Easy preparation of an MRI contrast agent with high longitudinal relaxivity based on gadolinium ions-loaded graphene oxide

Abstract

As far as the longitudinal relaxivity (r₁) is concerned, gadolinium ions (Gd³⁺)-based MRI contrast agents modified by traditional carriers do not appear to be far superior to the clinically used Magnevist. In this study, a type of MRI contrast agent (Gd³⁺@CGO) possessing a significantly high r₁ value was easily prepared using a carboxyl-functionalized graphene oxide (CGO) as a nanocarrier to directly interact with GdCl₃·6H₂O. With 2.8 wt% of Gd³⁺ loaded on CGO, the prepared Gd³⁺@CGO shows good dispersibility in water and possesses r₁ of 63.8 mM⁻¹ S⁻¹, which is 14 times higher than that of Magnevist. It is exciting to note that Gd³⁺ anchored on CGO remains stable at least for one year, probably relying on the electrostatic adsorption and physical encapsulation effect of CGO towards Gd³⁺.

---

1. Introduction

The clinical success of magnetic resonance imaging (MRI), a powerful noninvasive diagnostic modality, has fuelled an interest in techniques to prepare efficient contrast agents. In recent years, superparamagnetic iron oxide nanoparticles have attracted enormous attention due to their nontoxicity and cytocompatibility;¹ However, it has a dark contrast, which restrains its clinical use. Conversely, gadolinium ions (Gd³⁺) are characteristic of accelerating the longitudinal relaxivity (r₁), and therefore exhibit bright contrast, which have been clinically used on a widespread basis through coordination to a polydentate chelator, e.g., diethylene triamine pentaacetic acid (DTPA) and diethylene triamine tetraacetic acid (DTTA). However, these low-molecular weight Gd³⁺ complexes are not ideal for application in several cases, including cardiovascular and cancer imaging due to their limitations such as the weak signal of contrast enhancement and rapid extravasating from vasculature.² Considerable research has been done to improve MRI contrast enhancement by incorporating Gd³⁺ into various carriers, such as polyamidoamine dendrimers,³ quantum dots⁴ and silica nanoparticles,⁵ in which complicated reactions are required to introduce DTPA or DTTA to coordinate Gd³⁺.⁶ The Gd³⁺-loaded nanomaterials prepared through these methods display r₁ values ranging from 4.1 to 28.8 mM⁻¹ S⁻¹, which actually are unremarkable compared with clinical Magnevist MRI contrast agent (r₁ = 3.4–4.1 mM⁻¹ S⁻¹).

In virtue of the nanocavity, fullerenes and carbon nanotubes with good biocompatibility have been widely used as nanoplatforms for constructing high-performance endohedral gadofullerene- and gadonanotube-based MRI contrast agents. Gd@C82 (ref. 7) and Gd@C60 (ref. 8) can be categorized into the first generation of gadofullerenes. Later on, Gd@C60[C(COOH)₂]₂₀ (r₁ = 4.6 mM⁻¹ S⁻¹)⁹ and Gd₃N@C80[DiPEG5000(OH)_x] (r₁ = 143 mM⁻¹ S⁻¹) with improved water solubility were developed.¹⁰ Unfortunately, the yields of these water-soluble gadofullerenes were low. Using the current arc evaporation method, Zhang obtained two types of water-soluble gadofullerenes [Sc₂GdN@C80O₁₂(OH)₂₆, ScGd₂N@C80O₁₂(OH)₂₆] with high yield.¹¹ Nevertheless, gadofullerenes obtained by this method only show ordinary r₁ values (20.7 mM⁻¹ S⁻¹ and 17.6 mM⁻¹ S⁻¹, respectively), compared with Magnevist. Carbon nanotubes can also be used to encapsulate Gd³⁺, which lead to a high-performance MRI contrast agents. Wilson prepared superparamagnetic nanotubes with a high r₁ value of 170 mM⁻¹ S⁻¹ before hydrophilic modification, in which Gd³⁺ clusters were loaded and confined by soaking and sonicating single-walled carbon nanotubes in a GdCl₃ aqueous solution.¹² Generally, Gd³⁺ ions are loaded into fullerenes and carbon nanotubes merely through a physical encapsulation effect.

In comparison with fullerenes and carbon nanotubes, the graphene oxide (GO) has significant advantages, such as low cost and better water dispersibility, benefitted from its abundant hydrophilic oxygen-containing groups,^13,14 which make GO an attractive nanocarrier for MRI contrast agents. Furthermore, carboxyl-functionalized GO (CGO) possessing more carboxyl groups can show improved water solubility,¹⁵ which was usually prepared by converting the hydroxyl and epoxy groups of GO into carboxyl groups. These carboxyl groups can participate in esterification, amidation and coordination reactions,¹⁶ thus broadening the potential for the functionalization of GO. Nevertheless, there are only few reports about GO or CGO being used as a nanocarrier for Gd³⁺-based MRI contrast agents. Shen and coworkers were the first to obtain an MRI contrast agent using CGO as a nanocarrier through the electrostatic adsorption of Gd³⁺–DTPA chloride on CGO sheets. However, in this case, Gd³⁺ ions were adsorbed on CGO enduring complicated reactions and still relying on DTPA, and the r₁ value of the product was not reported by the author.¹⁷ Recently, an interesting finding was reported, i.e., carboxyphenylated graphene nanoribbons can be used as ligands to chelate Gd³⁺ and prepare a high-performance, T₁-weighted MRI contrast agent.¹⁸ However, the sizes of these nanoribbons were too large to ensure good biocompatibility,¹⁹ which were 125–280 nm, 7–15 nm and up to 20 mm in width, thickness, and length, respectively.

In this study, we have developed an easy method to prepare high-performance, Gd³⁺-based MRI contrast agents employing CGO as a nanocarrier without the assistance of polydentate chelator, i.e., DTPA or DTTA. The CGO is used without further treatment. GO sheets with a lateral size of ca. 200 nm are prepared and further carboxylated to form CGO by chloroacetic acid.

2. Experimental

2.1. Preparation of GO

GO was prepared from purified flake graphite (200 mesh) by a modified Hummers' method. Briefly, flake graphite (1 g) was mixed with concentrated H₂SO₄ (98 wt%, 33 mL) and NaNO₃ (0.5 g) at 0 °C for 20 min in an ice bath. Potassium permanganate (3 g) was gradually added with stirring while maintaining the temperature of the mixture below 5 °C. The ice bath was then removed, and the mixture was stirred at 35 °C for five days. Finally, 45 mL of deionised water was slowly added, and the temperature was maintained at 98 °C for 2 h, followed by the addition of H₂O₂ (6 mL, 30 wt% aqueous solution) at 60 °C. To remove the ions of the oxidant and other inorganic impurities, the resultant mixture was purified according to the protocol proposed by Zhang, L. et al.²⁰

2.2. Preparation of CGO

The dried GO (100 mg) powders were dispersed into deionised water (100 mL) and sonicated for 2 h to give a transparent solution. Sodium hydroxide (6 g) and chloroacetic acid (ClCH₂COOH) (5 g) were added to the GO solution and sonicated for 2 h under a temperature of 35 °C to convert the epoxy and hydroxyl groups of GO to carboxyl groups, giving CGO. The resultant CGO solution was purified by repeated rinsing and centrifugation until the product was totally dispersed in deionised water. As a control, GO was treated through the same previously mentioned procedure, but in the absence of ClCH₂COOH, and the product was named as rGO.

2.3. Preparation of Gd³⁺@CGO

Hexahydrate gadolinium chloride (GdCl₃·6H₂O) was dissolved in deionised water to obtain a solution with a concentration of 1 mg mL⁻¹. 4 mL of GdCl₃·6H₂O solution was then added dropwise into 50 mL of CGO dispersion (1 mg mL⁻¹) with stirring. After the reaction duration of 12 h at an approximate temperature of 35 °C and a pH value of 6.5, the reaction mixture was precipitated with a saturated NaCl aqueous solution. The supernatant was collected for following investigations, and the precipitate was repeatedly centrifugated and washed three times with deionised water, which was finally dried in a vacuum oven at 50 °C to produce pure Gd³⁺@CGO. As a comparison, GdCl₃·6H₂O aqueous solution of the same volume ratio was added to GO and rGO solution, respectively, and subjected to the same previously mentioned reaction process.

A uniform powder mixture (Mx) containing 2.8 wt% of Gd³⁺ was prepared by efficiently grinding CGO and GdCl₃·6H₂O together in an agate mortar.

2.4. Characterization methods

Two types of tests were performed to monitor whether Gd³⁺ ions from GdCl₃·6H₂O were all stably anchored on CGO after the reaction. On one hand, the collected supernatant was tested by xylenol orange and inductively coupled plasma optical emission spectrometer (ICP-OES, Varian 715). On the other hand, the reaction mixture of CGO with GdCl₃·6H₂O (after reaction of 12 h) was dialyzed against deionised water through a 500 D dialysis bag for 1 h, 5 h, 12 h and 24 h, respectively, and the Gd³⁺ concentration in these dialysates were tested by ICP-OES.

The interaction between Gd³⁺ and CGO was studied by zeta potential analysis, Fourier transform infrared (FT-IR) spectra, X-ray diffraction (XRD) patterns and X-ray photoelectron spectroscopy (XPS). The mass content of Gd³⁺ in Gd³⁺@CGO was determined by ICP-OES. To perform the ICP-OES test, Gd³⁺@CGO was dispersed into deionised water to form a uniformly stable dispersion with a concentration of 0.1 mg mL⁻¹. The lateral sizes and the morphologies of the GO, CGO and Gd³⁺@CGO were obtained by a high-resolution transmission electron microscope (HRTEM, 200 kV, JEOL JEM-2100). Zeta potentials of GO, CGO and Gd³⁺@CGO aqueous dispersion (1 mg mL⁻¹, pH = 7) were measured with a Nano-ZS90 Nanosizer (Malvern). FT-IR spectra with a resolution of 4 cm⁻¹ were recorded by a Bruker TENSOR 27 FT-IR spectrometer. XRD patterns were scanned in the range of 10–50° at a scan rate of 1° min⁻¹ by a D8Advance X-ray diffractometer (Bruker AXS Company, Germany) with Cu Kα radiation (40 kV, 20 mA, λ = 1.54051 Å). XPS (K-Alpha, Thermo Scientific Company, UK) was performed using a focused monochromatic Al Kα X-ray (1486.7 eV), which was also used to identify the element composition of Gd³⁺@CGO.

2.5. Relaxometric measurement

The T₁-weighted image and relaxation rate r₁ of Gd³⁺@CGO were measured and compared with those of a clinical Magnevist MRI contrast agent with the same concentration of Gd³⁺ (0.05 mM). To measure the r₁ of the Gd³⁺@CGO, the relaxation time (T₁) was measured using a 3.0 T MRI scanner (GE Medical Systems). T₁-weighted images were obtained with a SE sequence (TR/TE = 300 ms/20 ms, Matrix 256 × 256 pixel, FOV 24 cm × 24 cm, NEX 1, slice thickness of 5 mm, gap 1.5 mm). To acquire the T₁ values, inversion recovery sequence was used, applying three different TI (TI = 100, TI = 300, TI = 600, TR/TE 4000 ms/9.2 ms). The r₁ is estimated based on eqn (1) as follows:²¹

(1/T₁)_obsd = (1/T₁)_dia + r₁ × [M] (1)

where (1/T₁)_obsd is the observed relaxation rate of water proton in the presence of paramagnetic species, (1/T₁)_dia is the relaxation rate in the absence of the paramagnetic species, and [M] is the molar concentration of the paramagnetic species.

3. Results and discussion

3.1. The formation of CGO

After treatment with chloroacetic acid under a highly alkaline conditions, GO underwent rapid deoxygenations, and the π–π conjugated structure in GO was recovered. GO with a lateral size of approximately 200 nm (Fig. 1a) showed a high degree of oxidation (the atomic ratio of O/C is 0.43, determined by XPS, Fig. 1d), which was then treated by chloroacetic acid under highly alkaline conditions to obtain CGO. In this case, hydroxyl groups including those derived from hydrolyzed epoxy groups and the inherent hydroxyl groups on the GO sheet were partially eliminated under the highly alkaline conditions. As a result, the atomic ratio of O/C was reduced to 0.32 (determined by XPS, Fig. 1e) in CGO, and the π–π conjugated structure was partially recovered.²² As shown in the C1s core-level spectrum of GO (curve 1, Fig. 2a), the peaks at approximately 284.7 eV, 287 eV and 289 eV are attributed to the C–C (C=C), C–O and C=O (coming from ketone and carboxyl groups) bond, respectively. After GO was converted to CGO, the area of the C–O peak was reduced, whereas the area of C–C (C=C) peak was clearly increased. This was in consistence with the Raman spectra, in which the i(D)/i(G) belonging to CGO was lower than that of GO (Fig. S1†).²³

Fig. 1 TEM images and corresponding wide-scan XPS spectra of GO (a) and (d), CGO (b) and (e) and Gd³⁺@CGO (c) and (f).

Fig. 2 (a) C1s core-level XPS spectra (1: GO; 2: CGO; 3: rGO) and (b) XRD patterns (1: GO; 2: CGO; 3: Mx; 4: Gd³⁺@CGO).

After GO was converted to CGO, the layer distance decreased due to the restoration of the π–π conjugated structure, and the lateral size showed a slight increase. The peak in the XRD pattern of GO at 2θ of 12° corresponds to a layer-to-layer distance (d-spacing) of approximately 0.73 nm,²⁴ whereas a wide diffraction peak at 2θ of 24° (corresponding to d-spacing 0.37 nm) superimposed on a broad scattering background is observed in the XRD curve of CGO (Fig. 2b). The lateral size of CGO is approximately 220 nm, which is slightly higher than that of GO (200 nm) (Fig. 1).

Although a portion of epoxy groups were hydrolyzed to hydroxyl groups and further eliminated from GO sheets, considerable amount of epoxy and hydroxyl groups were converted into carboxyl groups during the carboxylation process of GO. To certify that GO was successfully carboxyl-functionalized, a control test was carried out as follows: GO was treated through the same procedure under an alkaline condition, as mentioned for the preparation of CGO except for the addition of ClCH₂COOH, the resulting product was named as rGO. The molar ratio of C=O to C–O bonds in CGO was determined to be approximately 0.96 : 1, which was larger than that in rGO (0.48 : 1) or in GO (0.46 : 1), as demonstrated in the C1s core-level spectra (Fig. 2a, Table S1†). The zeta potential of the CGO aqueous solution was −50.6 mV, which was more negative in comparison with −48.2 mV of GO aqueous solution at the same concentration and pH value. This further confirmed the carboxylation of GO and implied the good dispersibility of CGO. Because it is well known that an absolute value of zeta potential higher than 30 mV demonstrates sufficient charge repulsion to ensure a stable dispersion.

3.2. The formation of Gd³⁺@CGO

Compared to GO and rGO, CGO with smaller interlayer spacing and more carboxyl groups can stably anchor many free Gd³⁺ ions for at least one year. It is noticed that CGO, GO as well as rGO dispersions still remained uniform and stable after the addition of GdCl₃·6H₂O solution, which had to be precipitated with saturated NaCl aqueous solution. The supernatant of each solution was collected for xylenol orange test. With one drop of xylenol orange solution added to approximately 10 mL of each supernatant, the supernatant of the CGO solution appeared yellow, whereas the supernatant of either GO or rGO solution turned to purple. An ICP-OES test showed a particularly low concentration of Gd³⁺ (0.14 ppm) in the supernatant of CGO dispersion, which may contain a few Gd³⁺ ions that anchored on the smaller CGO sheets. Thus, to further demonstrate that few free Gd³⁺ ions existed in the CGO solution, the dialysate of the CGO dispersion was tested by ICP-OES, as well. No Gd³⁺ ions could be detected in the dialysate until the CGO dispersion was dialyzed for 24 h (0.023 ppm). Thus, we can say that the CGO can serve as a nanocarrier to anchor Gd³⁺, while the GO and rGO cannot. The Gd³⁺ ions were stably anchored to a great extent by CGO, which remained unchanged even after one year, as demonstrated by the xylenol orange test results (Fig. 3). Moreover, there was no free Gd³⁺ detected in the dialysate, even when the CGO dispersion (after reacting with GdCl₃·6H₂O for 12 h) was dialyzed against deionised water in an ultrasonic system (300 W) for 1 h.

Fig. 3 (a) and (b) images of Gd³⁺@CGO dispersion with a concentration of 1 mg mL⁻¹; (c) results of xylenol orange tests (1 and 2: GO, rGO solution reacted with GdCl₃·6H₂O for 12 h, respectively; 3 and 4: the freshly prepared Gd³⁺@CGO dispersion, and Gd³⁺@CGO dispersion stored for one year, respectively).

A portion of carboxyl groups in CGO were consumed because of the electrostatic interactions with Gd³⁺, leading to the intermolecular aggregation of CGO. After CGO dispersion interacted with GdCl₃·6H₂O for 12 h, pure products were collected, as described in Section 2.3. The pure products displayed zeta potential of −40.6 mV, which was higher than −50.6 mV of CGO when dispersed in deionised water to form a dispersion of 1 mg mL⁻¹, indicating that a portion of the carboxyl groups are consumed by the electrostatic adsorption of Gd³⁺.¹⁷ This type of electrostatic interaction between Gd³⁺ and CGO was further proven by the appearance of the absorption peaks belonging to the carboxylate at 1614 cm⁻¹ and 1373 cm⁻¹ in the FT-IR spectrum of the pure product (detailed analysis can be obtained from part three of the ESI, Fig. S3†). In addition, the lateral size of the CGO sheet was increased from 220 nm to 260 nm after encapsulating Gd³⁺ (Fig. 1b and c). The TEM image of the pure products also suggested that Gd³⁺ ion clusters were internally loaded in the confined space of CGO (Fig. 1c).

As previously described, CGO can be directly used as a nanocarrier for Gd³⁺ ion clusters without any further functionalization. The wide-scan XPS spectra confirm the existence of Gd³⁺ in the pure products (Fig. 1f). In addition, the mass concentration of Gd³⁺ in Gd³⁺@CGO was 2.8 wt%, as determined by ICP-OES, which was very close to the theoretical value (3.1 wt%). This result along with the results of xylenol orange test showed that all Gd³⁺ ions from GdCl₃·6H₂O were stably anchored on CGO after they reacted with each other for 12 h. The pure products derived from CGO and GdCl₃·6H₂O were named as Gd³⁺@CGO.

3.3. Relaxation properties of Gd³⁺@CGO

The T₁ relative signal intensity of Gd³⁺@CGO was significantly higher than that of clinical Magnovist containing the same amount of Gd³⁺. The imaging performance of the Gd³⁺@CGO was tested using 3.0 T MR (GE Medical Systems). It can be seen in Fig. 4 that Gd³⁺@CGO shows considerably brighter contrast than Magnovist, whereas the pristine GO and CGO do not show any brightness, but rather a little dark contrast. This indicates that the Gd³⁺ ions are anchored on CGO and show bright contrast. The dark contrast of GO and CGO is caused by the trace amounts of Mn²⁺ ions confined between their sheets.²⁵ The r₁ of Gd³⁺@CGO is 63.8 mM⁻¹ S⁻¹, which is approximately 14 times higher, compared with r₁ ∼ 4.6 mM⁻¹ S⁻¹ of Magnovist in water with the same Gd³⁺ concentration ([Gd³⁺] = 0.05 mM). Furthermore, both the brightness of the T₁-weighted image and the r₁ value of Gd³⁺@CGO do not show significant decrease after the Gd³⁺@CGO is dialyzed against deionised water through a 500 D dialysis bag with ultrasonic treatment (300 W) for 1 h. The r₁ of Gd³⁺@CGO dialyzed against deionised water is 54.1 mM⁻¹ S⁻¹. This further demonstrated that the Gd³⁺ ions are stably encapsulated in CGO sheets. Herein, the slight decrease in the r₁ value of 54.1 mM⁻¹ S⁻¹ may be caused by a small decrease of Gd³⁺ concentration after dialysis. During the dialysis of the Gd³⁺@CGO solution, impurities (e.g. MnO₂, NaOH) originated from GO and CGO preparation procedures penetrate into the dialysate, while the water molecules pass into the Gd³⁺@CGO solution, finally leading to Gd³⁺ content in the dialyzed Gd³⁺@CGO solution to be slightly lower than 0.05 mM.

Fig. 4 T₁-W_I reference images (1: H₂O; 2: Magnovist; 3: Gd³⁺@CGO; 4: GO; 5: CGO; and 6: Gd³⁺@CGO after dialysis against deionised water for 1 h with assistance of ultrasonic vibration).

The significantly enhanced longitudinal relaxivity of Gd³⁺@CGO can be attributed to its aggregation behavior and the efficient access of Gd³⁺ ion clusters to water molecules. It has been reported that either the accessibility of Gd³⁺ centers to water molecules or the intermolecular aggregation of MRI contrast agents plays an important role in the r₁ enhancement of gadonanotube or gadofullerene MRI contrasts.^11,12 As demonstrated by the TEM images, Gd³⁺@CGO did undergo slight aggregation in comparison with the original CGO. Furthermore, the rich hydrophilic groups (–COOH and –OH) on the CGO surface are expected to improve the proton conductivity, thus promoting the access of Gd³⁺-ion cluster centers in Gd³⁺@CGO to water molecules.²⁶

The fact that Gd³⁺ ions are stably anchored on CGO, showing low toxicity against HeLa cells, suggests the low cytotoxicity of Gd³⁺@CGO. According to the CCK-8 assay, we found out that CGO had few negative impacts on the viability of HeLa cells, although GO did produce a few negative impacts. We attributed this phenomenon to the better water dispersibility of CGO than GO. The viability of HeLa cells are reduced to lower than 80% when incubated with a concentration of GO higher than 80 μg mL⁻¹, while it remains more than 80% when incubated in a very high concentration of CGO of 200 μg mL⁻¹ for 24 h (Fig. S3†, testing method is shown in part 4 of the ESI). Because in Gd³⁺@CGO, Gd³⁺ ions are stably anchored on CGO, we assume that Gd³⁺@CGO will not have significant negative impacts on cell viability. The cytotoxicity study of Gd³⁺@CGO is still in progress. On the whole, Gd³⁺@CGO easily prepared in our work, which exhibit particularly high r₁ values and possess potential biocompatibility, is an attractive candidate for high-performance MRI contrast agents.

3.4. Possible mechanism for the interaction between Gd³⁺ and CGO

The present analysis results do not show direct evidence for the formation of coordination bonds between Gd³⁺ and carboxyl groups of CGO. XPS has been widely used to detect the formation of complexes according to the fact that the binding energy will increase for the ligand atom and decrease for the central atom after forming a complex.²⁷ The XPS analysis of Gd³⁺@CGO indicates that there is no detectable coordination bond between the carboxyl groups in CGO and the Gd³⁺. For instance, the gadolinium in Gd³⁺@CGO appears to play the role of the central atom, whose binding energies slightly reduce to 143.0 eV, 148.7 eV and 1187.4 eV from 143.2 eV, 149.1 eV and 1187.7 eV (corresponding to Gd4d_5/2, Gd4d_3/2 and Gd3d, respectively²⁸) in Mx. However, the oxygen atoms belonging to Gd³⁺@CGO do not perform the role of ligand atoms, and their binding energies do not increase, compared with that in CGO (Fig. 5). It is reported in ref. 18 that Gd³⁺ can be chelated by the carboxyl groups in GO functionalized with p-carboxyphenyldiazonium salt, but the coordination interaction was illustrated without any direct experimental evidence (spectral data). In addition, although several types of metal ions (e.g. Ca²⁺, Mg²⁺, and Eu³⁺) have been shown to be chelated by oxygen-containing functional groups on the GO sheets,^16,29 it is a well-known fact that chelating Gd³⁺ is difficult due to its large radius. Thus, the question of whether Gd³⁺ can be chelated by the carboxyl groups in CGO still remains.

Fig. 5 (a) Gd4d and (b) Gd3d core-level XPS spectra (1: Mx; 2: Gd³⁺@CGO), as well as (c) O1s core-level XPS spectra (2: Gd³⁺@CGO; 3: CGO).

From our point of view, because no apparent coordination bonds are detected, besides the electrostatic adsorption effect, the encapsulation effects may play a key role in stably anchoring Gd³⁺ on CGO for up to one year, similar to the case that Gd³⁺ is encapsulated in carbon nanotubes and fullerenes. Herein, The TEM image of Gd³⁺@CGO previously showed the internal load of Gd³⁺ in CGO. The XRD pattern of Gd³⁺@CGO also gives us a suggestion from another perspective. There are no obvious diffraction peaks attributed to GdCl₃·6H₂O in the XRD pattern of Gd³⁺@CGO. According to ref. 12, the absence of XRD diffraction patterns belonging to GdCl₃·6H₂O in gadonanotubes was attributed to the low Gd³⁺ content. However, we found that rather than the low Gd³⁺ content, the encapsulation effect of CGO may be the main reason for the absence of the XRD diffraction pattern of GdCl₃·6H₂O in Gd³⁺@CGO because obvious diffraction peaks assigned to GdCl₃·6H₂O appeared in the XRD curve of Mx containing the same mass content of Gd³⁺ as those in Gd³⁺@CGO.

In summary, although it has been shown that a certain amount of Gd³⁺ can be stably anchored on the CGO, the mechanism still needs further investigation. Several questions needed to be answered in our future work. For instance, do the oxygen containing groups (in particular carboxyl groups) in CGO form coordination complexes with Gd³⁺? Is there any other interaction in addition to the electrostatic and encapsulation effects, which contributes to anchoring the Gd³⁺?

4. Conclusions

An easy synthetic strategy, which leads to a high-performance MRI contrast agent, named Gd³⁺@CGO, is developed using CGO as a nanocarrier. Differing from the conventional carriers that relied on polydentate ligands, CGO can directly and stably anchor Gd³⁺. In addition, CGO with intrinsically high hydrophilic properties guarantees the accessibility of Gd³⁺ centers to water molecules and high r₁ value of Gd³⁺@CGO. The Gd³⁺@CGO possesses the capability of casting a bright T₁-weighted image with r₁ ∼ 63.8 mM⁻¹ S⁻¹, which is nearly 14 times higher as the present clinically used Magnevist. In addition, the low cytotoxicity of CGO indicates the potential biocompatibility of Gd³⁺@CGO. Although the load mechanism is not yet clear, it is exciting to find that Gd³⁺ ion clusters encapsulated in CGO remained stable for at least one year, thus implying that CGO has the ability to absorb other metals.

Acknowledgements

This work is supported by the National Natural Science Foundation of China (no. 81171318) and Health Department Foundation, China (no. 2010E07).

Notes and references

1. S. Srivastava, R. Awasthi, D. Tripathi, M. K. Rai, V. Agarwal, V. Agrawal, N. S. Gajbhiye and R. K. Gupta, Small, 2012, 8, 1099–1109.
2. X. Wang, Y. Feng, T. Ke, M. Schabel and Z.-R. Lu, Pharm. Res., 2005, 22, 596–602.
3. V. S. Talanov, C. A. Regino, H. Kobayashi, M. Bernardo, P. L. Choyke and M. W. Brechbiel, Nano Lett., 2006, 6, 1459–1463.
4. W. J. Mulder, R. Koole, R. J. Brandwijk, G. Storm, P. T. Chin, G. J. Strijkers, C. de Mello Donegá, K. Nicolay and A. W. Griffioen, Nano Lett., 2006, 6, 1–6.
5. F. Carniato, L. Tei, A. Arrais, L. Marchese and M. Botta, Eur. J. Chem., 2013, 19, 1421–1428.
6. J. S. Kim, W. J. Rieter, K. M. Taylor, H. An, W. Lin and W. Lin, J. Am. Chem. Soc., 2007, 129, 8962–8963.
7. T. Akasaka, S. Nagase, K. Kobayashi, T. Suzuki, T. Kato, K. Yamamoto, H. Funasaka and T. Takahashi, J. Chem. Soc., Chem. Commun., 1995, 1343–1344.
8. Y. Kubozono, H. Maeda, Y. Takabayashi, K. Hiraoka, T. Nakai, S. Kashino, S. Emura, S. Ukita and T. Sogabe, J. Am. Chem. Soc., 1996, 118, 6998–6999.
9. R. D. Bolskar, A. F. Benedetto, L. O. Husebo, R. E. Price, E. F. Jackson, S. Wallace, L. J. Wilson and J. M. Alford, J. Am. Chem. Soc., 2003, 125, 5471–5478.
10. P. P. Fatouros, F. D. Corwin, Z.-J. Chen, W. C. Broaddus, J. L. Tatum, B. Kettenmann, Z. Ge, H. W. Gibson, J. L. Russ and A. P. Leonard, Radiology, 2006, 240, 756–764.
11. E. Y. Zhang, C. Y. Shu, L. Feng and C.-R. Wang, J. Phys. Chem. B, 2007, 111, 14223–14226.
12. B. Sitharaman, K. R. Kissell, K. B. Hartman, L. A. Tran, A. Baikalov, I. Rusakova, Y. Sun, H. A. Khant, S. J. Ludtke and W. Chiu, Chem. Commun., 2005, 3915–3917.
13. Y. Sheng, X. S. Tang, E. W. Peng and J. M. Xue, J. Mater. Chem. B, 2013, 1, 512–521.
14. S. Sun and P. Wu, J. Mater. Chem., 2011, 21, 4095–4097.
15. L. Zhang, J. Xia, Q. Zhao, L. Liu and Z. Zhang, Small, 2009, 6, 537–544.
16. S. Park, K.-S. Lee, G. Bozoklu, W. Cai, S. T. Nguyen and R. S. Ruoff, ACS Nano, 2008, 2, 572–578.
17. A. J. Shen, D. L. Li, X. J. Cai, C. Y. Dong, H. Q. Dong, H. Y. Wen, G. H. Dai, P. J. Wang and Y. Y. Li, J. Biomed. Mater. Res., Part A, 2012, 100, 2499–2506.
18. A. Gizzatov, V. Keshishian, A. Guven, A. M. Dimiev, F. Qu, R. Muthupillai, P. Decuzzi, R. G. Bryant, J. M. Tour and L. J. Wilson, Nanoscale, 2014, 3059–3063.
19. H. Yue, W. Wei, Z. Yue, B. Wang, N. Luo, Y. Gao, D. Ma, G. Ma and Z. Su, Biomaterials, 2012, 4013–4021.
20. L. Zhang, J. Liang, Y. Huang, Y. Ma, Y. Wang and Y. Chen, Carbon, 2009, 47, 3365–3368.
21. S. Setua, D. Menon, A. Asok, S. Nair and M. Koyakutty, Biomaterials, 2010, 31, 714–729.
22. X. Fan, W. Peng, Y. Li, X. Li, S. Wang, G. Zhang and F. Zhang, Adv. Mater., 2008, 20, 4490–4493.
23. A. C. Ferrari and D. M. Basko, Nat. Nanotechnol., 2013, 8, 235–246.
24. D. A. Dikin, S. Stankovich, E. J. Zimney, R. D. Piner, G. H. Dommett, G. Evmenenko, S. T. Nguyen and R. S. Ruoff, Nature, 2007, 448, 457–460.
25. B. S. Paratala, B. D. Jacobson, S. Kanakia, L. D. Francis and B. Sitharaman, PLoS One, 2012, 7, e38185.
26. A. Enotiadis, K. Angjeli, N. Baldino, I. Nicotera and D. Gournis, Small, 2012, 8, 3338–3349.
27. F. Mercier, C. Alliot, L. Bion, N. Thromat and P. Toulhoat, J. Electron Spectrosc. Relat. Phenom., 2006, 150, 21–26.
28. D. Raiser and J. Deville, J. Electron Spectrosc. Relat. Phenom., 1991, 57, 91–97.
29. J. Y. Kim, J. B. Lee, H. J. Kim, K. Shin, Y. S. Yu and Y. Oh, Bull. Korean Chem. Soc., 2010, 31, 1485–1488.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c4ra09073j

---