report synthesis of superparamagnetic iron oxide nanoparticles using a DC thermal plasma. Ferrocene vapor and oxygen were injected into an argon/helium ...
Thermal Plasma Synthesis of Superparamagnetic Iron Oxide Nanoparticles for Biomedical Applications Pingyan Lei and Steven L. Girshick Department of Mechanical Engineering, University of Minnesota, Minneapolis, MN, USA Abstract: Superparamagnetic iron oxide nanoparticles are of interest as contrast agents for magnetic resonance imaging, and can be heated by an alternating magnetic field, facilitating tumor destruction by hyperthermia. We report synthesis of superparamagnetic iron oxide nanoparticles using a DC thermal plasma. Ferrocene vapor and oxygen were injected into an argon/helium plasma that was then expanded through a subsonic nozzle. Particles were collected on glass fiber filters located in the reactor exhaust. In-situ measurements of particle size distributions were made using an aerosol sampling probe interfaced to a scanning mobility particle sizer (SMPS). Collected powder was characterized by transmission electron microscopy (TEM), X-ray diffraction, and vibrating sample magnetometry (VSM). Synthesized particles consisted of the superparamagnetic maghemite (γ-Fe2O3) or magnetite (Fe3O4) phases and hematite (α-Fe2O3) impurities. TEM images show primary particle diameters of 5-8 nm, while SMPS measurements indicate that the aerosol at the reactor exhaust consisted of small agglomerates, with mobility diameters mostly in the 10-20 nm range. VSM measurements confirmed that the powder was superparamagnetic, with saturation magnetizations in the 15-30 emu/g range, depending on the oxygen flow rate. Keywords: Thermal plasma, iron oxide, nanoparticles, superparamagnetism, biomedical applications
1. Introduction Superparamagnetic iron oxide nanoparticles (SPIONs) have potential biomedical applications due to their nanoscale dimensions and magnetic properties [1, 2]. Iron oxide nanoparticles of two crystalline phases, magnetite (Fe3O4) and maghemite (γ-Fe2O3) exhibit superparamagnetism when the crystallite size is smaller than about 20 nm. SPIONs provide enhanced contrast for magnetic resonance imaging (MRI) [3], and can be heated by both Néelian and Brownian relaxation under an alternating magnetic field, producing a localized temperature increase, which can be used for hyperthermia therapy in cancer treatment [4]. Several previous studies reported synthesis of iron oxide nanoparticles using various types of plasmas [5-9]. However, in these studies either the magnetic properties were not characterized [5,8], or the particle size and/or phase composition were not conducive to superparamagnetism [6,7,9]. In the present work, a DC thermal plasma was used to synthesize iron oxide nanoparticles. Measurements were made of particle size distributions, morphology, phase composition and magnetic properties. The results demonstrate the first (to our knowledge) plasma synthesis of superparamagnetic iron oxide
nanoparticles.
2. Experimental Setup The experimental setup used to synthesize iron oxide nanoparticles is shown in Fig. 1. The DC plasma was operated at a current of 250 amperes, using as plasma gas a 30/5 slm mixture of argon and helium. Ferrocene (Fe(CH)10) vapor was used as the iron precursor. Ferrocene is a stable powder at room temperature, and sublimes upon heating. Ferrocene was sublimated in a heated bed at 120°C, controlled by a heating mantle. The ferrocene vapor was entrained in argon, which flowed through the heated bed at 0.5 slm. Oxygen was introduced downstream of the heated bed, and the mixture was injected into the plasma at the upstream end of a converging nozzle made of boron nitride. The flow exiting the nozzle expanded into a 250-mm-diameter chamber, maintained at a pressure in the range 40-53 kPa. A ceramic tube (32-mm OD, 25-mm ID, 356-mm length) was positioned 51 mm above the nozzle exit to provide a longer high temperature region for inflight particle annealing [10]. Cold argon quench gas was injected into the ceramic tube. The flow rates of quench gas and oxygen were treated as variable parameters.
3. Results and Discussion
Figure 1 Experimental apparatus for synthesizing superparamagnetic iron oxide nanoparticles.
The size distribution of plasma-synthesized iron oxide nanoparticles was measured on-line with SMPS, consisting of a differential mobility analyzer and a condensation particle counter, by extracting aerosol from the reactor exhaust using an ejector driven by high-pressure nitrogen. A sample collector located in the exhaust line was used to collect product nanoparticles on a glass fiber filter for offline characterization to study chemical composition and magnetic properties of the synthesized powder. Nanoparticle phase composition was investigated using X-ray diffraction (XRD), which was performed on a Siemens D-500 diffraction meter using a 2.2-kW sealed cobalt source. Magnetic measurements were conducted on a Princeton micro vibrating sample magnetometer using a maximum applied field of one Tesla at room temperature. Nanoparticle morphology was characterized by TEM, which was conducted on a Tecnai G² F30 electron microscope. Samples for TEM were collected on lacey carbon grids by an electrostatic precipitator with an applied voltage of 3 kV, downstream of a bipolar charger (Po radiation source) located in the sampling line.
The magnetic properties of plasma-synthesized iron oxide nanoparticles were characterized by VSM. Fig. 2 shows hysteresis loops measured at room temperature with various oxygen flow rates, with a constant ferrocene feeding rate of ~7.9 sccm. Magnetic moments were normalized by sample mass measured on the glass fiber filter. The oxygen flow rate is seen to have a strong effect on the saturation magnetization. The highest saturation magnetization, 28.4 emu/g, was achieved with an oxygen flow rate of 200 sccm. The lack of evident hysteresis (to within the line thickness) in the curves is a signature of superparamagnetism. For the case of 200-sccm oxygen flow rate, the coercivity and remanence equalled 22.1 Oe and 1.87 emu/g, respectively. The measured saturation magnetization is lower than reported bulk values for the maghemite and magnetite phases [11]. As discussed below, we hypothesize that the lower magnetization may result from the presence of phase impurities, particularly of the hematite phase, which is antiferromagnetic.
Figure 2 VSM measurements of hysteresis loops of iron oxide nanoparticles at various oxygen flow rates.
TEM images of nanoparticles produced at 200 sccm of oxygen, without argon quenching flow, are shown in Fig. 3. The TEM image in Fig. 3(a) shows a large agglomerate (believed to have formed during deposition on the TEM grid) of iron oxide nanoparticles, with the diameter of primary particles in the range 5-8 nm. The high-resolution TEM image in
Fig. 3(b) clearly shows the lattice fringes of a single iron oxide nanoparticle. The d-spacing measured in the image equals ~2.65 Å, corresponding to the maghemite (310) plane. (a)
(b)
Figure 4 XRD patterns of plasma-synthesized iron oxide nanoparticles compared to magnetite and maghemite standards (purchased from Sigma-Aldrich).
Figure 3 TEM images of plasma-synthesized iron cxide nanoparticles.
Fig. 4 compares the XRD pattern of plasmasynthesized iron oxide nanoparticles, produced at an oxygen flow rate of 200 sccm and zero argon quench flow, with standard magnetite and maghemite nanopowder purchased from Sigma-Aldrich. The standard maghemite nanopowder shows small peaks from hematite impurities, which may be caused by exposure to air. Magnetite and maghemite have the same cubic structure and similar crystal lattice parameters (magnetite a=0.8396 nm, maghemite a=0.83474 nm) [11]. Their XRD patterns are virtually identical. As seen in Fig. 4, the plasmasynthesized magnetic nanoparticles contain the maghemite (or magnetite) phase as well as the antiferromagnetic hematite phase. Hematite contamination decreases the average saturation magnetization of the product. The formation of the hematite phase may be caused by the high-temperature plasma process. Another possible cause of hematite formation is that the superparamagnetic iron oxide nanoparticles are not stable due to their small sizes. A thin layer of the hematite phase may form quickly on the surfaces of such small iron oxide nanoparticles when they are exposed to air. Measured size distributions of plasmasynthesized iron oxide particles with various argon quench gas flow rates are shown in Fig. 5. These particles were produced with ~7.9 sccm ferrocene
vapor and 200 sccm oxygen. SMPS measures the mobility diameter of particles, which may consist of agglomerates that contain several primary particles. At the same precursor feeding rate and oxygen flow rate, the average particle size is seen to decrease as the quench gas flow rate increases. The mode of the size distributions in Fig. 5 decreases from ~18 nm to ~12 nm as the quench gas flow rate increases from 0 to 30 slm.
Figure 5 SMPS size distribution measurements for various flow rates of argon quench gas.
4. Summary A DC thermal plasma system was used to synthesize superparamagnetic iron oxide nanoparticles using ferrocene as iron precursor. Plasmasynthesized magnetic iron oxide nanoparticles had primary particle sizes of 5-8 nm. The powders produced are predominantly the superparamagnetic ma-
ghemite (or magnetite) phase, and also show evidence of hematite impurities. To our knowledge this is the first report of plasma synthesis (by any type of plasma) of superparamagnetic iron oxide nanoparticles. The oxygen flow rate has a strong effect on the magnetic properties of the product powder, likely by affecting the iron oxide stoichiometry and phase composition. Introducing argon quench into the reaction chamber is an effective means to reduce the agglomerate size of plasma-synthesized magnetic iron oxide nanoparticles. At the highest quench gas flow rate tested the mean agglomerate size equalled ~16 nm, indicating that each agglomerate contains several primary particles.
5. Acknowledgments This research was primarily supported by the U.S. National Science Foundation under Award Number CBET-0730184, and partially supported by the Minnesota Futures Grant Program. Parts of the characterization work were carried out in the College of Science and Engineering Characterization Facility, and the Institute for Rock Magnetism, University of Minnesota.
References [1] J. Gao, H. Gu, B. Xu, Acc. Chem. Res. 42, 1097 (2009). [2] A. K. Gupta, M. Gupta, Biomaterials 26, 3995 (2005). [3] S. Kumano, T. Murakami, T. Kim, M. Hori, A. Okada, T. Sugiura, Y. Noguchi, S. Kawata, K. Tomoda, H. Nakamura, Am. J. of Roentgenol. 181, 1335 (2003) [4] G. F. Goya, V. Grazu M. R. Ibarra, Curr. Nanosci. 4, 1 (2008). [5] I. Bica, Mater. Sci. Eng. B 68, 5 (1999). [6] D. N. McIlroy, J. Huso, Y. Kranov, J. Marchinek, C. Ebert, S. Moore, E. Marji, R. Gandy, J. Appl. Phys. 93, 5643 (2005). [7] I. Banerjee, Y.B. Khollam, C. Balasubramanian, R. Pasricha, P.P. Bakare, K.R. Patil, A.K. Das, S.V. Bhoraskar, Scri. Mater. 54, 1235 (2006). [8] L. Zajickova, P. Synek, O. Jasek, M. Elias, B. David, J. Bursik, N. Pizurova, R. Hanzlikova, L. Lazar, Appl. Surf. Sci. 255, 5421 (2009). [9] V. Subramanian, R. Baskaran, H. Krishnan, Aerosol Air Qual. Res. 9, 172 (2009). [10] R. Jain, S. L. Girshick, J. V. R. Heberlein, R. Mukherjee, B. Zhang, T. Nakamura, A. Mochizuki, Plasma Chem. Plasma Process. 30, 795 (2010). [11] R.M. Cornell, U. Schwertmann, The Iron Oxides: Structure, Properties, Reactions, Occurrences and Uses, 2nd ed. (Wiley-VCH, Weinheim, 2003).
