reduction

The industrial manufacture of pharmaceuticals, agrochemicals and dyes produces nitro-aromatic compounds which are harmful to the environment and eventually to humans. These compounds can be reduced into their amine form, which is non-toxic, with the appropriate catalytic process. The nanoreactor with the optimum core–shell structure, i.e., ScP33%CeP as well as the pure ceria, was examined for the catalytic reduction of 4-nitrophenol to 4-aminophenol in the presence of NaBH4 as the reducing agent at ambient conditions. As shown in Fig. 9, the absorption peak at 317 nm indicates the presence of the organic pollutant, whereas the reduction of 4NP to 4-nitrophenolate ion, after the addition of the reducing agent, is indicated by the peak shift to 400 nm. This peak is gradually decreased over the time and a new peak at ~ 300 nm appeared indicating the reduction of 4NP to 4AP [57,58,59].

Fig. 9
figure 9

UV–Vis spectra of 4NP before and after adding NaBH4

The concentration of NaBH4 exceeded the concentration of 4-NP. The high concentration of the reducing agent results in high pH values and as a consequence in deceleration of the degradation of the borohydride ions. In addition, the oxidation of the 4-aminophenol product is prevented due to the presence of free hydrogen from the borohydride ions which purged out the air. After the addition of the catalyst, NaBH4 and 4-NP are absorbed onto the catalysts surface and the reduction process is started by transferring electrons from the BH4 donor to the 4NP acceptor. At this point, it must be pointed out that in the absence of catalyst the reaction was not taken place even after 24 h [39, 57,58,59]. Consequently, the presence of the catalytic nanoreactor is crucial.

Figure 10a, b presents UV–Vis spectra of ScP33%CeP and CeO2. As shown in Fig. 10c, ScP33%CeP achieved 47.8% conversion of the organic pollutant 4-NP whereas pure CeO2 achieved only 8.5% conversion. Therefore, it can be concluded that the new structure favors the catalytic performance as it is far exceeding that of pure cerium dioxide. The higher rate of conversion of the core–shell material can be attributed to the smallest particle size and the good dispersion of CeO2 on the silicas surface so no aggregates are present and thus more reactive sites are offered [60]. Significant benefit of core–shell structures, as the ScP33%CeP, is their bigger particle size, which simplifies their separation and collection by a heterogeneous system.

Fig. 10
figure 10

Time versus UV–Vis absorption spectra for the catalytic reduction of 4NP by NaBH4 in the presence of ScP33%CeP (a) and CeO2 (b), 4NP concentration versus time plots of 4NP reduction by NaBH4 in the presence of ScP33%CeP and CeO2 (c), comparative plots of ln(A/A0) for ScP33%CeP and CeO2 toward the reduction of 4NP (d)

The catalytic reduction of 4NP to 4-AP in excess of NaBH4 is commonly described by a first-order rate kinetic equation with respect to nitrophenolate ion concentration. Hence, the apparent rate constant (k) for each material was estimated from the slope of the straight lines in Fig. 10d [57,58,59]. The degradation rate for the core–shell structure was estimated at 0.003 min−1, whereas that for the pure cerium oxide was much lower (0.0002 min−1) in accordance with the conversion rates discussed above. Furthermore, in both cases, an induction time (t0) for the first 15 min is observed, indicating the surface atoms’ of the nanoparticle reorganization in order to be activated before the beginning of the reaction [61].

Further studies concerning the amount of the catalyst were conducted for the core–shell structure and pure ceria, and the results are presented in Table 3. As it can be observed, with the increase in the catalyst amount, the % conversion and as a consequence the rate constant k are reduced in both cases. At this point, it must be pointed out that in the case of pure CeO2, the rate constants for catalyst concentrations 0.4 and 0.6 g/l were not estimated due to the very low % conversion of 4NP to 4-AP. The decrease in the % conversion is possibly ascribed to the aggregation of the particles due to the higher amount of catalyst and hence the reduction in the total active surface area of the material. Generally, in a heterogeneous system, the dispersion of the catalyst in the aqueous system may influence the rate of the reaction [62, 63].

Table 3 % Conversion and the respective rate constants of the catalytic systems

From the above results, it is evident that the development of a new SiO2@CeO2 core–shell structure promotes the catalytic activity of ceria. Using only a small amount of CeO2 as a shell on silica cores, the catalytic activity of the final material is five times higher than that of pure CeO2.

5 Conclusions

Uniform SiO2@CeO2 core–shell nanoparticles were prepared through an environmentally friendly chemical precipitation method. The whole process was carried out with the objective to establish an economical and environmentally friendly process, so the temperature used was that of the environment, while only distilled water was employed as solvent. Important parameters proved the calcination and surface modification of silica cores by HBPEI, the w/w Ce(NO3)3·6H2O/SiO2/ΗΒΡΕΙ ratio employed (with the optimum one being at 0.33:1) and the CeO2 precipitation pH which was adjusted again by HBPEI. By this way, homogeneous, spherical SiO2@CeO2 core–shell nanoparticles of a mean size at 300 nm comprising a uniform ceria layer were developed. The CeO2 shell on top of SiO2 cores had a very fine nanostructure consisted of nanocrystals of the cubic CeO2 with sizes ranging between 2 and 6 nm. Measurements on pure ceria particles (not on top of silica cores) developed using the same procedure revealed significantly higher particle size, about 10–20 nm, compared to that of ceria particles in the core–shell material. The N2 porosimetry study showed that the material with the optimum coating had a reduced specific surface area compared to pure cerium dioxide, i.e., 27.5 m2/g versus 48.7 m2/g due to the much smaller amount of ceria particles impacted on much larger silica cores. The catalytic activity of the core–shell structure over 4-nitrophenol was found at 47.8% conversion, while that of the pure CeO2 was 8.5%, indicating the high efficacy of the core–shell structure and the good dispersion of CeO2 particles on the silica cores.