Magnetic induction heating and drug release properties of magnetic carbon nanotubes

Abstract

Aim

The use of magnetic carbon nanotubes for multi-modal cancer treatment, incorporating both hyperthermia and drug delivery functions, has drawn substantial interest. Yet, the present method of regulating hyperthermia temperature involves manually adjusting the magnetic field intensity, adding to the complexity and difficulty of clinical applications. This study seeks to design novel magnetic carbon nanotubes capable of self-temperature regulation, and investigate their drug loading and release characteristics.

Methods

Using the co-precipitation method, we synthesized magnetic carbon nanotubes with a Curie temperature of 43 °C. A comprehensive investigation was conducted to analyze their morphology, crystal structure, and magnetic characteristics. To enhance their functionality, chitosan and sodium alginate modifications were introduced, enabling the loading of the antitumor drug doxorubicin hydrochloride (DOX) into these magnetic carbon nanotubes. Subsequently, the loading and release properties of DOX were investigated within the modified magnetic nanotubes.

Results

Under alternating magnetic field, magnetic carbon nanotubes exhibit self-regulating properties by undergoing a magnetic phase transition, maintaining temperatures around 43 °C as required for hyperthermia. On the other hand, during magnetic induction heating, the release percentage of DOX reached 23.5% within 2 h and 71.7% within 70 h at tumor pH conditions, indicating their potential for sustained drug release.

Conclusions

The prepared magnetic carbon nanotubes can effectively regulate the temperature during hyperthermia treatment while ensuring controlled drug release, which presents a promising method for preparing nanomaterials that synergistically enhance magnetic hyperthermia and chemotherapy drugs.

Keywords:

• Magnetic carbon nanotubes
• magnetic induction heating
• magnetic hyperthermia
• drug release

1. Introduction

The combination of magnetic hyperthermia and chemotherapy drugs has gained increasing attention in tumor therapy research. Compared to the individual use of magnetic hyperthermia or chemotherapy, this combined approach offers not only the advantages of each treatment alone but also synergistic sensitization effects that can enhance treatment efficiency [1–3]. The selection of temperature range for magnetic hyperthermia plays a crucial role in the effectiveness of the treatment. Numerous studies have indicated that significant improvements in the efficacy of chemotherapy drugs are observed only when a specific temperature threshold is reached [4–6]. For instance, many cytostatic agents such as doxorubicin hydrochloride and bleomycin exhibit a ‘threshold behavior’ [5], where temperatures below 42 °C show minimal treatment enhancements, while temperatures exceeding 42 °C demonstrate marked sensitization effects. However, temperatures higher than 45 °C can induce apoptosis in normal cells, leading to undesirable side effects [7]. Therefore, to minimize damage to normal tissue, it is essential to control the hyperthermia temperature below 45 °C.

Due to the high toxicity of chemotherapy drugs on normal cells, it is also crucial to control their concentration within a specific range in the bloodstream [8]. However, current administration methods primarily involve injections, which can lead to temporary spikes in blood drug concentration and cause pain for patients due to toxic side effects [9]. This issue can be effectively addressed through the use of sustained release drug preparations. These preparations are typically achieved by combining chemotherapy drugs with carrier materials that slow down the dissolution of drugs into the bloodstream. Carbon nanotubes (CNTs) have emerged as widely utilized carrier materials for sustained drug release [10–12]. In comparison to single-wall CNTs, multi-wall CNTs exhibit stronger bonding capabilities with various chemotherapy drugs such as gemcitabine (GEM) or doxorubicin hydrochloride (DOX) [13, 14], which allows for better sustained drug release efficacy. Additionally, numerous studies have demonstrated that drug-loaded multi-wall CNTs can effectively suppress cell proliferation [15] and inhibit tumor metastasis. Consequently, multi-wall CNTs have garnered increasing attention as potential drug carrier materials. Surface functionalization plays a crucial role in improving the biocompatibility and influencing the drug release mechanism of CNTs, which is an essential factor in their application as drug carriers. Natural polysaccharides, known for their excellent biocompatibility, are commonly employed for functionalizing CNTs. Notably, chitosan and sodium alginate possess remarkable abilities to selectively absorb onto the surface of tumor cells, thereby enhancing the targeting efficacy of tumor therapy [16–18]. As a result, they are widely used for modifying CNTs.

The combination of CNTs and magnetic nanoparticles offers a promising avenue for further exploration in multi-modal tumor therapy involving magnetic hyperthermia and chemotherapy. Numerous studies have been conducted on the synthesis of magnetic carbon nanotubes [19–21]. However, most magnetic carbon nanotubes have been prepared using Fe₃O₄ nanoparticles, with hyperthermia temperature control primarily relying on manual adjustment of the magnetic field intensity [22]. In the event of a malfunction or failure of the magnetic field adjustment unit in the equipment, maintaining precise control over the magnetic field intensity becomes challenging, potentially leading to damage to normal tissue due to excessive temperatures.

In this study, we aim to develop magnetic carbon nanotubes with a Curie temperature of 43 °C using the co-precipitation method. These magnetic carbon nanotubes possess inherent self-regulating temperature properties within the range of 42–45 °C, achieved through magnetic phase transition. Such characteristic eliminates the need for additional temperature measurement and control methods during hyperthermia treatments, thereby enhancing safety. To explore the potential synergistic sensitization effect by combining magnetic hyperthermia with chemotherapy drugs, we loaded DOX onto the magnetic carbon nanotubes and conducted a comprehensive investigation into the release behavior of the drug. The findings from this research can serve as a valuable reference for future studies on magnetic carbon nanotubes based multi-modal tumor therapy.

2. Experimental section

2.1. Preparation of pristine magnetic carbon nanotubes (MCNTs)

The chemicals employed in this study included ferric (III) trichloride hexahydrate (FeCl₃·6H₂O, 99%, Sinopharm), chromium(III) chloride hexahydrate (CrCl₃·6H₂O, 99%, Sinopharm), cobalt dichloride hexahydrate (CoCl₂·6H₂O, 99%, Sinopharm), zinc (II) chloride (ZnCl₂, 99%, Sinopharm), ammonium hydroxide (NH₃·H₂O, 28%, Sinopharm), ethanol (99.7%, Sinopharm), and carboxylic multi-walled carbon nanotubes (CNTs, 98%, Tanfeng Tech). These chemicals were utilized as received without any additional purification steps.

The MCNTs were synthesized using the co-precipitation method. Initially, CrCl₃·6H₂O (0.879 g), FeCl₃·6H₂O (2.081 g), ZnCl₂ (0.405 g), CoCl₂·6H₂O (0.602 g), and CNTs (0.75 g) were dissolved in 80 ml of deionized water and stirred for 30 min. The resulting solution was then transferred to a three-neck bottle while continuously purging with nitrogen gas to prevent oxidation. An oil bath was used to raise the temperature to 90 °C, which was monitored using an optical fiber sensor. NH₃·H₂O (28%, 30 ml) was injected at a rate of 1 ml/min as an alkali agent for co-precipitation, accompanied by moderate stirring at 300 rpm. After maintaining the temperature at 90 °C for six hours, the homogeneous precursor was allowed to cool naturally to room temperature. The final product was washed with ethanol and deionized water until reaching a neutral pH and subsequently dried at 50 °C for 24 h in an air drying chamber. For comparison, pure magnetic nanoparticles (MNPs) were prepared using the same steps without the addition of CNTs.

2.2. Characterization

X-ray diffraction (XRD) analysis was performed using a PANalytical Empyrean X-ray Diffractometer with an X-ray wavelength of 0.15406 nm. Transmission electron microscopy (TEM) and high-resolution TEM (HRTEM) images were obtained using the FEI Tecnai G2 F30 transmission electron microscope. Raman analysis was conducted on a Renishaw Invia Laser Raman Spectrometer with a laser wavelength of 532 nm. The elemental composition was determined using an EPMA-1600 electron probe. Fourier transform infrared spectroscopy (FTIR) data were collected using a NEXUS 670 spectrometer. Hysteresis loops were recorded using a JDM-13 vibrating sample magnetometer. X-ray photoelectron spectra (XPS) was recorded on a MULT1LAB2000 photoelectron spectroscopy. The Zeta potentials were measured using a Zeta sizer Nano ZS analyzer.

The Curie temperature was determined by thermogravimetric analysis (TGA), which has been widely used in relevant studies [23–27]. In this work, TGA analysis were performed using the Mettler-Toledo TGA 851 instrument. MNPs and MCNTs were placed on a thermobalance, followed by positioning a Nd-Fe-B magnet 20 cm above the balance. Initially, at room temperature, the nominal weight of both MNPs and MCNTs appeared lower than their actual weight due to magnetic attraction. However, once the temperature surpassed the Curie temperature, the MNPs and MCNTs lost their magnetism, resulting in no further attraction from the magnet. At this point, the nominal weight became equal to their actual weight. By analyzing the weight changes observed in the TGA curves, it is possible to determine the Curie temperature.

The magnetic induction heating property was investigated using an alternating magnetic field generator, as illustrated in supplementary materials Figure S1. MCNTs were decentralized in deionized water to form suspension with the concentration of 20 mg/mL.

A PerkinElmer Lambda 750s UV-visible spectrophotometer was used to determine the concentration of DOX in phosphate buffer solution. The standard curves of DOX were measured and the results were given in supplementary materials Figure S2.

2.3. Surface modifications of MCNTs

In this study, we obtained magnetic carbon nanotubes with various modifications (ALG-MCNTs, CHI-MCNTs, and ALG/CHI-MCNTs) by incorporating chitosan and sodium alginate into the synthesis process.

Sodium alginate modified MCNTs (ALG-MCNTs): 120 mg of sodium alginate and 60 mg of MCNTs were dissolved in 120 ml of deionized water. The resulting mixture was subjected to ultrasonic processing for 30 min followed by stirring at room temperature for 24 h. After washing with deionized water, the ALG-MCNTs were obtained.

Chitosan modified MCNTs (CHI-MCNTs): 150 μL acetic acid and 120 mg chitosan were added into 120 ml deionized water to prepare chitosan solution. Then 60 mg MCNTs were dissolved in chitosan solution with ultrasonic processing for 30 min and stirring for 24 h at room temperature. After being washed by deionized water, CHI-MCNTs were obtained.

Sodium alginate/chitosan co-modified MCNTs (ALG/CHI-MCNTs): 50 mg ALG-MCNTs were added into the chitosan solution prepared in previous step. Then the mixed solution was ultrasonically processed for 30 min and stirred for 24 h at room temperature. After being washed by deionized water, ALG/CHI-MCNTs were obtained.

2.4. DOX loading experiments

MCNTs, ALG-MCNTs, CHI-MCNTs, and ALG/CHI-MCNTs along with DOX (purchased from Nanjing Oddfoni Biological Technology Ltd.) were dissolved in PBS with a pH of 6.86. The concentrations of DOX and magnetic carbon nanotubes were adjusted to 1.5 mg/mL and 0.5 mg/mL, respectively. The mixed solution was kept at room temperature, and then a magnet was used to separate the magnetic carbon nanotubes from the solution at different time intervals. Two milliliters of the solution were extracted each time to determine the concentration of DOX using a UV-visible spectrophotometer. To maintain a constant volume, an equal amount of PBS was immediately added after each extraction. The loading quantities of DOX onto the magnetic carbon nanotubes (Q_t) can be calculated using formula (1), (1) $${\mathrm{Q}}_t=\frac{(C_0-C_t)V}{m}$$(1) where C₀ represents the initial DOX concentration, C_t is the DOX concentration at time t, V denotes the volume of the solution, and m represents the mass of magnetic carbon nanotubes. The DOX loading curves can be determined by calculating Q_t.

2.5. DOX release experiments in vitro

The DOX loaded MCNTs, ALG-MCNTs, CHI-MCNTs, and ALG/CHI-MCNTs were dissolved in PBS solutions with pH values of 4.00 and 6.86, respectively, with a concentration of magnetic carbon nanotubes at 20 mg/mL. The mixed solution was then placed in a water bath at 37 °C. A magnet was used to separate the magnetic carbon nanotubes from the solution at different time intervals, and 2 ml of the solution was extracted to determine the DOX concentrations. After each extraction, an equal amount of PBS was added to maintain a constant volume of the solution. The DOX cumulative release can be calculated using formula (2), (2) $$\text{Cumulative release }(\text{\%})=\frac{(C_n{V}_0+V\sum _{i=1}^{n-1}{C}_i)}{W}\times 100\text{\%}$$(2) where V₀ represents the initial volume of the solution, V denotes the volume of solution extracted, C_n and C_i represent the DOX concentrations measured at times n and i respectively, W signifies the mass of DOX present in the solution.

To explore the DOX release behavior under an alternating magnetic field, a constant temperature environment was created using circulating water at 37 °C. An alternating magnetic field with a frequency of 100 kHz and intensity of 16 kA/m was applied for one hour at various time intervals. Subsequently, the temperature curves and cumulative DOX release curves were recorded during the application of alternating field.

2.6. Cytotoxicity assay

The HaCaT cells (obtained from the China Center for Type Culture Collection) utilized in this investigation were cultivated in high glucose DMEM supplemented with 10% fetal bovine serum and 1% penicillin-streptomycin. The cells were maintained in a 37 °C incubator with 5% CO₂. Before commencing the experiments, the MCNTs were subjected to sterilization using a 75% (v/v) ethanol and deionized water solution for 48 h, followed by rinsing with phosphate buffer saline three times. Subsequently, the MCNTs were dispersed in the cell culture medium at various concentrations ranging from 0 μg/ml to 200 μg/ml. For the in vitro assessment of cytotoxicity, HaCaT cells were seeded in 96-well plates.

The CCK-8 assay was employed to quantitatively evaluate the in vitro cytotoxicity of MCNTs against HaCaT cells. For the experiment, 5 × 10⁴ HaCaT cells were seeded in 96-well plates at a volume of 1 ml per well. After a 3-h incubation period, the adherent cells were exposed to fresh culture medium containing various concentrations of MCNTs ranging from 0 μg/ml (control) to 200 μg/ml. Cell viability was assessed at 24, 48, and 72 h post-exposure. To determine cell viability, the culture media were replaced with new media containing 10 μL of CCK-8 and incubated in the dark at 37 °C for 2 h. The resulting formazan product was measured at an absorbance of 450 nm using a microplate reader (SPECTRAFLUOR, TECAN, Sunrise, Austria) after transferring each set of MCNTs solutions (100 μL) into a new 96-well plate. The following formula (3) was utilized to calculate cell viability, (3) $$\mathit{Viability} (\%)=\frac{(O{D}_{\mathit{MCNTs}}-O{D}_{\mathit{blank}})}{(O{D}_{\mathit{control}}-O{D}_{\mathit{blank}})}\times 100\%$$(3)

The optical density of wells containing varying concentrations of MCNTs, DMEM medium, and cells is noted as OD_MCNTs. In the control group, consisting of only DMEM medium and cells, the optical density is noted as OD_control. Furthermore, the optical density of the well containing solely DMEM medium is noted as OD_blank.

To facilitate visual observation of cell viability, fluorescence analysis was conducted using calcein-AM, Hoechst 33258, and propidium iodide (PI) stains. After 72 h of culturing, cells were washed three times with PBS and subsequently stained with a solution containing 2 μmol/L calcein-AM, 5 μg/ml Hoechst 33258, and 4 μmol/L PI. The stained cells were then incubated in a dark environment at 37 °C for 15 min. Following three additional washes with PBS, a fluorescent microscope (OLYMPUS BX71) captured images of the live and dead cells.

3. Results and discussion

The morphology and crystal characteristics of MCNTs were examined using TEM, XRD, and Raman spectroscopy. Figure 1a and d display the TEM images of MCNTs and MNPs, respectively. The inset of Figure 1d presents the size distribution of MNPs, which was determined from almost 120 nanoparticles and found to have a mean size of 15.7 nm. Through the addition of CNTs, successful decoration of the nanoparticles on the surface of CNTs was achieved. The HRTEM image in Figure 1b shows interlayer distances of 0.258 and 0.351 nm, corresponding to the (311) lattice plane in spinel nanoparticles and the (002) lattice plane in CNTs, respectively. The selected area electron diffraction (SAED) image displays three distinct dense rings: the ring pattern (002) matches the crystal structure of CNTs, while (311) and (400) correspond to the spinel crystal structure of nanoparticles. XRD patterns of MCNTs, MNPs, and CNTs are presented in Figure 1c. All peaks in MNPs can be attributed to the standard body-centered cubic spinel structure (JCPDS-ICDD database 03-0864), with no impurity peaks detected. Compared to the MNPs pattern, MCNTs exhibit an additional peak at (002), originating from CNTs [28]. The Raman spectra of MCNTs shown in Figure 1e reveal three active vibration modes. Among them, the D band and G band are associated with sp³ and sp² hybridized carbons in CNTs [29]; the band at 685 cm⁻¹ corresponds to the A_1g Raman mode of the spinel structure [30], representing stretching vibrations of metal ions and O²⁻ in tetrahedral sites. EPMA analysis determined the molar ratio of Zn, Co, Cr, and Fe in MCNTs as 0.543:0.457:0.58:1.42, closely matching their initial stoichiometry (0.54:0.46:0.6:1.4). Based on these findings, it can be concluded that metal ions have successfully incorporated into the lattice of nanoparticles, and the ferrite nanoparticles on MCNTs can be determined as Co_0.54Zn_0.46Cr_0.6Fe_1.4O₄.

Figure 1. (a) TEM image of MCNTs; (b) HRTEM and SAED images of MCNTs; (c) XRD patterns of MCNTs, MNPs and CNTs; (d) TEM image of MNPs; (e) Raman spectra of MCNTs; (e) FTIR spectra of MCNTs and CNTs.

To obtain a clear understanding of the alterations in functional groups on MCNTs, FTIR analysis was conducted. As depicted in Figure 1e, a comparison between the spectra of CNTs and MCNTs reveals notable differences. On MCNTs, the disappearance of peaks at 1200 cm⁻¹ and 1705 cm⁻¹ is observed, while new band peak emerge at 3437 cm⁻¹. The appearance of a broad peak at 3437 cm⁻¹ suggests the presence of hydroxyl groups on MCNTs, likely induced by the alkaline solution used during the synthesis process [31]. The bands at 1200 cm⁻¹ and 1705 cm⁻¹ correspond to the carboxylic group on CNTs, representing the bending of the C-O bond and the stretching and bending vibrations of the C = O bond, respectively. The absence of these peaks on MCNTs indicates changes in the carboxylic group, which can be attributed to the formation of either a monodentate complex or a bidentate complex between the carboxyl group and metal atoms on the surface of nanoparticles [32]. These spectroscopic changes provide evidence that the nanoparticles are covalently bonded to the carboxylic CNTs.

The content of alginate or chitosan on functionalized carbon nanotubes were determined by TGA. As depicted in supplementary materials Figure S3, the weight of MCNTs remains constant as the temperature increases from 30 °C to 620 °C. However, CHI-MCNTs, ALG-MCNTs, and ALG/CHI-MCNTs exhibit significant weight loss in the temperature range of 60 °C to 550 °C, attributed to the thermal decomposition of chitosan and alginate [33, 34]. Consequently, the content of chitosan on CHI-MCNTs, alginate on ALG-MCNTs, and the combined chitosan and alginate on ALG/CHI-MCNTs can be determined as 69.6%, 53.3%, and 74.9%, respectively.

The hysteresis loops of MCNTs and MNPs were measured using a vibrating sample magnetometer, and the results are presented in Figure 2a. Both MCNTs and MNPs exhibit superparamagnetic properties, with coercivity values (Hc) of 15.5 and 7.7 Oe, respectively. The saturation magnetization (Ms) of MCNTs was found to be 23.6 emu/g, significantly lower than that of MNPs (Ms = 57.5 emu/g). This decrease in Ms can be attributed to the introduction of nonmagnetic CNTs, which reduces the magnetic moment per unit mass and accounts for the disparity between MCNTs and MNPs.

Figure 2. (a) Hysteresis loops of MCNTs and MNPs; (b) XPS Fe2p spectra of MNPs; (c) XPS Fe2p spectra of MCNTs; (d) TGA curves of MNPs; (e) TGA curves of MCNTs; (f) time dependent temperature curves of MCNTs at magnetic field intensity of 8 kA/m and 16 kA/m.

Furthermore, the transition of iron ion valence state in MCNTs can also impact the Ms. Figure 2b and c display the XPS Fe2p spectra of MNPs and MCNTs, respectively. It is evident that both spectra can be deconvoluted into seven distinct peaks representing different iron states [35, 36]. Peaks (1) and (5) correspond to the 2p3/2 and 2p1/2 states of Fe²⁺, while peaks (2) and (6) correspond to the 2p3/2 and 2p1/2 states of Fe³⁺; Peak (3) represents the satellite peak of Fe²⁺, whereas peaks (4) and (7) represent the satellite peak of Fe³⁺. By comparing the area under the Fe³⁺ and Fe²⁺ peaks with the total Fe2p area, the proportion of Fe³⁺ and Fe²⁺ can be determined. The calculated Fe³⁺ proportion for MNPs is approximately 89.7%, which is 14.6% higher than that of MCNTs (75.1%), which means an iron valence transition from tervalence to bivalence in MCNTs. This transition may result from the reduction of CNTs during the synthesis process. Since the magnetic moment of Fe²⁺ is significantly lower than that of Fe³⁺ [37], the transition in iron valence also contributes to the decrease in Ms.

Figure 2d and e present the TGA curves obtained from MNPs and MCNTs. As the temperature increased, the weight initially exhibited an upward trend before stabilizing once it surpassed 200 °C. This increase in weight can be attributed to a decrease in magnetization. Notably, the range of temperature over which magnetization was lost appeared broad. This phenomenon can be explained by variations in nanoparticle sizes. According to Nikolaev’s research [38], the Curie temperature of nanoparticles is proportional to their exchange bond density, with the surface exhibiting a lower density compared to the core. Consequently, smaller nanoparticles with larger surface areas tend to possess lower Curie temperatures. The size distribution of nanoparticles likely influenced the temperature range at which magnetization was lost. In the case of nanoparticles with a uniform size, each nanoparticle would possess the same Curie temperature, resulting in a narrow temperature range over which magnetization is lost. However, in our study, nanoparticles exhibited a size distribution where 23% fell within the 10–14 nm range, 51% within the 14–16 nm range, and 26% within the 17–19 nm range. This broad size distribution contributes to a wider temperature range over which magnetization is lost. When the temperature reached the Curie temperature of the majority of nanoparticles, there was a sharp increase in weight observed in the TGA curve. By identifying the maximum value of the first-order derivative, the Curie temperature of MNPs and MCNTs was determined as 42.8 °C and 43 °C, respectively (refer to the inset of Figure 2d and e).

The time-dependent temperature curves of MCNTs are presented in Figure 2f. By setting the magnetic field intensity and frequency to 8 kA/m and 100 kHz, the temperature of the MCNTs suspension increased from 21.0 °C to 42.6 °C within a span of 60 min (indicated by the pink dot-line in Figure 2f). To quantitatively assess the heating efficiency of MCNTs, the specific absorption rate (SAR) can be calculated using the formula below [39], (4) $$\mathit{SAR}=C\left(\frac{dT}{dt}\right)\left(\frac{m_s}{m_m}\right)$$(4)

In the given formula, C represents the specific heat capacity of the MCNTs suspension (4.18 J·g/K), and dT/dt denotes the slope of the time dependent temperature curve. Furthermore, m_s refers to the mass of the suspension, while m_m corresponds to the mass of MCNTs within the suspension. To accurately assess the SAR without significant heat loss affecting the value, we focused on the initial 60 s of the experiment, resulting in a SAR value of 5.68 W/g.

To explore the self-regulating temperature capabilities of MCNTs, we applied a magnetic field with a higher intensity (16 kA/m). This led to an achievement of four times higher SAR (21.32 W/g); however, the MCNTs suspension reached a stable temperature of 43.2 °C, close to the Curie temperature. These self-regulating temperature properties can be attributed to the magnetic phase transition exhibited by MCNTs. When the temperature is below the Curie temperature, magnetized MCNTs generate heat under the influence of an alternating magnetic field, causing the temperature to rise continuously. Once the temperature reaches the Curie temperature, the MCNTs lose their magnetism and cease generating heat. This mechanism enables the MCNTs to maintain a constant temperature around the Curie temperature, exhibiting self-regulating behavior at approximately 43 °C, thus meeting the temperature requirements for hyperthermia applications.

The DOX loading curves of MCNTs, ALG-MCNTs, CHI-MCNTs, and ALG/CHI-MCNTs are presented in Figure 3a. Initially, the DOX loading quantities (Q_t) for all magnetic carbon nanotubes exhibited a rapid increase with time (t). Subsequently, after reaching the 30-min mark, the loading quantities remained constant.

Figure 3. (a) DOX loading curves of MCNTs, ALG-MCNTs, CHI-MCNTs, ALG/CHI-MCNTs; (b) fitting curves of pseudo-second-order model; (c) Zeta potentials of MCNTs, ALG-MCNTs, CHI-MCNTs, ALG/CHI-MCNTs before and after loading DOX; (d) equilibrium DOX loading quantities against Zeta potentials.

In order to explore the DOX loading mechanism, we employed both the pseudo-first-order model and the pseudo-second-order model to fit the DOX loading curves depicted in Figure 3a. The corresponding formulas were given in Eq. (5)(5) $$\hspace{0.17em}\ln \hspace{0.17em}(Q_e-Q_t)=\hspace{0.17em}\ln \hspace{0.17em}{Q}_e-\frac{k_{1}t}{2.303}$$(5) (pseudo-first-order model) and Eq. (6)(6) $$\frac{t}{Q_t}=\frac{t}{Q_e}+\frac{1}{k_2{Q}_e^2}$$(6) (pseudo-second-order model) [20], (5) $$\hspace{0.17em}\ln \hspace{0.17em}(Q_e-Q_t)=\hspace{0.17em}\ln \hspace{0.17em}{Q}_e-\frac{k_{1}t}{2.303}$$(5) (6) $$\frac{t}{Q_t}=\frac{t}{Q_e}+\frac{1}{k_2{Q}_e^2}$$(6) where, t represents time, Q_t corresponds to the DOX loading quantities at a given time t, Q_e signifies the equilibrium DOX loading quantities, k₁ and k₂ denote the rate constants for the pseudo-first-order model and pseudo-second-order model, respectively. The fitting parameters for both models can be found in Table 1, while the fitting curves of the pseudo-second-order model are displayed in Figure 3b. Notably, the correlation coefficient R² of the pseudo-first-order model is lower compared to that of the pseudo-second-order model. Furthermore, the equilibrium DOX loading quantities obtained from the fitting results (Q_{e, cal}) of the pseudo-first-order model do not align with their corresponding experimental values (Q_{e, exp}). Conversely, the R² values of the pseudo-second-order model exceed 0.99, and the Q_{e, cal} values closely match the Q_{e, exp} values. Consequently, the pseudo-second-order model proves more suitable for describing the DOX loading processes on magnetic carbon nanotubes. This indicates that the loading rate of DOX is influenced by the existing loading quantities, and the presence of DOX on magnetic carbon nanotubes inhibits further DOX loading.

Table 1. Fitting parameters of pseudo-first-order and pseudo-second-order model for DOX loading curves of magnetic carbon nanotubes.

Magnetic carbon nanotubes	Qe, exp (mg/g)	Pseudo-first-order			Pseudo-second-order
		Qe, cal (mg/g)	k1 (min^−1)	R^2	Qe, cal (mg/g)	k2 (min^−1)	R^2
MCNTs	168.71	39.64	0.13818	0.88247	170.94	0.00346	0.99984
CHI-MCNTs	114.75	13.63	0.07793	0.87234	115.34	0.00806	0.99999
ALG-MCNTs	187.38	41.63	0.09030	0.90986	189.75	0.00255	0.99998
ALG/CHI-MCNTs	140.53	18.25	0.09260	0.86988	141.44	0.00646	0.99989

The experimental equilibrium DOX loading quantities Q_{e, exp} ranged from 114.75 to 187.38 mg/g, which closely aligns with the findings in related literature [20]. The presence of amino groups rendered DOX positively charged, while the magnetic carbon nanotubes carried a negative charge due to the existence of carboxylic groups. Consequently, electrostatic adsorption played a crucial role in the DOX loading mechanism. To assess the impact of surface potentials on magnetic carbon nanotubes, known to influence the strength of electrostatic adsorption [40], a Zeta potential analyzer was employed for characterization purposes. The results are presented in Figure 3c. The sorting order of magnetic carbon nanotubes based on their potentials, ranging from negative to positive, was as follows: ALG-MCNTs, MCNTs, ALG/CHI-MCNTs, and CHI-MCNTs. Following the loading of DOX, all magnetic carbon nanotubes exhibited positive shifts in their potentials. This can be attributed to the following reason: MCNTs possessed negative charges due to the presence of carboxylic groups; sodium alginate, being a polyanion electrolyte, contributed negative charges that caused a negative shift in potential upon modification; chitosan, on the other hand, acted as a polycation electrolyte, introducing positive charges that resulted in a positive shift in potential following modification; when chitosan and sodium alginate were co-modified, partial charges were neutralized, placing the potential of ALG/CHI-MCNTs between CHI-MCNTs and ALG-MCNTs. Upon loading DOX, the positive charges of the drug neutralized the negative charges present on the magnetic carbon nanotubes, leading to a positive shift in potentials. It is evident that surface potentials exerted a significant influence on the DOX loading quantities on magnetic carbon nanotubes, with more negative potentials resulting in stronger electrostatic adsorption and higher DOX loading quantities (Figure 3d).

Apart from the electrostatic adsorption mechanism mentioned above, previous studies have suggested that DOX, with its anthracene ring structure, possesses π electrons that enable its attachment to carbon nanotubes through π-π stacking interactions [41]. At present, it remains challenging to ascertain which loading mechanism, electrostatic adsorption or π-π stacking interactions, plays a more significant role in this particular study. Further investigation is required to determine the specific contributions of electrostatic adsorption and π-π stacking interactions to the quantities of DOX loaded onto the magnetic carbon nanotubes.

Figure 4 illustrates the cumulative release curves of DOX at pH 4.00 (Figure 4a) and pH 6.86 (Figure 4b) at 37 °C. At pH 4.00, after 70 h, the cumulative release percentages of DOX from CHI-MCNTs, ALG/CHI-MCNTs, MCNTs, and ALG-MCNTs were 74.7%, 45.8%, 31.2%, and 14.3% respectively. At pH 6.86, the cumulative release percentages of DOX from CHI-MCNTs, ALG/CHI-MCNTs, MCNTs, and ALG-MCNTs after 70 h were 9.4%, 6.5%, 4.5%, and 2.7% respectively. It is evident that lower pH environments promote the release of DOX. These pH-sensitive properties, which also reported by Liu et al. [42], are attributed to the protonation process of DOX. In PBS solutions with lower pH levels, an increased concentration of hydrogen ions facilitates the protonation of -NH₂ groups on DOX, thereby enhancing its hydrophilicity and solubility. Due to the Warburg effect, tumor tissues exhibit a lower pH compared to normal tissues. Consequently, the pH-sensitive properties of magnetic carbon nanotubes hinder the release of DOX in normal tissues but enable effective release in tumor tissues, thus improving the targeting potential for tumor therapy.

Figure 4. Cumulative release curves of DOX with pH at 4.00 (a) and 6.86 (b). error bars were calculated ± SD of three repetitions.

In clinical studies of magnetic hyperthermia, the process of magnetic induction heating typically lasts for one hour. To investigate the release behavior of DOX from magnetic carbon nanotubes during magnetic hyperthermia, DOX loaded magnetic carbon nanotubes were immersed in PBS solutions at a temperature of 37 °C and pH value of 4.00. An alternating magnetic field with a frequency of 100 kHz and intensity of 16 kA/m was applied for one hour at specific time intervals, while recording temperature curves and DOX release curves. Figures 5a, c, e, and g represent the temperature curves and cumulative DOX release curves of MCNTs, CHI-MCNTs, ALG-MCNTs, and ALG/CHI-MCNTs over a period of 70 h. Figures 5b,d,f, and h depict the temperature curves and cumulative DOX release curves of MCNTs, CHI-MCNTs, ALG-MCNTs, and ALG/CHI-MCNTs within the first 5 h. Upon application of the alternating magnetic field, the temperature of the magnetic carbon nanotube solutions rapidly increased from 37 °C to 43 °C and remained constant at 43 °C for one hour. Subsequently, upon removal of the alternating magnetic field, the temperature returned to 37 °C. The range of DOX release percentages from the magnetic carbon nanotubes within 2 h and 70 h was found to be 10.1–23.5% and 18.2–71.7%, respectively, which closely aligned with the results obtained without applying the magnetic field (DOX release percentage range within 2 h and 70 h was 7.5–22.4% and 14.3–74.7%). Notably, the DOX release percentages of CHI-MCNTs within 2 h and 70 h were 23.5% and 71.7%, respectively, meeting the requirements for sustained drug release as specified in the Chinese Pharmacopeia (2015 Edition). These findings indicate that CHI-MCNTs can achieve sustained release of DOX at the self-regulating temperature of 43 °C, thereby potentially enabling synergistic sensitization of magnetic hyperthermia and DOX therapy.

Figure 5. Temperature curves and DOX cumulative release curves of magnetic carbon nanotubes under alternating magnetic field (MCNTs: a, b; CHI-MCNTs: c, d; ALG-MCNTs: e, f; ALG/CHI-MCNTs: g, h). among them, a, c, e and g are temperature curves and DOX cumulative release curves in 70 h; b, d, f and h are temperature curves and DOX cumulative release curves in 5 h.

To evaluate the potential cytotoxicity and ensure the biocompatibility of MCNTs, we conducted a CCK-8 assay and fluorescence analysis. The cytotoxic effects of MCNTs were examined on HaCaT cells at various concentrations: 0 μg/ml, 12.5 μg/ml, 25 μg/ml, 50 μg/ml, 100 μg/ml, and 200 μg/ml, following co-cultures for 24, 48, and 72 h. In Figure 6a, it can be observed that the viability of HaCaT cells remains above 81% across all MCNTs concentrations, with a slight decrease as the MCNTs concentration increases.

Figure 6. (a) Cell viability evaluated using the CCK-8 assay following co-culture with MCNTs for 24, 48, and 72 h; (b) fluorescence images of HaCaT cells captured after 72 h of co-culture with MCNTs, showing staining with Calcein-AM (green), PI (red), and Hoechst 33258 (blue).

To provide a more comprehensive assessment of cellular conditions, we performed live-dead fluorescent microscopy analysis. Following a 72-h co-culture with MCNTs, the cells underwent staining using Calcein-AM, PI, and Hoechst 33258. PI emits red fluorescence upon intercalating into DNA, indicating cell death, which occurs only after the dye permeates the cell membrane. Hoechst 33258 generates blue fluorescence by binding to DNA, marking the cell nucleus. Additionally, green fluorescence arises from the enzymatic hydrolysis of Calcein-AM, a reaction exclusive to living cells due to esterase activity. Figure 6b demonstrates that after 72 h of co-culture with MCNTs, cells at all concentrations exhibit positive staining for Calcein-AM and Hoechst 33258, while showing negative staining for PI. This live-dead analysis supports the quantitative CCK-8 assay presented in Figure 3a. In essence, these findings highlight the minimal cytotoxicity of MCNTs for HaCaT cells in vitro, within the concentration range of 12.5 μg/ml to 200 μg/ml. Acknowledging the complexity and significance of cytotoxicity testing, we will conduct further extensive studies involving other cell lines.

4. Conclusions

The co-precipitation method was utilized to prepare magnetic carbon nanotubes with the aim of achieving a synergistic sensitization effect of magnetic hyperthermia and chemotherapy drugs. These magnetic carbon nanotubes exhibit a Curie temperature of 43 °C, allowing them to self-regulate their temperature around this point during alternating magnetic fields, meeting the requirements for hyperthermia treatments. The loading of DOX was investigated, and it was found that the quantity of DOX loaded is closely linked to the surface potentials of the magnetic carbon nanotubes, as determined through Zeta potential characterization. In DOX release experiments under alternating magnetic fields, chitosan-modified magnetic carbon nanotubes displayed release percentages of 23.5% and 71.7% after 2 h and 70 h, respectively, which indicates a suitable sustained-release profile for DOX. This study serves as a valuable reference for the preparation of nanomaterials to achieve the synergistic sensitization of magnetic hyperthermia and DOX chemotherapy.

Author contribution

Xudong Zuo: Original draft, Methodology, Formal analysis, Conceptualization. Dongmei Zhang: Project administration, Validation, Funding acquisition. Jiandong Zhang: Investigation, Supervision. Tao Fang: Investigation, Supervision.

Disclosure statement

No potential conflict of interest was reported by the authors.

Data availability statement

The data used for the current study are available from the corresponding author upon reasonable request.

Additional information

Funding

This work was supported by National Natural Science Foundation of China (Grant No. 52302348), The Natural Science Foundation of the Jiangsu Higher Education Institutions of China (Grant No. 21KJB430047), Project for Leading Innovative Talents in Changzhou (Grant No. CQ20210105) and Changzhou Science and Technology Bureau (CM20223017).