Rapid magnetic heating treatment by highly charged maghemite nanoparticles on Wistar rats exocranial glioma tumors at microliter volume.
ABSTRACT One of the most significant challenges implementing colloidal magnetic nanoparticles in medicine is the efficient heating of microliter quantities by applying a low frequency alternating magnetic field. The ultimate goal is to accomplish nonsurgically the treatment of millimeter size tumors. Here, we demonstrate the synthesis, characterization, and the in vitro as well as in vivo efficiency of a dextran coated maghemite (gamma-Fe(2)O(3)) ferrofluid with an exceptional response to magnetic heating. The difference to previous synthetic attempts is the high charge of the dextran coating, which according to our study maintains the colloidal stability and good dispersion of the ferrofluid during the magnetic heating stage. Specifically, in vitro 2 mul of the ferrofluid gives an outstanding temperature rise of 33 degrees C within 10 min, while in vivo treatment, by infusing 150 mul of the ferrofluid in animal model (rat) glioma tumors, causes an impressive cancer tissue dissolution.
Rapid magnetic heating treatment by highly charged
maghemite nanoparticles on Wistar rats exocranial glioma
tumors at microliter volume
Ioannis Rabias,1,a?Danai Tsitrouli,1Eleni Karakosta,1Thomas Kehagias,2
Georgios Diamantopoulos,1Michael Fardis,1Dimosthenis Stamopoulos,1
Thomas G. Maris,3Polykarpos Falaras,4Nikolaos Zouridakis,4
Nikolaos Diamantis,5Georgios Panayotou,5Dimitrios A. Verganelakis,6
Garyfalia I. Drossopoulou,7Effie C. Tsilibari,7and Georgios Papavassiliou1
1Institute of Materials Science, National Centre for Scientific Research
“Demokritos”, Athens 15310, Greece
2Department of Physics, Aristotle University of Thessaloniki, Thessaloniki 54124, Greece
3Department of Medical Physics, Faculty of Medicine, University of Crete,
Heraklion, Crete 71201, Greece
4Institute of Physical Chemistry, National Centre for Scientific Research “Demokritos”,
Athens 15310, Greece
5Institute of Molecular Oncology, Alexander Fleming Biomedical Sciences Research
Center, Vari 74145, Greece
6Encephalos- Euromedica, Rizariou 3, Halandri, Attiki 15233, Greece
7Institute of Biology, National Centre for Scientific Research “Demokritos”,
Athens 15310, Greece
?Received 17 March 2010; accepted 18 May 2010; published online 21 June 2010?
One of the most significant challenges implementing colloidal magnetic nanopar-
ticles in medicine is the efficient heating of microliter quantities by applying a low
frequency alternating magnetic field. The ultimate goal is to accomplish nonsurgi-
cally the treatment of millimeter size tumors. Here, we demonstrate the synthesis,
characterization, and the in vitro as well as in vivo efficiency of a dextran coated
maghemite ??-Fe2O3? ferrofluid with an exceptional response to magnetic heating.
The difference to previous synthetic attempts is the high charge of the dextran
coating, which according to our study maintains the colloidal stability and good
dispersion of the ferrofluid during the magnetic heating stage. Specifically, in vitro
2 ?l of the ferrofluid gives an outstanding temperature rise of 33 °C within 10
min, while in vivo treatment, by infusing 150 ?l of the ferrofluid in animal model
?rat? glioma tumors, causes an impressive cancer tissue dissolution. © 2010 Ameri-
can Institute of Physics. ?doi:10.1063/1.3449089?
Ferrofluids are stable colloidal dispersions of magnetic nanoparticles in any ordinary liquid
typically in aqueous media.1–3The magnetism of such nanoparticles has been an active field of
research for over 50 years.4–6Recently, ferrofluids attracted strong interest, as they exhibit unique
properties that can be finely tuned by adjusting their composition, coating, size, and shape at the
nanoscale. This raises possibilities for important biomedical applications, both in vitro and in vivo,
in the fields of pharmaceutics, medical diagnosis, and treatment.7–12
A remarkable property of these colloidal systems is the considerable heating effect due to
several loss mechanisms when subjected to an alternating ?ac? magnetic field. This has been
utilized for applications in cancer therapy. Specifically, ferrofluids are selectively deposited in
tumors, and subsequently irradiated with an ac magnetic field, in the radio-frequency ?RF? range
a?Author to whom correspondence should be addressed. Electronic mail: firstname.lastname@example.org.
BIOMICROFLUIDICS 4, 024111 ?2010?
4, 024111-11932-1058/2010/4?2?/024111/8/$30.00© 2010 American Institute of Physics
of 50 kHz–1 MHz ?magnetic heating?.13In this way the energy absorption and the respective heat
delivery are focused on the tumor without influencing the surrounding healthy tissue. However, by
scaling the irradiated volume down to a few millimeters, which is desirable for treating tumors at
the limit of the medical diagnosis resolution without affecting the healthy tissue, the efficiency of
the method decreases dramatically due to the extremely small quantity of ferrofluid that can be
adsorbed in tumors of such small size.14It is thus a compulsory task to find ways to increase the
efficiency of the magnetic heating of biocompatible ferrofluids that are produced with standard
wet chemistry methods.
In this paper, we report the synthetic route, the structural, and the magnetic properties of
highly charged dextran coated maghemite ??-Fe2O3? nanoparticles with remarkable energy ab-
sorption from an ac magnetic field. An important factor for efficient magnetic heating, especially
in microliter quantities, is the nanoparticle concentration in the aqueous solution. In principle, the
denser the ferrofluid ?in our case 20 mg/ml, see Sec. II?, the highest heating effect obtained,
providing that surface charges prohibit nanoparticle agglomeration. On the other hand, dense
ferrofluids have considerable interparticle interactions, which trend to particle clustering and
strong reduction of the magnetic heating efficiency. Hence, surface charge plays a pivotal role in
the stability of the ferrofluid, especially at elevated temperatures.
Synthetic efforts have thus been focused on the optimization of the surface charge of the
dextran coating, while keeping pH close to the physiological acidity value in the living organisms
?pH?7?, which is a prerequisite for biological applications. Specifically, 2 ?l of the fluid ?con-
taining 40 ?g of ?-Fe2O3? exhibits a temperature rise of 33 °C within 10 min, while 150 ?l
containing 3 mg of ?-Fe2O3reached to temperatures up to 99 °C within 3 min with no indication
of flocculation, when exposed to a moderate magnetic field with amplitude of 11 kA/m and
frequency of 150 kHz. We also demonstrate the in vivo efficiency of our ferrofluid by inserting
150 ?l in small rat glioma tumors, causing significant cancer tissue dissolution, after heating for
approximately 20 min.
II. MATERIALS AND METHODS
A. Ferrofluid preparation—structural and magnetic characterization
A dispersion of maghemite nanoparticles was synthesized by the reaction of ferric chloride
and ferrous chloride in the presence of KOH using coprecipitation method.15,16Three ferrofluid
suspensions were synthesized: One comprised of uncoated maghemite nanoparticles and two
comprised of nanoparticles coated with dextran, a biocompatible macromolecule which easily
adsorbs on the maghemite surface. The dextran molecule hinders cluster growth after nucleation
due to the reaction between iron cations and hydroxide anions. The nominal molecular weight of
the dextran coating used was 55 000.
Specifically, maghemite nanoparticles ?ferrofluid samples S1, S2, S3? were prepared by co-
precipitation mixing 100 ml acidic solutions of 0.66 M FeCl3and 0.33 M FeCl2. In the case of S2,
S3 ferrofluids with 1 wt % of dextran were also added. A 100 ml alkaline solution of KOH ?1 M?
was then added dropwise over 5–10 min with stirring on a magnetic stirrer while maintaining
constant temperature to prevent widening of nanoparticles’size distribution. The stirring continued
for a short period of 20 min under a nitrogen-gas atmosphere at 60 °C—a longer period would
introduce particle growth. The particles obtained were washed three times with ethanol and water
using ultracentrifugation ?5000 rpm for 10 min at 10 °C? with nitrogen purged water. The ferrof-
luids were prepared at high pH and were constantly stirred during this fast one step reaction to
avoid the appearance of larger particles. The pH of the final ferrofluid was decreased by carefully
numbered titrations of 0.01 M HCl ?three washes? to achieve neutrality. In the last titration 0.1%
of dextran was added for better dispersion. Last, and of extreme importance, continuous sonication
for 10 min and repetitive filtering of the final colloid through a 0.2 ?m porous membrane were
performed. For comparison and optimization reasons, three ferrofluids with different surface
charges were investigated: An ionic one, with no coating ?S1? and zeta potential value of 8 mV,
and two dextrans coated with zeta potential values of 70 mV ?S2? and 350 mV ?S3?, respectively.
024111-2Rabias et al.Biomicrofluidics 4, 024111 ?2010?
In the case of S1, nanoparticle agglomerations, with average size of ?200 nm according to
dynamic light scattering measurements, were formed right after the synthesis at neutral pH. The
highest values of zeta potential for S1 ?40 and –40 mV, respectively? were observed at low ?
?4? and high ??8? pH values, most likely due to the hydrodynamic forces introduced by soni-
cation, which were strong enough to prevent the formation of weak bonds. On the other hand, both
dextran coated S2 and S3 ferrofluids exhibited positive zeta potential values at neutral pH and
were nicely dispersed into the fluid due to strong electrostatic and steric repulsion forces.
In this way, zeta potential values as high as 350 mV ?S3? at pH?7 were achieved. The coated
maghemite nanoparticle ferrofluid ?S3? yield, determined by weighing of the lyophilized sample of
the preparation, was 20 mg/ml, a value which characterizes a dense colloidal suspension.
For the structural characterization, the crystal structure, size, and shape of the maghemite
nanoparticles were determined by high-resolution transmission electron microscopy ?HRTEM? in
a Jeol 2011 electron microscope operated at 200 kV with a point resolution of 0.194 nm and Cs
=0.5 mm. The maghemite nanoparticles were magnetically characterized on a Quantum Design
superconducting quantum interference device magnetometer. Magnetic hyperthermia experiments
were performed using an induction heating furnace equipped with water cooled coils, operating at
150 kHz. The method used for temperature measurements is based on a fiber optic thermometry
system that allows users to measure temperature in extremely harsh electromagnetic environments.
The technology involved is called fluoroptic thermometry ?FOT? and is based on probes that are
totally immune to electromagnetic interference and of entirely nonmetallic construction. These
qualities make FOT probes perfectly suited for measuring temperatures in harsh environments
often encountered during biomedical research, such as ?a? high static magnetic fields, kilohertz
range switching gradients, etc. used in magnetic resonance imaging and ?b? alternating magnetic
fields at the kilohertz range like induction heating devices. The commercially available system
used in this study for accurate temperature measurements is the m3300 Biomedical Laboratory Kit
fluoroptic thermometer ?Luxtron, LumaSence Technologies, Santa Clara, CA, USA?. The probe is
protected in a Tefzel jacket and is only 500 ?m in diameter, and thus suitable for measuring
microliter quantities of ferrofluids.
B. Glioma cell lines—animals
Male Wistar rats weighing 250–350 g were used in this study. The rats were fed ad libitum
with standard laboratory food and water. They were individually housed in a controlled environ-
ment ?18–22 °C; 50%–75% relative humidity? and maintained under a 12 h dark cycle. All
animal procedures adhered to standard principles of animal care and were approved by the local
animal welfare committee.
C6 glioma cells were maintained in Ham F12 minimum essential medium containing 10%
fetal bovine serum, 1% penicillin/streptomycin. The cells were grown to confluency in a humidi-
fied atmosphere of 5% CO2at 37 °C. Exponential growth cultures were harvested with a solution
of 0.05% trypsin and 0.02% EDTA and resuspended in Ham F12 medium. C6 cells were washed
in Ham F12 medium and viable cells counted by a hemocytometer. Finally, cells were suspended
in Ham F12 minimum essential medium to a final concentration of 8?106cells per 10 ?l for
Rats were anesthetized by IP injection of xylazine ?Bayer, Leverkusen, Germany? at 10 mg/kg
and ketamine ?Parke-Davis, Courbevoie, France? at 100 mg/kg in combination with Atropine at
0.05 mg/kg. Tissue covering the skull was gently elevated and the bregma identified. Approxi-
mately 8?106cells were inoculated in the area anterior of the bregma.
After 2 weeks, the development of a tumor was palpable and visible. On the fourth week
tumors had an average size of 5–10 mm ?Fig. 1?, and inoculation of nanoparticles took place. Rats
were anesthetized as above.
Tumors were infused with 150 ?l of S3 ferrofluid and the rats were subjected to magnetic
hyperthermia treatment for 20 min. Following the treatment, the rats were sacrificed and tumors
024111-3Heating by nanoparticles on glioma-tumorBiomicrofluidics 4, 024111 ?2010?
were excised and fixed in formaldehyde. Paraffin-embedded sections were stained with
hematoxylin-eosin and examined under a Nikon Eclipse E800 microscope using 2x and 20x
III. RESULTS AND DISCUSSION
A. Structural and magnetic data
Nanoparticles have been examined by HRTEM and readily compared with the corresponding
structural properties of bulk maghemite. In Fig. 2, HRTEM images of the bare ?S1? and highly
charged coated ?S3? samples, together with images of bulk maghemite, are shown, which illustrate
the atomic structure of individual nanoparticles and bulk material, oriented along the ?1¯12? and
?011? low-index zone axes of the cubic system. The experimental interplanar spacings resolved in
the studied samples were 0.48?0.01, 0.41?0.01, 0.29?0.01, and 0.25?0.01 nm ?Fig. 2?, which
were unambiguously attributed to the d111?0.482 nm?, d200?0.417 nm?, d220?0.295 nm?, and d311
?0.252 nm? d-spacings of maghemite. In all cases defect-free crystalline structures were deter-
mined ?Figs. 2?a?, 2?b?, and 2?d??. No extended defects, i.e., grain boundaries, were observed
within nanoparticles, which indicate that the great majority of nanoparticles consisted of single
crystalline grains. A few twin boundaries between small uncoated nanoparticles in S1 were de-
tected and analyzed ?Fig. 2?d??, while the analysis verified positively the cubic crystal structure of
S3 consisted of maghemite crystal particles that exhibited rounded corners and roughly ellip-
soid shapes having core sizes between 6 and 20 nm in diameter with the main crystal population
in the range of 10–12 nm ?Figs. 2?a? and 2?b??. Moreover, the uncoated maghemite crystal par-
ticles ?S1? had sizes varying from 3 to 25 nm in diameter, while its main crystal population was
measured in the range of 8–12 nm, respectively. The latter exhibited rectangular or cubic shapes,
whereas the smaller ones showed a spherical shape as expected for nanoparticles less than 5 nm in
diameter. In addition, the practically absolute match between experimental values of interplanar
spacings of nanocrystals and the corresponding structural properties of the bulk crystal suggested
strain-free coated and uncoated nanoparticles.
FIG. 1. T2weighted magnetic resonance image of rat head. The glioma tumor area marked with yellow is pointed with a
024111-4Rabias et al. Biomicrofluidics 4, 024111 ?2010?
On the basis of our HRTEM data we thus anticipate that ferrofluids S2 and S3 are monodis-
perse, well crystalline systems with average size of 10–12 nm. The magnetic moments of samples
S1 and S3 were measured at room temperature as a function of the magnetic field and showed
superparamagnetic behavior, as expected for maghemite nanoparticles of 10 nm diameter. The
obtained saturation magnetizations at room temperature were 61 emu/gr and 39 emu/gr for the S1
and S3 samples, respectively ?the S2 sample gave similar results to S3?.
B. In vitro and in vivo magnetically induced hyperthermia
The reason for synthesizing highly charged nanoparticles for magnetic hyperthermia applica-
tions is to endure constant heating for several minutes at high temperatures and keep the colloidal
stability of the ferrofluid intact. In a colloidal suspension during magnetically induced RF heating,
the resulting rising of temperature increases the average numbers of particle collisions. Hence,
keeping nanoparticles as far as possible, by increasing electrostatic repulsion, it is possible to
minimize the number of particle collisions ?and thus prevent agglomeration, which dramatically
induces the efficiency of RF heating? for a long period of time.
Figure 3 shows the temperature dependence as a function of the heating time for the three
ferrofluids. Heating was performed with a copper coil ?inset in Fig. 3?b??, which produced a
moderate ac magnetic field with amplitude of 11 kA/m and frequency of 150 kHz, much lower
FIG. 2. HRTEM micrographs. ?a? Dextran coated crystalline particles ?S3? of 10–12 nm in diameter ?optimally charged?.
?b? A single crystalline dextran coated nanoparticle ?S3? oriented along the ?011? zone axis. ?c? A part of bulk maghemite
crystals viewed along the ?1¯12? direction along with the corresponding selected area electron diffraction pattern. ?d? Two
uncoated ?S1? nanoparticles in twin orientation, where arrows signify the twin boundary. Both are oriented along their
corresponding ?011? zone axes. The experimental values of the angles between ?111?/?200? ?55°? and ?111?/?220? ?35°?
crystal planes are shown as well.
024111-5Heating by nanoparticles on glioma-tumorBiomicrofluidics 4, 024111 ?2010?
than the typical value of 400–500 kHz commonly used.17–19Samples were heated from room
temperature to the maximum temperature value and then left to cool down back to the initial
In Fig. 3?a? we observe the strong effect of the zeta potential upon heating: S1 with the lowest
zeta potential value does not produce any significant heating effect and started to flocculate right
after heating. S2 exhibits a respectable temperature rise, reaching 60 °C within 10 min. Similar to
the S1 sample, S2 showed problems on stability and endurance, and it was instantly flocculated
after the heating treatment. On the contrary, S3 exhibits an impressive heating effect, reaching
98 °C within 3 min, with no traces of flocculation and a remarkable long-term stability, despite
the fact that it reached almost the water boiling temperature. The corresponding specific absorp-
tion rate ?SAR? is calculated at 286 W/g. This is among the highest SAR values obtained under the
specific experimental conditions ?see supporting information?.20Most spectacular, by scaling
down the quantity of S3, an exceptional heating response is observed ?Fig. 3?b??: 2 ?l of this
ferrofluid containing 40 ?g of ?-Fe2O3?red line? exhibits a temperature rise of 33 °C within 10
min. As far as we know, this result is the highest increase in temperature for such a microgram
quantity of maghemite nanoparticles reported in the literature.
The experimental data for sample S3 in Fig. 3?b? have been analyzed according to T?t?
=Tinitial+?Tmax?1−exp?−t/???, where Tinitialis the initial temperature, ?Tmaxthe overall tempera-
ture change, and ? the time constant of heating.21The pertinent quantity in magnetic hyperthermia
is the initial rate ?dT/dt?t=0which in our case equals to ?Tmax/?. The fits of the above equation to
the experimental data during the heating process are shown as solid black lines in Fig. 3?b? and
yield similar heating rate ?Tmax/??17 °C/min for the 2 and 10 ?l samples. Reliable analysis of
the 150 ?l sample was not possible due to the fast attainment of the water boiling temperature.
However, by fitting the experimental data during the cooling process, the same cooling rate
?Tmax/??12 °C/min is obtained for all three samples. This demonstrates the accuracy and reli-
ability of the experiments. In addition, the 150 ?l sample had not been physically or chemically
destroyed by reaching the water boiling temperature, a sign of the high stability of the synthesized
nanodispersion. Evidently, the principal factor which controls the heating effect in the colloid state
is the prevention of agglomeration, i.e., the elimination of interparticle interactions, providing that
the size and magnetic properties of the maghemite core particles are “tuned” to maximum energy
absorption. The in vitro experiments of our ferrofluids markedly showed that it is the surface
charge that prevents agglomeration during the heating stage of the ferrofluid by minimizing the
collisions between the nanoparticles. It is anticipated that strong interactions are present in the
FIG. 3. Magnetic heating effect on aqueous dispersions of dextran coated maghemite nanoparticles. ?a? The influence of
the zeta potential on 150 ?l of ferrofluid S1 ?green line?, S2 ?blue line?, and S3 ?black line?. ?b? The influence of the
volume on 150 ?l ?black line?, 10 ?l ?blue line?, and 2 ?l ?red line? of ferrofluid S3. Solid black lines refer to fitting of
the experimental data ?see text for details?. The green line is the curve for 2 ?l water that was used as control sample. In
the inset, the coil of the magnetic heating apparatus with the 10 ?l sample and the optical fiber used for the temperature
monitoring are shown after removing the heat insulating cover.
024111-6Rabias et al.Biomicrofluidics 4, 024111 ?2010?
ferrofluids due to the high concentration of maghemite in the suspensions ?20 mg/ml?, and it is the
presence of the charged coating that prevents agglomeration and gives rise to the strong heating
effect in the 350 mV S3 sample. Indeed, the less charged 70 mV S2 sample, which has the same
high concentration ?20 mg/ml? as the S3 sample, presents less magnetic heating than the S3
ferrofluid ?Fig. 3?a?? due to the fact that the weak charge cannot compensate the large interparticle
interactions and the S2 material exhibits flocculation and no colloidal stability during magnetic
In order to demonstrate the ability of the S3 ferrofluid to function in vivo and assess its
medical potential, we infused 150 ?l in small rat tumors with average size of 5–10 mm and
applied an 11 kA/m ac field at 150 kHz for 20 min. Specifically, C6 rat glioma cells were injected
extracranially and allowed to form tumors, which were infused with ferrofluid. The rats’ heads
were then placed within the induction-heating coil ?lower inset in Fig. 4?b??. Infrared temperature
image ?upper inset in Fig. 4?b?? shows the ability to heat a precise area of the glioma tumor.
Following the treatment, the rats were sacrificed and tumors were excised and paraffin-embedded
sections were stained to reveal the nuclei and cytoplasm of cells using hematoxylin and eosin, a
standard staining protocol for examining tissue integrity. As can be seen in Figs. 4?b? and 4?d?,
treatment led to extensive tumor tissue damage and dissolution, indicating that the obtained tem-
peratures were close to boiling temperatures. Control experiments included the infusion of S3
ferrofluid without subsequent treatment as well as exposure to magnetic hyperthermia in the
absence of ferrofluid ?Figs. 4?a? and 4?c??. No tumor tissue damage was observed in either case.
Thus, the produced surface charged coated ferrofluid S3 performed excellently in the in vivo
studies due to the preservation of the good dispersion of the ferrofluid within the tumor during the
FIG. 4. For in vivo magnetic heating of small rat glioma tumors 150 ?l of the S3 ferrofluid was infused into rat glioma
tumors with size of 5–10 mm and subsequently subjected to magnetic hyperthermia treatment for 20 min. Panels ?a? and
?c? ?low and high magnifications, respectively? represent sections of control tumor tissue treated without nanoparticles.
Panels ?b? and ?d? ?low and high magnifications, respectively? show extensive damage of the tumor tissue after treatment
with ferrofluid. White bars=200 ?m. The lower inset in panel ?b? shows the experimental setup for in vivo magnetic
heating. The upper inset is an infrared image, which demonstrates the ability of the ferrofluid to produce strong localized
heating at the tumor position.
024111-7Heating by nanoparticles on glioma-tumorBiomicrofluidics 4, 024111 ?2010?
To summarize, we have investigated the synthesis, the structural magnetic properties of highly
charged coated maghemite nanoparticles with remarkable absorption of an ac magnetic field at
very small quantities and demonstrate their in vivo efficiency for hyperthermia treatment of small
size tumors. The exceptional heating performance of the produced ferrofluid is due to ?i? the
appropriate size and excellent magnetic properties of the core nanoparticles and ?ii? the high
surface charge of the dextran coating that produces homogeneous and stable nanoparticle disper-
sion resulting to minimal interparticle interactions, minimum collision rate, and maximum energy
absorption during heating. A balance between density and charge is the key for an effective and
stable ferrofluid at elevated temperatures. To the best of our knowledge the microliter ferrofluid
quantities used for magnetic heating are the smallest reported in the literature.
The authors are thankful to the Greek General Secretariat for Research and Technology for
providing financial support through National Projects EPAN ?YB-23? and PEP Attikis ?ATT-25?.
1J. L. Dormann, D. Fiorani, and E. Tronc, Adv. Chem. Phys. 98, 283 ?1997?.
2T. Lahaye, T. Koch, B. Fröhlich, M. Fattori, J. Metz, A. Griesmaier, S. Giovanazzi, and T. Pfau, Nature ?London? 448,
3S. A. Safran, Nature Mater. 2, 71 ?2003?.
4L. Néel, Ann. Geophys. 5, 99 ?1949?.
5L. E. Euliss, S. G. Grancharov, S. O’Brien, T. J. Deming, G. D. Stucky, C. B. Murray, and G. A. Held, Nano Lett. 3,
6J. Dobson, Nat. Nanotechnol. 3, 139 ?2008?.
7Q. A. Pankhurst, J. Connolly, S. K. Jones, and J. Dobson, J. Phys. D 36, R167 ?2003?.
8A. Ito, F. Matsuoka, H. Honda, and T. Kobayashi, Cancer Gene Ther. 10, 918 ?2003?.
9K. Okawa, M. E. Sekin, M. Maeda, M. Tada, and M. Abe, J. Appl. Phys. 99, 08H102 ?2006?.
10O. Veiseh, C. Sun, J. Gunn, N. Kohler, P. Gabikian, D. Lee, N. Bhattarai, R. Ellenbogen, R. Sze, A. Hallahan, J. Olson,
and M. Zhang, Nano Lett. 5, 1003 ?2005?.
11I. Rabias, M. Fardis, E. Devlin, N. Boukos, D. Tsitrouli, and G. Papavassiliou, ACS Nano 2, 977 ?2008?.
12I. Rabias, H. Pratsinis, G. Drossopoulou, M. Fardis, T. Maris, N. Boukos, N. Tsotakos, D. Kletsas, E. Tsilibary, and G.
Papavassiliou, Biomicrofluidics 1, 044104 ?2007?.
13I. A. Brezovich, Med. Phys. Monogr. 16, 82 ?1988?.
14R. Hergt and S. Dutz, J. Magn. Magn. Mater. 311, 187 ?2007?.
15R. Massart and V. Cabuil, J. Phys. Chem. 84, 967 ?1987?.
16C. G. Granqvist and R. H. Buhrman, J. Appl. Phys. 47, 2200 ?1976?.
17J. P. Fortin, C. Wilhelm, J. Servais, and C. Menager, J. Am. Chem. Soc. 129, 2628 ?2007?.
18R. E. Rosensweig, J. Magn. Magn. Mater. 252, 370 ?2002?.
19R. Hergt, R. Hiergeist, M. Zeisberger, G. Glöckl, W. Weitschies, L. P. Ramirez, I. Hilger, and W. A. Kaiser, J. Magn.
Magn. Mater. 280, 358 ?2004?.
20See supplementary material at http://dx.doi.org/10.1063/1.3449089 for the method for temperature measurements, SAR
values: Comparison with literature, and characterization techniques for the nanoparticles.
21A. Józefczak and A. Skumiel, J. Magn. Magn. Mater. 311, 193 ?2007?.
024111-8Rabias et al.Biomicrofluidics 4, 024111 ?2010?