High-efficiency plasma surface modification of graphite-encapsulated magnetic nanoparticles using a pulsed particle explosion technique

Recently, carbon-coated magnetic nanoparticles have attracted considerable interest in materials science research. The incorporation of both metallic nanoparticles and carbon in a stable core–shell system improves their advantageous properties, which make them potentially applicable in various applications such as magnetic data storage, magnetic fluid, magnetic inks,¹⁾ catalyst support,²⁾ magnetic separation, electrode, additives for many uses (i.e., as sintering agents and propellants), conductive paste, conductive coating, and biotechnological and biomedical applications.^3–10)

Bare metallic nanoparticles have high reactivity and high toxicity, which are limitations for realizing their practical applications, such as instability under oxidation and degradation conditions (i.e., in acids). The other disadvantages are their easy agglomeration and unsupported-surface structure for providing functional group attachment for absorbing appropriate molecules. Consequently, coating bare metal-magnetic nanoparticles with a protective shell is an appropriate technique to overcome those limitations.

Compared with polymers and silica, carbon with the graphite structure is a promising coating material for bioapplications because of its high stability at high temperature and pressure, and in various chemical and physical environments (i.e., acid or base media). Moreover, the graphite shell allows for further functionalization with specific functional groups and biomolecules. Unfortunately, graphite-encapsulated magnetic nanoparticles conventionally prepared by the arc-discharge method, generally disperse only in organic solvents. This phenomenon makes them unsuitable for bioapplications. To enhance nanoparticle biocompatibility, surface modification processing has become a necessary procedure before nanoparticles find practical applications.

One of the efficient methods of surface modification is plasma treatment, which has been commonly used for many industrial applications. Plasma surface modification is environmentally friendly with a short reaction time, and provides various functional groups.¹¹⁾ Plasma processing can markedly increase production if the system is optimized. Recently, the plasma processing of magnetic materials has drawn much attention with regard to nanoparticle surface treatment for medical uses, such as drug delivery systems or magnetic resonance imaging systems. There have been several plasma reactor systems already developed for particle treatment, such as the bell jar reactor,^12,13) downstream reactor,¹⁴⁾ rotary drum reactor,¹⁵⁾ plasma fluidized bed reactor,^16–18) circulating bed reactor,^19,20) plasma batch reactor,²¹⁾ plasma downer reactor,²²⁾ and plasma reactor with a mechanical vibrator such as an electromagnet¹²⁾ or a stirrer.¹³⁾

The purpose of these plasma systems is to interface particles with plasma species. An efficient interaction between the particle surface and the plasma is the key to achieve the maximum surface modification. Early attempts to improve the dispersion of pigment particles were carried out using plasma techniques.^23,24) In the case of polymer webs, the entire surface is exposed to plasma using conventional drum- or batch-type plasma reactors.²⁵⁾ However, such plasma reactors are often unsuitable for particle materials owing to the lack of solid mixing.²⁶⁾

As described in our previous paper,²⁷⁾ we successfully modified graphite-encapsulated iron compound magnetic nanoparticles deposited on a silicon substrate with amino groups using Ar and NH₃ plasmas in successive stages. To treat the particles homogeneously, they should be placed on the sample stage such that they are dispersed as widely as possible. When the placement of the particle sample is not performed well, the uniform treatment of the entire bulk of particles is difficult to achieve because the modification will likely take place only on the top layers of the sample, that is, particles inside the bulk will be less exposed to the plasma than particles at the surface of the bulk.

To enhance the interaction between the particles and the plasma, a modified setup is required. We consider that the particle explosion technique enables an enhanced surface interaction between the particle samples and the plasma species. Therefore, in the present study, we developed a plasma reactor for particle treatment to explode the particles inside the chamber by a negative pulsed biasing of the sample stage during the plasma processing.

2.1. Nanoparticle fabrication and plasma processing setup

Graphite-encapsulated iron compound magnetic nanoparticles were prepared by the arc discharge method,^28,29) which has already been described in previous papers.^27,30) The characteristics of the magnetic nanoparticles, such as magnetic properties or crystalline structures, have been described in a previous paper.³⁰⁾ Following nanoparticle synthesis, the nanoparticles are treated using a radio-frequency (RF) inductively coupled plasma device. A schematic view of the chamber is shown in Fig. 1 . The stainless-steel chamber is 200 mm in both diameter and height. The water-cooling copper pipe helical antenna with a coil diameter of 100 mm and a pipe diameter of 20 mm was wound around the quartz bell jar (110 mm in outer diameter and 260 mm in height) mounted on the stainless-steel chamber. The helical antenna was coupled to an RF power generator at 13.56 MHz via a matching network. The typical input RF power was about 80 W.

Zoom In Zoom Out Reset image size

The chamber used in this work was modified by adding the metal substrate for the sample stage inside the chamber. The metal substrate of 10 mm diameter was attached to the center of the glass dish placed at z = −2 cm (see Fig. 1), where z = 0 is defined as the center of the helical antenna on its axial axis. To confine the nanoparticles exploded by applying bias, a glass tube with a diameter of 80 mm and a height of 60 mm was placed on the glass dish.

In the experiment, firstly, we put the particle sample (about 5 mg) on the metal substrate. The chamber was evacuated to a base pressure of approximately 10⁻³ Pa. After the vacuum evacuation, NH₃ gas was introduced into the chamber and kept at 50 Pa. During the plasma processing, the gate chamber was closed to prevent the nanoparticles from flowing to the turbo pump system. The biasing conditions were as follows: a substrate pulse biasing of −1 kV was applied at a repetition frequency of 1 kHz and a duty ratio of 50%. The negative pulsed bias voltage was turned on immediately after switching the plasma on. Generally, it will take time to match the input and reflection powers before applying the pulsed bias to the substrate. The bias time t_b was varied from 0 to 60 s. After biasing off, the plasma was kept turned on up to the desired plasma treatment time t_p of 10 min. The explosion of the particles by applying pulsed biasing was visually observed and recorded using a high-resolution digital camera (Nikon D90) at a capture speed of 24 fps. The videos were then processed using the software VirtualDub to add the timestamp and obtain sequential images.

2.2. X-ray photoelectron spectroscopy and transmission electron microscopy analysis

2.3. Estimation of amino group population

2.4. Dispersion property

The present study was started by preparing nanoparticle samples by the arc discharge method. The TEM images and EDS profiles of successfully fabricated nanoparticles are shown in Fig. 2 . Using TEM, we confirmed that iron compound magnetic nanoparticles are clearly encapsulated by graphitic carbon. This result shows good agreement with the EDS profiles of the selected particles that reveal at least three possible phases: particles that exhibit the presence of O, Fe, and C. The interplanar distances of the graphite lattice and iron fringes are about 0.34 and 0.202 nm, respectively.

Zoom In Zoom Out Reset image size

Following the nanoparticle fabrication, we placed nanoparticle samples (about 5 mg) on the metal substrate of the sample holder located at the center of the glass dish. After applying a pulse bias with a voltage of −1 kV, a frequency of 1 kHz, and a duty ratio of 50%, the particles were exploded, as shown in the successive images in Fig. 3 . These sequential images were captured during the pulsed biasing. The first image (left top) in Fig. 3 shows the condition before starting the experiment (plasma OFF). The next image shows the situation just after turning the plasma on. Then, it generally took time to produce a stable plasma. Until the time of turning the bias on, the sample particles are still in the metal stage. Plasma generation will be easily achieved if the matching is adjusted at the desired power beforehand. The shorter time interval for turning on the bias provides an effective interaction between the plasma and the particle sample because the lifetime of an NH₂ radical is short (a few microseconds).^34–36) The lifetime of these plasma species is important because we used the no-flow gas condition, that is, the chamber gate valve was closed during plasma treatment, maintaining a pressure at 50 Pa. The negative pulsed bias voltage was turned on about 11.4 s after turning on the plasma. At the biasing-on time, the particles started to explode and dropped back to the substrate stage after half a second. After a certain biasing time (called t_b), the bias voltage was turned off, but the plasma treatment was continued without biasing up to the desired treatment time t_p. For clarity, red dashed lines are added in the pictures to show when and how the particles began to fly and ended up.

Zoom In Zoom Out Reset image size

In Fig. 4 , we summarized the phenomenon observed in Fig. 3 as the temporal behavior of the height changes of the exploding particles, at t_b = 15 s and t_p = 3 min. The height of the exploding particles was measured as the highest position taken from each image frame. The inset figure shows a magnified view of the peak area. The letters shown in Fig. 4 denote the main events. A and B represent the timer ON and plasma ON, which correspond to the first (0.000 s) and second (0.417 s) images in Fig. 3, respectively. C represents the time when the biasing was turned on, shown in the fourth image (11.791 s) in Fig. 3. The gap between B and C indicates the time required for the impedance matching between the RF source and the plasma. The particles started to explode at C, reached their maximum height at D (12.125 s), and ended up at E (12.5 s). F represents the time when the bias was turned off after 15 s of having the bias turned on, and G represents the time when the plasma was switched off after 180 s (3 min).

Zoom In Zoom Out Reset image size

The explosion event is explained by the ion bombardment mechanism. After turning on the biasing, the pulsed negative high voltage caused a high electric field in the sheath between the plasma and the substrate. The existence of a high electric field caused the ions in the plasma to accelerate toward the substrate where the particles were placed. Because the particles were placed on the stage in powder form, once the powder was bombarded by plasma ions, the particles spontaneously popped out and exploded upwards in a very fast manner. This phenomenon is similar to a conventional ion sputtering event. The ion bombardment energy depends on the difference in potential between the plasma and the substrate (ΔV = V_p − V_sub) and is subsequently expressed as e(V_p − V_sub). In this study, if the estimated plasma potential V_p is more or less on the order of 10 V, the ion bombardment energy when biasing is turned on is approximately 1 keV, estimated from a biasing voltage of −1 kV. This proposed mechanism also enables agglomerated nanoparticles to separate into fine particles during the explosion process. By using a small (diameter 10 mm) metal substrate with a high negative voltage, the particles explode and do not return back to the substrate but fall on the glass dish area surrounding the metal substrate owing to gravitational force. The dynamic behavior of the particles and the forces acting on them under biasing have been discussed in several papers,^37–41) which are beyond the scope of our present paper.

After the plasma treatment, the treated particles were characterized by XPS. Figure 5 shows the data set of the N 1s peak of the XPS profiles with various biasing times t_b referred to as the C → F range (see Fig. 4). The observed N 1s peaks located at approximately 399.8 eV are considered as a signal of the nitrogen-containing group for the amino group, which was successfully grafted on the surface. Figure 5 shows the N 1s spectra at different biasing times of t_b = 0 (no biasing), 2, 15, 30, and 60 s, each of which was obtained for various plasma treatment times t_p of up to 30 min. Comparing the biasing system with the nonbiasing system (t_b = 0 s), the intensity of the N 1s peak of the biasing system (t_b = 2, 15, 30, and 60 s) is significantly increased. For example, the intensity of the N 1s peak of the nanoparticles treated in the biasing system (t_b = 15 s) is raised to about 3–4 times higher than those in the nonbiasing system (t_b = 0 s) for short plasma treatment times t_p of up to 3 min. This indicates that the plasma treatment with applied biasing had a significant effect of increasing N 1s peak intensity. These increases in the N 1s peak intensity are supposed to be due to the enhancement of the efficient interaction between the particles and the plasma species, particularly nitrogen-containing species, during an explosion event.

Zoom In Zoom Out Reset image size

To discuss the XPS profile in more detail, we compare the atomic concentration of nitrogen to the atomic concentration of carbon in terms of the N/C ratio shown in Fig. 6 . The N/C ratios of the five data sets with t_b = 0, 2, 15, 30, and 60 s are shown in sequence on comparable scales. Each data set was obtained for various plasma treatment times within 30 min. Figure 6 shows that all the data sets have similar behaviors. The N/C ratio steeply increased within a treatment time t_p range of 2–3 min, reached their saturation values, and then decayed slightly thereafter. Comparing the powders with and without a biasing event, the N/C ratios of the nonbiasing (t_b = 0 s) and biasing (t_b > 0 s) systems were significantly different. For example, the N/C ratio of the nonbiasing system was only about 2.6% (t_b = 0 s, t_p = 3 min), but that for the biasing system increased to be about 7% (t_b = 15 s, t_p = 3 min), which became the maximum N/C ratio. Note that the results of the N/C ratios indicate roughly similar trends to those of N 1s peak intensities shown in Fig. 5. The present results are very analogous to our previous results of amino group introduction into the polymer surface using low-pressure microwave NH₃ plasma.⁴²⁾ The N/C ratio is maximum at a short plasma treatment time of 30 s and then decreases owing to the surface damage by ion bombardment and/or the hydrogen etching effect as treatment time increases.

Zoom In Zoom Out Reset image size

From Figs. 5 and 6, note that the short plasma treatment times of about 3–5 min are suitable for achieving the nearly maximum N/C ratio of 6% for various biasing times of 2–30 s. In the present study, we obtained almost the same results for the maximum N/C ratio, roughly 6% for different biasing times. These results suggest that the powders exploded by applying the negative pulsed bias were essential to improve the N/C of the particles.

To quantitatively evaluate the absolute values of the number of amino groups grafted to the graphite-encapsulated iron composite nanoparticles, we have performed a conventional chemical derivatization reaction method using sulfosuccinimidyl 6-[3'(2-pyridyldithio)-propionamido] hexanoate (sulfo-LC-SPDP).^31,32) Details of the derivatization steps are presented in the experimental section.

For the amino group derivatization, we examined three samples: (a) untreated nanoparticles, (b) treated particles in the nonbiasing system (t_b = 0 s; t_p = 3 min), and (c) treated nanoparticles in the biasing system (t_b = 15 s; t_p = 3 min), shown in Fig. 7 . The latter represented the sample that has the highest N/C ratio among all the samples. The amino group analysis gave results showing that no absorbance was observed at 343 nm for the untreated nanoparticles, indicating that no amino groups were grafted groups to the surface of pristine graphite-encapsulated iron composite nanoparticles. On the other hand, a very large number of amino groups were obtained from the plasma-treated samples. The populations of amino groups for the treated nanoparticles under the experimental conditions of (t_b = 0 s, t_p = 3 min) and (t_b = 15 s, t_p = 3 min) were estimated to be about 1.9 × 10⁴ and 8.0 × 10⁴ molecules per particle, respectively. The present results also indicate the efficient enhancement of amino group modification by roughly fourfold by the negative pulsed biasing during the NH₃ plasma processing.

Zoom In Zoom Out Reset image size

Furthermore, to confirm the NH₃ plasma effect on the hydrophilicity, we examined the dispersion of the nanoparticles in water, the results of which are shown in Fig. 7. Comparing three vials with the same amount of particles (i.e., a, b, and c), we observed significant differences. As shown in Fig. 7, the dispersion of the nanoparticles is significantly improved after the plasma treatment. The untreated nanoparticles (a) did not disperse in water, and all of the nanoparticles remained on the top surface of the water. However, the dispersion of the nanoparticles treated in the biasing system (c) is better than those treated in the nonbiasing system (b). This can be observed by looking at the dark color of the solution, which corresponds to the dispersed nanoparticles in water. The improvement in the dispersion of the treated nanoparticles in the biasing system is assumed to be caused by the optimum interaction between NH₃ plasma and nanoparticles. The explosion event in the biasing system allows the amino group to attach not only to a portion but also to the entire nanoparticle surface. Consequently, the hydrophilicity of the nanoparticles after plasma treatment greatly improved because a wider surface area of the nanoparticles is available for grafting by amino groups. The amino groups on the outmost graphite layer are likely to play a key role in enhancing the hydrophilicity.

Figure 8 shows an illustration of the interaction of water molecules with the aminated surface of the nanoparticles. The grafted amino group can bind water molecules through intermolecular forces such as dipole–dipole forces and hydrogen bonds, as shown by the yellow line in Fig. 8. The larger number of amino groups may allocate more areas for attracting more water molecules, which contributes to the improvement in the dispersion property of the nanoparticles.

Zoom In Zoom Out Reset image size

HR-TEM was also used to analyze the damaging effects on the morphological and structural properties of the nanoparticles before and after plasma treatment. Figure 9 shows images of the plasma-treated nanoparticles observed by HR-TEM in low and high magnifications. (a) to (f) correspond to the treated nanoparticles under various (t_b, t_p) conditions of (0 s, 3 min), (2 s, 3 min), (15 s, 3 min), (30 s, 3 min), (60 s, 3 min), and (15 s, 30 min). All of the images confirm that no damage or destruction was induced in the nanoparticle structure after performing the plasma treatment in the nonbiasing or biasing system. The structure of the graphite layers was found to be stable under all the experimental conditions. Similarly to the graphite coating, the iron compound core also remained encapsulated inside the graphite layers even when the particles were subjected to long-term plasma treatment, as shown in Fig. 9(f). This result indicates that the pulsed-biasing particle explosion technique performed in the present study is highly efficient in amino group functionalization. Furthermore, it is also suitable for surface modification, particularly for powder samples because of its ability to retain the structural stability of nanoparticles.

Zoom In Zoom Out Reset image size

The surface of graphite-encapsulated iron compound magnetic nanoparticles fabricated by arc discharge was successfully modified by a pulsed particle explosion technique. This technique was performed by applying a high negative bias of −1 kV to the substrate stage for 2–60 s at a repetition frequency of 1 kHz and a duty ratio of 50% in ammonia plasma generated using a radio frequency inductively coupled plasma device. The intensity of the N 1s peak in the XPS spectrum of the nanoparticles treated in the biasing system was three to four times higher than that treated in the nonbiasing system owing to the enhancement of the interaction between the nanoparticles and the plasma species. The present results also indicate the efficient enhancement of amino group modification by approximately fourfold from about 1.9 × 10⁴ molecules/nanoparticle in the case of (t_b = 0 s, t_p = 3 min) to 8.2 × 10⁴ molecules/nanoparticle in the case of (t_b = 15 s, t_p = 3 min) by the negative pulsed biasing during the NH₃ plasma processing. Moreover, the results also showed that the dispersion of the treated nanoparticles in biasing system was improved compared with those of the untreated and treated samples in the nonbiasing system.

In addition, analysis by HR-TEM showed no significant damage on the nanoparticle structures, indicating that the present technique is suitable mainly for the surface modification of particle samples owing to its high efficiency in surface modification without causing any significant change or destruction of the structural and morphological properties.

This work was supported in part by a Grant-in-Aid for Scientific Research (No. 2110010) from the Japan Society for the Promotion of Science (JSPS). The authors would like to thank Associate Professor A. Ogino of Shizuoka University for technical assistance in the plasma chamber development and UV–vis absorption spectroscopy measurement.