Transducing magnetism to soft responsive shells: smart nanosystems for magneto-triggered release of ibuprofen

Abstract

Smart nanocarriers have emerged as powerful platforms for externally controlled drug delivery, enabling precise spatial and temporal regulation of release while minimizing systemic side effects. Ibuprofen (IBU), a hydrophobic nonsteroidal anti-inflammatory drug with low aqueous solubility, particularly benefits from delivery systems that improve stability and bioavailability. In this work, we report on the design, synthesis, and evaluation of a smart hybrid nanosystem (SHN) for magnetically induced thermal release of IBU. The SHN consists of a superparamagnetic iron oxide nanoparticle (SPION) core acting as the magnetic module, a mesoporous silica (SiO^MP₂) intermediate shell providing drug-loading capacity, surface protection, functional integration, and an outer thermoresponsive poly(di(ethylene glycol) methyl ether methacrylate) (PDEGMA) layer enabling remote magneto-actuated temperature-triggered release. Structural and physicochemical characterization by FTIR, TEM, SAXS, and DLS confirmed the successful formation of a well-defined (core@shell)-g-polymer architecture. Magnetization measurements further demonstrated the preservation of superparamagnetic behavior with zero coercivity and a saturation magnetization of approximately 60 emu g⁻¹. IBU was efficiently loaded into the SHN through a multistep impregnation procedure. In situ release studies performed in tris(hydroxymethyl)aminomethane buffer (pH 7.8) showed pronounced thermoresponsive behavior, with up to 88% of the loaded IBU released at 50 °C, whereas significantly lower release was observed under physiologically relevant conditions. Notably, magnetic field-triggered release was approximately 7-fold greater than conventional thermal release at equivalent experimental times. This enhanced release occurred despite a macroscopic cooling of the dispersion during the radiofrequency magnetic fields experiment, suggesting that localized nanoscale heating at the SPION–polymer interface, rather than bulk heating, is the primary driver of the magneto-thermal transduction mechanism. Overall, these findings demonstrate that (SPION@SiO^MP₂)-g-PDEGMA SHN represents a promising platform for remotely triggered IBU delivery, combining magnetic responsiveness, controlled release capability, and structural versatility with strong potential for biomedical applications.

Smart hybrid nanosystems a promising platform for remotely triggered drug delivery, combining magnetic responsiveness, controlled release capability, and structural versatility with strong potential for biomedical applications. Graphical abstract image partly generated using Google Gemini AI tool.

1. Introduction

Over recent decades, controlled drug delivery nanosystems have emerged as a key strategy to enhance therapeutic efficacy by enabling localized and regulated drug release within biological environments.^1,2 Such nanoformulations aim to improve spatial control over drug distribution, thereby reducing systemic exposure and associated side effects. This approach is important in the medical field, particularly in tissue engineering and bone implant applications.³ For this purpose, employing support materials that combine biocompatibility with biological functionality is essential and a central challenge.⁴

Among these materials, superparamagnetic iron oxide nanoparticles (SPIONs) are widely used in biomedicine specifically owing to the proven biocompatibility of iron oxides, low production costs, and tunable magnetic properties.⁵ Nanoscale iron oxide phases such as magnetite (Fe₃O₄), maghemite (γ-Fe₂O₃), or hematite (α-Fe₂O₃) are widely studied SPIONs in which the tunable superparamagnetic behavior and biocompatibility are key advantages.^6,7 Previous work from our group has demonstrated the controlled synthesis and robust magnetic characterization of silica-coated SPION, highlighting their potential as versatile platforms for biomedical applications.^8,9

Due to their small size and high surface area, these nanoparticles can be modified with targeting ligands, such as polymers and proteins, to impart additional characteristics beyond the intrinsic properties of the synthesized material, creating promising tools for drug delivery to the body.¹⁰ The ability of SPION to swiftly respond to external magnetic fields enables control of their targeting to specific areas of the body, increasing the precision and effectiveness of drug release and reducing side effects.¹¹ However, the primary limitation of SPION is their strong tendency to aggregate in the absence of an external surface coating. When exposed to physiological media, uncoated SPION tend to absorb plasma proteins, promoting opsonization and subsequent aggregation. This process can be mitigated or even prevented by applying a hydrophilic coating, which reduces aggregation either through electrostatic interactions or steric hindrance.¹² Another concern is that high iron concentrations in the body can lead to oxidative stress and other forms of toxicity, particularly affecting the heart and liver. Thus, despite their unique theragnostic properties, the strong tendency of SPION to aggregate often limits their usefulness in biomedical applications.¹³ To address these limitations, the incorporation of mesoporous coating has proven particularly effective. This not only prevents aggregation but also ensures the nanoparticles are non-toxic, biocompatible, and allows for surface modification with biologically active agents.¹⁴

Mesoporous materials are characterized by an interconnected pore network in the nanometric range, whose organization and dimensions can be tailored to regulate mass transport and surface interactions.¹⁵ Their structural organization and pore size can be modified through the incorporation of ions, functional groups, or molecules capable of interacting with the material and its surface, thereby altering their properties.^16,17 The large surface area (up to 1000 m² g⁻¹), the high pore volume, and the relatively narrow mesopore size distribution (typically in the 2–15 nm diameter range) are key characteristics of mesoporous materials, enabling them to play a significant role in drug delivery.^18,19 Mesoporous silica (SiO^MP₂) is gaining significant attention in biomedicine due to its versatility in functionalization, allowing it to be tailored to meet the specific requirements of various applications.^20,21 It is biocompatible, chemically stable, and has well-established and tested synthesis parameters. Magnetic nanoparticles can be coated with a silica layer, resulting in the core@shell system, which can be used in hyperthermia studies in conjunction with localized drug delivery. Among the various methods used to coat SPION with silica,^22–26 the sol–gel method is the most widely employed. This approach relies on the hydrolysis and polycondensation of an alkoxysilane compound, such as tetraethyl orthosilicate (TEOS), in the presence of water and an appropriate catalyst.

Lu et al.²⁷ synthesized triporous microspheres (m-Fe₃O₄@dm-SiO₂) consisting of a mesoporous magnetite core coated with a SiO₂ shell with dual porosity, capable of functioning in multimodal therapies that combine hyperthermia and drug release. These microspheres exhibited a high surface area (426 m² g⁻¹) and a saturation magnetization of 52 emu g⁻¹, enabling efficient heating under an alternating magnetic field (AMF). Additionally, they demonstrated a high drug loading capacity and sustained release of the model drug 5-fluorouracil (5-Fu). SPION coated with different silica morphologies (non-porous (@SiO₂), mesoporous (@SiO^MP₂), and a combination of non-porous and mesoporous layers (@SiO₂@SiO^MP₂)) were investigated by Reczyńska et al.²⁸ These coatings enhanced the surface area of the nanoparticles, allowing for the loading of drugs or other bioactive molecules. Tests showed that both SPION and SPION coated with a single layer of mesoporous silica are cytocompatible with lung epithelial cells. As a result, these carriers can be loaded with anticancer drugs and administered via inhalation directly to pulmonary tumors. Besides, Garciá et al.²⁹ synthesized Fe₃O₄@SiO^MP₂ core@shell nanoparticles using solvothermal and sol–gel approaches. These nanoparticles exhibited superparamagnetic behavior, a high surface area (231 m² g⁻¹), and a thermal response induced by an AC magnetic field (SAR = 63 W g⁻¹ Fe₃O₄). Using ibuprofen as a model drug, they demonstrated a high loading capacity (10% by weight) and controlled release (67%) over short periods upon exposure to an external AC magnetic field (5 min), achieving greater efficiency compared to sustained release at fixed temperatures. The localized thermal effect within the particles suggests an effective mechanism for magnetically triggered drug release.

Despite all efforts, silica still presents some disadvantages, such as its dissolution in aqueous media^30,31 and the present limitations on attaining strict control over drug release.^32–34 However, silica facilitates the utilization of well-established and versatile hybridization chemistry. In this sense, to further improve the efficiency of drug release through hyperthermia, silica nanoparticles can be coated with responsive polymers to create smart nanomaterials. The term “smart” refers to the ability to detect and respond to an external stimulus.³⁵ Different triggering mechanisms, including thermal, photothermal, magnetic, and biochemical stimuli, have been explored to achieve controlled therapeutic responses.² Among these approaches, MXene-based photothermal nanoplatforms have shown promising performance by converting near-infrared irradiation into localized hyperthermia for externally controlled therapy,³⁶ while piezoelectric biomaterials have demonstrated potential for tissue regeneration and wound healing through the conversion of mechanical stimuli into electrical signals.³⁷ In parallel, multifunctional wound dressings integrating stimuli-responsive drug release, antibacterial activity, and in situ sensing have emerged as promising platforms for advanced wound management.^38,39 Stimuli-responsive nanoplatforms have also been explored for cancer therapy, including systems designed to synergize with siRNA and enhance antitumoral immune responses through immunogenic cell death.⁴⁰ Overall, these systems provide complementary advantages depending on the intended application, particularly by enabling remote activation, controlled release, and dynamic interactions with biological environments. In this context, hybrid nanosystems combining magnetic nanoparticles, mesoporous structures, and functional polymers have attracted considerable interest because they integrate magnetic targeting, high drug-loading capacity, and externally triggered therapeutic responses within a single platform.

Thermoresponsiveness is defined as the ability of a coating to function as an on–off switch. In the case of polymers, this thermoresponsiveness is governed by phase transitions in aqueous solutions, resulting in a drastic change in solubility. This behavior is commonly described by the lower critical solution temperature (LCST), above which polymer–solvent interactions are reduced, leading to polymer collapse and decreased solubility. Certain polymers exhibit transition temperatures close to body temperature (35–37 °C), which is a key factor for their application in biomedicine.^33,34,41,42 SPION-based thermoresponsive hybrid nanosystems represent a promising strategy for multifunctional biomedical applications by combining magnetic hyperthermia, controlled release, and tunable surface functionality. Choi et al.⁴³ investigated a temperature-sensitive antibacterial coating composed of PDEGMA (poly(diethylene glycol methacrylate) ester) applied to titanium implants. The material was immersed in a levofloxacin solution, allowing the drug to be incorporated and trapped within the expanded polymeric layers. The results showed that the LCST (lower critical solution temperature) of PDEGMA is closer to 32 °C, which is close to human body temperature. Above this temperature, the polymeric layers collapsed, triggering the release of the encapsulated drug and demonstrating antibacterial activity in vitro tests against Staphylococcus aureus. Améndola et al.³⁴ synthesized smart hybrid nanosystems (SHN) designed for the loading of positively charged drugs and their targeted release under controlled pH and temperature conditions. The nanoparticles were prepared with a core@shell architecture, consisting of a silica nanoparticle (NP) core and a copolymer shell with dual responsiveness, incorporating PNIPA (poly(N-isopropylacrylamide)) and PAA (poly(acrylic acid)). By combining DLS and NMR analyses, the group demonstrated that these nanosystems can convert external stimuli into mechanical contraction of the polymer shell, enabling a dual, remotely controlled response, that was evaluated through the release of doxorubicin, a chemotherapeutic drug. Studies on the synthesis of silica nanoparticles coated with the polymer PDEGMA have been reported in the literature, such as the work of Penelas et al.,⁴¹ in which core@brush hybrid nanoparticles were obtained by a grafting approach using monomers. In this study, different synthesis methods were investigated, and it was shown that although all methods were effective in immobilizing PDEGMA on the surface, the choice of synthesis route is crucial to control key properties of the core–brush nanoparticles, such as density of the grafted polymer, dispersion and colloidal stability. However, to date, no studies have reported the use of magnetic cores, such as SPION, in combination with PDEGMA.

In this work, we report on the design of a hybrid nanoarchitecture that couples magnetic actuation with a stimuli-responsive soft interface, allowing controlled magneto-triggered release particularly predesigned for ibuprofen (IBU). Notice that IBU is a hydrophobic nonsteroidal anti-inflammatory molecule with low aqueous solubility; therefore, its delivery through smart nanocarriers is necessary to improve its stability, bioavailability, and release it on the site of action. Specifically, we combine the properties of three different nanobuilding blocks: SPION, mesoporous silica (SiO^MP₂) and the PDEGMA polymer to produce responsive nanosystems with a (core@shell)-g-polymer architecture. The SPION core enables a response to external alternating radio frequency magnetic fields generating heat dissipation, which is used as heat nanosource. Besides, exposure of the SPION based-nanosystems to static magnetic fields facilitated their manipulation and targeting. The SiO^MP₂ shell, in turn, provides a porous structure functionalized by a layer of PDEGMA, a thermoresponsive polymer. It is important to highlight that the PDEGMA was carefully selected considering its properties as well as the nature of the molecule to be transported. Additionally, although (core@shell)-g-polymer systems responsive to magnetic stimuli have been investigated in the last years, there are still some crucial bottlenecks that have not been yet solved. For example, as was mentioned even though all efforts, silica still presents disadvantages such as its dissolution in aqueous media. In our work, the coating with PDEGMA, in addition to acting as a gatekeeper, acts as a coating presenting at the same time valuable biocompatibility, for biomedical applications are key features of oligo PEG-containing nanoparticles. These three integrated nanobuilding blocks offer an externally tunable smart behavior, enabling the nanosystems to respond in a controlled manner to environmental stimulus, such as temperature changes by magnetic field induction. The (core@shell)-g-polymer design offers advantages in colloidal stability, control over molecular loading and release, and the ability to modulate nanosystems surface properties, providing a versatile platform for applications in controlled drug delivery, hyperthermia magnetic therapy, imaging diagnostics, and tissue engineering.

2. Materials and methods

2.1. Materials

Tetraethyl orthosilicate (98%, TEOS), 3-trimethoxysilylpropyl methacrylate (97%, MEMO), diethylene glycol methacrylate ester (97%, DEGMA), N,N′-methylenebis(acrylamide) (98%, BIS), 2,2-dimethoxy-2-phenylacetophenone (99%, DMPA), ferric chloride hexahydrate (FeCl₃·6H₂O), ferrous chloride tetrahydrate (FeCl₂·4H₂O) were purchased from Sigma-Aldrich. Absolute ethanol, concentrated aqueous ammonia, NaOH and tris(hydroxymethyl)aminomethane (TRIS) were obtained from Biopack, and methanol was purchased from Merck. Ibuprofen (IBU) (500 mg, 100% w/w) was obtained from Farmácia Fórmulas (Santo André, São Paulo, Brazil). All reagents were used as received. The water used was deionized (18 MΩ cm) and filtered.

2.2. Synthesis of smart hybrid nanosystem (SHN)

2.2.1. Synthesis of nanosystem core: magnetic nanoparticles (SPION)

The superparamagnetic nanoparticles (SPION) were synthesized using the chemical coprecipitation method⁴⁴ with ferric chloride hexahydrate (FeCl₃·6H₂O) and ferrous chloride tetrahydrate (FeCl₂·4H₂O) as precursor salts, in a molar ratio of 2 : 1. Ammonium hydroxide (NH₄OH), a weak base, was used to induce the precipitation of the SPION.

Initially, a mixture of 10 mL of an aqueous solution containing FeCl₂·4H₂O (1.0 M) and FeCl₃·6H₂O (0.5 M) in a volumetric ratio of 2 : 8 was subjected to vigorous stirring on a magnetic stirrer at room temperature. Meanwhile, 100 mL of NH₄OH solution (0.7 M) was slowly added with a burette over a period of 30 minutes. The solution immediately changed from brown to black, resulting in the formation of a black solid corresponding to the SPION.

The precipitate formed was magnetically decanted using a neodymium magnet and the supernatant was discarded. Then a coating process was performed with cetyltrimethylammonium bromide (CTAB) to increase the stability of the SPION, prevent aggregation, and to have a template for the formation of the mesoporous SiO₂ shell. For this, 15 mL of a CTAB solution (10% in water) was added to the precipitate, and the system was maintained at 35 °C for 14 hours under constant stirring. The solution was then centrifuged (4186 × g for 1 hour) and the resulting precipitate was resuspended in 10 mL of water.

2.2.2. Coating of SPION with mesoporous silica: SPION@SiO^MP₂ NPs

For the coating of the SPION with a mesoporous silica layer, an additional centrifugation step was performed to remove excess CTAB from the SPION-CTAB solution previously prepared. Specifically, 100 µL of the SPION-CTAB solution was diluted in 12 mL of water and centrifuged at 7441 × g for 1 hour. After discarding the supernatant, the precipitate was resuspended in 12 mL of water.

The mesoporous silica layer (SiO^MP₂) was grown using the sol–gel approach with TEOS as SiO₂ precursor and CTAB as mesoporous template.⁴⁵ First, 300 µL of a 0.1 M NaOH solution (in water) was added dropwise to the aqueous SPION-CTAB suspension. Then, three aliquots of 100 µL of TEOS solution (20% in methanol) were gradually added to the crude reaction mixture, with a 30-minute interval between each addition. After complete reagent incorporation, the solution was stirred at 500 rpm for 24 hours. Subsequently, the mixture underwent thermal treatment in a glycerin bath at 60 °C for 1 hour to consolidate the SiO₂ shell, followed by successive dispersion–centrifugation cycles (2 MeOH and 2 EtOH; 11 627 × g, 45 min), yielding SPION@SiO^MP₂ NPs dispersed in 10 mL of ethanol. The CTAB extraction by the washing steps was corroborated by zeta potential measurement, see SI.

2.2.3. Photoinduced PDEGMA polymerization: (SPION@SiO^MP₂)-g-PDEGMA SHN production

As a preliminary step for the polymerization of SPION@SiO^MP₂, the surface of the NPs was functionalized via a silanization process using 3-trimethoxysilylpropyl methacrylate (MEMO), a coupling agent essential for the polymerization process. To achieve this, 50 µL of concentrated NH₃ was added dropwise to the solution containing the SPION@SiO^MP₂. After 5 minutes of reaction, 78 µmol of a MEMO (10% v/v in ethanol) was added drop by drop to the mixture, and the system was kept under stirring overnight at room temperature. After two consecutive dispersion–centrifugation cycles (7441 × g, 45 min), the MEMO-functionalized SPION@SiO^MP₂ were dispersed in absolute ethanol to obtain a stock suspension of approximately 10% w/v.

Finally, the polymerization step was carried out using DEGMA as monomer and a photoinduced mechanism, all SHN samples were prepared with the same MEMO-functionalized SPION@SiO^MP₂ batch. To control and optimize the size and the architecture of the polymer shell different concentrations of the monomer and the crosslinking agent (BIS) was studied. The DEGMA concentrations evaluated were: 0.025, 0.050, and 0.100 M to control the length of the polymer shell. Besides, two different architectures were explored by changing the amount of the crosslinking agent, microgel like via the addition of BIS, or brush like in absence of BIS. Table 1 detail the experimental conditions evaluated for SHN, (SPION@SiO^MP₂)-g-PDEGMA, production with the quantities of each reagent. Notice that the SHN samples were labeled as SHN_y^x, superscript indicates DEGMA concentration (x = 0.025, 0.500 or 0.100 M) and subscript x specifies BIS presence (y = 1) or absence (y = 0).

Table 1. Experimental conditions evaluated for SHN production (SPION@SiO^MP₂)-g-PDEGMA. SHN samples were labeled as SHN_y^x, superscript (x) indicates DEGMA concentration and subscript is indicating BIS presence (y = 1) or absence (y = 0).

Sample	DEGMA (M)	DEGMA (µL)	BIS 0.03 M (µL)	DMPA 0.017 M (µL)	EtOH (µL)	H2O (µL)
SHN00.025	0.025	67	0	500	7000	7500
SHN10.025		134	500		6500
SHN00.050	0.050	268	0		7000
SHN10.050		67	500		6500
SHN00.100	0.100	134	0		7000
SHN10.100		268	500		6500

The photopolymerization was carried out according to a previously reported method.³³ In a glass vial, the monomer, DEGMA, and 1 mL of stock suspension of MEMO-functionalized SPION@SiO^MP₂ were dissolved in ethanol. For crosslinked SHN, 500 µL of BIS solution in ethanol (0.03 M) was added, microgel like architecture was reached for polymeric shell. While in absence of BIS solution, brush-like polymer was produced. Then, deionized water was incorporated to obtain a mixture of ethanol/water 50 : 50 (final volume 15 mL). The reaction mixture was bubbled with nitrogen (N₂) for 10 minutes to degas the system and after that DMPA solution in ethanol (0.017 M) was poured. Following this, the reaction mixture was bubbled with N₂ for an additional 5 minutes to remove any residual oxygen. The reaction system was then irradiated for 3 hours under stirring using a 15 W black light lamp, 18 inches in length (λ_max = 352 nm). Upon completion of the polymerization, the synthesized smart SHN were centrifuged (7441 × g, 45 min) and redispersed in absolute ethanol for washing. This process was repeated four times. Finally, the SHN were dispersed in absolute ethanol to obtain an approximately 10% w/v stock suspension.

2.3. Physicochemical characterizations of the synthesized materials

2.3.1. Vibrational spectroscopy in the infrared region (FTIR)

FTIR characterization was conducted using a Nicolet iN10 spectrometer, equipped with an MCT-A detector cooled with liquid nitrogen. The samples were prepared on an aluminum slide, drop by drop, with intervals between additions to allow complete evaporation of ethanol before each subsequent drop. A total of 10 applications were made for each sample. Then, they were dried in a vacuum oven at 35 °C for 72 hours. Measurements were performed in reflectance mode, with 64 scans and a resolution of 4 cm⁻¹. For each sample a minimum of four different regions were measured, and the resulting data were processed using Origin 8.5 software.

2.3.2. Transmission electron microscopy (TEM)

The morphology of the SPION@SiO^MP₂ and SHN was characterized using a transmission electron microscope (TEM-ZEISS-EM109T) operating at 180 kV. The samples were prepared by drop-casting ethanol suspensions onto copper grids (400 mesh) coated with a standard carbon film. For each sample a minimum of four different regions were explored, and the size distribution and estimated diameter was calculated on average number of particles using the ImageJ public domain image processing program.

2.3.3. Dynamic light scattering (DLS)

The hydrodynamic diameter (D_h) and the polydispersity index (PDI) of the SHN in each step synthesis were determined using dynamic light scattering (DLS) using the Nanobrook Omni instrument (Brookhaven Instrument Corp.), operating at a 90° angle (λ = 637 nm) and equipped with a controlled temperature cell, results were analyzed using its built-in software package. The samples by diluting a 200 µL aliquot of the synthesized SHN suspensions in absolute ethanol, resulting in a final volume of 2 mL. For each sample, ten measurements were performed, and the parameters hydrodynamic diameter (D_h) and polydispersity index (PDI) were acquired on average. Autocorrelation functions (ACF) were obtained and fitted using the cumulative method via SOP instrument. The D_h was calculated via the Stokes–Einstein equation, assuming a spherical shape, where the size scales in an inverse way with the rate of diffusion. PDI values were obtained from the cumulants analysis, as a dimensionless value of the broadness of the particle size distribution. All measurements were made in triplicate.

Furthermore, to investigate the SHN thermoresponsiveness behavior, D_h was studied as a function of temperature, above and below the LCST. For this, the samples were redispersed in 2 mL of Milli-Q water, requiring a solvent exchange process. The exchange was performed in two centrifugation steps, each one following the next speed ramp: 5867 × g for 25 minutes and then at 16 708 × g for 5 minutes. After redispersing the SHN in water, D_h values were obtained through DLS measurements at 90°, by setting the temperatures to 15 and 60 °C, as extreme temperatures of PDEGMA LCST, with a stabilization time of 15 min. For each temperature, ten measurements were performed, and the parameters hydrodynamic diameter (D_h) and polydispersity index (PDI) were acquired on average. All measurements were made in triplicate.

2.3.4. Small angle X-ray scattering (SAXS)

Small-angle X-ray scattering (SAXS) measurements were performed on colloidal dispersions of SPION-based samples in ethanol using a SAXS Point 5.0 instrument (Anton Paar) equipped with a Cu Kα radiation source. Samples were loaded into 1.5 mm borosilicate capillaries and sealed to prevent solvent evaporation. Data were collected at room temperature at a sample-to-detector distance of 1.6 m, providing a scattering vector range of q = 0.03–2.1 nm⁻¹, where q = 4π sin(θ)/λ and 2θ is the scattering angle. Two-dimensional scattering patterns were azimuthally integrated to obtain one-dimensional intensity profiles I(q). Background contributions from the corresponding solvent (ethanol) and the capillary were measured under identical conditions and subtracted from the sample data. Standard instrumental corrections (detector response, transmission, and geometry) were applied using the instrument software, and the resulting intensity curves were used for structural analysis.

SAXS data analysis and model fitting were carried out in SasView. To describe the hierarchical structure of the dispersions, the experimental profiles were fitted using a model comprising (i) a concentric core–shell–shell spherical form factor representing the primary hybrid nanoparticles and (ii) an additive power-law term to account for low-q excess scattering associated with mesoscale heterogeneities/aggregation.

2.3.5. N₂ adsorption–desorption measurements

The specific surface area of the samples was determined by nitrogen (N₂) physisorption measurements carried out at 77 K using a Quantachrome Nova surface area and porosity analyzer. Prior to the analysis, the samples were degassed under vacuum at 40 °C for 24 h to remove physically adsorbed species. Nitrogen adsorption isotherms were recorded over a wide range of relative pressures (p/p₀), and the specific surface area was calculated according to the Brunauer–Emmett–Teller (BET) method.⁴⁶ The BET surface area was obtained from the linear region of the adsorption isotherm in the relative pressure range of 0.05–0.28 and expressed in m² g⁻¹.

2.3.6. Magnetization measurements

The specific magnetization (M) as a function of the applied magnetic field (H) at room temperature was measured using a Lake Shore 7404 vibrating sample magnetometer (VSM), with a maximum applied field of µ₀H_max = 1.9 T. This experiment was carried out for SHN at each step of the sequence of synthesis, SPION-CTAB, SPION@SiO^MP₂ and (SPION@SiO^MP₂)-g-PDEGMA to investigate whether the magnetics properties were affected by the surrounding shell of the SPION core. Moreover, the iron concentration in the samples was determined using the thiocyanate method.⁴⁷ Briefly, the samples were chemically digested to dissolve the iron species, followed by complexation with thiocyanate ions to form Fe(iii)–thiocyanate complexes. The Fe(iii) concentration was then quantified by UV-vis spectrophotometry at 480 nm using a previously established calibration curve.

2.4. Evaluation of SHN as ibuprofen (IBU) carrier

2.4.1. IBU loading experiments

Ibuprofen (IBU) loading onto the (SPION@SiO^MP₂)-g-PDEGMA (SHN) was performed using the impregnation method. For this, 1 mL of the SHN suspension was centrifuged at 16 708 × g for 10 minutes. After centrifugation, the supernatant (ethanol) was removed, and the resulting precipitate was treated with 0.5 mL of an ethanol IBU solution (10 mg mL⁻¹). The sample was stirred at 200 rpm at room temperature for 1 hour employing a MaxQ Mini 4450 shaker. After this period, the sample was dried at 60 °C for 3 hours to remove the ethanol. This ibuprofen impregnation process was repeated three more times (adding a total of 2 mL of ibuprofen solution to the SHN), resulting in a total of four impregnation steps. After impregnation steps, the IBU-SHN sample was washed with 0.5 mL of ethanol to remove any excess ibuprofen. The ethanol was collected by magnetic decantation using a neodymium magnet, which facilitated the separation of the IBU-SHN from the supernatant. The same procedure was performed using the SPION@SiO^MP₂ for comparison. All measurements were performed in triplicate.

To determinate the total IBU loaded, a release assay was performed using a solvent in which IBU presents a good solubility and, then the IBU amount was calculated from the release solution. Thus, the IBU-SHN sample were redispersed in 2 mL of ethanol and stirred at 150 rpm using the MaxQ Mini 4450 shaker for 72 hours at 50 °C to ensure complete IBU release. For the quantification, the solution was then centrifuged, and the supernatant was collected for analysis. Then, the IBU concentration in the supernatant quantification was performed using UV-vis absorption spectroscopy (Thermo Scientific™ Evolution™ 201/220 UV-visible spectrophotometers) at a wavelength of 212 nm. A calibration curve (1.25–24 µg mL⁻¹) was done and the IBU loading content into SHN was analyzed using eqn (1):

1

2.4.2. IBU in situ release studies

To evaluate the efficiency of (SPION@SiO^MP₂)-g-PDEGMA as a smart carrier for controlled drug release, IBU in situ release assays were performed at three different temperatures. To simulate the human body environment, the IBU-SHN were dispersed in 2 mL of TRIS buffer solution (20 mM, pH 7.8). The samples were exposed to 25, 37, and 50 °C respectively over a period of 3 hours under stirring in MaxQ Mini 4450 shaker. Then, the SHN was separated by centrifuged (16 708 × g for 10 minutes) and the supernatant was collected for analysis. IBU release was quantified using UV-vis spectrophotometry at a wavelength of 264 nm, with a new calibration curve in TRIS buffer solution (20 mM, pH 7.8) (25–450 µg mL⁻¹). All measurements were performed in triplicate, statistical differences between IBU release according to temperature were calculated by applying ANOVA and Tukey post hoc test, p < 0.05. The IBU released amount was calculated using eqn (2):

2

Furthermore, fluorescence experiments were carried out on the centrifugated SHN particles to corroborate the absence of IBU after its release. For this experiment, samples were resuspended into absolute ethanol (2 mL) and fluorescence emissions of the solutions were recorded by using the Fluorescence Spectrometer Edinburgh FS5 equipped with 150 W CW ozone-free xenon arc lamp and quartz cell (1 cm). Samples were excited at λ_ex = 270 nm, and the fluorescence emission were recorded in the range 280–350 nm. Besides, fluorescence emission of the particles before the experiment, IBU solution and ethanol were acquired as controls.

2.5. Magnetic remotely triggered IBU release experiments

2.5.1. Magneto calorimetric assays

Time-dependent calorimetric experiments for determining the Specific Absorption Rate (SAR) were conducted by exposing an aqueous colloidal suspension to radiofrequency (RF) magnetic fields, held in inside a transparent glass Dewar to minimize heat losses. The samples studies were SPION-CTAB (control) and (SPION@SiO^MP₂)-g-PDEGMA. The RF magnetic field generator consisted of a power source–resonator system (Hüttinger TIG 2.5/300). The experiments were carried out at a frequency of 295 kHz and magnetic field amplitude of 52 kA m⁻¹. The temperature was monitored throughout the experiments using an optical fiber sensor positioned at the center of the sample. The sensor was connected to a calibrated signal conditioner (Neoptix) with an accuracy of ±0.1 °C.

The SAR values were calculated from the initial slope of the experimental heating curves using eqn (3), where C is the volumetric heat capacity of the solvent 4.18 J cm⁻³ K⁻¹ and [x] is the iron concentration in the colloid.

3

Additionally, the iron concentration in the samples was determined using the same protocol in Section 2.3.5.

2.5.2. Remotely triggered IBU release assays

Remotely triggered drug release experiments were performed using the same conditions described in magnetic calorimetric assay (Section 2.5.1). The experiments were carried out at a frequency of 295 kHz and a magnetic field amplitude of 52 kA m⁻¹. It is important to notice that the same IBU loading protocol and IBU-SHN concentration sample than in situ conventional thermal release experiments were followed to compare if magnetic trigger induce similar releases than external conventional heating (see Section 2.4.2). A portion of 0.5 mL of the IBU-SHN suspension was placed inside a transparent glass Dewar to minimize heat losses. The sample was exposed to the magnetic field for 15 min and then the particles were separated from the supernatant via centrifugation (13 201 × g, 30 min). The supernatant was used to quantify the IBU release by UV-vis absorption spectroscopy at a wavelength of 264 nm, employing the calibration curve specified in Section 2.4.2.

3. Results and discussion

3.1. Synthesis and characterization of smart hybrid nanosystems, SHN

The SHN were synthetized combining and integrating three building blocks into a unique nanosystem with an architecture (core@shell)-g-polymer, an SPION core coated by a mesoporous silica layer (SiO^MP₂) further functionalized with a thermoresponsive polymer (PDEGMA). The chemical coprecipitation method,⁴⁴ using iron salts (Fe²⁺ and Fe³⁺), enabled the rapid, cost-effective, and efficient synthesis of the SPION core. Its stabilization with CTAB resulted in lower polydispersity, and a template to deposit a SiO^MP₂ layer. The SiO^MP₂ shell was subsequently functionalized with a surface coupling agent and subjected to polymerization via free radical incorporation, initiated by light (photoinduced),^33,34 leading to the formation of (SPION@SiO^MP₂)-g-PDEGMA SHN capable of responding to external stimuli.

Scheme 1 presents a simplified description of the synthesis process of (SPION@SiO^MP₂)-g-PDEGMA. Notice that the polymer layer has been represented by two different structures corresponding to the amount of BIS used for the synthesis. When no crosslinked agent was added to the polymerization reaction, the structure was obtained like brushes, associated to “tethered polymer chains” or “end-grafted polymers structure”. However, when BIS was present during polymerization, the structure produced was a crosslinked network, microgel-like.

Scheme 1. Schematic of the synthesis of (SPION@SiO^MP₂)-g-PDEGMA with a (core@shell)-g-polymer structure produced, the two polymers nanoarchitecture brush-like and microgel-like are showed.

Specifically, the growth of the mesoporous silica (SiO^MP₂) layer around the SPION core was carried out through a template-assisted sol–gel method⁴⁵ using tetraethyl orthosilicate (TEOS) as the precursor under alkaline hydrolysis–condensation conditions, in the presence of methanol, and CTAB as a pore-directing template. Subsequently, CTAB was removed by extraction with methanol, leading to free pores for further modification, which was corroborated by zeta potential (Z_pot), see Table S1, SI. During the synthesis of SPION@SiO^MP₂, CTAB acts as a structure-directing agent and is strongly adsorbed onto the silica surface through electrostatic interactions between its positively charged quaternary ammonium groups (CTA⁺) and the surface silanol groups (Si–OH). This adsorption leads to the formation of a positively charged layer surrounding the particles, which is reflected in markedly positive Z_pot values prior to surfactant extraction, indicating the presence of CTAB still associated with the surface and confined within the mesoporous structure.⁴⁸ After CTAB extraction, Z_pot of the SPION@SiO^MP₂ sample was −33.9 ± 3.4 mV, which is associated to negatively charged silica surface due to Si–O⁻ groups (pK_a,SiOH,Q3 = 2).⁴⁹ The efficient removal of CTAB exposes the intrinsic silanol groups of the silica matrix, which tend to deprotonate in aqueous media, generating negatively charged siloxide groups (Si–O⁻). Consequently, the surface charge of the nanoparticles shifts from positive to negative, explaining the observed inversion of the zeta potential after methanol washing. This negative surface charge promotes electrostatic repulsion between particles, thereby contributing to the colloidal stability of the system.⁵⁰

Complementary information on the surface and textural properties of SPION@SiO^MP₂ was obtained by N₂ adsorption–desorption measurements (Fig. 1). The isotherms reveal distinct behaviors for the SPION@SiO^MP₂ and SPION-CTAB samples (control), reflecting differences in porosity and structural organization. In the low relative pressure region (p/p₀ = 0.05–0.28), used for the calculation of the specific surface area by the BET (Brunauer–Emmett–Teller) method, the SPION@SiO^MP₂ sample exhibits the highest N₂ adsorption. The SPION-CTAB sample presents a specific surface area value of 58 m² g⁻¹, which is lower than that of SPION@SiO^MP₂ (67 m² g⁻¹). This result is associated with the presence of the SiO^MP₂ shell, which introduces intrinsic porosity into the system, typically within the mesoporous range, thereby increasing surface accessibility to the adsorbate and contributing significantly to the enhancement of the BET surface area.^28,51

Fig. 1. N₂ adsorption–desorption isotherms of SPION@SiO^MP₂ and SPION-CTAB (control) measured at 77 K. The shaded region (p/p₀ = 0.05–0.28) indicates the pressure range used for BET surface area calculation. SPION@SiO^MP₂ exhibits a higher specific surface area (67 m² g⁻¹) than SPION-CTAB (58 m² g⁻¹).

In contrast, in the high relative pressure region (p/p₀ > 0.8), a more pronounced increase in the adsorbed volume is observed for the SPION-CTAB sample. This behavior is characteristic of interparticle porosity, also referred to as textural porosity, arising from voids between particles and aggregates.^52,53 Although CTAB acts as a surface-stabilizing agent, it is not expected to promote the formation of an organized and permanent porous network on the SPION surface, resulting in a lower specific surface area despite the higher adsorption volume at high relative pressures.^54,55 After the SiO^MP₂ coating, the contribution of interparticle porosity is significantly reduced, as evidenced by the decreased volume adsorbed in the high relative pressure region. This effect can be attributed to an increase in the effective particle size, the partial filling of interparticle voids, and the formation of a continuous SiO^MP₂ shell, which promotes more compact particle packing. Consequently, N₂ adsorption in the SPION@SiO^MP₂ material becomes dominated by the intrinsic porosity of the silica layer, leading to a higher specific surface area even though a lower total adsorbed volume is observed at high relative pressures.

The polymerization process was performed in two stages following an adapted version of protocol previously reported by the research group.^33,34,41 In the first step, the SPION@SiO^MP₂ was surface functionalized through a silanization process in a basic medium by directly adding trimethoxypropyl methacrylate (MEMO) to the suspension, resulting in MEMO-functionalized SPION@SiO^MP₂ yearling vinyl groups. In the second step, the immobilized vinyl groups acted as coupling agent to introduce reactive groups on the SPION@SiO^MP₂ surface, allowing the incorporation of the polymer through these anchoring sites. The polymerization was carried out using the DEGMA (di(ethylene glycol) methacrylate) as a monomer, BIS as the crosslinking agent, DMPA as the photoinitiator for free radical generation, and deionized water/absolute ethanol as the solvent system at 25 °C under UV irradiation. This protocol allowed the synthesis of (SPION@SiO^MP₂)-g-PDEGMA SHN via an uncontrolled free radical polymerization mechanism making these factors key for a low-cost, scalable industrial synthesis. As mentioned above, to control the polymer architecture and thus optimize the thermoresponsiveness behavior, three monomer concentration were studied in presence or absence of BIS. The monomer concentration determines the thickness of the polymer shell; the BIS presence leads to a crosslinked network microgel-like structure, and its absence produces a brush-like associated to “tethered polymer chains” or “end-grafted polymers structure” layer. In total six SHN samples were produced and labeled as SHN_y^x, superscript x indicates DEGMA concentration (0.025, 0.500 or 0.100 M) and subscript is indicating BIS presence (y = 1) or absence (y = 0), see Table 1.

To evaluate the composition and confirm the synthesis of the (core@shell)-g-polymer nanostructures, FTIR analyses were performed at the different step synthesis. Fig. 2a shows the FTIR spectra of SPION@SiO^MP₂ (black spectrum) and MEMO-functionalized SPION@SiO^MP₂ (red spectrum). In both spectra, main bands corresponded to the SiO^MP₂ shell were observed: an intense band 1084 cm⁻¹ corresponding to the Si–O–Si bond vibration, a band at 1640 cm⁻¹ attributed to the bending vibration of OH group from adsorbed water, and a broad band between 3000 cm⁻¹ and 3600 cm⁻¹ associated with adsorbed water and the stretching vibrations of the Si–OH hydroxyl groups. Moreover, signals corresponding to –CH₂ bonds were observed in the SPION@SiO^MP₂ spectrum due to the presence of residual ethoxy groups resulting from the partial hydrolysis of TEOS during the sol–gel synthesis process. This FTIR information confirmed the successful synthesis of SPION@SiO^MP₂. Additionally, in the MEMO-functionalized SPION@SiO^MP₂ spectra the appearance of two new bands was detected: 1165 cm⁻¹ and 1150 cm⁻¹, characteristic of methoxy groups bonded to silicon (Si–O–C), indicating the presence of active silane groups. The increased intensities of the bands corresponding to the asymmetric (2929 cm⁻¹) and symmetric (2858 cm⁻¹) stretching vibrations of the –CH₂ group further support successful silanization. Thus, the MEMO-functionalized SPION@SiO^MP₂ preparation was evidenced. This stepwise spectral evolution is consistent with our previous FTIR studies on silica-coated nanoparticles, reinforcing the reliability of the synthetic route and the robustness of the inorganic core@shell interface.

Fig. 2. FTIR spectra of the synthesized systems (a) SPION@SiO^MP₂ (black spectrum), MEMO-functionalized SPION@SiO^MP₂ (red spectrum), (b) (SPION@SiO^MP₂)-g-PDEGMA SHN without BIS (SHN₀^0.050 – black spectrum) and (SPION@SiO^MP₂)-g-PDEGMA SHN with BIS (SHN₁^0.050 – red spectrum).

After the polymerization step, new bands characteristic of PDEGMA appeared in the FTIR spectra of the (SPION@SiO^MP₂)-g-PDEGMA SHN synthetized in BIS presence (SHN₁^0.050 – red spectrum) and without BIS (SHN₀^0.050 – black spectrum), as shown in Fig. 2b. The band at 1722 cm⁻¹ corresponds to the stretching of the –C O bond, as well as the significant increase in the bands associated with the asymmetric (2929 cm⁻¹) and symmetric (2858 cm⁻¹) stretching of the –CH₂ group further confirms the presence of the PDEGMA shell around the SPION@SiO^MP₂.⁴¹ Additionally, for samples synthesized in BIS presence an extra band at 1540 cm⁻¹ corresponding to BIS was detected. This band was associated with the deformation of the –NH group in the amide structure of acrylamide. This FTIR results confirmed the successful PDEGMA polymer layer formation on the MEMO-functionalized SPION@SiO^MP₂ in both cases microgel (BIS presence) and brush-like (BIS absence) shape. It is important to notice that same spectra were obtained for the rest of the SHN samples, see SI, Fig. S1. Furthermore, the band assignments for all FTIR spectra data are indicated in Table S2 (SI).

The morphology of the synthesized SPION, as well as the thickness of SiO^MP₂ and PDEGMA layers, were determined using Transmission Electron Microscopy (TEM). Fig. 3a shows the TEM image and the core size distribution histogram of SPION@SiO^MP₂. It can be observed that the nanoparticles have an almost isotropic shape, with a dark gray region corresponding to the SPION core, surrounded by a light gray continuous layer corresponding to the SiO^MP₂ shell. The observed aggregation is very common in iron oxide nanoparticles subjected to TEM analysis due to the attractive interactions between the SPION particles and the carbon film on the copper grids, as well as the agglomeration caused by the high vacuum applied during the analysis. The size distribution histogram showed that the SPION@SiO^MP₂ system presents a SPION core with an average diameter of 10 ± 2 nm and a SiO^MP₂ shell thickness of about 3 ± 1 nm. These values are in good agreement with previous TEM studies from our group on similar SPION-based core@shell systems, confirming the reproducibility of the synthetic approach and the formation of well-defined nanostructures.⁵⁶ Regarding the SHN, the TEM image for sample SHN₀^0.050 (Fig. 3b) also revealed nearly spherical shape with a dark gray SPION core and a light gray layer surrounding it. In this case, the light gray layer exhibited a greater thickness than the SPION@SiO^MP₂ sample and it could be attribute to the presence of PDEGMA polymer shell. According to the core size distribution histogram, the SPION core had a diameter of 9.6 ± 1.8 nm, and the layer corresponding to the SiO^MP₂ coated with PDEGMA shell had a thickness of 5.9 ± 0.8 nm, where 3.0 ± 0.5 nm corresponds to the SiO^MP₂ shell and 3 ± 1 nm to the external PDEGMA shell. It is important to mention that the PDEGMA was grafted inside and outside the SiO^MP₂, because the functionalization with MEMO and so with the PDEGMA took place both inside and outside the pores. Thus, the 3 nm layer corresponds to the polymer on the external surface, while part of the PDEGMA is confined into the pores.

Fig. 3. (a) TEM image of SPION@SiO^MP₂ with its respective SPION core size distribution histogram. (b) TEM image of (SPION@SiO^MP₂)-g-PDEGMA SHN for the SHN₀^0.050 system with SPION core respective size distribution histogram.

The DLS technique is an essential tool for the analysis of colloidal suspensions, providing vital information on the particle size via the hydrodynamic diameter (D_h) and about the particles aggregation or size distribution by polydispersity (PDI). DLS measurements were performed using absolute ethanol as solvent because it is well-known that the polymers have the maximum possible stretching under these conditions.⁴¹Table 2 presents the values of D_h and PDI for SHN samples and controls without the polymer.

Table 2. Hydrodynamic diameter (D_h), polydispersity index (PDI), for each of the synthesized SHN and controls. The SHN samples were labeled as SHN_y^x, superscript x indicates the molar concentration of DEGMA, and the subscript indicates BIS presence (y = 1) or absence (y = 0). Besides, SAXS primary diameters are presented and calculated as D_primary = 2 × (R_core + tSiO^MP₂ + t_PDEGMA).

Sample	[DEGMA] (M)	Polymer structure	D h (nm)	PDI	SAXS primary diameter Dprimary (nm)
SPION	—	—	338 ± 9	0.23 ± 0.01	—
SPION-CTAB	—	—	225 ± 4	0.24 ± 0.01	—
SPION@SiOMP2	—	—	204 ± 3	0.28 ± 0.02	—
MEMO-functionalized SPION@SiOMP2	—	—	227 ± 4	0.21 ± 0.02	—
SHN00.025	0.025	Brush-like	240 ± 6	0.23 ± 0.02	25.241
SHN10.025	0.025	Microgel-like	269 ± 4	0.26 ± 0.01	26.518
SHN00.050	0.050	Brush-like	266 ± 3	0.24 ± 0.01	29.994
SHN10.050	0.050	Microgel-like	294 ± 14	0.24 ± 0.04	29.147
SHN00.100	0.100	Brush-like	302 ± 4	0.28 ± 0.02	32.553
SHN10.100	0.100	Microgel-like	300 ± 3	0.25 ± 0.03	27.307

The naked SPION have a relatively large D_h (338 ± 9) nm, this could be attributed to the strong attractive interactions such as van der Waals forces, depletion forces, and hydrophobic interactions occur between the oxide molecules and the surrounding medium, leading to the formation of large aggregates.^57,58 For the SPION-CTAB sample D_h was (225 ± 4) nm due to the stabilization of the core with CTAB, which reduces aggregation. The CTAB molecule interacts with the SPION surface, generating electrostatic repulsion between them due to the positive charge in its hydrophobic tail. In contrast, the SPION@SiO^MP₂ sample exhibited an even smaller D_h, (204 ± 3) nm, compared to SPION-CTAB. This is a consequence of the SiO^MP₂ layer around the SPION providing a physical barrier between them, leading to greater stabilization. As well as the silica is polar and enhances its interaction with the polar solvent. Additionally, the PDI values remained constant at each stage, indicating that the colloidal system was not significantly affected by the synthesis steps.

During the formation of the SHN, an increase in D_h was observed due to the polymeric layers deposited, while PDI values remaining consistent, indicating colloidal stability was not affected by the MEMO functionalization and the photopolymerization process. On one hand, D_h increased when DEGMA concentration was higher, which was associated with a larger polymer layer mainly detected for brush-like structure. The D_h was (240 ± 6) nm, (266 ± 3) nm and (302 ± 4) nm for samples SHN₀^0.025, SHN₀^0.050 and SHN₀^0.100, respectively. While, for microgel-like structure, this trend was less evident due to the swelling of the crosslinked polymer layer. On the other hand, for the systems synthesized in BIS presence, the D_h increased compared to the others synthesized with the same monomer concentration but without BIS. For example, D_h of sample SHN₁^0.025 was (269 ± 4) nm, while for sample SHN₀^0.025 was (240 ± 6) nm. Also, D_h of sample SHN₁^0.050 was (294 ± 14) nm, while for SHN₀^0.050 was (266 ± 3) nm. As mentioned, the BIS presence allowed to produce a crosslinked polymer network (microgel-like),³⁴ which could cause the increase detected in the D_h values due to its swelling. However, for systems synthesized with a higher amount of DEGMA (0.100 M), no significant difference in D_h was observed between those with (SHN₁^0.100 (300 ± 3) nm) or without BIS (SHN₀^0.100 (302 ± 4) nm). This could be attributed to as the brushes that grew from inside the pores are the longest and the excesses polymer surrounding the SPION@SiO^MP₂ can form interactions that create a dense and compact shell, physical crosslinking mimic a microgel structure. As a result, the polymer chains can occupy three-dimensional spaces, making it difficult to observe the BIS effect.

Furthermore, Fig. S2 shows the autocorrelation functions (ACF) together with the corresponding intensity-averaged size distributions obtained by DLS for the samples listed in Table 2. In all cases, the ACF decayed to zero at times longer than 80 µs, allowing the exclusion of the presence of large aggregates as well as additional relaxation processes. Furthermore, comparing the DLS data with TEM results, we observe a significant difference. This discrepancy can be attributed to the hydration of SHN in suspension. When the polymer is in contact with polar solvents like ethanol or water, as was the case during the DLS analysis, its polymer chains are stretched and there is a predominance of polymer–solvent interactions, leading to an increase in polymer thickness.

SAXS profiles for the SHN and control samples display a smooth decay of scattered intensity with a clear low-q upturn, Fig. 4. Considering the accessible q-range (0.03–2.1 nm⁻¹), the experiment probes characteristic real-space lengths from about 210 nm down to about 2 nm. Therefore, the measured SAXS curves are primarily sensitive to the internal architecture of the primary hybrid nanoparticles (core and shells, tens of nanometers), while larger-scale organization (clusters/aggregates, hundreds of nanometers) is captured mainly through the low-q upturn. This hierarchical response is consistent with DLS results, which report D_h in the 200–300 nm range for the SHN dispersions (Table 2), indicating that DLS is dominated by solvated entities and/or clusters rather than the primary particle size alone.

Fig. 4. Overlay of SAXS I(q) curves for SPION@SiO^MP₂ (control) and the SHN series.

To quantify the primary particle architecture while accounting for mesoscale heterogeneity, SAXS data were fitted using a spherical core–shell–shell form factor combined with an additive power-law term, see Fig. S3. This approach is commonly used for dispersions where a well-defined primary object coexists with aggregate-like contributions at low q.^59,60 In order to reduce parameter correlation and maintain consistency with microscopy, the inorganic dimensions were constrained using TEM. Specifically, the SPION core radius was constrained to 5.0 nm (consistent with a 10.0 ± 1.8 nm core diameter) and the SiO^MP₂ thickness was constrained to 3.0 nm (3.0 ± 0.5 nm). These constraints anchor the most reliable geometric features (TEM) and allow SAXS to quantify the solvated PDEGMA corona as an effective outer shell thickness (t_PDEGMA) and an effective polymer scattering length density (SLD) that reflects polymer/ethanol volume fraction within the corona.

With the inorganic (SPION@SiO^MP₂) dimensions constrained, the fitted external PDEGMA-shell thickness t_PDEGMA spans 4.620–8.276 nm across the SHN series (Table S3). This corresponds to primary particle diameters in the 25–33 nm range, computed as: D_primary (nm) = 2 × (R_core + tSiO^MP₂ + t_PDEGMA) = 16 + 2 × t_PDEGMA for the TEM-constrained geometry (R_core = 5.0 nm; tSiO^MP₂ = 3.0 nm). These primary sizes are much smaller than the DLS hydrodynamic diameters, supporting a hierarchical picture in which the dispersion contains primary SHN that is organized into larger solvated entities/clusters.

TEM analysis for SHN₀^0.050 indicates an external PDEGMA thickness of 2.9 ± 1.0 nm, while SAXS yields larger effective shell thicknesses for the solvated corona. This difference is physically expected because TEM probes a dry, vacuum state and tends to underestimate the extension of soft polymer layers, whereas SAXS probes the radial average electron-density profile in solution, where the polymer corona is swollen and solvent-permeable. In addition, the PDEGMA was grafted both inside and outside the SiO^MP₂ shell, so the “external polymer thickness” from TEM should not be interpreted as the total polymer contribution sensed by SAXS; rather, SAXS provides an effective, contrast-dependent shell representation.

In the brush-like (without BIS) series SHN₀^0.025, SHN₀^0.050 and SHN₀^0.100, t_PDEGMA increases monotonically from 4.620 nm to 6.997 nm and 8.276 nm. This trend is consistent with increased polymer content and/or a more developed solvated corona at higher DEGMA concentration. The effective polymer SLD values for brush-like samples (SLD_2,eff = 8.3–8.6 in units of 10⁻⁶ Å⁻², as used in SasView) remain closer to ethanol SLD (7.45), indicating a relatively solvent-rich corona.

In the BIS-crosslinked series (microgel-like polymer structure), t_PDEGMA values are 5.259 nm, 6.573 nm, and 5.653 nm, for SHN₁^0.025, SHN₁^0.050 and SHN₁^0.100, respectively. The effective polymer SLD is higher (SLD_2,eff = 9.3–9.6), indicating a more compact, PDEGMA-rich corona with reduced ethanol volume fraction, consistent with the expected effect of crosslinking. These static structural features provide a nanoscale basis to rationalize the architecture-depending on chain length, grafting density, and polymer architecture.

The low-q upturn was modeled by a power-law contribution I(q) proportional to q^−m, with fitted exponents m between 2.75 and 3.05 (Table S3).⁶¹ Exponents close to 3, as observed here, are consistent with relatively compact aggregate-like heterogeneities (D_f close to 3) and/or crossovers between mass-fractal and rough-interface regimes. Given the limited low-q window and the presence of multiple structural levels, m is used here as a comparative descriptor of mesoscale organization rather than as a unique, model-independent fractal dimension.⁶⁰ Importantly, the presence and magnitude of this low-q component provide a structural rationale for why DLS reports D_h in the 200–300 nm range while SAXS simultaneously resolves primary particle dimensions in the tens-of-nanometers regime. It is worth noting that when combined with TEM analysis, SAXS and DLS data, the results supports a model of compact inorganic cores surrounded by a swollen, silica and thermoresponsive polymer corona in dispersion, a structural feature crucial for tuning colloidal stability and drug-release behavior.

An important characteristic of SPION as drug carriers in biomedical applications is their superparamagnetic behavior, which takes place when single magnetic domains are present in nanoparticles with diameters ranging from 10 nm to 20 nm. This behavior enables the SPION to show magnetization only when an external magnetic field is applied; once the field is removed, no residual magnetization is observed. Fig. 5 shows the magnetization curves of bare SPION (black curve), SPION@SiO^MP₂ (red curve) and (SPION@SiO^MP₂)-g-PDEGMA (blue curve) measured under isothermal magnetic conditions at room temperature. The bare SPION exhibited a typical superparamagnetic response at room temperature, with negligible remanence and zero coercivity, reaching a saturation magnetization of approximately 60 emu g⁻¹, in good agreement with values reported for magnetite nanoparticles in the literature.^62,63 After surface modification, SPION@SiO^MP₂ and (SPION@SiO^MP₂)-g-PDEGMA presented saturation magnetization at 53 and 65 emu g⁻¹, respectively. Both samples maintained the characteristic superparamagnetic behavior observed for the bare SPION, as evidenced by the absence of magnetic hysteresis, remanent magnetization and coercive field. Despite the presence of non-magnetic SiO^MP₂ and PDEGMA shells, the overall magnetic response of the SHN remained preserved, indicating that the surface functionalization processes did not compromise the magnetic nature of the magnetite core.

Fig. 5. Magnetization curves of the SPION, SPION@SiO^MP₂ and (SPION@SiO^MP₂)-g-PDEGMA samples with a maximum applied field of µ₀H_max = 1.9 T.

3.2. Thermoresponsiveness studies of the synthesized SHN

To evaluate thermoresponsive behavior of the SHN, the D_h was analyzed before and after exposure of the samples to temperature stimuli using DLS measurements. It is well established that the LCST of PDEGMA depends on the chain length and the architecture polymer architecture (e.g., brushes, chains, or gels), as well as the presence of comonomers.⁴³ Below the LCST, PDEGMA exhibits hydrophilic behavior, remains stretched and soluble in the surrounding medium promoting polymer swelling. Conversely, at temperatures above the LCST, PDEGMA becomes hydrophobic and insoluble, causing a collapse of the polymer chains. Wassel et al. and Voß et al. investigated the thickness-dependent thermoresponsive LCST behavior of PDEGMA brushes and demonstrated a linear relationship between brush thickness and LCST values.^64,65 Their studies further showed that shorter PDEGMA chains exhibit LCST values close to 35 °C. Considering these previous findings, D_h measurements were performed at 15 and 60 °C, temperatures selected as the lower and upper extremes relative to the LCST behavior of PDEGMA. Fig. 6 presents the results of the thermoresponsive behavior analysis, together with a schematic representation of each SHN architecture, which supports the response observed.

Fig. 6. (a) SHN D_h variation in response to temperature, 15 °C (orange) and 60 °C (blue). (b) Highlighting SHN D_h with the best thermoresponsiveness (SHN₀^0.050). (c) Schematic representation of the nanoarchitecture of the six SHN.

In the SHN synthesized with at the lowest DEGMA concentration (0.025 M), no thermoresponsive behavior was observed by DLS. The D_h values obtained at 15 °C and 60 °C were similar for SHN samples with and without BIS. This behavior may be associated with a low PDEGMA content predominantly confined within the mesopores, resulting in a narrow shell that incorporates a limited solvent volume; therefore, the D_h remains practically unaffected by temperature changes. When the highest DEGMA concentration (0.100 M) was employed, thermoresponsiveness was only observed for the SHN synthetized in BIS presence, corresponding to a microgel-like structure. As previously mentioned, a high DEGMA concentration promotes the formation of a dense and disorganized polymer layer surrounding the SPION@SiO^MP₂ which may adopt a three-dimensional conformation even in the absence of BIS, see schematic representation in Fig. 6c. Although the sample SHN₀^0.100 could form physical crosslinks, the mobility of the PDEGMA brushes is restricted by their confinement, thereby inhibiting a thermoresponsive behavior. However, for the microgel-like sample (SHN₁^0.100), the formation of a crosslinked network promotes a better organization of the polymer chains, allowing them to stretch and collapse in response to temperature variations. Interestingly, when an intermediate DEGMA concentration (0.050 M) was used, excellent thermoresponsive behavior was observed in the SHN without BIS (SHN₀^0.050). In this case, the optimized amount of PDEGMA ensures proper organization of the polymer chains around the SPION@SiO^MP₂, allowing them to freely expand and contract in response to temperature changes, Fig. 6c. Conversely, in the presence of the crosslinking agent (SHN₁^0.050), no thermoresponsive behavior was detected, as the polymer chains became crosslinked, restricting their mobility and limiting their thermal response. Furthermore, it is important to highlight that this thermoresponsive behavior is fully reversible. When the temperature returns to ambient conditions, the samples recover their initial D_h values. Notably, sample SHN₀^0.050 presented the best thermoresponsive behavior, Fig. 6b, and additionally the porosity of SiO^MP₂ shell may be less hindered than in systems coated with a crosslinked polymer network.

These results are supported by the structural information obtained by FTIR data, DLS and SAXS results at room temperature, Fig. 6c represents the nanoarchitecture expected according to these data. It is important to notice that SAXS measurements in this work were acquired at a single temperature and in ethanol; therefore, SAXS does not directly demonstrate LCST-driven coil-to-globule transitions (polymer collapse/expansion), which would require variable-temperature scattering measurements in the same medium. Nevertheless, the room-temperature SAXS results provide independent structural constraints on the solvated corona, including its thickness (t_PDEGMA) and compactness (SLD_2,eff), and reveal clear architecture-dependent differences between brush-like and microgel-like shells. These static nanoscale descriptors are consistent with the literature describing how polymer architecture and chain organizations govern thermoresponsive behavior in PDEGMA-based systems.

3.3. SHN as potential ibuprofen drug carriers, in situ loading and release experiments

Ibuprofen (IBU) loading capacity and controlled release of the synthesized SHN was studied as a model drug for magneto-triggered delivery. IBU is a nonsteroidal anti-inflammatory drug (NSAID) with low water solubility, ranging from 0.06 mg mL⁻¹ at pH 1.2 to 2.3 mg mL⁻¹ at pH 7.4, due to its hydrophobic nature.⁶⁶ Considering that the SHN₀^0.050 sample exhibited the highest thermoresponsiveness and additionally the porosity of SiO^MP₂ shell is likely less restricted than in systems coated with a crosslinked polymer network, all experiments in this section were conducted using this SHN.

The incorporation of drugs into SHN is a challenging process due to the repulsion caused by the hydrophobic groups of the drug, which arises from the presence of the polymer layer within the pores and the confined charges.^17,67 In this work, IBU loading was achieved by using a protocol consisting of four consecutive impregnations^68,69 each with 0.5 mL of an IBU solution in ethanol (10 mg mL⁻¹), on the synthesized SHN₀^0.050. It is important to notice that during the impregnation protocol, IBU is preconcentrated into the SHN surface, and then, when the solvent (ethanol) is evaporated, the PDEGMA brushes collapse, as detected by DLS, thereby opening the mesoporous, allowing the drug diffuse into the pores. When the system returns to room temperature, the brushes stretch fully, encapsulating and trapping the drug inside the SHN pores. According to the quantification performed using UV-vis spectrophotometry, the amount of IBU loaded onto SPION@SiO^MP₂ was 1185 µg of drug per gram of sample, while SHN₀^0.050 sample, showed 1593 µg of drug per gram of SHN. After all impregnation steps, both systems demonstrated excellent efficiency in IBU loading, with the SHN system showing an approximately 30% higher IBU loading capacity, due to the PDEGMA chains being able to retain more of the drug within the mesoporous silica structure.

In situ release experiments were performed redispersing by the IBU-SHN into fresh buffer TRIS solution and studied at 25, 37 and 50 °C, which corresponds to temperatures typically of laboratory storage, body application and above the PDEGMA LCST. Fig. 7 and Table S4 (SI) present the results of IBU release by sample amount (µg IBU per mg of sample) and the percentage of IBU released, respectively. The same protocol was carried out with IBU loaded onto SPION@SiO^MP₂ as control.

Fig. 7. (a) IBU released (µg) per sample amount (SPION@SiO^MP₂ and SHN₀^0.050), at different temperatures (25, 37, and 50 °C). Statistical differences between IBU release according to temperature were calculated by applying ANOVA and Tukey post hoc test, p < 0.05. (b) Fluorescence emission spectrum of SHN₀^0.050 sample before and after IBU release experiment. (c) Fluorescence emission spectrum of SPION@SiO^MP₂ sample before and after IBU release experiment.

The results clearly demonstrate the influence of the polymer's thermoresponsiveness on IBU release. Although SPION@SiO^MP₂ can load IBU due to its mesoporous structure, no significant changes were observed (p > 0.05) in IBU release at different temperatures (see Table S4, SI). This agrees with a previous report by Charnay et al. in which rapid IBU release from MCM-41 nanoparticles, reaching a plateau after 45 min at 91%, which was associated with a free diffusion.⁶⁸ In contrast, for SHN₀^0.050, a significant increase in the percentage of IBU release was observed with the increase in temperature (p > 0.05), demonstrating that the thermoresponsiveness of the PDEGMA (see Table S4) is involved in the drug release. The lowest percentage of IBU release was observed at the lowest temperature (25 °C), which is below the LCST of PDEGMA. Under these conditions, the PDEGMA chains remain stretched and act as a physical barrier, limiting the release of the drug. As the temperature increases, these chains begin to collapse, reducing the diffusion barrier for IBU and facilitating drug release from the SHN pores. The highest percentage of IBU release was observed at 50 °C (above the LCST), where the polymer is fully collapsed, allowing for more efficient drug diffusion.

Furthermore, fluorescence emissions spectra of the samples before and after IBU release were acquired (Fig. 7b and c). Fig. S4 presents the fluorescence emission spectrum of an IBU solution (5 mg mL⁻¹) and ethanol as controls. Samples were excited at λ_ex = 270 nm, and the fluorescence emission was recorded in the range from 280 nm to 350 nm. The fluorescence emission obtained for IBU solution agreed with the literature presenting a maximum of emission at λ_em = 286 nm.⁷⁰ Before the release experiment, the characteristic emission band at 286 nm was clearly detected for both IBU loaded samples (SHN₀^0.050 and the control SPION@SiO^MP₂), confirming successful drug incorporation into the nanocarriers. However, the initial fluorescence intensity of SHN₀^0.050 was considerably higher, indicating greater loading efficiency compared to control sample SPION@SiO^MP₂, which agrees with the IBU quantification. After the release process, a pronounced decrease in emission intensity was observed, particularly for SHN₀^0.050. The emission profile of this sample showed clear temperature dependence: at 25 °C, a residual fluorescence signal was still detected, consistent with partial drug retention; at 37 °C and 50 °C, the signal was drastically reduced, indicating efficient thermally triggered drug release. It is important to note that the residual emission bands detected after release were attributed to ethanol, as confirmed by the control spectra (Fig. S4). In contrast, SPION@SiO^MP₂ exhibited only minor changes in fluorescence intensity after the release experiment and no evident temperature-dependent behavior, suggesting limited responsiveness and lower release efficiency. Together, these results reinforce that the PDEGMA shell in the SHN enhances IBU loading, as well as enables an efficient temperature-controlled release mechanism compared to the SPION@SiO^MP₂ system.

3.4. Studies on magnetic-triggered IBU release, proof-of-concept

To characterize the potential of the SHN colloids for magnetic heating, the temperature of the samples was measured while they were exposed to RF magnetic fields, and the Specific Absorption Rate (SAR) values were subsequently calculated from the obtained heating curves. Fig. 8 shows the temperature vs. time curve for (SPION@SiO^MP₂)-g-PDEGMA (SHN₀^0.050) and SPION-CTAB (control).

Fig. 8. Temperature vs. time dependence for SAR values calculation of SPION-CTAB as control (a) and (SPION@SiO^MP₂)-g-PDEGMA, SHN₀^0.050 (b) colloids exposed to RF magnetic fields using H = 52 kA m⁻¹, and f = 295 kHz. Notice that the [Fe₃O₄] concentration for both samples indicated was calculated by UV-vis. (c) Comparison of the amount of IBU released according to the type of trigger considering the same experiment time. Inset the experimental set-up of trigger condition.

In both cases, a clear temperature increase was observed, enabling the measurement of the heating ramps and the subsequent determination of the SAR values. For SPION-CTAB the SAR measured value was 10(1) W g⁻¹ (Fig. 8a), while for (SPION@SiO^MP₂)-g-PDEGMA (SHN₀^0.050) the SAR was 1.9(1) W g⁻¹ (Fig. 8b). The obtained SAR values of the SHN₀^0.050 indicate moderate heating efficiency under the applied magnetic field conditions. As was expected, the SAR value for SHN sample is smaller than the control, associated with the shells surrounding the SPION core, SiO^MP₂ and PDEGMA.

A proof of concept was demonstrated for magnetic remotely triggered IBU release. In this experiment, the IBU-loaded SHN samples were prepared following the impregnation protocol described in Section 3.3, yielding 1593 µg of IBU loaded per gram of SHN. Prior to the experiment, the IBU-loaded SHN was redispersed in fresh TRIS buffer solution and preheated to 37 °C. The dispersion was then transferred into a Dewar vessel at room temperature, which caused an initial cooling of the sample to 33 °C before the RF magnetic field was applied. Upon application of the alternating RF magnetic field (H = 52 kA m⁻¹, f = 295 kHz) for 15 minutes, the macroscopic temperature of the dispersion continued to decrease, reaching approximately 28 °C at the end of the experiment, following a quasi-exponential cooling profile (Fig. S5, SI). This macroscopic cooling confirms that the heat dissipated by the SPION under the RF field was insufficient to overcome the thermal losses to the Dewar vessel and surroundings. Critically, throughout the entire experiment, the bulk temperature of the dispersion remained well below 37 °C and never approached the LCST of PDEGMA.

Despite this macroscopic cooling, after the magnetic field application, 45% of the loaded IBU was released, corresponding to 723 µg per mg of SHN. As shown in Fig. 8c, IBU release triggered by magneto-thermal transduction was approximately 7 times greater than that produced by conventional warm-up heating under the same experimental time. This striking difference cannot be attributed to bulk heating of the medium, since the macroscopic temperature remained below 33 °C throughout the RF experiment, while the warm-up comparison was performed at a temperature above the PDEGMA LCST. Instead, these results provide direct evidence that the RF magnetic field induces localized nanoscale heating at the SPION surface, which is efficiently transduced to the surrounding PDEGMA shell, triggering its collapse and the consequent IBU release. The fact that PDEGMA chain collapse and drug release occur without any detectable macroscopic temperature increase above the LCST strongly supports a nanoscale thermal transduction mechanism, in which the local temperature at the SPION–polymer interface transiently exceeds the LCST, even though this is not reflected in the bulk temperature of the dispersion.

Accordingly, it may be inferred that increasing the duration of magnetic field exposure could promote the complete release of the loaded IBU within a shorter time frame compared to heating in the absence of a magnetic field. Although the applied magnetic field amplitude and frequency exceed classical safety thresholds for clinical hyperthermia, they are appropriate for a proof-of-concept study and allow us to directly evidence the efficient nanoscale transduction of magnetic heating into a polymer-mediated release response.

Together, these findings demonstrate that a controlled IBU release process can be engineered by exploiting the responsiveness of the PDEGMA shell on the SiO^MP₂ surface, induced by magneto-thermal trigger exerted on the SPION core, which is then transduced as localized heat affecting the conformation of the PDEGMA shell. The direct observation of efficient drug release under conditions of macroscopic cooling constitutes unambiguous evidence of nanoscale magneto-thermal transduction and clearly distinguishes the magnetic triggering mechanism from simple bulk heating. These proof-of-concept results confirm the potential of this SHN for application as a smart carrier for IBU and its remotely magnetic trigger release. While biological evaluation is beyond the scope of the present study, the rational design principles demonstrated here open clear avenues for future in vitro and in vivo validation.

4. Conclusions

In this work, we report the rational design and synthesis of smart hybrid nanosystems (SHN) based on the integration of three complementary building blocks into a well-defined (core@shell)-g-shell architecture, with potential application as IBU carriers with magnetically triggered release behavior. SPION were selected as core owing to their magnetic responsiveness and were synthesized by a co-precipitation method using CTAB as a stabilizing agent. The SPION core was subsequently coated with a SiO^MP₂ shell via a sol–gel method, in which CTAB acted as a structure-directing agent for mesopore formation. This SiO^MP₂ layer stabilizes the SPION core along with provides a versatile interface for the incorporation of a functional polymeric component. PDEGMA was chosen as the external layer due to its well-defined thermoresponsive behavior and solubility compatibility with IBU. Polymer integration was achieved through a two-step strategy involving silica surface functionalization followed by photopolymerization. SAXS experiments revealed that PDEGMA is distributed both within the mesoporous silica network and on the outer surface of the SPION@SiO^MP₂. Detailed structural analysis enabled the reproducible preparation of six distinct (SPION@SiO^MP₂)-g-PDEGMA SHN with controlled particle size and well-defined polymer architectures, reaching from crosslinked networks to brush-like architecture. Notably, the thermoresponsive behavior of the SHN was strongly dependent on the polymer architecture, with the most pronounced response observed for brush-like systems. This enhanced responsiveness is attributed to the increased flexibility of polymer chains confined within the mesoporous structure. A proof-of-concept experiment showed that exposure to an alternating magnetic field triggers IBU release approximately 7-fold more effectively than conventional external heating at equivalent experimental times. Remarkably, this enhanced release occurred despite a macroscopic cooling of the dispersion during RF field application, suggesting that localized nanoscale heating at the SPION–polymer interface, rather than bulk heating of the medium, is the primary driver of the magneto-thermal transduction mechanism. Overall, this study highlights the critical role of nanoscale design in controlling the collective properties and functionality of hybrid nanomaterials. By precisely controlling the composition, architecture, and spatial distribution of each building block within the SHN, it is possible to finely tune their magnetic, structural, and thermoresponsive behavior, ultimately governing their performance as externally triggerable carriers. The findings demonstrate how subtle variations in polymer architecture and confinement at the nanoscale can lead to markedly different macroscopic responses, highlighting the importance of structure–property–function relationships in complex hybrid systems. This rational design strategy provides a versatile platform for the development of multifunctional SHN with significant potential for biomedical applications, including remotely triggered drug delivery, stimuli-responsive therapeutics, and advanced nanomedicine concepts.

Conflicts of interest

There are no conflicts to declare.

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

In addition, the supporting data has been provided as part of the supplementary information (SI). Supplementary information: data of experimental conditions evaluated for SHN production: zeta potential for mesoporous silica layer incorporation, bands assignments from the FTIR, autocorrelation functions of DLS measurements, SAXS core–shell–shell fitting parameters. Detailed information of the in situ IBU release experiments. See DOI: https://doi.org/10.1039/d6ra02747d.