Control of gas phase nanoparticle shape and its effect ...

Home

Search

Collections

Journals

About

Contact us

My IOPscience

Control of gas phase nanoparticle shape and its effect on MRI relaxivity

This content has been downloaded from IOPscience. Please scroll down to see the full text. 2015 Mater. Res. Express 2 035002 (http://iopscience.iop.org/2053-1591/2/3/035002) View the table of contents for this issue, or go to the journal homepage for more

Download details: IP Address: 87.147.217.30 This content was downloaded on 20/02/2015 at 11:46

Please note that terms and conditions apply.

Mater. Res. Express 2 (2015) 035002

doi:10.1088/2053-1591/2/3/035002

PAPER

RECEIVED

11 December 2014

Control of gas phase nanoparticle shape and its effect on MRI relaxivity

ACCEPTED FOR PUBLICATION

19 January 2015 PUBLISHED

17 February 2015

Sıtkı Aktaş1, Stuart C Thornton1, Chris Binns1, Leonardo Lari3,4, Andrew Pratt4, Roland Kröger4 and Mark A Horsfield2 1 2 3 4

Departments of Physics and Astronomy, University of Leicester, Leicester LE1 7RH, UK Cardiovascular Sciences, University of Leicester, Leicester LE1 7RH, UK York JEOL Nanocentre, University of York, YO10 5DD, UK Department of Physics, University of York, YO10 5DD, UK

E-mail: [email protected] Keywords: nanoparticle shape control, MRI relaxivity, core–shell nanoparticles

Abstract We have used a sputtering gas aggregation source to produce Fe@FeO nanoparticles with different shapes, by annealing them at different temperatures in the gas phase. Without annealing, the most common shape found for the nanoparticles is cubic but annealing the nanoparticles at 1129 °C transforms the cubes into cuboctahedra. Measurements of the MRI relaxivity show that the cubic nanoparticles have a higher performance by a factor of two, which is attributed to a higher saturation magnetization for this shape. This indicates that the shape-control enabled by gas-phase synthesis is important for obtaining optimal performance in applications.

1. Introduction The use of magnetic nanoparticles in biomedical applications has attracted intense interest due to their unique properties and their potential to solve certain persistent problems in conventional diagnosis and treatment. Applications include magnetic nanoparticle hyperthermia [1, 2], drug delivery [3, 4], magnetic particle imaging [5, 6] and contrast enhancement in magnetic resonance imaging (MRI) [7, 8]. Here we present a method for the shape control of gas phase produced Fe nanoparticles and demonstrate the effect of shape on MRI relaxivity. MRI is an important medical diagnostic tool due to the high soft tissue contrast, good spatial resolution and penetration depth [9]. In general, it is based on the precession of nuclear magnetism of hydrogen atoms in tissue water and lipids. The signal is detected as an induced voltage in a receiver coil, and spatial information is encoded using spatial magnetic field gradients to alter the precession frequency. Contrast is generated due to the different relaxation rates of different tissues. The main problem of MRI measurements in detecting tumours is that relaxation rates are not pathologically specific for normal and malignant tissues [9]. Hence, in order to improve the diagnostic quality, contrast agents are usually employed [10], which do not directly contribute to the image but their effect on the relaxation (longitudinal or transverse) of surrounding water hydrogen nuclei is observed. The currently used contrast agents are Gd chelates and superparamagnetic Fe oxide nanoparticles [11]. The transverse relaxation rate (R2) of nanoparticles has been linked with the saturation magnetization, Ms, of the material and the diameter of the particles, d, by equation (1) [5] R i α Ms 2d 2.

Accordingly, in order to improve the image contrast, due to the differential uptake a of contrast agent, it is necessary to maximize the saturation magnetization of nanoparticles that are as large as possible within the single domain size range. Several techniques are available for producing nanoparticles with a narrow size distribution, including chemical, biological and physical methods [1]. They have some synthesis advantages such as the ability to © 2015 IOP Publishing Ltd

(1)


S Aktaş et al

Figure 1. Synthesis of annealed or un-annealed Fe@FeO nanoparticles and deposition (a) on TEM grids and (b) in liquid suspension under the UHV conditions.

produce large amounts of biocompatible nanomaterial cheaply and some disadvantages such as the production of only oxide nanoparticles or synthesis as a powder that must then be made soluble in water to make them suitable for medical applications.

2. Experimental In order to remove the synthesis disadvantages and obtain higher saturation magnetization, pure Fe nanoparticles were produced by a sputter gas aggregation source in an ultra-high vacuum (UHV) environment as described elsewhere [1]. The main advantage of this technique is that it enables the production of pure metal core nanoparticles with a narrow size distribution. These can then be deposited into water while maintaining UHV conditions in the source [1]. It is also possible to form core–shell nanoparticles with a variable shell thickness. The optical, magnetic and electrical properties of the nanoparticles depend not only on the size but also the shape, which influences the performance in biomedical applications. Indeed, the nanoparticle shape has a significant impact on the mechanism of reaction with the environment such as oxidation as was previously shown for cuboid Fe/Fe oxide core–shell nanoparticles [12]. Here we present a new method for controlling the shape of the gas phase particles by annealing them under UHV conditions. The specific nanoparticles studied were Fe@FeO with a core–shell structure containing a pure Fe core and a thin oxide shell, which forms naturally when the pure Fe nanoparticles are exposed to air or deposited into water. The synthesis technique is sketched briefly in figure 1. It comprises a UHV compatible sputtering gas aggregation source, which is based on the NCU200 source built by Oxford Applied Research Ltd and whose design is similar to the originally reported sputter source [13]. This was used to produce the Fe nanoparticles and for shape control, the gas phase beam was passed through an empty tubular heated crucible. It is easy to show that the nanoparticle temperature equilibrates with the tube temperature within a short distance of the entrance aperture. After passing through the heater the Fe nanoparticles were deposited onto holey carbon transmission electron microscopy (TEM) grids under UHV conditions for the shape analysis. In order to examine the effect of shape on the MRI relaxivity, nanoparticles were annealed and deposited into water under UHV conditions using the method described by Binns et al [1]. The problem caused by the high vapour pressure of water at room temperature, is overcome by cooling the deposition surface to 77 K prior to starting the deposition. The water is then injected as a molecular beam onto the same substrate as the nanoparticles so that they are embedded in an ice matrix, which is melted at the end of the process to produce the liquid suspension [1]. Since the vapour pressure of crystalline or amorphous ice is in the range 10−12–10−14 mbar at 77 K, 2


S Aktaş et al

Figure 2. (a) and (b) are TEM images taken with the JEOL 2100 of Fe@FeO nanoparticles deposited directly onto the TEM grid with crucible temperatures at 25 °C and 1129 °C, respectively. Insets show the size distribution of nanoparticles deposited at 25 °C and 1129 °C crucible temperature.

[14] the entire synthesis can be carried out under UHV conditions. Molecular dynamic simulations [1] have shown that deposition into ice under these conditions has no effect on the particle shape.

3. Results and discussion Two sets of TEM and ice-matrix samples were prepared at two different crucible temperatures which were 25 °C (crucible off) and 1129 °C. The TEM study was carried out using the JEOL 2100 instrument in Leicester and also, for atomic resolution work, the aberration corrected JEOL 2200 FEG-TEM at the York-JEOL Nanocentre. The liquid suspensions were analyzed in the MRI unit at the University Hospitals of Leicester NHS Trust to determine their performance as contrast agents in MRI diagnosis. Figure 2 shows TEM images of iron nanoparticles deposited in vacuum directly onto a TEM grid with two different crucible temperatures, with the other deposition conditions (source power and gas pressure) kept constant. Log-normal size distributions were determined by analyzing the TEM images (see insets of figures 2(a) and (b)). With the crucible at 25 °C, the most probable particle size was 19 nm with the standard deviation of 0.3 nm. At 1129 °C, the most probable diameter was also 19 nm with a standard deviation of 0.3 nm. Within the error of the analysis, the size distribution was not affected by annealing. A detailed shape analysis of the unheated and heated nanoparticles was carried out on the TEM images and is illustrated in figure 3(a). Both the unheated and heated nanoparticle beam consist of a mixture of shapes which was determined by observing nanoparticles TEM images taken by the aberration corrected JEOL 2200 FEGTEM, and included cubes (figure 3(b)), cuboctahedral (figure 3(c)), spheres (figure 3(d)) and more complex shapes. The cuboctahedral shaped nanoparticles show either a hexagonal or octagonal shape in the twodimensional TEM images depending on which facet is in contact with the surface. The proportion of cubic nanoparticles was higher than other shapes in un-annealed nanoparticles and it decreased with the annealing temperature. It is clearly demonstrated in figure 3(a) that passing the nanoparticles through the hot crucible at 1129 °C converted the majority of cubes to cuboctahedra. As expected, the proportion of the spherical particles did not change significantly. The main reason for the shape changes by heating can be found by comparing the surface energy per unit area of the different shaped nanoparticles, which have the same volume. Since size distributions of annealed and un-annealed nanoparticles (see figures 2(a) and (b)) are the same, both contain a similar number of Fe atoms. If we assume the cubic nanoparticle surface energy per unit area is 2123 J m−2, (the bulk value), the surface energies of cuboctahedral and spherical nanoparticles can be determined to be 1761 and 1714 J m−2, respectively. Thus the cubic nanoparticles, which are very common during the synthesis process, are metastable and overcoming an internal energy barrier by heating enables them to transform to a shape with a lower surface energy. 3


S Aktaş et al

Figure 3. (a) Nanoparticle shapes as a function of the crucible temperature used to anneal the free particles. (b), (c) and (d) cubic, cuboctahedral and spherical shapes observed by the JEOL 2200 FEG-TEM after depositing the nanoparticles in vacuum. The oxide shell around the iron core formed after removing the substrate from vacuum. (e) Shell thickness as a function of crucible temperature. (f) Transverse (R2) relaxation rate versus concentration for Fe@FeO nanoparticles synthesized with the crucible temperature at 25 °C (red circles) and at 1129 °C (blue circles).

Figure 3(e) shows the average shell thickness of nanoparticles as a function of crucible temperature, which demonstrates that the oxide shell is entirely due to post deposition exposure and not influenced by any outgassing of the crucible. The details of the oxidation process with time have been reported by Pratt et al [12]. Thus it is appropriate to describe the heating as changing only the shape, allowing the study of the effect of this parameter on performance in applications. Figure 3(f) shows the transverse relaxation rate (R2) versus concentration for Fe@FeO nanoparticles synthesized with the crucible temperature at 25 °C (red circles) and at 1129 °C (blue circles). It is evident that the MRI relaxivity of the nanoparticles significantly decreases with annealing and the un-annealed sample, in which 4


S Aktaş et al

cubes are much more common (see figure 3(a)) shows the highest relaxivity. It has been reported that the saturation magnetization of cubic nanoparticles is higher than that of spherical nanoparticles of the same volume [15]. This was ascribed to the change in the magnetic spin distributions in a cube and a sphere of the same volume. Simulations showed that the ratio of disordered spins in cubic nanoparticles (4%) is smaller than that in spherical (8%) nanoparticles [16]. The drop observed in the transverse relaxivity is higher than would be predicted by this change when included in equation (1), but since the disordered spins are at the surface, the result could indicate a disproportionate influence of the surface spins on MRI relaxivity. The calculation of the transverse relaxivity value of the un-annealed Fe@FeO nanoparticles from the slope of the curve in figure 3(f) requires a knowledge of the concentration of the nanoparticles. In these samples we could only obtain an upper limit of 0.02 mg ml−1 [17], which gives a lower limit of the relaxivity to be 35 mM−1 s−1. The actual value could however be much larger than this, and recent measurements of similar un-annealed samples for which as estimate of concentration was available yielded a relaxivity of 425 mM−1 s−1 [17].

4. Conclusion These results show that controlling the shape of the nanoparticles is possible by passing the nanoparticles through a hot tubular crucible under UHV conditions. Cubic nanoparticles have the highest transverse MRI relaxivity, which is ascribed to a lower proportion of disordered surface spins than in particles of other shapes. This work shows that increasing the proportion of cubes by, for example, changing the synthesis conditions could produce a yet higher relaxivity since even without annealing only 45% of the nanoparticles are cubic. In addition, changing the shape will change the magnetic anisotropy so the method could be used to optimize performance in applications where the nanoparticle anisotropy is important, for example magnetic nanoparticle hyperthermia [17].

References [1] Binns C et al 2012 J. Nanoparticle Res. 14 1136 [2] Johannsen M, Gneveckow U, Taymoorian K, Thiesen B, Waldoefner N, Scholz R, Jung K, Jordan A, Wust P and Loening S 2007 Int. J. Hyperth. 23 315–23 [3] Dobson J 2006 Drug Dev. Res. 67 55–60 [4] Veiseh O, Gunn J W and Zhang M 2010 Adv. Drug Deliv. Rev. 62 284–304 [5] Binns C 2014 Medical applications of magnetic nanoparticles Nanomagnetism: Fundamentals and Applications (Amsterdam : Elsevier) pp 217–58 [6] Weizenecker J, Gleich B, Rahmer J, Dahnke H and Borgert J 2009 Phys. Med. Biol. 54 L1–10 [7] Pankhurst Q A, Thanh N T K, Jones S K and Dobson J 2009 J. Phys. D: Appl. Phys. 42 224001 [8] Tong S, Hou S, Zheng Z, Zhou J and Bao G 2010 Nano Lett. 10 4607–13 [9] Chaughule R S, Purushotham S and Ramanujan R V 2012 Proc. Natl Acad. Sci. India 82 257–68 [10] Lok C 2001 Nature 412 372–4 [11] Reimer P and Balzer T 2003 Eur. Radiol. 13 1266–76 [12] Pratt A, Lari L, Hovorka O, Shah A, Woffinden C, Tear S P, Binns C and Kröger R 2014 Nat. Mater. 13 26–30 [13] Haberland H, Karrais M, Mall M and Thurner Y 1992 J. Vac. Sci. Technol. A 10 3266–71 [14] Kouchi A 1987 Nature 330 550 [15] Zhen G et al 2011 J. Phys. Chem. C 115 327–34 [16] Noh S-H, Na W, Jang J-T, Lee J-H, Lee E J, Moon S H, Lim Y, Shin J-S and Cheon J 2012 Nano Lett. 12 3716–21 [17] Aktas S 2014 PhD Thesis University of Leicester http://hdl.handle.net/2381/29160

5