Property Tuning through Fine Size Control and Hierarchical Nanostructuring of Metal Oxide Nanoparticles Supported in Porous Matrices: a Review

Abstract

In the nanoscale, matter presents different properties compared to its bulk counterparts, owing to the considerable increase of the surface/bulk ratio, as well as the occurrence of quantum confinement effects. One method to fine control the growth of nanoparticles, their size and nanostructure is to use metalorganic deposition combined with a porous matrix as host, where pores act as nanoreactors for nanoparticle growth, resulting in chemically integrated systems. In this review, we look at the history and achievements of the chemical method of impregnation-decomposition cycles for the precise size control, property tuning and tailored synthesis of metal oxide nanoparticles. We give an overview of the various oxide nanoparticles and nanomaterials developed over the years, and how the method of impregnation-decomposition cycles allowed the synthesis of pure, doped and core-shell oxide nanoparticles, as well as the tuning of their properties through the combined fine control of nanoparticle size, nanostructure, and composition.

Keywords: porous materials; nanoparticles; impregnation-decomposition cycles; core-shell; integrated chemical systems

1. Introduction

To this day nanomaterials remain in the highlight of numerous technological advances. Comprised of matter with an average size in the range of 1-100 nm, these small materials have found applications in several facets of science and life, for example in catalysis,¹,² biomedicine,³,⁴ sensors,⁵ textiles,⁶ energy⁷,⁸,⁹ and environmental applications.¹⁰,¹¹ Their wide range of applications stem from the special properties presented by matter in the nanoscale. Compared to their bulk counterparts, nanomaterials have a much larger surface-to-bulk ratio, which means that surface effects become more pronounced. For example, in catalysis, the much higher surface area and/or the higher density of relevant surface defects are advantageous. Other significant effect, observed especially for nanoparticles (NP) below 10 nm, are size-related confinement effects. For semiconductor NP an increase in the band gap is observed, similarly to how plasmons suffer confinement in the case of metallic NP. Both effects may be exploited, for example, to tune the optical absorption or emission profile of the final material for applications in sensing, light-emitting diodes (LED) or photocatalysis. The control of NP size to obtain different light emission profiles was also the research theme of the Chemistry Nobel Prize laureates of 2023, M. G. Bawendi, L. E. Brus and A. Yekimov.¹² Such small nanomaterials, however, typically require some form of stabilization, since their increased and highly energetic surface area is typically unstable and may result in their growth through coalescence, Ostwald’s ripening or other growth mechanisms, effectively changing the materials’ properties in the process.

On the other hand, while many nanomaterials present interesting properties on their own, the fast pace at which technology advances makes necessary the use of ever more complex materials to fulfil increasingly sophisticated requirements for nowadays’ applications. This means that, in most cases, different compounds and materials must be combined to obtain the specific properties required. This aspect, together with the necessity of NP stabilization, infer that a system consisting of NP and stabilizers could be one such complex material. Going a step further, one may also combine NP with functional stabilizers, where both parts of the system have their own unique and equally relevant properties for a given application, as in the case of NP supported on a functional host for catalysis. When each of the individual species, compounds or materials involved perform specific relevant roles for a given application, the complete system may be considered a chemically integrated system.¹³,¹⁴,¹⁵,¹⁶,¹⁷ New properties may arise of the specific combinations utilized, further enhancing the integration aspect of the system. As an example, a porous material with a functional surface may be used to grow and stabilize NP in its pores, while its pores may function as nanoreactors for localized catalysis performed by NP hosted inside, an aspect that will be further elaborated along the present review.

In this review, we present the synthesis of chemically integrated systems composed by metal oxide NP grown inside porous hosts using the method of impregnation-decomposition cycles (IDC) and the development of this strategy over the past two decades. This method was developed as a layer-by-layer growth strategy to construct such systems based on the combination of a porous material as host matrix with the use of single-source precursors for the bottom-up growth of nanoparticles. The host’s pore size, pore network and strong chemical bonds with the grown oxide NP both limit the nanoparticles’ growth beyond a certain threshold (given by the pore sizes) and prevent their growth through mechanisms such as coalescence, sintering or Ostwald’s ripening. On the other hand, the layered growth allowed control over the amount of metal oxide introduced, the size of the NP, as well as their crystalline phase, morphology and hierarchical nanostructure, allowing to fine tune the oxides’ properties by exploiting effects occurring at the nanoscale.

2. The IDC Method: a Layer-by-Layer Growth Strategy

The method of impregnation-decomposition cycles (IDC) was the first application of the typical metal-organic decomposition (MOD) method in confined spaces, by using a porous host as an impregnation matrix.¹⁸ A cycle consists of an impregnation step, where the porous host as either slabs or as powder is immersed in a solution of a specific metalorganic precursor for a set amount of time (typically 8 h) and then washed with a compatible solvent to ensure that the precursor molecules are only inside the porous structure, followed by a decomposition step, where the porous material imbedded with precursor are submitted to thermal decomposition to obtain the desired metal oxide. This process, or cycle, may be repeated at will (Figure 1), with the possibility of changing the precursors used, until the desired nanoparticle structure or size has been reached, or until the porous host can no longer reasonably intake more reagent, due to filled pores and obstructed pore necks, as well as incomplete decomposition resulting of these pore constrictions. This extreme case was observed for a test performed with continuous IDC of ZnO until the impregnation limit of commercially available porous Vycor glass (PVG 7930, or just PVG) was reached at 29 IDC.²⁰ Another important factor is that, to keep the thermal history of the samples equal, all samples undergo all thermal treatments applied. This means that all samples are thermally treated on every decomposition step, regardless if they undergo impregnation or not, i.e., a sample with 3 IDC and a sample with 10 IDC undergo 10 thermal treatments (from the decomposition steps).

Figure 1.
Schematic representation of the impregnation-decomposition cycles (IDC) method, using PVG for the synthesis of either pure TiO₂ or TiO₂@CeO₂ core@shell NP (adapted from reference ¹⁹).

The reagents used to obtain metal oxides are single-source metalorganic precursors, where both the metal and the oxide are provided by the reagent, resulting in the desired metal oxide compound. This reagent may also be combined with smaller amounts (%) of another single-source precursor as a source of a desired dopant for one or more IDC layers. Examples of such reagents are M-2-ethylhexanoates (M^x+(hex)_x) and M-acetylacetonates (M^x+(acac)_x).

The next step was then the choice of porous host. From the wide range of materials available at the time, potential candidates had to be chosen based on what the final chemically integrated systems were supposed to achieve: restriction of nanoparticle growth through the porous structure, stabilization of said NP, and chemical and thermal stability against the grown metal oxide NP, as well as to withstand their synthesis and typical application requirements. The first porous host tested was a α-NbPO₅ porous glass-ceramic,¹⁸ which presented good chemical and thermal stability, but a large average pore size of ca. 130 nm. The glass-ceramic presented a relatively complex synthesis method, and because of its pore size also resulted in larger NP.¹⁸ The improved properties and either better availability or ease of synthesis, however, quickly brought mesoporous silica glasses such as the PVG and the SBA-15 silica (Santa Barbara Amorphous, or just SBA-15) to the spotlight. With average pore diameter below 30 nm, high chemical and thermal stability, high porosity and high surface reactivity, these porous hosts were promising hosts for the synthesis and stabilization of small metal oxide NP.

The remaining step is the activation of the surface of said porous hosts, optimizing the impregnation of metalorganic compounds. For porous glasses, the composition plays a role as to how these materials should be activated. In the case of the porous silica glass PVG, the most used host for IDC, this is done through two simple steps: the acidic treatment of the PVG slabs with HCl, followed by a treatment with acetone, after which they are stored under vacuum and/or drying conditions before being used. This ensures the availability of active Si–OH sites at the surface, which readily react with the single-source precursors used for metal oxide synthesis. This provides a chemical bond anchoring point for the NP, ensuring their stability within the porous structure and preventing NP coalescence. Furthermore, their thorough distribution throughout the host matrix also prevents other aggregation mechanisms like Ostwalds’ ripening. Since such coalescence mechanisms are prevented, the temperature is constant and the concentration of the solution remains approximately constant, NP growth is then governed by the minimization of the high surface/interface free energy of the formed NP,²¹ resulting in the desired controlled growth with increasing number of IDC.²²

3. Typical Characterization Methods

For almost all materials synthesized using IDC a base combination of characterization techniques was used for their structural analysis, and the specifics regarding such IDC-based systems are as follows:

3.1. Cumulative mass gain

The mass gain of the material is linear with the number of IDC, until the impregnation limit of the porous host is reached, corroborating with the controlled growth promoted by the method (Figure 2I).

Figure 2.
Examples of the typical characterization techniques applied to materials synthesized using the IDC method: (I) cumulative mass gain obtained for the growth of CeO₂@TiO₂ core@shell NP inside PVG (reproduced with permission from reference ¹⁹, Copyright 2011, American Chemical Society); (II) N₂ sorption-desorption obtained for PVG/5CeO₂@xTiO₂ (reproduced with permission from reference ¹⁹, Copyright 2011, American Chemical Society); (III) Raman spectroscopy measurements of PVG/5TiO₂@xMoO₃ (reproduced from reference ²³, Copyright 2010, with permission from Elsevier); (IV) diffuse reflectance spectroscopy (DRS) of PVG/xTiO₂ and respective applied Kubelka-Munk F(R) applied in the inset (reproduced from reference ²⁴, Copyright 2014, with permission from Elsevier); (V) X-ray diffraction (XRD) patterns obtained for free Fe₂O₃ (higher intensity and peak resolution) and PVG/10Fe₂O₃ (reproduced from reference ²⁵, Copyright 2009, with permission from AIP Publishing); (VI) TEM obtained for PVG/7CeO₂, where the inset shows the nanoparticle size distribution obtained for all TEM images collected (reproduced from reference ²⁶, Copyright 2007, with permission from Elsevier).

3.2. N₂ sorption-desorption

Applied to evaluate the pore structure of the host, showing a decrease in available surface area and pore volume with an increasing number of IDC. No change is observed for the curve profiles, proving there is no appreciable clogging of the pore space over as many as 10 IDC (Figure 2II).

3.3. Raman spectroscopy (Raman)

A very powerful analysis technique, given it can measure typical metal oxide vibrational bands as well as those of the porous hosts, where shifts in both wavelength and intensity are meaningful when evaluating the material’s properties. This becomes especially important when evaluating which crystalline phases may be present in the structure (this may be either difficult or unreliable through X-ray diffraction or high resolution transmission electron microscopy alone), as well as for the detection and analysis of core@shell structures (Figure 2III), or even for the determination of nanocrystal size using the phonon confinement model.

3.4. Diffuse reflectance spectroscopy in the UV-Vis (DRS)

Considering the porous hosts used present either no absorption or low absorption bands in the low UV-Vis region (200-800 nm), the Kubelka-Munk approximation may be applied on DRS spectra to obtain the metal oxides’ characteristic absorption bands within this range, including their band gap (Figure 2IV). Such measurements are indispensable to track the growth of semiconductor oxide NP with the number of IDC, observed by the shift of its band gap due to quantum confinement effects.

3.5. X-ray diffraction (XRD)

The typical XRD profile presents the characteristic glass/amorphous halo of the porous host and generally very broad and low intensity diffraction peaks characteristic of the crystal structure(s) present, and in some cases may only be seen when using a high intensity XRD source. This is a consequence of the very small size of the particles (< 10 nm) and their thorough distribution throughout the porous structure of the host (Figure 2V). Scherrer’s law may be applied to obtain an estimate of the average crystallite size, but given the broadening effect the best and most reliable option for such measurements is through electron microscopy.

3.6. Electron microscopy

Transmission (TEM), high-resolution transmission (HRTEM) and either dark-field TEM or high-angle annular dark-field scanning TEM (HAADF-STEM) are techniques that allow the direct visualization of the NP for evaluation of their morphology, their spread throughout the sample and their size distribution. Through HRTEM or HAADF-STEM one can evaluate the NP crystallinity and possible crystal structure(s) through the diffraction fringes observed. From a series of images (TEM, STEM or HRTEM, depending on the size and visibility of the NP), counting normally at least 200 NP, one can obtain the average size distribution of the synthesized NP (Figure 2VI).

Other techniques such as photoluminescence and magnetic measurements may also contain important structural information, and their individual contributions, both for structural analysis as well as for application tests, will be discussed along this review.

4. Overview of the Metal Oxide Nanoparticle Systems Developed with IDC

A range of different metal oxides and combinations between them was explored, with varying application purposes. From single metal oxides such as TiO₂ and CeO₂ for catalysis, to complex multilayered doped NP for light emitting diode (LED) applications, all systems were planned and tailored seeking to provide a solution to a current challenge at the time of their development. Perhaps the most important properties obtained were the stabilization of very small (< 10 nm) crystalline metal oxide NP inside a porous host, providing micro-chambers, for example, for catalysis, the control of the obtained crystalline phase as well as the fine control of the NP size with varying IDC, both optimizing their chemical activity. In Figure 3, a detailed summary of the metal oxide systems developed using the IDC method is presented, organized according to their composition, main purpose of the work and a short summary of what each work achieved.

Figure 3.
List of the metal oxide systems developed using the IDC related to their composition and/or structure, porous host used, the main purpose of the work and a short description of what each one achieved.

In the following sections each one of the main system categories: single oxide, double-oxide and core@multishell, is discussed in detail.

4.1. Single oxide systems

At first the IDC method was a strategy developed to obtain nanosized metal-oxide inside porous matrices with control over their size and structure, generating materials as chemically integrated systems, where the materials’ properties depend not only on the individual composing species, but also on how they interact with each other. The nanoparticles are grown inside the porous host, thoroughly distributed within the porous structure, and they present high crystallinity. The average nanoparticle size is directly proportional to the number of IDC (the higher the number of IDC, the larger the nanoparticles), and presented in general values below 10 nm. A few examples are shown in the TEM images in Figure 4.

Figure 4.
TEM bright field images of γ-Fe₂O₃ obtained with (I) 2 IDC and (II) 8 IDC. The insets show the histograms obtained for the respective particle size distributions (reproduced from reference ²⁵, Copyright 2009, with permission from AIP Publishing); (III) TEM dark field image of NiO NP after 7 IDC (reproduced with permission from reference ²⁹, Copyright 2010, American Chemical Society); (IV) HRTEM images of sample PVG/5CeO₂@15TiO₂ with respective interplanar distances d₁₀₁ = 3.52 Å (anatase TiO₂) and d₁₁₁ = 3.13 Å, d₂₀₀ = 2.82 Å (cubic CeO₂) (reproduced with permission from reference ¹⁹, Copyright 2011, American Chemical Society); HRTEM images of samples (V) PVG/10ZnO and (VI) PVG/ZEZ 5/2/10, where the insets show the observed diffraction fringes characteristic of wurtzite ZnO (reproduced with permission from reference ²⁰, Copyright 2015, American Chemical Society); (VII) dark field image and (VIII) HRTEM image of SBA-15/7CeO₂ (reproduced from reference ³², Copyright 2011, with permission from Royal Society of Chemistry).

The first NP synthesized using this strategy were of TiO₂, an extensively used compound in the fields of catalysis and elimination of environmental pollutants,³⁶,³⁷,³⁸,³⁹,⁴⁰,⁴¹ where the control over size, as well as the stabilization of the TiO₂ anatase crystalline phase, which is normally unstable but much more catalytically active, was made possible.¹⁸ This synthesis was performed in site inside an α-NbPO₅ porous glass-ceramic as porous host, previously studied in the group. This first study¹⁸ evaluated all the relevant aspects involved with the method: the impregnation of the metalorganic precursor was confirmed via DRS, and the appropriate decomposition temperature was determined through TGA (thermogravimetric analysis), since the nanoporous environment imposes different conditions on the decomposition process, the process begins more easily since the precursor is well distributed within the matrix, but takes longer to finish because of the pore space restriction, when compared to the free precursor. It was noted that, opposed to the rutile TiO₂ obtained for the free precursor, decomposition performed at up to 750 °C within the porous host resulted in the formation and stabilization of anatase TiO₂ NP, which were well distributed within the porous structure of the host. The cumulative mass gain obtained was linear, which, combined with the analysis of the typical Raman TiO₂ Eg bands for the samples synthesized with increasing number of IDC, have shown that the amount of metal oxide, as well as NP growth are controllable through the number of IDC, and that no sintering or NP coalescence occurs. Furthermore, the porous structure was preserved, and no solid-state reaction was observed between TiO₂ and the α-NbPO₅ porous host, given no such characteristic bands were observed through XRD or Raman spectroscopy.

The investigation of such chemically integrated systems then continued with the work by Mazali et al.,²⁷ this time aiming to obtain quantum confined NP using not only the same α-NbPO₅ matrix, but also the then commercially available porous Vycor glass 7930 (PVG), which presented higher chemical and thermal resistances, higher porosity and smaller average pore size (< 20 nm). Both SnO₂ and CdS were studied, and we will focus on the results obtained for the former. The crystallite size of the NP was estimated through the Scherrer equation applied to the XRD data, and was smaller for NP synthesized inside PVG compared to α-NbPO₅, as would be expected considering the former has a pore size of 4-20 nm, much smaller than the estimated 130 nm of the latter. Furthermore, PVG also presented linear mass increase with the number of IDC, allowing tailored control of the NP size, where no coalescence effects were also observed, owing to the size limitation of the pores in combination with the thorough distribution of the NP over the porous matrix. Lastly, no solid-state reactions were observed. This consolidated PVG as a highly interesting host for such chemically integrated systems, allowing the controlled growth of very small metal oxide NP within nanopores that could present a double function of nanoreactors for catalysis or in gas sensing.

The synthesis of TiO₂ NP was then revisited, this time using PVG as porous host.²² Once more, anatase TiO₂ was stabilized inside the pores of the host, but this time with a much smaller average size of about 5 nm for 3 IDC. The XRD profile now also presented the glass halo characteristic of PVG together with a broad, lower intensity diffraction peak associated with anatase TiO₂, characteristic of the formation of very small (< 10 nm) NP within a porous host, as mentioned before in this review. Raman spectroscopy analysis of the PVG/xTiO₂ samples corroborate with the phonon confinement model, where the intensity and frequency shift of the TiO₂ E_2g mode is a function of the NP size, as can be seen in Figure 5. Nanocrystal growth (related to the Raman band profile) was a linear function of the mass increase alone via IDC, for all samples had the same thermal treatment history regardless of how many impregnation cycles they underwent. This once more proved that the IDC method enabled the tailoring of the NP size, crystalline phase and morphology, and thus of its properties, now within a quantum confinement regime, by using an appropriate porous host.

Figure 5.
Size-dependent Raman scattering profile of TiO₂ E_2g (reproduced from reference ²², Copyright 2006, with permission from Springer Nature). The solid points are experimental data obtained through 10 IDC. The solid, dotted, and hashed lines were obtained by plotting the phonon confinement model equation first formulated by Richter et al.⁴² for nanocrystal sizes of 20, 10 and 5 nm, respectively. The inset shows the average observed NP diameter obtained through TEM for 3 IDC, and through experimental data fit using the same equation for 5, 7 and 10 IDCs.

The next metal oxide synthesized inside PVG was CeO₂.²⁶ CeO₂ is a very interesting material especially for catalytic applications,⁴³,⁴⁴,⁴⁵,⁴⁶,⁴⁷ where the higher the materials’ surface area and better distribution and availability for reacting, the better. It is a natural conclusion then to synthesize stable and very small CeO₂ NP over large surfaces, especially in a porous host support like PVG, which could in principle be directly used without the need for additional silica added as bedding material. Mazali et al.²⁶ applied the Scherrer’s equation to the data obtained through the (220) diffraction peak of CeO₂ for the PVG/xCeO₂ systems, obtaining an estimated crystallite size between 3.7 nm for 3 IDC and 5.6 nm for 10 IDC. TEM images of the PVG/7CeO₂ sample have shown an average NP size of about 4 nm, with a distribution between 2 and 6 nm observed. The Raman spectral signature of CeO₂, the T_2g mode at around 460 cm^–1 undergoes redshift and a decrease in intensity with decreasing IDC. This corroborates with the observed increasing crystallite size with increasing IDC number. A similar fit of the phonon confinement model was performed for CeO₂, where 7 IDC has shown an excellent fit with the theoretical model, and diameters of 3 and 10 nm were obtained for 3 and 10 IDC, respectively. Furthermore, plotting the Raman frequency as a function of the reciprocal nanocrystal size (Figure 6), even combined with results previously obtained in the literature by Spanier et al.⁴⁸ has resulted in a quadratic dependence of frequency on 1/d, as can be seen in Figure 6. The discrepancies observed between the estimated nanocrystal sizes by TEM, Raman and XRD are a consequence of the different phenomena involved in each. In the case of TEM, what is measured is highly dependent on the resolution, and corresponds to the size of the NP, and not of the crystallites, as is the case for Raman and XRD. The crystallite sizes obtained through these two techniques is also highly dependent on the resolution of the diffractogram or spectrum in question. These in turn are sensitive to the formation and presence of defects or strain in the crystal lattice, ultimately affecting the observed signals and thus the estimated crystallite values obtained.

Figure 6.
Raman frequency as a function of the reciprocal nanocrystal size (1/d) for CeO₂ nanocrystals, obtained from the work by Spanier et al.⁴⁸ and the referenced study²⁶ (reproduced from reference ²⁶, Copyright 2007, with permission from Elsevier).

The synthesis of a PVG/xWO₃ system was also performed, as a part of a larger work seeking the construction of core@shell NP.³⁰ The latter will be discussed in the respective following section, but a few observations about the pure WO₃ system could be made: as with the previous systems, the mass increment was linear with the number of IDC, indicating control over NP size, but presented a mixture of the monoclinic and triclinic WO3. Oliveira²⁸ synthesized other systems such as CuO and Bi₂O₃ in PVG. Both oxides have shown a linear mass increase and, although microscopipc analysis of the samples was not performed, most likely resulted in controlled size of the CuO NP, when one compares the XRD data for PVG/4CuO and PVG/10CuO, for example. Two interesting observations were made for Bi₂O₃, however: first, the stabilized crystalline phase was the normally unstable δ-Bi₂O₃, demonstrating a similar effect of an unexpected phase stabilized in the nanoscale similarly to the observed for TiO₂ NP; second, the cumulative mass gain was almost 4 times higher than for CuO and CeO₂ systems, and with a relatively high decomposition temperature for Bi₂O₃ (600 °C) after 7 IDC the formation of a mixed silicate phase Bi₂SiO₅ was observed. This was the sole example observed for a solid-state reaction involving metal oxide NP grown on silica porous matrices using the IDC method, which in retrospect could be attributed to the high decomposition temperature applied (600 °C), considering the already low melting point of 817 °C for bulk Bi₂O₃.

Another line of research sought the study of magnetic nanoparticles, based on γ-Fe₂O₃²⁵ and on NiO,²⁹ both presenting interesting magnetic properties and effects strongly related to the NP size.⁴⁹,⁵⁰,⁵¹,⁵²,⁵³,⁵⁴,⁵⁵,⁵⁶,⁵⁷ With the size of these materials reduced to the nanoscale regime, surface and finite size effects become prominent and play a pivot role in the nanoparticles’ magnetic behavior. Finite size effects correspond to the reduced number of exchange-coupled spins in the particle, while surface effects are related to the lower symmetry at the surface of the NP, generating surface anisotropy and thus magnetic frustration and spin disorder. Such effects may also be enhanced depending on the magnetic behavior of the NP. In the work by Cangussu et al.,²⁵ γ-Fe₂O₃ NP were synthesized in PVG using the IDC method to study the size effect over the properties of the final integrated chemical system. Similarly, to TiO₂, a different crystalline phase was stabilized for the NP in comparison with the expected α-Fe₂O₃ for the bulk structure, and, like in the previous studies, the NP size could be precisely tuned by the number of IDC used. The obtained γ-Fe₂O₃ NP presented magnetic behavior typical of the expected by well-established non-interacting particle systems, while interacting-particle aspects such as the increasing blocking temperature T_B, the temperature limit above which the system behaves as superparamagnetic, the behavior observed for the NP in question, become more important with an increasing number of IDC as can be seen from the relation between the relaxation time (1/w) at T_B and the inverse T_B in Figure 7. This further corroborates with the well-spread synthesis of the NP within the porous host.

Figure 7.
Plot of log (1/w) vs. 1/T_B for a wide range of measured frequencies of the PVG/x γ-Fe₂O₃ samples, with x = 4, 6, 8 or 10 IDC (reproduced from reference ²⁵, Copyright 2009, with permission from AIP Publishing).

The second work for the synthesis of magnetic metal oxide NP was performed by Carneiro et al.²⁹ with NiO NP in PVG. Once more an increase in the blocking temperature T_B was observed with the number of IDC, as well as a dependance on the magnetic behavior with the number of IDC and thus the size of the nanoparticles. An interesting observation here regarding the magnetic properties measured was the formation of a structure akin to core@shell, composed by an antiferromagnetic ordered core with an uncompensated magnetic moment, coated with magnetically disordered surface clusters. Evidence of this was the intriguing behavior of the field shift hysteresis H_E and the coercivity H_C vs. temperature for an increasing number of IDC (Figure 8).

Figure 8.
Evolution of the values obtained for the field shift of hysteresis (H_E) and coercivity (H_C) with temperature, for samples with 3, 5 and 7 IDC NiO. The lines serve only as guidance (reproduced with permission from reference ²⁹, Copyright 2010, American Chemical Society).

Following a different kind of investigation, this time seeking other means of evaluating the growth of metal oxide NP, Strauss et al.³¹ were the first to do a systematic study evaluating the potential of investigating the controlled growth of < 10 nm SnO₂ oxide NP by using Eu³⁺ as an indirect luminescence probe. By establishing a relationship between this ion’s ⁵D₀ → ⁷F₁:⁵D₀ → ⁷F₂ asymmetric ratio and the number of IDC based on Figure 9, they have shown the possibility of using lanthanide luminescence as a highly sensitive method for nanoparticle size at the quantum confinement regime.

Figure 9.
Eu³⁺ emission spectra with l_exc = 260 nm obtained for xSnO₂:Eu³⁺/bulk samples where x = number of IDC, and SnO₂:Eu³⁺ bulk sample (reproduced with permission from reference ³¹, Copyright 2011, American Chemical Society).

The first work in this research line where an application of such chemically integrated porous host/metal oxide systems was evaluated in addition to the semiconductor properties was in the work by e Silva et al.³² Here, NP of CeO₂ were synthesized inside the pores of an SBA-15 mesoporous silica, and the materials oxygen storage capacity, a very important property regarding metal oxides applied in catalysis, was evaluated versus the number of IDC and NP size. This was also the first work in this research line where SBA-15 was used as a porous host for metal oxide NP. The authors have shown it was possible to control the amount of CeO₂ inserted in the system, as well as the size of the CeO₂ NP grown, through the number of IDC, as confirmed per XRD and Raman peak/band broadening and shifting, and per UV-Vis band gap measurements, with a gradually smaller band gap with an increasing number of IDC, a behavior characteristic of the quantum confinement regime. This corroborates with the observed for the previous systems with PVG as host, as well as with CeO₂ or other oxides. N₂ adsorption-desorption data also corroborated with a lowering surface area and pore volume with an increasing number of IDC. While no blockage of the pore structure was observed, a contraction of the porous host structure was occurring with every IDC, the cause of which was attributed to the thermal treatment of the decomposition step(s), a behavior typically expected for such ordered mesoporous silicas. Furthermore, due to the ordered nature of the porous host, TEM dark field HRTEM measurements were necessary to observe the CeO₂ NP inside the SBA-15 structure. The oxygen storage capacity (OSC) of the SBA-15/xCeO₂ was measured through hydrogen uptake by temperature programmed reduction (H₂-TPR) (Figure 10). One can immediately see the effect of the small size of NP at 1 IDC when compared to the samples with 5 and 10 IDC: the absence of a second H₂ consumption peak. The second peak at higher temperature is a result of bulk CeO₂ reacting with H₂, showing that at 1 IDC the NP size is indeed so small that the “surface” CeO₂ response is the only one observed. With the growth of the NP with 5 and 10 IDC, the second peak then appears, showing the increasing presence of “bulk” effects. One could also infer that the surface CeO₂ is the most reactive, since the normalized uptake of H₂ is much higher for 1 IDC, a justified observation since the smaller the NP, the higher the crystal lattice distortion and thus the higher the resulting oxygen non-stoichiometry.

Figure 10.
H₂-TPR measurements obtained for SBA-15/xCeO₂ samples, where (a), (b) and (c) correspond to 1, 5 and 10 IDC, respectively (reproduced from reference ³², Copyright 2011, with permission from Royal Society of Chemistry).

In a following work using SBA-15, the authors sought to grow TiO₂ NP, similarly to what was done before with PVG, and tested the applicability of the samples in the photocatalytic degradation of salicylic acid.³³ N₂ sorption results corroborated with the observed for the SBA-15/xCeO₂ systems, and the same control of NP size and oxide amount was possible with the number of IDC, but one important difference was observed: the formation of mixed phases of rutile and anatase TiO₂ NP, in contrast to the pure anatase phase obtained when using PVG. Another important difference was observed on the DRS spectra, where the band gap energy was much higher than expected for TiO₂ NP. While the increase in band gap energy is normally expected and attributable to the quantum confinement effect for NP < 5 nm, this abnormal increase was attributed not only to this effect, but to a combination also involving the formation of strong Ti–O–Si bonds. This second effect can strongly modify the electronic structure of the Ti atoms by increasing its effective positive charge, further increasing the band gap energy. One could infer, however, that some other effect related to the difference in porous host might play a role, since PVG also forms Ti–O–Si bonds and this abnormal band gap energy increase is not observed in the case of PVG/xTiO₂ systems. The performance of the synthesized samples in the photocatalytic degradation of salicylic acid has shown that the sample with the smallest size NP and thus also lowest relative amount of TiO₂ presented the highest activity (PVG/1TiO₂) (Figure 11). This result shows that, despite the larger band gap, and thus lower photon absorption rate of the sample, the surface area and surface catalytic performance play a much more important role, which can also be observed by the rather similar performance of the following higher IDC samples. This result also corroborated with the OSC performance of CeO₂ NP, where the sample with 1 IDC has shown the most relative activity.

Figure 11.
Catalytical performance of SBA-15/xTiO₂ for the photodegradation of salicylic acid, where C is the concentration of acid as a function of time and C₀ is its initial concentration. Through the Langmuir-Hinshelwood model applied to a substrate with low initial concentration, the equation may be simplified to the pseudo-first order expression –ln(C/C₀) = kt, and thus the rate constant k may be determined from a curve of –ln(C/C₀) vs. t (reproduced from reference ³³, Copyright 2013, with permission from Royal Society of Chemistry).

The last work in this line of investigating metal oxide NP/porous host chemically integrated systems constructed via IDC for catalytical-related applications was the work by Strauss et al.²⁴ In this work, the authors sought to isolate the size effects from other possible effects affecting TiO₂ NP catalytical performance, and to understand the former in detail through the application of PVG/xTiO₂ systems in the photocatalytic degradation of salicylic acid and methylene blue decolorization. Samples with TiO₂ NP with different sizes obtained through different numbers of IDCs were tested, and similarly to the previous systems an inverse relationship between size and photocatalytic activity was observed, as can be seen in Figure 12. However, the formation of one of the most reactive species involved in photocatalytic oxidation processes, the hydroxyl radical •OH, was observed to be directly proportional to the size of the TiO₂ NP. This is expected, since the generation of these radicals is enhanced with the improved absorption of light by the NP, which gradually increases with gradually smaller band gaps with an increasing NP size. These two aspects (the surface-to-bulk ratio and the formation of hydroxyl radicals) seemed at first at odds against one another, since the photocatalytic activity was highest for the smallest NP, even with the less efficient •OH formation. This has shown that surface effects played a more significant role in the photocatalysis mechanisms involved, with the higher surface area favoring adsorption-desorption mechanisms for the organic compounds studied, and the higher surface-to-bulk ratio reducing the amount of recombination volume (“bulk” TiO₂). Since the recombination of photogenerated electrons and holes occurs mostly in the “bulk”, and since this mechanism is the key suppression pathway for the photocatalytic activity of TiO₂, as they are responsible for the formation of strongly oxidizing and reducing species involved in photocatalysis, it makes sense that preventing it by constructing higher surface-to-bulk ratio NP plays the major role in the material’s photocatalysis performance.

Figure 12.
Catalytical performance of PVG/xTiO₂ for (I) salicylic acid photodegradation and (II) methylene blue decolorization. Evaluation of OH radical formation by TiO₂ irradiation through hydroxyl reaction with terephthalic acid (TPA), where (III) relates the normalized intensity of TPAOH emission against the light absorption response and (IV) against the inverse of NP size determined from Raman spectroscopy (reproduced from reference ²⁴, Copyright 2014, with permission from Elsevier).

4.2. Core@shell systems

Through the single metal oxide NP chemically integrated systems described so far it was clear to see how the control of NP growth and thus size, as well as their stabilized crystalline phase through the number of IDC lead to the fine tuning of the material properties. It was also observed that the layer-by-layer growth promoted by this method could create a new degree of complexity in the system by building core@shell NP, exchanging the single-source precursor after a number of IDC to grow a shell of a different metal oxide.

The first attempt was performed by Santos et al.²³ through the synthesis of TiO₂@MoO₃ and MoO₃@TiO₂ NP in PVG. The formation and development of the core@shell structure was done through Raman spectroscopy (Figure 13), by monitoring the anatase TiO₂ E_2g band at around 150 cm^–1 and the α-MoO₃ characteristic υ_symMo=O and corner sharing υMo–O–Mo bands at around 990 and 820 cm^–1, respectively. With an increasing number of IDC, a shift to lower energy and higher intensity are observed for both the pure TiO₂ and MoO₃ NP, a characteristic behavior as per phonon confinement effect due to the small NP size. With a MoO₃ core, adding subsequent layers of TiO₂ results in the observation of TiO₂ characteristic bands, displaying phonon confinement, while the MoO₃ slowly lose intensity. In the other way around, with a TiO₂ core and a MoO₃ shell, the reverse is observed. This effect was one of the indicators of the formation of a core@shell structure. Furthermore, the broadening of the bands, as well as disappearance or low definition of the corner sharing υMo–O–Mo bands together with the dislocation of the υ_symMo=O band to lower energies were also remarkable signs of the formation of a core@shell structure, both in the case of coating with a TiO₂ shell and for a thinner shell of MoO₃, resulting of distortions in the corner sharing octahedral structure in MoO₃. Another strong indicator was the phonon confinement effect observed when varying “core” and “shell” thickness: for larger cores with a similar shell size, a stronger confinement was observed, corroborating with an expected thinner shell in the case of core@shell formation. A work³⁰ reporting the synthesis of Ti O₂@WO₃ in PVG soon followed, since a readily dispersion of WO₃ over TiO₂ is expected owing to the strong Ti–O–W bonding. The authors observed the formation of mixed triclinic and monoclinic phases in the case of WO₃, both for the pure core and for WO₃ shells, and similar effects to TiO₂@MoO₃ systems were observed regarding Raman analysis. The direct visualization of the core@shell structure through HRTEM measurements for these samples was not possible, which was attributed to a very thin shell being formed, together with the mixed phase profile of WO₃, and the fact the NP are contained within a porous host, hindering the visualization of diffraction fringes.

Figure 13.
Raman spectra obtained for (I) PVG/5MoO₃@xTiO₂; (II) PVG/xMoO₃@5TiO₂; (III) PVG/5TiO₂@xMoO₃ and (IV) PVG/xTiO₂@5MoO₃, where x = 0, 3, 5 or 7 IDC (reproduced from reference ²³, Copyright 2010, with permission from Elsevier).

The work by Correa et al.¹⁹ expanded on the strategy from Santos et al.²³,³⁰ for the synthesis of core@shell NP of TiO₂ and CeO₂ in PVG, tuning core and shell sizes not only with fixed core or shell but also simultaneously changing the number of IDC for both, for TiO₂@CeO₂ and CeO₂@TiO₂ systems. The analysis of the TiO₂ E_g and CeO₂ T_2g bands in the Raman spectra led to similar conclusions as the observed for the previous core@shell systems, and the observed results also fit very well with the theory of phonon confinement according with the size/thickness of the core/shell analyzed. One important aspect to consider is the different, albeit expected, behaviors of bands E_g and T_2g with increasing particle size: E_g is shifted to lower frequencies, while T_2g is shifted to higher frequencies, a result of the different phonon dispersion relations for the phonon branch to which each Raman mode belongs. Another interesting observation was the suppression of CeO₂ surface defects when coated with a TiO₂ shell, noted by the absence of defect-related bands in the region of 360 to 480 cm^–1, which also become very prominent for very thin CeO₂ shells. The authors also sought to image the core@shell structure of the NP by synthesizing a PVG/5CeO₂@15TiO₂ sample, since with fewer IDC the shell size is too thin to be imaged with the HRTEM used. This attempt proved successful, and an exemplar core@shell NP can be seen on Figure 14.

Figure 14.
Raman shifts observed for PVG/xCeO₂@yTiO₂, where x and y = 0, 3, 5 or 7 (reproduced with permission from reference ¹⁹, Copyright 2011 American Chemical Society).

A study by Kalinke et al.³⁴ followed a different approach, aiming to synthesize mixed oxides and mixed-phase oxides of Cu, Ce and Co using PVG as porous host and bimetallic complexes as single-source precursors, where the impregnation step was performed twice with different metal complexes before the decomposition step. Although partial cation substitution was observed for CuO/CeO₂ systems, these did not form a mixed oxide, but the formation of spinel copper cobaltite ((Cu,Co)Co₂O₄) with average size slightly above twice that of the individual oxides was observed (9 nm compared to 4 nm). This study implied the possibility of synthesizing mixed oxides or mixed oxide phase composites with fewer decomposition steps, using specific precursors that would allow the insertion of different metals before the decomposition, where the oxide formation is promoted.

4.3. Core@multishell systems: application-oriented design of hierarchical multilayer nanoparticles

The studies shown so far had demonstrated the IDC method’s potential and flexibility in the synthesis of tailored single-oxide and oxide@oxide core@shell and composite nanoparticles. The following studies took this a step further: the application-oriented design of nanoparticles by tailoring their hierarchical multilayer structure.

The first work by Oliveira et al.²⁰ followed that approach by designing single-oxide core@multishell ZnO:Eu^III NP in an attempt to incorporate Eu^III into the ZnO wurzite crystal structure, a very difficult task considering the much larger size of Eu^III,⁵⁸,⁵⁹,⁶⁰,⁶¹,⁶²,⁶³,⁶⁴ where the higher concentration of structural defects in NP was considered as a means for the successful doping. The strategy was to induce the forced thermal migration of Eu^III by making a “sandwich” structure with Eu^III-doped ZnO IDC layers in-between pure ZnO oxide IDC layers. The behavior of ZnO, however, was different from the expected based on the previous oxide systems. It seemed to interact much more effectively with the surface of SiO₂ than with itself, promoting the formation of Si–O–Zn groups instead of ZnO nanoparticles. A consequence of this was that no ZnO band gap was observed at first (Figure 15), only a shoulder at higher energy (225-250 nm wavelength) attributed to these surface groups, although NP were still observed through TEM. Interestingly, after the addition of the first ZnO:Eu^III layers, the formation of the ZnO band gap is then observed with ever increasing and red-shifting intensity with ↑IDC. This has shown that the insertion of a different species has promoted ZnO NP growth, much like a “seeding” process.

Figure 15.
UV-Vis absorption spectra obtained by DRS using the Kubelka-Munk function for PVG/ZnO@ZnO:Eu^III@ZnO samples. The inset shows the (F(R).hν)² vs. hn plots (reproduced with permission from reference ²⁰, Copyright 2015, American Chemical Society).

Looking at the photoluminescence curves (Figure 16), the presence of a ZnO → PVG energy transfer, where corresponding ZnO bands are observed when looking at the excitation spectra of a PVG-specific emission band (λ_em = 435 nm), corroborates with the strong interaction observed between SiO₂ and ZnO. A clear and definitive sign that Eu^III had been successfully inserted into ZnO was not observed, but the presence of “bulges” in the Eu^III λ_em = 613 nm excitation spectra exactly in the same region as the observed ZnO band gap bands for the PVG excitation spectra (in depth analysis can be found in the supplementary material of Oliveira et al.²⁰) suggests a ZnO → Eu^III energy transfer, and thus the very likely successful doping.

Figure 16.
Excitation spectra of the PVG emission (λ_em = 435 nm) and most intense Eu^III emission (⁵D₀ → ⁷F₂, λ_em = 613 nm) obtained for ZnO@ZnO:Eu^III@ZnO samples, showing energy transfers from Zn–O–Si groups and ZnO to PVG, and possibly from ZnO to Eu^III (380-400 nm region) (reproduced with permission from reference ²⁰, Copyright 2015, American Chemical Society).

The following work by de Oliveira et al.³⁵ consolidated all the previous experience using the IDC method to design highly tailored NP for LED applications. The goal was to design NP as a possible solution to an ongoing challenge at the time: obtaining stable warm-white-light-emitting materials.⁶⁵,⁶⁶,⁶⁷,⁶⁸,⁶⁹ This required a precise combination of the different light-emitting species, as well as the structure in which they were to be arranged, to prevent the various possible effects occurring between the different involved species, such as energy transfers and non-radiative relaxation, from interfering with the desired result.

To reach this goal, the full potential of the IDC method was explored. The commercially available PVG was used as a porous host to control nanoparticle growth and size. ZrO₂ was selected as oxide host for its high chemical and thermal resistances, as well as good solubility for large interstitial species such as Ln^III ions. On that basis, for a precise control of luminescence, lanthanide ions Eu^III and Tb^III were chosen as dopants to obtain red and green light emission, complementing the native-defect blue-emission from PVG. As a silica-based glass, however, the presence of Si–OH groups on PVG’s internal surface presented both an advantage and a disadvantage, since –OH groups are a well-known source of non-radiative deactivation of Ln-based phosphors, especially of Eu^III. For this, a thin layer of ZnO was coated on the PVG’s internal pore surface, based on the previous work by Oliveira et al.²⁰ Another effect to be prevented was the energy transfer between Tb^III and Eu^III. Considering their overall relative proximity inside a < 5 nm nanoparticle, the best strategy proved to be the physical separation between them through interstitial layers of pure oxide. Finally, tailoring the specific number of doped and pure layers allowed the fine-tuning of the light emission of the nanoparticles (Figure 17). In Figure 17 it is also clear to see the role that the nanostructure plays on the material performance, where the absence of pure layers in-between lead to Eu^III → Tb^III energy transfers occurring, the extreme case being when both are mixed for every layer (Figure 17f). The reader is referred to de Oliveira et al.³⁵ work for the in-depth analysis regarding the tuning of the light emission properties via the composition and nanostructure of the NP in these systems.

Figure 17.
Emission spectra and corresponding CIE diagrams obtained for the Ln^III-doped ZrO₂ nanoparticles synthesized inside ZnO-coated PVG with varying layered compositions: (a,b) ZnZrH: with every Ln^III-doped layer separated by pure ZrO₂ layers; (c,d) ZnZrC: without pure ZrO₂ layers; ZrM: without ZnO coating and with both dopants (Eu^III and Tb^III) mixed for every IDC (reproduced with permission from reference ³⁵, Copyright 2017, American Chemical Society).

Herein we presented the metal oxide-based NP-porous host chemically integrated systems developed using the IDC method. This method provided a combined solution for synthesizing chemically integrated systems where < 10 nm NP may be stabilized, while also allowing control over their growth, composition and in some cases of their crystal phase. The use of a porous host also opened the doors for further exploiting the contribution of another component in the chemically integrated systems, for example where pores could function as nanoreactors for catalysis, or where a functional surface could be applied to prevent luminescence suppression effects over doped NP.

5. Conclusions

In this review we went through the research history of the synthesis of tailored chemically integrated systems composed of metal oxide nanoparticles grown inside porous matrices using the method of IDC. This layer-by-layer growth method combined with the pore structure size limitations allowed the fine tuning of the nanoparticle size, shape, crystalline phase and nanostructure through their hierarchical construction. The number and composition of the IDC layers allowed the tailoring of nanoparticle properties, such as the band gap energy or Raman absorption bands through quantum confinement effects, the fine control of the material’s light emission through the number of doped layers and their order within the nanoparticles, as well as the construction of core@shell nanoparticles. This fine control of properties enabled the synthesis of chemically integrated systems for applications in: catalysis, where NP stabilization in a different crystal structure as the obtained for bulk, as well as with very small sizes due to the porous host stabilization, presented the highest activity for photocatalysis due to the much higher surface area, even with the increased optical band gap; magnetic nanoparticles presented a special core@shell structure with different magnetic behaviors for core and shell; and for light emission the role of the hierarchical nanostructure of the nanoparticles played a pivotal role in the NP performance for white light emission. We believe this method still has a lot of unexplored potential for the scientific community, and this series of studies have shown there can be endless possibilities in terms of metal oxide combinations and hierarchical nanostructure designs for a variety of applications. A couple examples would be: multi-oxide hierarchical structures containing 3 or more oxides of interest for example in catalysis, which combined with specific metallic NP could generate chemically integrated systems with special performance; or the assembly of hierarchical metal-oxide NP with multifunctional pollutant degradation potential, where each layer could be responsible for a given set of pollutant molecules.

Acknowledgments

The authors gratefully acknowledge the financial support provided by the CNPq, FAPESP, CAPES and INCT/INOMAT (National Institute of Science Technology in Complex Functional Materials). Contributions from the Brazilian Center for Research in Energy and Materials (CNPEM), Brazilian Synchrotron Light Laboratory (LNLS), Brazilian Nanotechnology National Laboratory (LNNano), Laboratory of Molecular Spectroscopy (IQ-USP), Multiuser Laboratory of Advanced Optical Spectroscopy (LMEOA/IQ-UNICAMP) analyses are also gratefully acknowledged. This is a contribution of the National Institute of Science and Technology in Complex Functional Materials.

This work is dedicated to Professor Oswaldo Luiz Alves from the Institute of Chemistry, UNICAMP, a recognized and enthusiastic Professor/Researcher of the Solid State Chemistry and Nanoscience fields who suddenly and prematurely passed away on July 10^th, 2021. The authors would like to acknowledge the immeasurable contribution that Prof Oswaldo Luiz Alves made not only to science, but to the lives and welfare of those that, in one way or another, could benefit from his work, wisdom and unyielding will.

References

• ¹ Khalil, M.; Kadja, G. T. M.; Ilmi, M. M.; J. Ind. Eng. Chem. 2021, 93, 78. [Crossref]
  » Crossref
• ² Astruc, D.; Chem. Rev. 2020, 120, 461. [Crossref]
  » Crossref
• ³ Ramos, A. P.; Cruz, M. A. E.; Tovani, C. B.; Ciancaglini, P.; Biophys. Rev. 2017, 9, 79. [Crossref]
  » Crossref
• ⁴ Shin, M. D.; Shukla, S.; Chung, Y. H.; Beiss, V.; Chan, S. K.; Ortega-Rivera, O. A.; Wirth, D. M.; Chen, A.; Sack, M.; Pokorski, J. K.; Steinmetz, N. F.; Nat. Nanotechnol. 2020, 15, 646. [Crossref]
  » Crossref
• ⁵ Willner, M. R.; Vikesland, P. J.; J. Nanobiotechnol. 2018, 16, 95. [Crossref]
  » Crossref
• ⁶ Shah, M. A.; Pirzada, B. M.; Price, G.; Shibiru, A. L.; Qurashi, A.; J. Adv. Res. 2022, 38, 55. [Crossref]
  » Crossref
• ⁷ Pomerantseva, E.; Bonaccorso, F.; Feng, X.; Cui, Y.; Gogotsi, Y.; Science 2019, 366, 8285. [Crossref]
  » Crossref
• ⁸ Wang, H.; Liang, X.; Wang, J.; Jiao, S.; Xue, D.; Nanoscale 2020, 12, 14. [Crossref]
  » Crossref
• ⁹ Kim, T.-Y.; Kim, M. G.; Lee, J.-H.; Hur, H.-G.; Front. Microbiol. 2018, 9, 2817. [Crossref]
  » Crossref
• ¹⁰ Guerra, F. D.; Attia, M. F.; Whitehead, D. C.; Alexis, F.; Molecules 2018, 23, 1760. [Crossref]
  » Crossref
• ¹¹ Beni, A. A.; Jabbari, H.; Results Eng. 2022, 15, 100467. [Crossref]
  » Crossref
• ¹² The Nobel Prize in Chemistry 2023, https://www.nobelprize.org/prizes/chemistry/2023/summary/, accessed in May 2024.
  » https://www.nobelprize.org/prizes/chemistry/2023/summary/
• ¹³ Rostami, S.; Ghaffarkhah, A.; Isari, A. A.; Hashemi, S. A.; Arjmand, M.; Mater. Adv. 2023, 4, 2698. [Crossref]
  » Crossref
• ¹⁴ Ouhenia-Ouadahi, K.; Yasukuni, R.; Yu, P.; Laurent, G.; Pavageau, C.; Grand, J.; Guérin, J.; Léaustic, A.; Félidj, N.; Aubard, J.; Nakatani, K.; Métivier, R.; Chem. Commun. 2014, 50, 7299. [Crossref]
  » Crossref
• ¹⁵ Li, W.; Wang, C.; Lu, X.; J. Mater. Chem. A 2021, 9, 3786. [Crossref]
  » Crossref
• ¹⁶ Chen, Z.; Wu, C.; Yuan, Y.; Xie, Z.; Li, T.; Huang, H.; Li, S.; Deng, J.; Lin, H.; Shi, Z.; Li, C.; Hao, Y.; Tang, Y.; Yo u , Y.; Al-Hartomy, O. A.; Wageh, S.; Al-Sehemi, A. G.; Lu, R.; Zhang, L.; Lin, X.; He, Y.; Zhao, G.; Li, D.; Zhang, H.; J. Nanobiotechnol. 2023, 21, 141. [Crossref]
  » Crossref
• ¹⁷ Zhao, G.; Yang, F.; Chen, Z.; Liu, Q.; Ji, Y.; Zhang, Y.; Niu, Z.; Mao, J.; Bao, X.; Hu, P.; Li, Y.; Nat. Commun. 2017, 8, 14039. [Crossref]
  » Crossref
• ¹⁸ Mazali, I. O.; Alvez, O. L.; J. Phys. Chem. Solids 2005, 66, 37. [Crossref]
  » Crossref
• ¹⁹ Corrêa, D. N.; e Silva, J. M. S.; Santos, E. B.; Sigoli, F. A.; Souza Filho, A. G.; Mazali, I. O.; J. Phys. Chem. C 2011, 115, 10380. [Crossref]
  » Crossref
• ²⁰ Oliveira, C. S.; Bettini, J.; Sigoli, F. A.; Mazali, I. O.; Cryst. Growth Des. 2015, 15, 5246. [Crossref]
  » Crossref
• ²¹ Thanh, N. T.; Maclean, N.; Mahiddine, S.; Chem. R ev. 2014, 114, 7610. [Crossref]
  » Crossref
• ²² Mazali, I. O.; Souza Filho, A. G.; Viana, B. C.; Mendes Filho, J.; Alves, O. L.; J. Nanopart. Res. 2006, 8, 141. [Crossref]
  » Crossref
• ²³ Santos, E. B.; e Silva, J. M. S.; Mazali, I. O.; Vib. Spectrosc. 2010, 54, 89. [Crossref]
  » Crossref
• ²⁴ Strauss, M.; Pastorello, M.; Sigoli, F. A.; e Silva, J. M. S.; Mazali, I. O.; Appl. Surf. Sci. 2014, 319, 151. [Crossref]
  » Crossref
• ²⁵ Cangussu, D.; Nunes, W. C.; Corrêa, H. L. S.; Macedo, W. A. A.; Knobel, M.; Alves, O. L.; Souza Filho, A. G.; Mazali, I. O.; J. Appl. Phys. 2009, 105, 013901. [Crossref]
  » Crossref
• ²⁶ Mazali, I. O.; Viana, B. C.; Alves, O. L.; Mendes Filho, J.; Souza Filho, A. G.; J. Phys. Chem. Solids 2007, 68, 622. [Crossref]
  » Crossref
• ²⁷ Mazali, I. O.; Romano, R.; Alves, O. L.; Thin Solid Films 2006, 495, 64. [Crossref]
  » Crossref
• ²⁸ Oliveira, M. M.: Obtenção de Nanocompósitos de Óxidos Semicondutores Inseridos em Vidro Poroso Vycor via Decomposição de Precursores Metalorgânicos; MSc Dissertation, University of Campinas, Campinas, Brazil, 2000. [Link] accessed in April 2024.
  » Link
• ²⁹ Carneiro, N. M.; Nunes, W. C.; Borges, R. P.; Godinho, M.; Fernandez-Outon, L. E.; Macedo, W. A. A.; Mazali, I. O.; J. Phys. Chem. C 2010, 114, 18773. [Crossref]
  » Crossref
• ³⁰ Santos, E. B.; e Silva, J. M. S.; Sigoli, F. A.; Mazali, I. O.; J. Nanopart. Res. 2011, 13, 5909. [Crossref]
  » Crossref
• ³¹ Strauss, M.; Destefani, T. A.; Sigoli, F. A.; Mazali, I. O.; Cryst. Growth Des. 2011, 11, 4511. [Crossref]
  » Crossref
• ³² e Silva, J. M. S.; Strauss, M.; Maroneze, C. M.; Souza, E. R.; Gushikem, Y.; Sigoli, F. A.; Mazali, I. O.; J. Mater. Chem. 2011, 21, 15678. [Crossref]
  » Crossref
• ³³ e Silva, J. M. S.; Pastorello, M.; Strauss, M.; Maroneze, C. M.; Sigoli, F. A.; Gushikem, Y.; Mazali, I. O.; RSC Adv. 2012, 2, 5390. [Crossref]
  » Crossref
• ³⁴ Kalinke, L. H. G.; Stumpf, H. O.; Mazali, I. O.; Cangussu, D.; Mater. Res. Bull. 2015, 70, 743. [Crossref]
  » Crossref
• ³⁵ de Oliveira, C. S.; Bettini, J.; Sigoli, F. A.; Mazali, I. O.; Cryst. Growth Des. 2017, 17, 5398. [Crossref]
  » Crossref
• ³⁶ Kocsis, G.; Szabó-Bárdos, E.; Fónagy, O.; Farsang, E.; Juzsakova, T.; Jakab, M.; Pekker, P.; Kovács, M.; Horváth, O.; Molecules 2022, 27, 8041. [Crossref]
  » Crossref
• ³⁷ Chu, J.; Sun, Y.; Han, X.; Zhang, B.; Du, Y.; Song, B; Xu, P.; ACS Appl. Mater. Interfaces 2019, 11, 18475. [Crossref]
  » Crossref
• ³⁸ Sampaio, M. J.; Yu, Z.; Lopes, J. C; Tavares, P. B.; Silva, C. G.; Liu, L.; Faria, J. L.; Sci. Rep. 2021, 11, 21306. [Crossref]
  » Crossref
• ³⁹ Oi, L. E.; Choo, M.-Y.; Lee, H. V.; Oong, H. C; Hamid, S. B. A.; Juan, J. C; RSC Adv. 2016, 6, 108741. [Crossref]
  » Crossref
• ⁴⁰ Macino, M.; Barnes, A. J.; Althahban, S. M.; Qu, R.; Gibson, E. K.; Morgan, D. J.; Freakley, S. J.; Dimitratos, N.; Kiely, C. J.; Gao, X.; Beale, A. M.; Bethell, D.; He, Q.; Sankar, M.; Hutchings, G. J.; Nat. Catal. 2019, 2, 873. [Crossref]
  » Crossref
• ⁴¹ Zhang, H; Zuo, S.; Qiu, M.; Wang, S.; Zhang, Y; Zhang, J.; Lou, X. W.; Sci. Adv. 2020, 6, eabb9823. [Crossref]
  » Crossref
• ⁴² Richter, H.; Wang, Z. P.; Ley, L.; Solid State Commun. 1981, 39, 625. [Crossref]
  » Crossref
• ⁴³ Zhou, J.; Guo, R.-t.; Zhang, X.-f.; Liu, Y.-z.; Duan, C.-p.; Wu, G.-l.; Pan, W.-g.; Energy Fuels 2021, 35, 2981. [Crossref]
  » Crossref
• ⁴⁴ Liyanage, A. D.; Perera, S. D.; Tan, K., Chabal, Y; Balkus Jr., K. J.; ACS Catal. 2014, 4, 577. [Crossref]
  » Crossref
• ⁴⁵ Luo, S.; Barrio, L.; Nguyen-Phan, T.-D.; Vovchok, D.; Johnston-Peck, A. C; Xu, W; Stach, E. A.; Rodriguez, J. A.; Senanayake, S. D.; J. Phys. Chem. C 2017, 121, 6635. [Crossref]
  » Crossref
• ⁴⁶ Wu, L.; Fang, S.; Ge, L.; Han, C; Qiu, P.; Xin, Y; J. Hazard Mater. 2015, 300, 93. [Crossref]
  » Crossref
• ⁴⁷ Jiang, D.; Wang, W; Sun, S.; Zhang, L.; Zheng, Y; ACS Catal. 2015, 5, 613. [Crossref]
  » Crossref
• ⁴⁸ Spanier, J. E.; Robinson, R. D.; Zhang, F.; Chan, S.-W; Herman, I. P.; Phys. Rev. B 2001, 64, 245407. [Crossref]
  » Crossref
• ⁴⁹ Coduri, M.; Masala, P.; Del Bianco, L.; Spizzo, F; Ceresoli, D.; Castellano, C; Cappelli, S.; Oliva, C; Checchia, S.; Allieta, M.; Szabo, S. V.; Schlabach, S.; Hagelstein, M.; Ferrero, C; Scavini, M.; Nanomaterials 2020, 10, 867 [Crossref]
  » Crossref
• ⁵⁰ Cao, D.; Li, H.; Pan, L.; Li, J.; Wang, X.; Jing, P.; Cheng, X.; Wang, W; Wang, J.; Liu, Q.; Sci. Rep. 2016, 6, 32360. [Crossref]
  » Crossref
• ⁵¹ Wu, W; Xiao, X. H.; Zhang, S. F; Peng, T. C; Zhou, J.; Ren, F; Jiang, C. Z.; Nanoscale Res. Lett. 2010, 5, 1474. [Crossref]
  » Crossref
• ⁵² Guivar, J.; Martínez, A.; Anaya, A.; Valladares, L.; Félix, L.; Dominguez, A.; Adv. Nanopart. 2014, 3, 114. [Crossref]
  » Crossref
• ⁵³ Tadic, M.; Nikolic, D.; Panjan, M.; Blake, G. R.; J. Alloys Compd. 2015, 647, 1061. [Crossref]
  » Crossref
• ⁵⁴ Janus, W; Ślęzak, T.; Ślęzak, M.; Szpytma, M.; Dróżdż, P.; Nayyef, H.; Mandziak, A.; Wilgocka-Ślęzak, D.; Zając, M.; Jugovac, M.; Menteş, T. O.; Locatelli, A.; Kozioł-Rachwał, A.; Sci. Rep. 2023, 13, 4824. [Crossref]
  » Crossref
• ⁵⁵ Jafari, A.; Jahromi, S. P.; Boustani, K.; Goh, B. T.; Huang, N. M.; J. Magn. Magn. Mater. 2019, 469, 383. [Crossref]
  » Crossref
• ⁵⁶ Al Boukhari, J.; Khalaf, A.; Awad, R.; Appl. Phys. A 2020, 126, 74. [Crossref]
  » Crossref
• ⁵⁷ Swiatkowska-Warkocka, Z.; Pyatenko, A.; Shimizu, Y.; Perzanowski, M.; Zarzycki, A.; Jany, B. R.; Marszalek, M.; Nanomaterials 2018, 8, 790. [Crossref]
  » Crossref
• ⁵⁸ Mukherjee, S.; Katea, S. N.; Rodrigues, E. M.; Segre, C. U.; Hemmer, E.; Broqvist, P.; Rensmo, H.; Westin, G.; Small 2023, 19, 2203331. [Crossref]
  » Crossref
• ⁵⁹ Haj-Hmeidi, Y.; Röder, R.; Cammi, D.; Ronning, C.; Nanotechnology 2019, 30, 095201. [Crossref]
  » Crossref
• ⁶⁰ Aneesh, P. M.; Jayaraj, M. K.; Bull. Mater. Sci. 2010, 33, 227. [Crossref]
  » Crossref
• ⁶¹ Jiang, N.; Ye, S.; Qiu, J.; J. Appl. Phys. 2010, 108, 083535. [Crossref]
  » Crossref
• ⁶² Shannon, R. D.; Prewitt, C. T.; Acta Crystallogr. , Sect. B: Struct. Sci., Cryst. Eng. Mater. 1969, B25, 925. [Crossref]
  » Crossref
• ⁶³ Lima, S. A. M.; Sigoli, F. A.; Davolos, M. R.; J. Solid State Chem. 2003, 171, 287. [Crossref]
  » Crossref
• ⁶⁴ Liu, Y.; Luo, W.; Li, R.; Liu, G.; Antonio, M. R.; Chen, X.; J. Phys. Chem. C 2008, 112, 686. [Crossref]
  » Crossref
• ⁶⁵ Oh, J. H.; Yang, S. J.; Do, Y. R.; Light: Sci. Appl. 2014, 3, e141. [Crossref]
  » Crossref
• ⁶⁶ Kuse, Y.; Ogawa, K.; Tsuruma, K.; Shimazawa, M.; Hara, H.; Sci. Rep. 2014, 4, 5223. [Crossref]
  » Crossref
• ⁶⁷ Rai, M.; Kaur, G.; Singh, S. K.; Rai, S. B.; Dalton Trans. 2015, 44, 6184. [Crossref]
  » Crossref
• ⁶⁸ Borik, M.; Kulebyakin, A.; Kyashkin, V.; Larina, N.; Lomonova, E.; Milovich, F.; Myzina, V.; Nezhdanov, A.; Ryabochkina, P.; Tabachkova, N.; Chernov, E.; Materials 2022, 15, 7722. [Crossref]
  » Crossref
• ⁶⁹ Lovisa, L. X.; Andrés, J.; Gracia, L.; Li, M. S.; Paskocimas, C. A.; Bornio, M. R. D.; Araujo, V. D.; Longo, E.; Motta, F. V.; J. Alloys Compd. 2017, 695, 3094. [Crossref]
  » Crossref

Edited by

• Editor handled this article: Célia M. Ronconi (Associate)
  “Our tribute to Prof. Oswaldo Luiz Alves for his contribution to science, exemplary work, profound wisdom, and unwavering determination.”

Publication Dates

• Publication in this collection
  14 June 2024
• Date of issue
  2024

History

• Received
  26 Jan 2024
• Accepted
  20 May 2024