Magnetic Nanocomposite Sol–Gel Systems for Remote Controlled Drug Release


  • Ashley M. Hawkins,

    1. Department of Chemical and Materials Engineering, University of Kentucky, 177 F. Paul Anderson Tower, Lexington, KY 40506-0046, USA
    Search for more papers by this author
  • Chelsie E. Bottom,

    1. Department of Chemical and Materials Engineering, University of Kentucky, 177 F. Paul Anderson Tower, Lexington, KY 40506-0046, USA
    Search for more papers by this author
  • Zhi Liang,

    1. Department of Chemical and Materials Engineering, University of Kentucky, 177 F. Paul Anderson Tower, Lexington, KY 40506-0046, USA
    Search for more papers by this author
  • David A. Puleo,

    1. Center for Biomedical Engineering, Unviersity of Kentucky, Lexington, KY 40506-0070, USA
    Search for more papers by this author
  • J. Zach Hilt

    Corresponding author
    1. Department of Chemical and Materials Engineering, University of Kentucky, 177 F. Paul Anderson Tower, Lexington, KY 40506-0046, USA
    • Department of Chemical and Materials Engineering, University of Kentucky, 177 F. Paul Anderson Tower, Lexington, KY 40506-0046, USA.

    Search for more papers by this author


original image

The remote heating of iron oxide nanoparticles in an alternating magnetic field is used to drive a thermoresponsive sol-gel block copolymer, Pluronic® F-127, through the upper phase transition temperature. This phase change triggers an accelerated release rate of a model drug. Actuation and return to baseline levels are demonstrated for multiple AMF doses.

Pulsatile and controlled drug release systems are widely studied due to their superior pharmaceutical efficacy and potential for improved patient compliance.1, 2 In many physiological ailments and diseases, controlling drug delivery can provide better therapy because it can mimic biological rhythms, delivers drug at times the body is best prepared to accept it, and maintains concentrations in the therapeutic range while avoiding unnecessary side effects.1–3 Previous work has examined different methods of pulsatile control, including thermosensitive polymers, chemically responsive systems, and external stimuli (i.e., electrical current, magnetic field, light, etc.).1, 4, 5 In this study, pulsatile control of drug delivery was achieved through the actuation of nanocomposite sol–gel block copolymers in an alternating magnetic field (AMF).

Thermosensitive sol–gel block copolymer systems are a unique class of polymers that exhibit reversible phase changes from a liquid (sol) state to a solid (gel) state upon a change in temperature.6 These systems have been extensively studied for use as injectable drug delivery depots because of their ability to exist as a solution at room temperature and gel upon reaching physiological temperatures.6, 7 Also, due to their amphiphilic polymer structure, they can provide an improved method for delivering hydrophobic drug molecules.6 The phase transition from a liquid to solid is due to the arrangement of the amphiphilic block copolymers into micelles at concentrations above the critical micelle concentration. When heated, the micelles increase in size, and after reaching a critical micelle volume fraction, the system is able to undergo hard sphere packing to form a solid gel state.8–10 There is a second, less studied transition that occurs at higher temperatures in which the gel returns to the solution phase. The exact mechanism for this phase transition has been debated, but it is most likely a result of micelles changing shape from spheres to ellipsoids or cylinders.6, 8 This change in phase can be used as a mechanism to trigger the release of entrapped drug molecules.

Remote controlled (RC) drug-delivery methods have been studied in recent years for delivering and maintaining the optimal therapeutic dose to patients.11 Several types of remotely actuated nanoparticles (e.g., systems that can absorb certain wavelengths of electromagnetic radiation and dissipate it as heat) have been studied, including magnetic nanoparticles,12–16 metallic nanoparticles,17–19 and carbon nanotubes.20, 21 In most cases, the nanoparticles create a change in temperature of the system, which then leads to a change in the physicochemical properties, thus modulating the system's functions. Magnetite nanoparticles (Fe3O4) are some of the most commonly used nanoscale heating sources. These particles have the ability to produce heat when exposed to an AMF due to hysteresis, Néel relaxation, and Brownian relaxation, with the specific mechanism depending on the size of particles and surrounding material properties.22 Modulation of drug release has been demonstrated using iron oxide heating to change the swelling state,13–16 degradation rate,12 drug diffusivity,23 and nanovalve orientation24.

More recently, controlled drug release from sol–gel materials using magnetic nanoparticles has been demonstrated.25 In this case, applying an external magnetic field causes the micelles to squeeze together, thereby increasing the local concentration of drug and hence concentration gradient, resulting in an increased rate of drug release.25 However, that method caused only a shift in the release curve of a hydrophobic model drug with no on–off control, and the magnetic field was applied throughout the duration of the experiment. In the present study, we demonstrate on-off control of the drug release, and the total AMF dosing exposure time accounts for less than 4% of the total experiment duration.

In this proof-of-concept study, the upper phase transition of sol-gel materials was utilized as a mechanism to allow controlled drug release. Pluronic® F-127 (BASF) was selected as the sol-gel polymer system due to its relatively fast dissolution time as well as its low toxicity in small doses.26 Iron oxide nanoparticles were chosen as the nanoscale heating source to heat the polymer to temperatures near or above the gel to solution transition. This phase change decreases the diffusion limitations for incorporated drug molecules and therefore allows for controlled drug release as illustrated in Scheme 1. The nanocomposite sol–gel can be easily injected percutaneously into a patient where the system will automatically gel upon reaching physiological temperatures. Upon application of the AMF, the nanoparticles will heat the system, thereby increasing drug release to the surrounding tissue. When the AMF trigger is removed, the sample will return to physiological temperatures, and the gel state will be reformed. The process of actuating drug release followed by the return to baseline levels could be used several times throughout the lifetime of the implant. The biocompatible polymer depot will dissolve and will continue to dissolve after all the drug is released, thereby eliminating the need to remove the implant at a later date.

Scheme 1.

Proposed application and mechanism of remote controlled drug delivery using nanocomposite sol-gel systems. A) Percutaneous injection–Polymer solution gels at body temperature to entrap drug, B) Application of AMF–Composite heats to upper transition and releases entrapped drug, C) Removal of AMF–Composite returns to a gel at body temperature, entrapping the remaining drug, D) Application of AMF–Composite heats to upper transition, releasing remaining drug. E) Expected temperature and release rate of the sol-gel system upon actuation. The vertical dotted line represents the start of experimentation following the initial gelation.

The phase change behavior of several Pluronic concentrations and iron oxide nanoparticle loadings were analyzed. For each composition, three samples were analyzed in test tubes, and the values reported are the midpoint of the range in which the transition was observed (Figure 1A). It was difficult to obtain exact temperature points due to the transition behavior of the samples, in that there is not an exact point of transition but a small range over which the phase change occurs. As observed in previous work, the temperature range of gelation was found to increase with polymer concentration.27 The iron oxide nanoparticles also appeared to decrease the range of the gel state, especially in lower polymer concentration systems. Other groups have reported alterations in the phase change temperature due to the presence of solutes.27–31 In most, but not all cases, hydrophobic solutes and additives tended to shift the temperature to a lower value.29 Salts were also shown to lower the temperature transition due to a “salting-out” effect in which salt anions compete for the surrounding water with the outer hydrophilic blocks of the micelles.27, 28 In other work, the effect of hydrophilic polymer addition was studied; the presence of PEG molecules increased the lower transition temperature. This is most likely due to a change in structure of the micelles and a disruption in their association in forming the gel state.30, 31 The presence of the iron oxide nanoparticles in this work may similarly alter the packing and arrangement of the polymer micelles, causing the differences in the transition temperatures. Another possible explanation is the presence of the PVP surfactant on the iron oxide nanoparticles, since this may act similarly to the hydrophilic PEG groups in previous studies and cause an increase in the lower transition temperature by disrupting the micelle structure and association.

Figure 1.

Characterization of sol–gel systems. A) Phase change behavior observed through test tube inversion, the data points mark the temperature at which a solution-to-gel or gel-to-solution phase change was observed. Phase behavior for several different concentrations of the Pluronic F-127 system are reported for the three different loadings of iron oxide nanoparticles 0% (diamond), 1.25% (triangle), and 2.5% (square). B) Temperature response of the 16% system to AMF exposure measured using the fibre optic temperature probe embedded in the sol-gel sample. Data is shown for 3 different iron oxide loadings: 0% (solid line), 1.25% (dotted line), and 2.5% (dashed line). AMF Settings: 27.90 kA m−1 at a frequency of 296 kHz. N = 3 ± 1 Standard Deviation.

The phase change behavior provides insight into potential systems that could be applicable in physiological applications. Ideally, the system would be a solution at room temperature and gel at physiological temperature, thereby allowing the system to be administered via injection. Secondly, for our intended application of heating to the upper transition, the system must be carefully selected to allow enhanced drug release with minimal temperature rise, so that damage to surrounding tissue due to hyperthermia can be limited. With these requirements, the 16% system with 1.25% iron oxide loading was chosen for drug-release studies. The lower phase transition occurred at approximately 26°C, which would allow for injection, and the upper phase transition at approximately 53°C. There is concern that the exposure of human tissue to elevated temperatures could cause thermal necrosis. Previous studies on thermal pathology indicate that responses to elevated temperatures were dependent on the particular type of tissue.32 For the present system, the most likely exposure will be to soft tissue, which can experience temperatures of 45°C for up to 30 minutes with few adverse effects.32, 33 In addition, several factors will further mitigate the local hyperthermic temperatures caused by the described nanocomposites. First, the heat will be generated by the nanoparticles embedded in the sol-gel depot, and thus, the heat will be dissipated and likely not reach ablative temperatures outside of the polymer depot itself. Secondly, the AMF dosing and subsequent heating will be short compared to the duration of the depot, and thus the surrounding cells may be able to recover from any transient hyperthermia effects. Finally, the temperature profiles inside the nanocomposite gel and in the surrounding tissues can be simulated to allow for appropriate selection of dosing parameters as done in prior studies34.

In order to remotely control a phase transition in the sol–gel, the iron oxide nanoparticles must heat the system to levels at or above the transition temperature. The temperature changes that occurred in a 16% F-127 gel with two different iron oxide loadings are shown in (Figure 1B). The pure (0% iron oxide loaded) system cooled from its initial temperature of 37°C during the 5 min exposure time period. In contrast, the iron oxide loaded systems exhibited significant heating during the same time period. The 2.5% loaded system heated to a higher temperature and at a faster rate than the 1.25% system. For both systems, the upper transition temperature was approximately 53°C, and therefore a ΔT of between 15 and 20°C would cause a phase transition. In the 2.5% loaded samples, there was a period of rapid heating from 40 to 90 s followed by a region with a deceased slope. This could be an indication that a phase transition occurred and the PBS sink was allowed to inundate the sol–gel sample, thereby decreasing the rate of temperature increase. Then, due to the presence of iron oxide in the combined solution of sol-gel and PBS, the solution continued to heat for the remainder of the experiment. The 1.25% loaded system heated gradually and reached the upper phase transition temperature between 125 and 150 s. Due to its steady rate of heating, the 1.25% loaded system was chosen for continued study, and the AMF exposure was selected to be 150 s for all future work. However, the demonstration of different heating rates at different nanoparticles concentrations indicates that the system could be easily tuned by a simple adjustment in nanoparticle loading.

The heating studies confirmed that the presence of iron oxide nanoparticles could be used to remotely heat the sol–gel to temperatures at and above the upper transition temperatures. Thus, drug release studies were carried out to observe if: A) application of the AMF could enhance the rate of drug release, and B) if the rate would return to baseline levels after the exposure, thereby demonstrating on/off control over drug release. Three control sets were run simultaneously: 1) an iron oxide loaded system kept in the 37°C bath, 2) a pure (0% iron oxide loaded system) system kept in the 37°C bath, and 3) a pure system exposed to the AMF. The last group was created to ensure that any enhancement in the release was due to the iron oxide nanoparticles heating and not an effect of a cooling to room temperature and subsequent transition to the lower solution phase.

It was observed that the iron oxide loaded samples exposed to the AMF showed a much faster rate of drug release than any of the other systems, as evidenced by the higher slope of the release curve (Figure 2A,B). For the first two time points, the fraction of drug released from the loaded AMF samples was within experimental error of the controls but then showed a significant shift at both the 30 and 90 min time points of AMF dosing. At the end of the study, the loaded AMF samples had released approximately 85% of the loaded model protein drug, lysozyme, compared to less than 50% in the control systems. Statistical analysis confirmed that the iron oxide loaded, AMF exposed system exhibited a significantly higher fraction of drug release beginning at 45 min and extending through the remainder of the study (p < 0.01). Comparisons of the three control systems indicated non-significant differences in the release curves throughout the experiment. The fraction released is also an indirect way of observing the rate of dissolution of the system. Thus, application of the field increases rate of drug release through a change in temperature, but it also enhances the rate of dissolution and disintegration of the polymer system. This must be taken into consideration when designing a drug delivery depot for in vivo application.

Figure 2.

Fraction of total drug released (A,B) and rate of drug release (μg min−1) (C,D) in 16% Pluronic F-127 systems. Samples were exposed to the alternating magnetic field for 2.5 min at the 30 and 90 min time points, as indicated by the arrows. A,C) Data for 1.25% iron oxide loaded systems; B,D) release for the 0% iron oxide loaded systems. The graphs show exposed samples (diamond) compared to control samples that were kept in the 37°C water bath (square). AMF Settings: 27.9 kA m−1 at a frequency of 296 kHz. N = 3 ± 1 Standard Deviation.

Another representation of the modulation of drug release through AMF exposure is the rate of release for each system (μg min−1) (Figure 2C,D). The results show a spike in release rate following each dose to values statistically significantly higher than any of the control systems (p < 0.0001). For both AMF doses, the release rate was more than double that of the controls in the time period immediately following the dosing. Another important observation is that the release rate of the loaded, AMF exposed samples returned to a rate that was not significantly different from the control systems following each dose, and therefore on-off control was achieved. There are some inherent limitations associated with this experiment because samples were taken every 15 min, which does not provide real-time data of release. Thus, the actual rate of release was likely even higher in the time during and immediately following the AMF exposure.

This study successfully demonstrated remote controlled drug release from Pluronic F-127 sol–gel systems using the heat generated by iron oxide nanoparticles exposed to an AMF. Though the system presented here is limited, the concept of utilizing remote heating and the resulting phase change to modulate drug release can be extended to other sol-gel systems and nanoscale heating sources. In this case, the major limitation is the duration of the depot and the non-degradable nature of the Pluronic chains. There are several methods of increasing the lifetime, including changing the polymer structure (i.e., block size, composition, etc.) or varying the concentration of polymer.6, 35, 36 There are also several sol-gel systems made of degradable polymer chains that will both dissolve and degrade in vivo.37, 38 A second improvement that could be made is to demonstrate a lower “off” state in the on/off dosing. In this report, the off state release was due to passive diffusion and dissolution of the Pluronic. To lower the baseline rate of release, there are also several options, including lowering the rate of dissolution, as described above, or decreasing the passive diffusion rate by choice of drug (i.e., larger molecular weight, hydrophobic drug, etc.). Options exist for adjusting the release rate of the polymer–nanoparticle system as well, including adjusting the iron oxide loading to tune heating or altering the AMF dosing schedule and duration.

In conclusion, this proof-of-principle demonstration of RC drug release from sol-gel polymeric materials opens a new area of study for RC nanocomposite materials. The demonstration of “on-off” delivery is an improvement to the previously demonstrated work with nanocomposite sol-gel materials.25 In comparison to other magnetically triggered remote controlled systems the sol-gels have an added benefit of being injectable and degrading or dissolving with time. With proper tuning, this approach is a practical option for RC in vivo drug delivery that would provide better control of therapeutic drug concentrations as well as improved patient compliance.

Experimental Section

Samples were prepared by dissolving varying weight percentages of Pluronic F-127 in phosphate-buffered saline (PBS, pH 7.4, 10.6 mM sodium phosphate,136 mM sodium chloride, 2.7 mM potassium chloride, and 1.5 mM potassium phosphate) using the cold method.39 For release studies the drug was dissolved in the PBS prior to the polymer addition. Iron oxide nanoparticles (purchased from Nanostructured & Amorphous Materials Inc., 25 nm, 0.2% PVP coating) were added to the mixtures and sonicated to disperse the particles. The phase behavior of the systems was characterized using test-tube inversion and observing flow.6, 31 A fibre optic temperature probe (Luxtron FOT Lab Kit) was used to measure the temperature rise in samples exposed to the AMF (Taylor-Winfield 3kW). The AMF was run at an amplitude of 27.90 kA m−1 and frequency 296 kHz for all experiments. Fluorescently labeled lysozyme was released from the sol–gel depots into a PBS sink above the sample. The sink was exchanged for fresh 37°C PBS every 15 min, and the samples were kept in a 37°C water bath when not in AMF exposure. AMF doses of 2.5 min were given at the 30 and 90 min time points.

Further experimental details are provided in the Supporting Information.

Supporting Information

Supporting Information is available from the Wiley Online Library or from the author.


The authors acknowledge NSF-IGERT for their support.