Water-dispersible sugar-coated iron oxide nanoparticles. An evaluation of their relaxometric and magnetic hyperthermia properties

Article abstract of DOI:10.1021/ja111448t

Synthesis of functionalized magnetic nanoparticles (NPs) for biomedical applications represents a current challenge. In this paper we present the synthesis and characterization of water-dispersible sugar-coated iron oxide NPs specifically designed as magnetic fluid hyperthermia heat mediators and negative contrast agents for magnetic resonance imaging. In particular, the influence of the inorganic core size was investigated. To this end, iron oxide NPs with average size in the range of 4-35 nm were prepared by thermal decomposition of molecular precursors and then coated with organic ligands bearing a phosphonate group on one side and rhamnose, mannose, or ribose moieties on the other side. In this way a strong anchorage of the organic ligand on the inorganic surface was simply realized by ligand exchange, due to covalent bonding between the Fe³⁺ atom and the phosphonate group. These synthesized nanoobjects can be fully dispersed in water forming colloids that are stable over very long periods. Mannose, ribose, and rhamnose were chosen to test the versatility of the method and also because these carbohydrates, in particular rhamnose, which is a substrate of skin lectin, confer targeting properties to the nanosystems. The magnetic, hyperthermal, and relaxometric properties of all the synthesized samples were investigated. Iron oxide NPs of ca. 16-18 nm were found to represent an efficient bifunctional targeting system for theranostic applications, as they have very good transverse relaxivity (three times larger than the best currently available commercial products) and large heat release upon application of radio frequency (RF) electromagnetic radiation with amplitude and frequency close to the human tolerance limit. The results have been rationalized on the basis of the magnetic properties of the investigated samples.

Full text of DOI:10.1021/ja111448t

ARTICLE
pubs.acs.org/JACS
Water-Dispersible Sugar-Coated Iron Oxide Nanoparticles.
An Evaluation of their Relaxometric and Magnetic
Hyperthermia Properties
Lenaic Lartigue,†,‡ Claudia Innocenti,‡ Thangavel Kalaivani,§,^ Azzam Awwad,|| Maria del Mar Sanchez Duque,†
Yannick Guari,*,† Joulia Larionova,† Christian Guꢀerin,† Jean-Louis Georges Montero,|| Vꢀeronique Barragan-Montero,*,||
Paolo Arosio,§ Alessandro Lascialfari,§,^,# Dante Gatteschi,‡ and Claudio Sangregorio*,‡,r
†Institut Charles Gerhardt Montpellier, UMR 5253 CNRS-UM2, Chimie Molꢀeculaire et Organisation du Solide, Universitꢀe Montpellier
II, Place E. Bataillon, 34095 Montpellier Cedex 5, France
‡Dipartimento di Chimica, Universitꢁa di Firenze and INSTM Research Unit, via della Lastruccia 3, 50019 Sesto F.no Firenze, Italy
§Dipartimento di Scienze Molecolari Applicate ai Biosistemi, Universitꢁa degli Studi di Milano and Consorzio INSTM, Milano Unit,
I-20134 Milano, Italy
^Centro S3, CNR-Istituto di Nanoscienze, I-41125 Modena, Italy
Equipe SyGReM, Institut des Biomolꢀecules Max Mousseron, UMR 5247, CNRS-UM1-UM2, B^atiment de Recherche Max Mousseron,
Ecole Nationale Supꢀerieure de Chimie de Montpellier, 8 Rue de l’Ecole Normale, 34296 Montpellier Cedex, France
#Dipartimento di Fisica “A. Volta”, Universitꢁa degli Studi di Pavia, Via Bassi 6, I-27100 Pavia, Italy
rCNR-ISTM, Via C. Golgi 19, I-23310 Milano, Italy
S Supporting Information
b
ABSTRACT: Synthesis of functionalized magnetic nanoparticles
(NPs) for biomedical applications represents a current challenge. In
this paper we present the synthesis and characterization of water-
dispersible sugar-coated iron oxide NPs specifically designed as
magnetic fluid hyperthermia heat mediators and negative contrast
agents for magnetic resonance imaging. In particular, the influence of
the inorganic core size was investigated. To this end, iron oxide NPs
with average size in the range of 4ꢀ35 nm were prepared by thermal
decomposition of molecular precursors and then coated with organic
ligands bearing a phosphonate group on one side and rhamnose,
mannose, or ribose moieties on the other side. In this way a strong anchorage of the organic ligand on the inorganic surface was
simply realized by ligand exchange, due to covalent bonding between the Fe3þ atom and the phosphonate group. These synthesized
nanoobjects can be fully dispersed in water forming colloids that are stable over very long periods. Mannose, ribose, and rhamnose
were chosen to test the versatility of the method and also because these carbohydrates, in particular rhamnose, which is a substrate of
skin lectin, confer targeting properties to the nanosystems. The magnetic, hyperthermal, and relaxometric properties of all the
synthesized samples were investigated. Iron oxide NPs of ca. 16ꢀ18 nm were found to represent an efficient bifunctional targeting
system for theranostic applications, as they have very good transverse relaxivity (three times larger than the best currently available
commercial products) and large heat release upon application of radio frequency (RF) electromagnetic radiation with amplitude and
frequency close to the human tolerance limit. The results have been rationalized on the basis of the magnetic properties of the
investigated samples.
’ INTRODUCTION
physicochemical properties enable them to operate as probes
and delivery vectors suitable as candidates for the next generation
of diagnostic and therapeutic techniques. In this regard, an
exciting potential of magnetic NPs resides in the possibility of
preparing multifunctional theranostic NPs which may be used as
contrast agents (CA) for magnetic resonance imaging (MRI) and
Nanometer-scale magnetic nanoparticles (NPs) have at-
tracted widespread attention due to the rapidly increasing
number and variety of their applications in the biomedical
sciences, including imaging1 and therapy.2 Indeed, they possess
unique nanoscale size-dependent physical and chemical proper-
ties that can be controlled in a manner that is not possible in
the corresponding bulk materials.3 When these tiny materials
are introduced into biological systems, their small size and
Received: December 20, 2010
Published: May 23, 2011
r
2011 American Chemical Society
10459
dx.doi.org/10.1021/ja111448t J. Am. Chem. Soc. 2011, 133, 10459–10472
|

Journal of the American Chemical Society
ARTICLE
as heat nanosources to achieve targeted treatment of malig-
nant cells.
others. Further, magnetic iron oxide NPs have been engineered by
conjugation with biomolecules to target specific gene expression24
or to be internalized by cells and thereby to allow cell migration to be
monitored in vivo.11a
Magnetic hyperthermia, which is based on the exothermic
properties of magnetic materials under the influence of an alternat-
ing current (ac) magnetic field, is a promising cancer thermother-
apy. Since the pioneering work of Gilchrist et al.,4 magnetic
hyperthermia has been the aim of numerous in vitro and in vivo
studies. In the quest for the most efficient heat mediators several
different classes of materials have been investigated such as magnetic
nanocomposite microparticles,5 surface-modified superparamag-
netic NPs (ferrofluids)6 and ligand-targeted magnetoliposomes.7
Surface-modified super-paramagnetic NPs based on not targeted
dextran or carbodextran coated magnetite NPs8 have been proposed
for magnetic fluid hyperthermia (MFH) since the early 1980s in
cancer treatment, and successful clinical trials have been performed
on human patients in the past decade.9
Furthermore, thanks to the capability of iron oxide NPs to
enhance proton relaxation at their sites of accumulation10 and to
their high biocompatibility, this kind of material has rapidly
become an important tool in the field of contrast enhancement in
MRI images applied to the noninvasive diagnosis of many
diseases in human soft tissues.11 During the past 2 decades,
super-paramagnetic iron oxides NPs (SPIONs) and ultrasmall-
sized super-paramagnetic iron oxides (USPIONs) dispersed
in aqueous phase have been investigated to explore their
potential as contrast-enhancing probes for MRI. Some of these
materials are under clinical evaluation or already used in clinical
practice: for instance, Endorem (Guerbet), also called Feridex,
Feraheme (AMAG Pharmaceuticals), Primovist and Eovist
(Bayer Schering Pharma AG) are used as CAs while other
products such as Supravist-SHU555C (Bayer Schering Pharma
AG), VSOP-C184 (Ferropharm GmbH), Ferumoxtran-AMI-
227 and Ferumoxide-AMI-25 (AMAG Pharmaceuticals) will
enter the market soon.12 Many efforts have also been made to
design iron oxide NPs with improved properties in order to
decrease the CA concentration, increase the circulation time in
the organism, and enhance selective NPs’ transport into the cell
membrane .13
Synthetic chemistry has now developed to a stage where it is
possible to produce magnetic NPs for in vivo biomedical
applications such as medical imaging and magnetic thermother-
apy achieving the so-called “find, fight, and follow” concept of
early diagnosis and therapy control. Consequently, NPs may be
injected intravenously, utilizing the blood circulation to transport
them to their target sites, and their progression in the body may be
followed by MRI. After penetrating the tumor cell, its eradication
should be achieved for the lowest possible NP concentration.
Nevertheless, the difficulty in designing such nanoprobes consists
in the fact that they should exhibit remarkable hyperthermal and
relaxometric efficiencies at physiological conditions and have ade-
quate biocompatibility and targeting properties. Thus, the use of
magnetic NPs for theranostic applications requires better control of
the NP core, size, shape, chemical composition, degree of aggrega-
tion, and surface state,14 whereas to ensure the circulation within the
living organism, effective vectorization of these nanoprobes is
required to direct them toward the desired target allowing a
reduction in the total amount of targeted NPs compared with
untargeted NPs. Along this line of thought, numerous studies have
been devoted to the investigation of iron oxide NPs stabilized with
different biocompatible ligands including various polymers,15
albumin,16 folic acid,17 dextran,18 pluronic,19 starch,20 herceptin,21
nucleic acid molecules (aptamer),22 peptides (cRGDyK),23 and
Among the different biocompatible tumor cell-targeting
ligands investigated to date, saccharides represent promising
molecules for the delivery of such nanoprobes. In fact, they are
involved as recognition markers in numerous physiological and
pathological processes, mainly occurring on the surfaces of the cell
and can be recognized by carbohydrate-binding proteins, such as
antibodies, enzymes, and lectins.25 Very recently, magnetic glycol-
NPs based nanosensors bearing carbohydrate ligands were shown to
strongly interact with endogenous lectins present on the cancer cell
surface.26 Moreover, membrane-bound receptors, such as lectins,
are known to assist material internalization inside the cell via
endocytosis following the binding of their ligand.27 Some examples
of saccharide-coated iron oxide NPs have been reported: these
include the attachment on Fe3O4 NPs of dihydroxyborylphenyl
groups functionalized with different mono- and oligosaccharides,28
the grafting of silica-coated Fe3O4 NPs with functionalized
D-mannose through a triazol linker,29 the noncovalent coating of iron
oxide NPs with poly(D,L-lactide-co-glycoside) via emulsion
techniques,30 or the co-precipitation of iron salts in the presence
of D-mannose.31
We have recently reported a new approach to achieve size
controlled water-dispersible magnetic iron oxide NPs covalently
linked to monosaccharide molecules via a phosphonate moiety.32
These water-dispersible biocompatible rhamnose-coated iron
oxide NPs of 4.0 nm were shown to have super-paramagnetic
behavior and nuclear relaxivities in the same order of magnitude
as Endorem. These characteristics make them potential third
generation MRI CAs because the saccharides also represent
specific ligands able to target lectins on skin cells. We first
focused on rhamnose, as it has been shown that fluorescent
rhamnose-coated liposomes could transfer fluorescence from
liposomes to skin cells. Moreover, ex vivo experiments on human
skin section have demonstrated specific delivery to the epidermis
or the horny layer due to the presence of rhamnose ligands on the
surface of liposomes.33 These preliminary results encouraged us
to further investigate this approach and to evaluate the potential
multifunctional theranostic efficiency of the sugar-coated iron oxide
NPs combining the MFH therapeutic action with the MRI diag-
nostic. To this end, our strategy for the synthesis of NPs was
extended to other sugar derivatives of interest, such as mannose
which is an efficient vector to target macrophages and hepathic
cells34 and, in order to demonstrate the general applicability of the
synthetic methodology, to the pentose derivative, ribose. Further, to
investigate the role of the inorganic core, rhamnose-coated iron
oxides NPs with various sizes in the range of 4ꢀ35 nm were
synthesized and their magnetic, relaxometric, and hyper-thermal
properties systematically studied.
’ EXPERIMENTAL SECTION
Nanoparticle Labeling. Molecules derived from carbohydrates
bearing ester phosphonate groups were referred to as Rha for rhamnose,
Man for mannose, and Rib for ribose (see Schemes 1 and 2). The iron
oxide NPs functionalized with these ligands are named as follows:
the prefix Rha, Man, or Rib denotes the type of carbohydrate deriva-
tive grafted on the surface of the nanoparticles of ca. 4, 7, 10, 16, 18,
or 35 nm. As an example, Rha-4 denotes iron oxide nanoparticles of
4 nm in size grafted with the rhamnose derivative. Thus, we obtained a
10460
dx.doi.org/10.1021/ja111448t |J. Am. Chem. Soc. 2011, 133, 10459–10472

Journal of the American Chemical Society
ARTICLE
Scheme 1. Synthesis of the Rhamnose Derivative Functio-
nalized at the Anomeric Position, RhaAC
(dichloromethane/methanol, 9/1) to give the expected compound in
64% yield. Rf = 0.52 (CH2Cl2/MeOH, 95/5). [R]D: ꢀ38.2 (c = 1.03,
CHCl3). SM (ESI > 0/MeOH/30 eV): m/z 445.32 [M þ Na]þ 1H.
NMR (400 MHz, CDCl3; ppm): δ 1.18 (3H, d, J6ꢀ5 = 6,3 Hz, CHꢀCH3),
1.95, 2.01, 2.11 (9H, 3s, CH3CO), 3.57ꢀ3.71 (12H, m, CH2ꢀO), 3.89
(1H, dd, J5ꢀ4 = 9.8 Hz, J5ꢀ6 = 6.3 Hz, CHꢀCH3), 4.75 (1H, d, J1ꢀ2 = 1.7
Hz, OꢀCHꢀO), 5.03 (1H, dd, J4ꢀ5 = 9.9 Hz, J4ꢀ3 = 9.9 Hz, CH3ꢀ
CHꢀCH), 5.22 (1H, dd, J2ꢀ3 = 3.5 Hz, J2ꢀ1 = 1.7 Hz, CHꢀCHꢀO), 5.27
(1H, dd, J3ꢀ4 = 10.1 Hz, J3ꢀ2 = 3.5 Hz, CH3ꢀCHꢀCHꢀCH). 13C NMR
(100.6MHz, CDCl3; ppm): δ17.3 (CHꢀCH3), 20.6, 20.7, 20.8 (COCH3),
61.7 (CꢀCH2O), 66.3 (CꢀCH2O), 67.1 (CꢀCH2O), 69.0 (CHꢀCH3),
69.8, 70.0, 70.3 (CꢀCH2O), 70.7 (CHꢀCHꢀO), 71.0 (CH3ꢀCHꢀ
CHꢀCH), 72.5 (CH3ꢀCHꢀCH), 97.5 (OꢀCHꢀO), 169.9, 170.0,
170.1 (COCH3).
Synthesis of 2-[2-(2-Bromoethoxy)ethoxy]ethyl-2,3,4,-tri-O-acetylr-
hamnopyranoside. A 4.02 mmol amount of 2-[2-(2-hydroxyethoxy)-
ethoxy]ethyltri-O-acetylrhamnopyranoside and 4.82 mmol (1.2 equiv) of
tetrabromomethane were dissolved in dichloromethane (20 mL) before
adding 5.63 mmol (1.4 equiv) of triphenylphosphine. The mixture was
magnetically stirred under argon at room temperature for 150 min. The
mixture was diluted in dichloromethane (50 mL), then washed (2 ꢁ 50 mL
of brine), and dried over anhydrous sodium sulfate, filtrated, and concen-
trated in vacuo. The residue was chromatographed on silica gel (ether) to
give the expected compound in 92% yield. SM (ESI > 0/MeOH/30 eV): m/z
507.29 and 509.2 [M þ Na]þ. 1H NMR (400 MHz, CDCl3; ppm): δ 1.14
(3H, d, J6ꢀ5 = 6.3 Hz, CHꢀCH3), 1.91, 1.98, 2.08 (9H, 3s, CH3CO),
3.39ꢀ3.43 (2H, m, CH2ꢀBr), 3.56ꢀ3.77 (10H, m, CH2ꢀO), 3.86 (1H,
dd, J5ꢀ4 = 9.9 Hz, J5ꢀ6 = 6.4 Hz, CHꢀCH3), 4.71 (1H, d, J1ꢀ2 = 1.5 Hz,
OꢀCHꢀO), 4.99 (1H, dd, J4ꢀ5 = 9.8 Hz, J4ꢀ3 = 9.8 Hz, CH3ꢀCHꢀCH),
5.18 (1H, dd, J2ꢀ3 = 3.5 Hz, J2ꢀ1 = 1.8 Hz, CHꢀCHꢀO), 5.23 (1H, dd,
Scheme 2. Phosphonate-Functionalized Sugars RhaAC
ManAC, and RibAC
,
J
3ꢀ4 = 10.1 Hz, J3ꢀ2 = 3.6 Hz, CH3ꢀCHꢀCHꢀCH). 13C NMR (100.6
MHz, CDCl3; ppm): δ 17.3 (CHꢀCH3), 20.6, 20.7, 20.8 (COCH3), 30.3
(CH2ꢀBr), 63.4 (CꢀCH2O), 66.1 (CꢀCH2O), 67.0 (CꢀCH2O), 68.9
(CHꢀCH3), 69.9, 70.4 (CꢀCH2O), 70.6 (CH2ꢀCHꢀO), 70.9
(CH3ꢀCHꢀCHꢀCH), 71.0 (CH3ꢀCHꢀCH), 97.4 (OꢀCHꢀO),
169.7, 169.8, 169.9 (COCH3).
Synthesis of Bis(trimethysilyl)[2-{2-[2-(2,3,4-tri-O-acetylrhamno-
pyranosyl)oxy]ethoxy}ethyl]phosphonate (RhaAC). A 17.32 mmol (4
equiv) amount of tris(trimethylsilylphosphite) was added to 4.33 mmol
of 2-[2-(2-bromoethoxy)ethoxy]ethyl-2,3,4,-tri-O-acetylrhamnopyranoside
under argon. The mixture was magnetically stirred under argon at 140 ꢀC
for 5 h. The tris(trimethylsilylphosphite) was then removed under vacuum
to give the expected compound in quantitative yield. 1H NMR (400.0 MHz,
CDCl3;ppm):δ0.23 (18H, m, Me3Si), 1.18 (3H, d, J =6.3 Hz, CHꢀCH3),
1.95, 2.05, 2.13 (9H, 3s, CH3CO), 2.07ꢀ2.12 (2H, m, CH2ꢀP), 3.63ꢀ3.78
(10H, m, OꢀCH2), 3.98 (1H, dd, J = 9.6 Hz, J = 6.1 Hz, CH3ꢀCH), 4.82
(1H, s, OꢀCHꢀO), 5.00 (1H, dd, J = 9.6 Hz, J = 9.9 Hz, CH3ꢀCHꢀCH),
5.23 (2H, m, CHꢀCHꢀCHꢀCH). 13C NMR (100.6 MHz, CDCl3;
ppm): δ 15.5 (CH3ꢀCH), 18.8ꢀ18.9ꢀ19.0 (COCH3), 63.5 (CH2ꢀP),
64.4 (CH3ꢀCH), 65.2 (CꢀCH2O), 67.9ꢀ67.1 (CH3ꢀCHꢀCHꢀCH),
68.1ꢀ68.2, 68.6ꢀ68.7 (CꢀCH2O), 69.2 (OꢀCHꢀCH), 95,6 (O-
CHꢀO), 168.0ꢀ168.1ꢀ168.2 (COCH3). For hydrolyzed product: SM
(ESI > 0/MeOH/30 eV): m/z 509.24 [M þ H]þ. [R]D: ꢀ27.1 (c = 1,
MeOH).31P NMR (81.0 MHz, CDCl3; ppm): δ 22.4.
Synthesis of the Mannose Derivative (ManAC). Synthesis of 2-[2-(2-
Hydroxyethoxy)ethoxy]ethyltetra-O-acetylmannopyranoside. A meth-
odology similar to the synthesis of 2-[2-(2-hydroxyethoxy)ethoxy]
ethyltri-O-acetylrhamnopyranoside utilized 64.1 mmol of boron tri-
fluoride etherate, 12.8 mmol of D-mannose tetraacetate, and 38.4 mmol
of triethylene glycol to give the expected compound in 78% yield.
SM (ESI > 0/MeOH/30 eV): m/z 503.22 [M þ Na]þ. Rf (5/5 AcOEt/
EP, v/v) = 0.45. 1H NMR (400 MHz, CDCl3; ppm) δ 1.99ꢀ2.04,
2.10ꢀ2.16 (12H, 4s, CH3CO), 3.61ꢀ3.84 (m, 11H, CH2CH2O),
4.09ꢀ4.12 (3H, m, AcOꢀCH2-b, CHꢀCH2ꢀOAC, 1 CHꢀOꢀCH2),
series of iron oxide nanoparticles of varying size (4ꢀ35 nm) bearing
different sugars, which are Rha-4, Man-4, Rib-4, Rha-7, Rha-10, Rha-16,
Rha-18, and Rha-35.
Synthesis. The syntheses were carried out using standard anaerobic
procedures and commercially available reagents. Absolute ethanol,
hexane, methanol, and dichloromethane (99%) were used as received.
Tetrahydrofuran, THF, was purified by distillation over sodium. 1,2-
Hexadecanediol (90%), oleic acid (90%), 2.0 M ammonia solution in
methanol, iron(III) oxide hydrated catalyst grade 30ꢀ50 mesh, docosan
(99%), tris(trimethylsilyl)phosphate, triethylene glycol (TEG), boron
trifluoride etherate, tetrabromomethane, and triphenylphosphite were
purchased from Aldrich Chemical Co. Phenyl ether (99%), oleylamine
approximate C18 content 80ꢀ90% (97%), and iron(III) acetylacetonate
(99þ%) were purchased from Acros Organics. 1-Methylpyrrole (99%)
was purchased from Alfa Aesar. Rhamnose, mannose, and ribose were
purchased from Benn Chemicals.
Synthesis of the Rhamnose Derivative (RhaAC). Synthesis of
2-[2-(2-Hyroxyethoxy)ethoxy]ethyltri-O-acetylrhamnopyranoside. A
90.35 mmol amount of boron trifluoride etherate was added dropwise
at 0 ꢀC to 18.07 mmol of L-(þ)-rhamnose pentaacetate dissolved in
dichloromethane (14 mL) and 54.21 mmol of triethylene glycol. The
mixture was magnetically stirred under argon, at 0 ꢀC for 15 min, and
then under ambient conditions overnight. The mixture was diluted in
dichloromethane (25 mL) and then washed (2 ꢁ 25 mL of water, 2 ꢁ
25 mL of a solution of 5% sodium bicarbonate, and 2 ꢁ 25 mL of water).
The organic layer was dried over anhydrous sodium sulfate, filtered, and
concentrated in vacuo. The residue was chromatographed on silica gel
10461
dx.doi.org/10.1021/ja111448t |J. Am. Chem. Soc. 2011, 133, 10459–10472

Journal of the American Chemical Society
ARTICLE
4.26ꢀ4.3 (1H, dd, AcOꢀCH2-a), 4.88 (1H, s, OꢀCHꢀO), 5.27 (1H, dd,
CH2ꢀCHꢀCH), 5.30 (1H, dd, OꢀCHꢀCH), 5.35 (1H, dd,
OꢀCHꢀCHꢀCH). 13C NMR (100.6 MHz, CDCl3; ppm): δ 20.7, 20.8,
20.9 (COCH3), 61.8 (CꢀCH2O), 62.4 (CHꢀCH3), 66.2 (CꢀCH2O),
67.4 (CꢀCH2O), 68.4 (CH3ꢀCHꢀCHꢀCH), 69.1 (CHꢀCH3), 69.6,
70.0, 70.4 (CꢀCH2O), 70.8 (CHꢀCHꢀO), 72.6 (CH3ꢀCHꢀCH), 97.7
(OꢀCHꢀO), 169.8, 170.0, 170.1, 170.7 (COCH3).
of triphenylphosphite to give the expected compound in 90% yield. SM
(ESI > 0/MeOH/30 eV): m/z 493.07 [M þ Na]þ. 1H NMR (400 MHz,
CDCl3; ppm): δ 2.20ꢀ2.28 (9H, m, OꢀCH3), 3.52ꢀ3.66 (12H, m,
CH2CH2O), 4.10ꢀ4.14 (1H, dd, AcOꢀCH2ꢀCH), 4.18ꢀ4.22 (2H, m,
AcOꢀCH2ꢀCH), 4.25ꢀ4.30 (2H, m, CHꢀOAc), 5.01 (0.6H, s,
OꢀCHRꢀO), 5.19 (0.4H, s, OꢀCHβꢀO). 13C NMR (100.6 MHz,
CDCl3; ppm): δ 21.1ꢀ21.3ꢀ21.5 (H3CꢀCO), 61.7 (CꢀCH2OH), 62.4
(CHꢀCH2), 67.1 (CꢀCH2O), 70.1 (CH2ꢀOAc), 70.1ꢀ70.4ꢀ70.9
(CꢀCH2O), 72.6 (CH2ꢀBr), 73.2 (CH2ꢀOAc), 84.5 (CHꢀCH2ꢀ
OAc), 108.0 (OꢀCHꢀO), 107.2ꢀ107.4, 107.5 (CdO).
Synthesis of 2-[2-(2-Bromoethoxy)ethoxy]ethyltetra-O-acetylman-
nopyranoside. A methodology similar to the synthesis of 2-[2-(2-bro-
moethoxy)ethoxy]ethyltri-O-acetylrhamnopyranoside utilized 10 mmol of
2-[2-(2-(hydroxyethoxy)ethoxy]ethyltetra-O-acetylmannopyranoside,
12 mmol (1.2 equiv) of tetrabromomethane and 14 mmol (1.4 equiv) of
triphenylphosphite to give the expected compound in 87% yield. SM (ESI >
0/MeOH/30 eV): m/z 565.25, 567.22 [M þ Na]þ. 1H NMR (400 MHz,
CDCl3; ppm): δ 1.92ꢀ1.98, 2.04ꢀ2.10 (12H, 4s, CH3CO), 3.4 (2H, t,
CH2Br), 3.6ꢀ3.77 (m, 9H, CH2CH2O), 3.98ꢀ4.06 (3H, m, AcOꢀCH2-b,
CHꢀCH2ꢀOAc, 1 CHꢀO-CH2), 4.2ꢀ4.25 (1H, dd, AcOꢀCH2-a),
4.81(1H, s, OꢀCH-O), 5.2ꢀ5.24 (1H, t, CH2ꢀCHꢀCH), 5.2 (1H, d,
OꢀCHꢀCH), 5.27ꢀ5.3 (1H, dd, OꢀCHꢀCHꢀCH). 13C NMR (100.6
MHz, CDCl3; ppm): δ 20.7, 20.8, 20.9 (COCH3), 30.4 (CH2ꢀBr), 62.4
(CHꢀCH3), 66.2 (CꢀCH2O); 67.4 (CꢀCH2O), 68.4 (CH3ꢀCHꢀ
CHꢀCH), 69.1 (CHꢀCH3), 69.6, 70.1, 70.6 (C-CH2O), 70.7 (CHꢀ
CHꢀO), 71.2 (CH3ꢀCHꢀCH), 97.7 (OꢀCHꢀO), 169.7, 169.9, 170.1,
170.7 (COCH3).
Synthesis of Bis(trimethylsilyl)[2-(2-{2-[(2,3,5-tri-O-acetylribofur-
anosyl)oxy]ethoxy}ethoxy)ethyl]phosphonate (RibAC). A methodol-
ogy similar to the synthesis of bis(trimethylsilyl)[2-(2-{2-[(2,3,4-tri-O-
acetylrhamnopyranosyl)oxy]ethoxy}ethylphosphonate] utilized 14.4 mmol
(4 equiv) of tris(trimethylsilylphosphite) and 3.6 mmol of 2-[2-(2-bro-
moethoxy)ethoxy]ethyl-2,3,5-tri-O-acetylribofuranoside to give the expected
compound in quantitative yield. For the hydrolyzed product SM (ESI > 0/
MeOH/30 eV): m/z 495.31 [M þ Na]þ. 1H NMR (400 MHz, CDCl3;
ppm): δ 2,22ꢀ2.29 (9H, m, OꢀCH3), 3.52ꢀ3.60 (10H, m, CH2CH2O),
5.42ꢀ5.45 (1H, dd, AcOꢀCH2ꢀCH), 4.18ꢀ4.22 (2H, m, AcOꢀC
H2ꢀCH), 4.85ꢀ5.20 (2H, m, CHꢀOAc), 5.54 (0.6H, s, OꢀCHRꢀO),
5.63 (0.4H, s, OꢀCHβꢀO). 13C NMR (100.6 MHz, CDCl3; ppm): δ
21.2ꢀ21.4ꢀ21.6 (H3CꢀCO), 38.3(-CH2PO3H2), 55.2 (CH2ꢀCH2ꢀ
PO3H2), 62.2 (CHꢀCH2), 69.5ꢀ70.3ꢀ70.8, 71 (CꢀCH2O), 73.2
(CH2ꢀOAc), 80.9 (CꢀCH2O), 84.5 (CHꢀCH2ꢀOAc), 108.0 (Oꢀ
CHꢀO), 107.4ꢀ107.6ꢀ107.8 (CdO). 31P NMR (81.0 MHz, CDCl3;
ppm): δ 9.21.
General Procedure for Sugar-Coated NPs. Oleic acid-/
oleylamine-coated NPs with sizes of ca. 4,14b 7,35 and 10 nm 14m and oleic
acid-coated NPs with sizes of ca. 16,36 18,36 and 35 nm35 NPs were
performed according to published procedures. The coating of the NPs with
different sugars was performed according to the procedure described below:
First Step. In a typical experiment, 0.46 mmol of oleic acid-/
oleylamine-coated iron oxide nanoparticles and 0.78 mmol of sugar
were dissolved in anhydrous THF (5 mL). The mixture was magnetically
stirred under argon to reflux for 24 h. Pentane (30 mL) was added to the
mixture at room temperature, and the solution was precipitated and
separated via centrifugation. The product was dissolved, and centrifuga-
tion (120 000 rpm, 10 min) was applied to remove undispersed residue.
The solution was precipitated with ethanol (40 mL), and organic
compound was removed via centrifugation.
For mannose-coated NPs (ManAC-4). IR: 2925, νas(CꢀH); 2856,
νs(CꢀH); 1747, ν(COO); 1226, ν(CO); 1080, ν(COC); 1051,
ν(CHcycle); 1015, ν(PꢀOꢀFe); 984, ν(PꢀOꢀFe); 588, ν(FeꢀO).
For ribose-coated NPs (RibAC-4). IR: 2925, νas(CꢀH); 2872,
νs(CꢀH); 1750, ν(COO); 1228, ν(CO); 1084, ν(COC); 1049,
ν(CHcycle); 1013, ν(PꢀOꢀFe); 976, ν(PꢀOꢀFe); 598, ν(FeꢀO).
For rhamnose-coated NPs (RhaAC-4). IR: 2925, νas(CꢀH); 2856,
νs(CꢀH); 1747, ν(COO); 1226, ν(CO); 1080, ν(COC); 1051,
ν(CHcycle); 1015, ν(PꢀOꢀFe); 984, ν(PꢀOꢀFe); 588, ν(FeꢀO).
Second Step. A 0.2 mmol amount of the NPs obtained in the first step
was dissolved in a solution of 2 M ammonia methanol (10 mL). The
solution was magnetically stirred under argon at room temperature for 4
h. Dichloromethane (30 mL) was then added to the mixture, and the
solution was precipitated and separated via centrifugation. The product
was dissolved, and centrifugation (120 000 rpm, 10 min) was applied
twice to remove organic compounds.
Synthesis of Bis(trimethylsilyl)[2-(2-{2-[(2,3,4,6-tetra-O- acetyl-
mannopyranosyl)oxy]ethoxy}ethoxy)ethyl]phosphonate (ManAC).
A methodology similar to the synthesis of bis(trimethylsilyl)[2-(2-{2-[-
(2,3,4-tri-O-acetylrhamnopyranosyl)oxy]ethoxy}ethylphosphonate uti-
lized 19.68 mmol (4 equiv) of tris(trimethylsilylphosphite) and 3.68
mmol of 2-[2-(2-bromoethoxy)ethoxy]ethyltri-O-acetylrhamnopyrano-
side to give the expected compound in quantitative yield. Rf (Et2O 100%) =
0.6. 1H NMR (400 MHz, CDCl3; ppm): δ 0.23 (18H, m, Me3Si,),
1.92ꢀ1.98, 2.04ꢀ2.10 (12H, 4s, CH3CO), 2.07ꢀ2.12 (2H, m, CH2P),
3.6ꢀ3.77 (9H, m, CH2CH2O), 3.98ꢀ4.06 (m, 3H, AcOꢀCH2-b, CHꢀ
CH2ꢀOAc, 1 CHꢀOꢀCH2), 4.2ꢀ4.25 (dd, 1H, AcOꢀCH2-a), 4.81(s,
1H, OꢀCHꢀO), 5.2ꢀ5.24 (t, 1H, CH2ꢀCHꢀCH), 5.2 (d, 1H,
OꢀCHꢀCH), 5.27ꢀ5.3 (dd, 1H, OꢀCHꢀCHꢀCH). 13C NMR
(100.6 MHz, CDCl3; ppm): δ 20.7, 20.8, 20.9 (COCH3), 62.4 (CHꢀ
CH3), 65.4 (CH2ꢀP), 65.8 (CH3ꢀCHꢀCH), 66.2 (CꢀCH2O), 67.4
(CꢀCH2O), 68.4 (CH3ꢀCHꢀCHꢀCH), 69.1 (CHꢀCH3), 69.6, 70.0,
70.1 (CꢀCH2O), 70.5 (CHꢀCHꢀO), 97.7 (OꢀCHꢀO), 169.7, 169.8,
170.0, 170.6 (COCH3). For hydrolyzed product SM (ESI > 0/MeOH/30
eV): m/z 567.45 [M þ Na]þ. 31P NMR (81.0 MHz, CDCl3; ppm): δ 22.4.
Synthesis of the Ribose Derivative (RibAC). Synthesis of 2-[2-(2-
Hydroxyethoxy)ethoxy]ethyl-2,3,5-tri-O-acetylribofuranoside. A
methodology similar to the synthesis of 2-[2-(2-hydroxyethoxy)-
ethoxy]ethyltri-O-acetylrhamnopyranoside utilzied 78.61 mmol of boron
trifluoride etherate, 15.72 mmol of acetylated ribose, and 47.16 of triethy-
lene glycol to give the expected compound in 51% yield. SM (ESI >
1
0/MeOH/30 eV): m/z 431.15 [M þ Na]þ. H NMR (400 MHz,
CDCl3; ppm): δ 2.18ꢀ2.21 (9H, t, CH3ꢀCO-), 3.54ꢀ3.67 (12H, m,
CH2CH2O), 4.09ꢀ4.13 (1H, dd, AcOꢀCH2ꢀCH), 4.18ꢀ4.22 (2H, m,
AcOꢀCH2ꢀCH), 4.26ꢀ4.31 (2H, m, CHꢀOAc), 5.05 (0.6H, s,
OꢀCHRꢀO), 5.22 (0.4H, s, OꢀCHβꢀO), 2.21ꢀ2.30 (9H, m,
OꢀCH3). 13C NMR (100.6 MHz, CDCl3; ppm): δ 20.5ꢀ21.0ꢀ21.3
(CH3ꢀCO-), 61.4 (CꢀCH2OH), 62.4 (CHꢀCH2), 67.1 (CꢀCH2O),
70.0 (CH2ꢀOAc), 70.1ꢀ70.4ꢀ70.8 (CꢀCH2O), 71.9 (CH2ꢀOH),
73.0 (CH2ꢀOAc), 82.8 (CHꢀCH2ꢀOAc), 102.0 (OꢀCHꢀO),
169.9ꢀ170.1ꢀ170.4 (CdO).
For Man-4. Anal. Found (% wt): Fe, 38.27; P, 5.12. IR: 3383 and
3199, ν(OH); 2921, νas(CꢀH); 2852, νs(CꢀH); 1243, ν(CO); 1095,
ν(COC); 1058, ν(CHcycle); 1015, ν(PꢀOꢀFe); 976, ν(PꢀOꢀFe);
590, ν(FeꢀO).
Synthesis of 2-[2-(2-Bromoethoxy)ethoxy]ethyl-2,3,5-tri-O-acet-
ylribofuranoside. A methodology similar to the synthesis of 2-[2-(2-
bromoethoxy)ethoxy]ethyltri-O-acetylrhamnopyranoside utilized 6.21
mmol of 2,3,5-tri-O-acetyl-R-ribofuranoside of 30,60,80-trioxanonanyle,
7.45 mmol (1.2 equiv) of tetrabromomethane, and 8.7 mmol (1.4 equiv)
For Rib-4. Anal. Found (% wt): Fe, 39.87; P, 4.99. IR: 3425 and 3183,
ν(OH); 2922, νas(CꢀH); 2856, νs(CꢀH); 1235, ν(CO); 1086, ν-
(COC); 1033, ν(CHcycle); 1001, ν(PꢀOꢀFe); 581, ν(FeꢀO).
10462
dx.doi.org/10.1021/ja111448t |J. Am. Chem. Soc. 2011, 133, 10459–10472

Journal of the American Chemical Society
ARTICLE
For Rha-coated. Anal. Found (% wt): (Rha-4) Fe, 41.87; P, 4.34; (Rha-7)
Fe, 48.71; P, 3.87; (Rha-10) Fe, 29.21; P, 5.01; (Rha-16) Fe, 50.11; P,
3.82; (Rha-18) Fe, 49.62; P, 3.72; (Rha-35) Fe, 61.23; P, 2.62. IR: 3377
and 3228, ν(OH); 2923, νas(CꢀH); 2868, νs(CꢀH); 1245, ν(CO);
1090, ν(COC); 1052, ν(CHcycle); 1014, ν(PꢀOꢀFe); 987, ν-
(PꢀOꢀFe); 584, ν(FeꢀO).
of 10 kHz e ν e 10 MHz, and (ii) a Stelar Spinmaster for ν > 10 MHz.
Standard radio frequency excitation pulse sequences CPMG- (T2) and
saturation-recovery (T1) were used. From the measured T1 and T2
values we have calculated the longitudinal and transverse relaxivities
using the usual formula: ri = [(1/Ti)meas ꢀ (1/Ti)dia ]/c (i = 1, 2), where
(1/Ti)meas is the value measured for the sample of magnetic center
concentration c (mmol Lꢀ1) and (1/Ti)dia refers to the nuclear relaxa-
tion rate of the diamagnetic host solution. GIBCO ultrapure water
DNase RNase Free Invitrogen was used for the preparation of two buffer
solutions. The first one is 50 mL of 10 mM Hepes and the second 50 mL
of 1 M Trima. These two solutions were mixed to obtain an aqueous
solution at pH = 7.4. The concentrations were 0.558, 0.218, 0.051, and
0.186 mmol Lꢀ1 of Fe, respectively, for samples Rha-4, Rha-7, Rha-10,
and Rha-18. The evaluation of heat generation was performed with an
in-house-built magnetothermal setup based on a Nova Star 5 kW
(Ameritherm Inc.) generator. A solution of coated magnetic NPs was
put in a thermal insulating support inside a coil and the temperature was
measured with a VR18_CR digital temperature recorder (Ceam Group)
connected to an optical fiber directly dipped into the sample. Although
the system is optimized for working in adiabatic conditions, temperature
dissipation with the surrounding air was observed. Ultrapure Milli-Q
water (Millipore Inc.), minimum resistivity of 18 MΩ, was used to
solubilize the NPs. The concentration was 21 g/L for Rha-4, 21 g/L for
Rha-7, 28 g/L for Rha-10, 6 g/L for Rha-16, and 21 g/L for Rha-35. The
measurements were performed on 0.3 g of these solutions.
RhamnoseꢀPhosphonate Leaching Studies for Rha-4.
Rhamnoseꢀphosphonate leaching studies from iron oxide NPs were
performed on Rha-4 in water for physiological conditions (pH = 7.4):
5.8 mg of Rha-4 was solubilized in 50 mL of water (0.649 mmol/L in
iron), and the solution was sonicated for 10 min. Then, the NPs were
precipitated by adding ethanol and washed thoroughly with the same
solvent. This procedure was repeated twice. The IR spectra of the
resulting NPs are unchanged, and the Fe/P ratio from elemental and
EDX analyses is shown to be unchanged with a value of 87:13, demon-
strating the strong anchorage of the organic rhamnoseꢀphosphonate
moieties at the NPs surface.
1
Physical Measurements. H, 13C, and 31P NMR spectra were
recorded using a Br€uker AC-400 spectrometer. Chemical shift values (δ)
are reported in parts per million from internal standard tretramethylsi-
lane. Coupling constants (J) are measured in hertz. Multiplicity is
reported as follows: s (singlet), d (doublet), t (triplet), m (multiplet),
and combination of this signal. Optical rotations were recorded with a
Perkin-Elmer 241 polarimeter. Fast atom bombardment mass spectra
(FAB-MS) were recorded in positive mode using a JEOL DX 300
spectrometer using NOBA (nitrobenzylic alcohol) as matrix. Column
chromatographies were performed on silica gel 60. Dynamic light
scattering (DLS) measurements were recorded with a Malvern high-
performance particle sizer. IR spectra were recorded using a Perkin-
Elmer 1600 spectrometer at 4 cmꢀ1 resolution. The surface composition
was monitored by X-ray photoelectron spectroscopy (XPS) on an
ESCALAB 250 (Thermo Electron). The X-ray excitation was provided
by a monochromatic Al KR (1486.6 eV) source. The diameter of the
analyzed surface was 400 μm. The background signal was removed using
the Shirley method. The surface atomic concentrations were determined
from photoelectron peak areas using the atomic sensitivity factors
reported by Scofield. Binding energies (BE) of all core levels were
referred to the CꢀC of C1s carbon at 284.8 eV. Elemental analyses were
performed by the Service Central d’Analyze (CNRS, Vernaison,
France). The samples were heated at 3000 ꢀC under He. Oxygen was
transformed in CO and detected by using an IR detector. Iron was
determined by using the emission spectroscopy technique PLASMA
ICP. An evaluation of the Fe/P ratio was also performed by using an
environmental secondary electron microscope FEI Quanta 200 FEG
coupled with an electron dispersive spectroscope Oxford INCA detec-
tor. Powder X-ray diffraction patterns were measured with a PanAna-
lytical diffractometer equipped with an ultrafast X’celerator detector
X’pert Pro with nickel-filtered copper radiation (1.5405 Å) and a Bruker
D8 Advance diffractometer equipped with Cu KR radiation and
operating in θꢀ2θ BraggꢀBrentano geometry at 40 kV and 40 mA.
Magnetic susceptibility data were collected with a Quantum Design
MPMS-XL SQUID magnetometer working in the temperature range of
1.8ꢀ350 K and the magnetic field range of 0ꢀ5 T. Data were not
corrected for the diamagnetic contribution because, when measured
separately, it was found to be negligible with respect to the sample
signals. All magnetic measurements were performed on dry powders
except when indicated. The samples were pressed into pellets in order to
prevent nanocrystal orientation under the external magnetic field.
Samples for transmission electron microscopy measurements were
prepared by depositing the NPs on carbon-coated copper grids. The
measurements were carried out with a microscope JEOL 1200 EXII
operated at 100 kV. NMR data were collected by using two different
pulsed FT-NMR spectrometers: (i) a Smartracer Stelar relaxometer
(with the use of fast-field-cycling technique) for frequencies in the range
’ RESULTS AND DISCUSSION
Synthesis and Characterization. The synthesis of iron oxide
NPs coated with carbohydrates was carried out using iron oxide
NPs, stabilized by oleic acid or a mixture of oleic acid/oleylamine,
by the displacement of the stabilizing agents, and by covalent
grafting of the carbohydrate derivatives via the phosphonate
function.
The synthesis of the rhamnose derivative functionalized at the
anomeric position, RhaAC, depicted in Scheme 1, was carried out
in three steps, leading mainly to the R-anomer (95/5). The
reaction is constrained to start at the anomeric position by
selectively deprotecting the sugar at this site.
The hydroxyl protection was obtained by the action of acetic
anhydride in pyridine. The selective deprotection of the anome-
ric position was achieved by use of hydrazinium acetate.37 The
glycosylation of the deprotected sugar, by the TEG moiety, was
obtained in an excellent yield using boron trifluoride as catalyst.
Primary alcohol was then halogenated by the couple triphenyl-
phosphine/carbon tetrachloride and an Arbuzov reaction with
the sorting trimethylsilyl phosphite38 led to the desired phos-
phonate. A similar methodology was applied for the synthesis of
the mannose ManAC and ribose RibAC derivatives functionalized
at the anomeric position, leading to mainly the R-anomer for the
mannose and the β-anomer for the ribose (see Scheme 2).
The postsynthetic anchorage of the phosphonate-functiona-
lized sugars at the surface of the iron oxide NPs was performed in
THF by heating for 24 to 48 h. The deprotection of the acetate
groups was performed in a methanol/ammonia mixture. The
synthesis of the NPs coated with the rhamnose derivative is
shown as an example in Scheme 3.
It is noteworthy that while the starting NPs are nonsoluble in
methanol and water, they become soluble in methanol but not water
when coated with acetate-protected sugar, and then become fully
soluble in water after sugar deprotection, as expected considering the
different organic groups covering the NPs surface at each step. By
following this procedure to synthesize the coated NPs, the sugar
10463
dx.doi.org/10.1021/ja111448t |J. Am. Chem. Soc. 2011, 133, 10459–10472

Journal of the American Chemical Society
ARTICLE
Scheme 3. Synthesis of the Rhamnose-Coated Iron Oxide NPs Rha-4ꢀRha35
moiety is proposed to be covalently bound to the iron oxide NPs via
the phosphonate moiety, which precludes sugar leaching and
consequently formation of uncoated iron oxide NPs in the living
organism. Such assessment was tested by intense sonication of an
aqueous solution of Rha-4 at physiological pH followed by pre-
cipitation and intense washing in order to remove all improperly
anchored organic molecules from the NPs surface. IR analysis
performed after such treatment still shows the presence of the
rhamnoseꢀphosphonate molecule; the elemental and the energy
dispersive spectroscopy (EDS) analysis showed an unchanged Fe/P
ratio confirming the absence of rhamnoseꢀphosphonate release
after such harsh treatment. Furthermore, the approach used ensures
the presence of rhamnose moieties pointing out from the iron oxide
NPs, which is an essential requirement for optimum recognition of
the appropriate lectin and for cell internalization.
Each step of the synthesis was followed by infrared spectros-
copy. The IR spectra of Rha-4, Man-4, and Rib-4 display, when
compared with the spectrum of the same NPs coated with oleic
acid and oleyl amine, the appearance of strong bands at ca. 1750
and 1226 cmꢀ1 attributed to the ν(COO) and ν(CO) stretching
modes together with the bands at ca. 1080 and at 1050 cmꢀ1
assigned to ν(COC) and ν(CHcycle), respectively (see the
Supporting Information available, Figure 1S). The band of the
ν(PꢀOꢀFe) mode at 1031 cmꢀ1 is consistent with the anchor-
ing of the phosphonate moiety to the surface (see the Supporting
Information).39 The characteristic broad envelop at 588 cmꢀ1 for
Man-4 and Rha-4 and at 598 cmꢀ1 for Rib-4, attributed to
ν(FeꢀO) bands of iron oxide, indicates that the iron oxide
remains intact after sugar anchorage. After deprotection of the
sugar acetate groups, the band at ca. 1750 cmꢀ1 (ν(COO))
completely disappears, and the intensity of the bands at ca.
1226 cmꢀ1 (ν(CO)) are strongly decreased (see the Supporting
Information). Consequently, this study highlights the covalent
coordination of the sugar derivative via the phosphonate group
on the iron oxide surface. Moreover, it confirms the efficient
deprotection of sugar acetate groups during treatment in a
solution of ammonia/methanol, leading to water-soluble NPs.
Note that the change of the NP size does not modify the IR
spectra. X-ray photoelectron spectroscopy (XPS) analysis was
further used to validate the successful coating of the sugar
molecules on the magnetite surface (see the Supporting Infor-
mation available, Figure 2S).40
Figure 1 (see also the Supporting Information available, Figure
3S, for Rib-4 and Man-4 derivatives). Monodisperse NPs of
controlled size in the range of 4ꢀ35 nm were obtained by varying
the synthetic parameters, and the coating process was demon-
strated not to modify the morphology, size, or size distribution of
the inorganic core (Supporting Information available, Figure 4S).
The average size and size distribution obtained from the analysis
of a large number of NPs are given in Table 1. A spherical shape is
obtained for the 4ꢀ18 nm NPs (Rha-4ꢀRha-18), while the
largest ones (Rha-35) present a cubic morphology. Indeed the
tendency to assume the more stable cubic habit on decreasing the
surface to volume ratio is a well-documented behavior for iron
oxide nanoparticles.41 In all cases the particles exhibit a narrow
size distribution. Dynamic light scattering (DLS) shows that the
NPs form colloidal solutions in water that are stable over long
time periods. The hydrodynamic size determined by this tech-
nique are on average ca. 4 nm larger than that obtained by TEM,
the difference being ascribed to the sugar layer surrounding the
magnetic core. Moreover, the NPs can be suspended in phos-
phate-buffered saline (PBS) and in physiological solution or in
sterile water even at high concentrations forming dispersion
stable for long periods.
The powder X-ray diffraction patterns within the 2θ range of
20ꢀ70ꢀ for Rha-4, Rha-7, Rha-10, and Rha-35 are given as
examples in Figure 2. Similar patterns were obtained for all the
investigated samples. The patterns exhibit diffraction peaks
which were indexed as (220), (311), (222), (400), (422),
(511), and (440) reflections, characteristic of the cubic spinel
structure of magnetite and maghemite, or any intermediate compo-
sition, which could not be discriminated due to significant line
broadening due to the small size of the crystalline domains. As
expected the line broadening increases upon decreasing the nano-
crystal size. The mean values of the cell parameters obtained by
using the Bragg law (see the Supporting Information available) vary
between 8.381 (Rha-4) and 8.411 Å (Rha-35), which is of the same
order as the literature value (8.396ꢀ8.3515 Å).42
The crystalline domain size, determined from the full width at
half-height of the most intense peaks, has been calculated using
the DebyeꢀScherrer formula. The values obtained (Table 1) are
very similar to those obtained by TEM, which is indicative of
crystallographic single domains and high crystallinity of the
samples.
Magnetic Properties. The magnetic behavior of the NPs was
first investigated by zero field cooled/field cooled experiments,
i.e., by measuring the temperature dependence of the magnetiza-
tion under a 5 mT field after cooling the sample in zero field
(ZFC) or with the probe field applied (FC). The magnetic
The transmission electronic microscopy (TEM) measure-
ments performed on the as-obtained NPs show in all cases that
they are nonaggregated and uniform. The TEM images of the
NPs of Rha-4ꢀRha-35 samples are shown as an example in
10464
dx.doi.org/10.1021/ja111448t |J. Am. Chem. Soc. 2011, 133, 10459–10472

Journal of the American Chemical Society
ARTICLE
Figure 1. TEM images for NPs of various sizes coated with rhamnose (Rha-4ꢀRha-35).
Table 1. Chemical Composition and Average Size of the Core (d) and (dH) Hydrodynamic Diametersa
sample
wt % Fe
wt % iron oxide
wt % P
d b (nm)
d c (nm)
dH d (nm)
dNMR e (nm)
Rha-4
41.9
36.4
29.3
55.6
44.4
55.2
57.9
50.3
40.5
76.8
61.4
76.3
4.3
5.0
8.5
2.3
3.9
2.6
4.1 (0.6)
6.7 (0.8)
10.0 (3.3)
16.2 (1.0)
18.2 (1.1)
35.2 (3.0)
4.5
6.7
8.2 (1.3)
10.4 (1.5)
16.7 (2.6)
21.3 (2.2)
23.6 (2.3)
43.5 (3.3)
4.6 (0.02)
10.16 (0.04)
10.30 (0.06)
Rha-7
Rha-10
Rha-16
Rha-18
Rha-35
10.9
16.0
18.5
35.4
18.54 (0.03)
a Average diameters were obtained from TEM, XRD, and DLS analyses for rhamnose-coated iron oxide NPs of different sizes (samples Rha-4ꢀRha-35).
Values in parentheses denote the standard deviations. b TEM analysis. c XRD analysis. d DLS analysis. e d = 2r are obtained by fitting the 1H-NMR data.
behavior also at room temperature (see the Supporting Informa-
tion available, Figure 6S). The Tmax values are plotted in
Figure 3b as a function of the average diameter of the NPs and
are summarized in Table 2. Tmax increases continuously with the
NPs size but with a d dependence slower (ca. d1.5) than the d3 law
predicted by the Nꢀeel model for the relaxation of not interacting NPs.
The prime cause of behavior is the increasing contribution of the
surface to the total effective magnetic anisotropy as d decreases
(Figure 4). A similar complex d dependence has been observed in
other systems described in the literature: for instance, for noninter-
acting Fe3O4 nanoparticles in the size range of 2.5ꢀ14 nm a d1.7 law
Figure 2. X-ray diffraction patterns within the 2θ range of 20ꢀ70ꢀ for
Rha-4, Rha-7, Rha-10, and Rha-35.
has been observed,43a whereas interacting Fe3O4 nanoparticles in the
range of 7.8ꢀ17.9 nm obtained from FeOOH show a quasi linear
TB vs d trend.43b
properties of the 4 nm NPs coated with different ligands are
essentially similar (see the Supporting Information available,
Figure 5S), indicating the nature of the sugar does not signifi-
cantly influence the physical behavior of the inorganic core. Note
also that the FC/ZFC curves of the nanoparticles before and after
sugar coating are identical. For this reason, in the following
discussion we will focus on the investigation of the Rha-4ꢀRha-
35 series only. Figure 3a shows the ZFC/FC magnetization from
2 to 300 K for Rha-4 and Rha-16 as representative of the whole
series. The ZFC magnetization curves increase with temperature,
reaching a maximum at temperature defined as Tmax. On the
contrary, the ZFC curve of sample Rha-35 (35 nm NPs) in-
creases continuously as the temperature increases and does not
show a maximum, suggesting the presence of a ferrimagnetic-like
The FC curves of Rha-4, Rha-7, and Rha-10 increase as the
temperature decreases and never reach saturation at low tempera-
ture, suggesting that interparticle interactions do not significantly
affect the relaxation dynamics. On the contrary, the FC magnetiza-
tion of Rha-16 and Rha-18 decreases below Tmax as the temperature
decreases, which can be interpreted as the fingerprint of strong
magnetostatic interparticle interactions.44 Accordingly, these fea-
tures only appear for larger NPs, where the strength of the magnetic
interactions is expected to be much greater.
The magnetization as a function of the applied field was
measured at room temperature and 2.5 K. The hysteresis loops
measured at 300 K display nonzero coercivity (10 mT) only for
sample Rha-35, confirming that it is largely blocked at room
temperature. All of the other samples are in the super-paramagnetic
10465
dx.doi.org/10.1021/ja111448t |J. Am. Chem. Soc. 2011, 133, 10459–10472

Journal of the American Chemical Society
ARTICLE
Figure 3. (a) ZFC/FC curves for Rha-4 (black dots) and Rha-16 (empty dots) with an applied field of 5 mT; (b) size dependence of Tmax for samples
Rha-4ꢀRha-18. The continuous line represents the best fit to a power law.
Table 2. Magnetic Characteristics of Samples Rha-4ꢀRha-35
Nꢀeel law
τ0 (s)
VogelꢀFulcher law
τ0 (s)
sample
Tmax(K)
MS(2.5K) (emu/g)
MS (300K) (emu/g)
HC(2K) (mT)
Ea/kB (K)
Ea/kB (K)
T0 (K)
Rha-4
19
71
96
94
83
82
80
83
77
75
70
68
65
76
30
33
337
1525
1.13 ꢁ 10ꢀ12
1.15 ꢁ 10ꢀ13
1.28 ꢁ 10ꢀ12
1.79 ꢁ 10ꢀ21
3.71 ꢁ 10ꢀ24
216
959
9.28 ꢁ 10ꢀ11
2.24 ꢁ 10ꢀ11
6.66 ꢁ 10ꢀ11
1.33 ꢁ 10ꢀ11
1.16 ꢁ 10ꢀ11
3
12
Rha-7
Rha-10
Rha-16
Rha-18
Rha-35a
112
200
241
>300
38
2702
2441
2020
2012
7
71
9471
120
150
100
54.5
12260
a AC susceptibility data of Rha-35 were not measured because the sample is blocked at 300 K in the investigated frequency range.
multidomain, as is the case of Rha-35, indicates that flower and
vortex magnetic states can further decrease HC.48
The temperature dependence of the in-phase (χ0) and out-of-
phase (χ00) components of the AC susceptibility were measured
for samples Rha-4ꢀRha-18. Figure 5a shows the thermal
dependence of χ0 and χ00 in a zero static field in the 1ꢀ1000
Hz range for Rha-16, here taken as a representative example. At 1
Hz, χ0 and χ00 exhibit peaks at 265 and 247 K, respectively, which
shift toward higher temperature with increasing frequency. A
similar frequency-dependent behavior was observed for all of the
NPs. Depending on the nature and strength of the magnetic
interactions, the observed frequency-dependent behavior may be
attributed to a blockingꢀunblocking process for isolated or
weakly interacting super-paramagnetic NPs, or to a spin glasslike
transition for the case of strongly interacting NPs, as is discussed
as follows.49 According to the Nꢀeel model, the temperature
dependence of the relaxation of the magnetization of noninter-
acting super-paramagnetic systems follows an Arrhenius law,
τ = τ0 exp(Ea/kBT), where Ea is the average energy barrier for
magnetization reversal, τ0 is the attempt time, and kB is the
Boltzmann constant.50 The best fits of the blocking temperatures
obtained from the χ00 maxima for different observation times
τ = 1/2πν to the Arrhenius law give the Ea and τ0 values listed in
Table 2. Figure 5b shows the Arrhenius law fit for Rha-16, as an
example. For Rha-4 and Rha-10 only, the obtained τ0values are
in the 10ꢀ8 to 10ꢀ12 s range typical for isolated nanoparticles in
the super-paramagnetic regime.51 In the other cases, τ0 is outside
this range, indicating the model is not appropriate to describe the
dynamics of the systems.49 Nevertheless, such small τ0values are
usually interpreted as the signature of magnetic moment correla-
tions caused by significant dipoleꢀdipole interparticle interac-
tions.52 For this reason, the temperature dependence of τ was
Figure 4. Effective magnetic anisotropy values obtained from Ea (red
circles) and TB (blue stars).
state. The MS values obtained by extrapolation of the high-field
data45 are reported in Table 2.
Conversely, at low temperature all samples display open
hysteretic loops. The coercive field, HC, reported in Table 2
and shown in Figure 7S (Supporting Information), increases
with the average size up to Rha-18 and then markedly decreases
for Rha-35. The observed evolution of HC with the average size
agrees well with the empirical model suggested by Hergt et al.,46
which summarizes the approaches and experimental data reported in
the literature over the past few decades:47 in the pure single-domain
regime HC increases with the size due to the progressively lower
efficacy of the thermal demagnetization process; further increase in
the size above the super-paramagnetic threshold leads to incoherent
reversal mechanisms becoming increasingly favored lowering the
field for reversal of the magnetization. Moreover, micromagnetic
calculation for cubic NPs in the transition states from single to
10466
dx.doi.org/10.1021/ja111448t |J. Am. Chem. Soc. 2011, 133, 10459–10472

Journal of the American Chemical Society
ARTICLE
Figure 5. (a) AC susceptibility of sample Rha-16 performed with various logarithmic spaced frequencies in the 0.1ꢀ1000 Hz range. (b) Linear plot of
the inverse of maximum temperature of each frequency according to the Arrhenius law (full dots) and according to the VogelꢀFulcher law (empty dots).
also fitted to the VogelꢀFulcher law, τ = τ0 exp(Ea/kB(T ꢀ T0)),
in which an additional parameter T0 is introduced to take into
account the strength of the interparticle interaction. The para-
meters so obtained are reported in Table 2. As expected for
systems where nonnegligible dipolar interactions occur, the
VogelꢀFulcher model produces τ0 values within the proper
range and with much lower energy barriers. Moreover, the strong
increase of T0 values indicates that interparticle interactions affect
only slightly the magnetization dynamics for Rha-7, while, for larger
particles (Rha-16, Rha-18), they cannot be neglected. This con-
clusion is in agreement with the behavior observed in the ZFC/FC
experiments. The consistency of the results obtained from AC and
Figure 6. Size dependence of the SAR values measured on water
DC techniques is also exemplified by the agreement between the
solution with a magnetic field of 168 kHz frequency and 21 kA/m for
effective anisotropy constant, K, evaluated from the energy barriers,
NPs Rha-4ꢀRha-35.
Ea = KV, extracted by the Arrhenius plot and from TB, evaluated as
the ZFC curve maxima, assuming K = 25kBTB/V. The decreasing
values of K upon increasing size, which ranges from 12.9 ꢁ 104 J/m3
that this value is close to the tolerance threshold limit which, after
the experiments by Brezovich et al. on human volunteers, is
for Rha-4 to 5.4 ꢁ 104 J/m3 for Rha-18 (values from AC measure-
assumed to be 4.85 ꢁ 108 A mꢀ1
s .
ꢀ1 54 However, Hergt et al.
estimated that for a small exposed region H0ν values up to 5 ꢁ
ments) is consistent with the increasing importance of the surface
3
3
contribution as the particle size is reduced, as mentioned above. The
observed K trend can also justify the anomalous TB vs d dependence
(Figure 3b).
109 A mꢀ1 ꢀ1 could be safely applied,46 and the Jordan group
s
3
3
at Berlin’s Charitꢀe Hospital has exposed human patients to fields
of 3.8ꢀ13.5 kA/m amplitude and 100 kHz frequency with only
minor side effects reported.55 Larger amplitudes or frequencies
can produce tissue heating due to induced eddy currents.
The obtained SAR values are plotted in Figure 6 as a function
of particle size.
Magnetic Hyperthermia Properties. Heat dissipation of
ferrofluid produced by the delay in the relaxation of the magnetic
moment may be observed when the NPs are exposed to an
alternating magnetic field of proper frequency. To experimen-
tally measure this effect, the sample is placed in an alternating
magnetic field and the temperature increase is recorded as a
function of the application time. The hyperthermal efficiency of a
material is normally evaluated by the so-called specific absorption
rate, SAR, defined as
Clearly, particles with sizes up to 7 nm do not produce
significant heating for the experimental conditions adopted.
When the size of the NPs exceeds this limit a sudden increase
of the SAR value is observed, reaching 76 W/g for Rha-35. The
SAR can be further increased by increasing the frequency. For
example, upon doubling the frequency, the SAR of Rha-16
changes from 61 to 185 W/g. This value corresponds to an
important heating efficiency and, although obtained under con-
ditions slightly above the tolerance limit, makes the material
suitable for application as heat mediators for magnetic fluid
hyperthermia. Indeed, it is usually accepted that a heat deposition
!
1
dT
dt
SAR ¼
cimi
ð1Þ
∑
me
i
where me is the total mass of the iron, ci is the specific heat of the
different species in solution, mi the weight, and dT/dt the slope of
the T(t) curve. Because the experimental setup we used is not
perfectly adiabatic (see Experimental Section), the dT/dt value
was extrapolated by taking the initial slope of the temperature
increase obtained from the linear term of a polynomial fit of the
whole curve. The ci values were taken from the literature,53 while
the mi values were obtained from elemental analysis. In all cases, a
rate of 100 mW cmꢀ3 in tissue is enough to have a sizable
3
effect,2b which could be reached for a Rha-16 concentration
between 0.5 and 1.6 g/cm3, depending on the frequency.
It should be also noted that, although the observed SAR lies in
the medium-high range of data reported in the literature to date for
maghemite/magnetite MNP,2f,g a direct comparison is difficult due
to the lack of a clear definition of a protocol for SAR evaluation, which
actually requires the control of many environmental parameters. In
ca. 0.3 g of 10 g Lꢀ1 solution is placed in an AC magnetic field of
3
frequency ν = 168 kHz and amplitude H0 = 21 kA/m, correspond-
ing to a product H0ν = 3.5 ꢁ 109 A mꢀ1
s
ꢀ1. It should be noted
3
3
10467
dx.doi.org/10.1021/ja111448t |J. Am. Chem. Soc. 2011, 133, 10459–10472

Journal of the American Chemical Society
ARTICLE
fact, for a given NP the SAR should be proportional to the frequency
of the AC magnetic field and to the square of its intensity. Even
though scaling procedures have been proposed, as for example
normalization with respect to the tolerance limit56 or to H0 ν,57
2
the lack of accuracy in the calorimetric measurements (dissipation,
initial temperature, etc.) can produce significant differences even for
samples of very similar composition, morphology, and size. In
particular, when the experimental conditions are not perfectly
adiabatic, the SAR is always underestimated since its value depends
on the thermal dissipation rate, which in turn strongly depends on the
environmental temperature.
To rationalize these results, we compare them with the
theoretical values estimated using the Rosensweig model58 for
super-paramagnetic NPs:
Figure 7. Nꢀeel, (O), Brown, (b), and total (ꢁ) relaxation times
estimated at room temperature in water assuming a viscosity η = 1.003 ꢁ
10ꢀ3 kg mꢀ1
s
ꢀ1 for samples Rha-4ꢀRha-18. For sample Rha-35, only τB
3
3
is included because its blocking temperature is greater than room tempera-
ture, hampering τN evaluation. The solid line represents the time corre-
sponding to the working frequency τ = 1/(2πν).
2
P
μ0χ0H0
2F
2πντ
SAR ¼ ¼ 2πν
2
F
1 þ ð2πντÞ
2
Theoretical SAR values where then estimated using eq 2,
where χ0 is obtained from the room-temperature MS values and τ
from AC powder data. The calculated trend follows that experimen-
tally observed confirming that, for our system in the single-domain
regime, the main mechanism for heat release is the relaxation losses
due to Nꢀeel reorientation. This result can be explained by consider-
ing that, on one side, the increase in size induces an increase of the
NP magnetic moment and, on the other, it causes the relaxation time
to approach the working frequency (see Figure 7). The numerical
simulation by Purushotham and Ramanujan,59 who determined that
the size corresponding to the SAR maxima for magnetite NPs is
between 19 and 22 nm for a frequency of 100ꢀ200 kHz, agrees well
with this conclusion.
However, a significant discrepancy between the theoretically
and experimentally obtained numerical values is observed. As
already pointed out by Rosensweig,58 the size distribution, which
has been neglected to a first approximation, must be taken into
account for a more realistic estimate of the SAR value. To assess the
importance of the size distribution effect, the SAR was evaluated by
replacing τ ≈ τN(V) (τN , τB) in eq 2 and integrating it over V,
whose distribution is considered log-normal with average and
standard deviation values reported in column 5 of Table 1. All the
other parameters, such as K and MS, for which a weaker dependence
on V is expected, are kept constant. The values obtained, reported in
brackets in column 6 of Table 3, are much more reasonable and
comparable to those obtained experimentally. As expected,2g,58 the
difference between the two theoretical SAR values highlights the role
of the size distribution, which tends to smooth the SAR values,
decreasing those which correspond to relaxation times close to the
working frequency, Rha-16, and increasing those far away, Rha-10.
Finally, sample Rha-35 shows the largest SAR. Since its
blocking temperature is well above 300 K, Brownian relaxation
is longer than the working frequency and the field amplitude used
is larger than the room-temperature coercivity, it can be argued
that the mechanism responsible for heat generation is different,
as in this case hysteresis losses are predominant. Indeed, even if
on a different time scale, a room-temperature minor hysteresis
loop between (25 mT on dry powder shows HC = 6.5 mT and a
maximum M = 0.3MS (see the Supporting Information available,
Figure 8S). Therefore, this result suggests that, at least for the
field amplitude and frequency used, hysteresis losses are more
efficient than super-paramagnetic losses. This behavior is the
opposite of that commonly found for iron oxide NPs, where
power losses are reported to be largest for sizes between 14 and
Vðμ0MSH0Þ
2πντ
¼ πν
ð2Þ
2
FkBT
1 þ ð2πντÞ
where P is the mean volumetric power dissipation, μ0 is the
vacuum magnetic permeability, F is the mass density, τ is the
relaxation time, and χ0 is the chord susceptibility; χ0 is defined as
2
μ0MS V/(kBT)L(ξ)/ξ, V being the average particle volume, kB
the Boltzmann constant, and L(ξ) the Langevin function, where
ξ = μ0MSVH/(kBT).
Two different mechanisms compete to determine the relaxa-
tion of the magnetization: the Nꢀeel relaxation, corresponding to
the magnetic moment reversal over the energy barrier and
characterized by τN = τ0 exp(KV/kBT), and the Brown relaxation
corresponding to the mechanical rotation of the whole particle
and described by τB = 3ηVH/kBT (η is the viscosity of the media
and VH is the hydrodynamic volume). The total relaxation time is
determined by the fastest process according to 1/τ = 1/τN þ 1/τB.
An evaluation of the SAR for the same conditions used for the
calorimetric measurements can then be attempted by use of the
hydrodynamic diameters obtained by DLS, namely, τB and Ea,
and τ0 obtained from AC measurements using the Nꢀeel model
for τN. The size dependence of Nꢀeel, Brown, and total relaxation
times for Rha-4ꢀRha-18 at room temperature is shown in
Figure 7. τN could not be estimated for Rha-35 since it has a
TB greater than room temperature.
Comparison of the relaxation times τN and τB shows that the
latter are always longer, so that mechanical rotation is never fast
enough to compete with the reversal of the magnetization
through the overcoming of the energy barrier. It should be noted
that τN was evaluated from AC data collected on powder samples
while SAR measurements were carried out on water solutions
where lower interactions are expected, even though the concen-
tration was high (around 10 mg mLꢀ1). This is not important
3
for Rha-4ꢀRha-10, where interparticle interactions are negligi-
ble, but it can affect the relaxation behavior for larger NPs. To
gain insight into this point, AC susceptibility measurements on
Rha-16 and Rha-7 solutions of the same concentration used for
magnetothermal experiments were also performed. A fit to an
Arrhenius law of the ln τ vs 1/Tmax plot gave energy barriers and
τ0 values close to those obtained with the respective powders
(see the Supporting Information available, Table 1S). This
suggests that at the high concentrations used for SAR measure-
ments particle aggregation occurs, which can allow for remark-
able interactions still being operative.
10468
dx.doi.org/10.1021/ja111448t |J. Am. Chem. Soc. 2011, 133, 10459–10472

Journal of the American Chemical Society
ARTICLE
Table 3. Relaxation Times and Theoretical and Experimental SAR Valuesa
sample
τN
τB
τ
Hν ꢁ 109 A mꢀ1 sꢀ1
theoretical SAR W/g
experimental SAR W/g of Fe
3
3
Rha-4
3.47 ꢁ 10ꢀ12
1.86 ꢁ 10ꢀ11
1.04 ꢁ 10ꢀ08
9.19 ꢁ 10ꢀ08
2.09 ꢁ 10ꢀ07
4.27 ꢁ 10ꢀ07
1.77 ꢁ 10ꢀ06
3.67 ꢁ 10ꢀ06
3.47 ꢁ 10ꢀ12
1.86 ꢁ 10ꢀ11
1.03 ꢁ 10ꢀ08
8.97 ꢁ 10ꢀ08
3.5
0 (0)
0 (0)
0
3
Rha-7
3.5
3.5
3.5
7.0
3.5
Rha-10
Rha-16
6 (37)
32
61
185
76
100 (90)
390 (187)
Rha-35
3.13 ꢁ 10ꢀ05
e3.13 ꢁ 10ꢀ05
a Columns 2ꢀ4 report the Nꢀeel, Brown, and total relaxation times estimated as described in the main text. In the fifth column the Hν product used to
measure the SAR is reported, while the two last columns give the experimental and calculated SAR values. Experimental SAR values were obtained using
eq 1 with cNPs = 0.67 J/(g K), cwater = 4.18 J/(g K), and csugar = 1.2 J/(g K); mi values were deduced from the relative weight percentage of particle/
3
3
3
S
sugar found by elemental analysis. Theoretical values were evaluated by eq 2 using the experimentally determined values for M and V, and the relaxation
times reported in column 4. The values in brackets refer to SAR values determined by taking into account the size distributions as mentioned in the text.
Figure 8. (a) Longitudinal, r1, and (b) transverse, r2, relaxivities presented through their NMR-D profile for Rha4 (0), Rha-7 (4), Rha-10 (ꢁ) and
Rha-18 (O). For comparison the corresponding data for the commercial contrast agent Endorem (b) are also shown. The arrows indicate the position
of the maximum in r1 for Rha4, Rha7, and Rha10. Rha-18 displays no maximum due to very high anisotropy.
60
20 nm
and then decrease for larger sizes. Nevertheless, a
Interestingly, the r2 value increases upon increasing the core
diameter, thus suggesting that the particles are in the so-called
“motional averaging regime” (MAR), where water diffusion
occurs on a much faster time scale than the resonance frequency
shift.63 With regard to sample Rha-35, we observed when the
sample was put inside the NMR electromagnet and irradiated
with RF pulses to measure T1 and T2 values, precipitation
occurred after 5ꢀ10 min, probably because of the presence of
a static magnetic field. Consequently, to ensure that the solution
was homogeneous such as in the case of the other samples,
sonication before each measurement at different frequency was
required. The r2 values obtained are comparable to those of Rha-4.
It should be remarked that the precipitation problems ob-
served for Rha-35 exclude from use in a MRI spectrometer at
the clinical level, because of clear risks of thrombosis. Moreover,
we should stress that in the hyperthermia measurements no such
precipitation was observed, due to the absence of a static
magnetic field.
The frequency r1(ν) behavior of the investigated samples
gives, as usual for super-paramagnetic contrast agents, informa-
tion on the physical mechanisms that are the main sources of the
nuclear relaxation times shortening. First we note that the data
for both r1 and r2 are consistent with those typical of super-
paramagnetic nanostructures,64ꢀ67 where the dimensions of the
magnetic core crucially influence the mechanism of relaxation.
Second (and as a further confirmation), the behavior of r1(ν) is
qualitatively the same as the Endorem sample, being represented
approximately by the sum of two contributions dominating,
respectively, at low (e1 ÷ 5 MHz) and high (g1 ÷ 10 MHz)
frequencies.65 At low frequencies, the mechanism driving the
nuclear relaxation depends directly on the magnetic anisotropy.
remarkable exception to this trend was also noted by Hergt et al., who
reported an impressive SAR (960 W/g) at 410 kHz and 10 kA/m for
monodisperse 38 nm iron oxide bacterial magnetosomes with open
hysteresis loops at room temperature.61 Further investigation of
these systems, including an assessment of the SAR dependence on
frequency and amplitude and measurements in solvents of different
viscosity with a more sensitive experimental apparatus, is required in
order to gain a better understanding of this behavior.
In summary, the 16 and 35 nm sugar-coated NPs are the best
candidates for hyperthermic applications; however, to be suitable
for theranostic purposes a satisfactory relaxometric efficacy is also
essential.
1
Relaxometric Properties. The H NMR relaxometry char-
acterization (i.e., NMR dispersion profile) of the samples has
been performed at room temperature by measuring the long-
itudinal and the transverse nuclear relaxation times T1 and T2 in
the frequency range of 10 kHz e ν e 60 MHz, corresponding to an
external magnetic field H= 0.00023ꢀ1.5 T.62 The range was chosen
in order to cover the typical fields for MRI tomographs, used both in
clinics (H = 0.2, 0.5, and 1.5 T) and research laboratories. It should
be noted that the measurements at room and physiological tem-
peratures gave the same results to within 10%.
The relaxivities r1 and r2 as function of frequency are reported
in Figure 8a,b. For comparison, the values obtained for a commercial
super-paramagnetic contrast agent produced by Guerbet Group, i.e.,
Endorem, have also been plotted. The size dependence of r2 and of
the ratio r2/r1 at 1.5 T is given in Figure 9. It is evident that Rha-7,
Rha-10, and Rha-18 have transverse r2 relaxivities that are higher
than Endorem by a factor as large as 3, over the whole investigated
frequency range.
10469
dx.doi.org/10.1021/ja111448t |J. Am. Chem. Soc. 2011, 133, 10459–10472

Journal of the American Chemical Society
ARTICLE
Rha-7, Rha-10, and Rha-18, which have a magnetic core greater
than 6 nm, display r2 values greater than the first-generation
contrast agent Endorem at fields (frequencies) compatible with
clinical use.
Thus, the developed NPs present properties which are inter-
esting for both magnetic resonance imaging and magnetic
hyperthermia and have optimum characteristics for inorganic
core sizes of ca. 16ꢀ18 nm. They are further functionalized with
carbohydrates identified as recognition vectors for certain lectins.
Therefore, these nanobjects can be regarded as third generation
multifunctional sensors for theranostics. Biological studies in
vitro and in vivo are currently under way to confirm the potential
applications of these nanoprobes. In particular, preliminary
cytotoxicity tests by MTT assays has revealed that the MNPs
are not toxic, while the first assay on in vitro labeling of melanoma
and breast cancer cell lines for Rha-16, carried out by the group
of Prof. P. Marzola at the University of Verona, Italy, has
provided evidence regarding the marked selectivity of our
samples. These experiments also proved that the sugar-coating
does not significantly increase nonspecific cellular uptake of the
particles. All these data will be the subject of a future publication.
Figure 9. Size dependence of r2 (b) and the ratio r2/r1 ()) determined
at 1.5 T (the most common field applied for medical MRI). The lines
represent the r2 (black line) and r2/r1 (dash line) values obtained for
Endorem.
At high frequencies, for the smallest particles the nuclear relaxa-
tion is dominated by the Curie-relaxation function that takes into
account the diffusion of water molecules (with diffusion correla-
tion time τD = r2/D, where r is the distance of closest approach
and D the diffusion coefficient of water molecules) in the
presence of magnetic centers. On the other hand, the maximum
disappears for the compound with the largest diameter.
The experimental data on our samples confirm (as for a few
other experimental cases, e.g., refs 66 and 65) qualitatively the
predictions of the theory presented by Roch and co-workers.65
We are currently using this theory to predict the disappearance of
the low-frequency dispersion when larger diameters (i.e., higher
anisotropy) are considered, to explain the disappearance of the
high-frequency maximum in r1(ν) for the largest diameter
sample, and to estimate the distance r of closest approach to
the magnetic core. A first estimate of the NMR diameter obtained
in this way, given in Table 1, suggests that the coating of our
samples is fully penetrated by the diffusing water molecules. This
work is currently in progress and will be published in the near
future.
’ ASSOCIATED CONTENT
S
Supporting Information. Figures showing: FTIR spectra
b
of Man-4 NP coated by oleic acid and with mannose derivative
before and after acetate group deprotection; X-ray photoelectron
spectra of sugar-coated magnetite Rha-16; TEM images of Man-
4 and Rib-4; TEM micrographs of Rha-4 before and after the
removal of the acetate group; ZFC/FC magnetizations of NPs
coated with mannose, rhamnose, and ribose derivatives; ZFC/
FC of Rha-35; coercive field vs NPs diameter for samples Rha-4-
Rhaꢀ35; minor hysteresis loop recorded at room temperature
on a powder of Rha-35 and a table listing magnetic properties of
Rh-7 and Rh-16 measured as dry powder and as solution with
the same concentration used for hyperthermia experiment.
This material is available free of charge via the Internet at
http://pubs.acs.org.
’ CONCLUSION
In this paper we describe the synthesis of water-dispersible
magnetite NPs of different sizes and coated with a crown of
sugars. To achieve this goal, we have developed a novel method
on the basis of the displacement of oleylamine/oleic acid ligands,
used as stabilizers during the synthesis of these NPs, by formation
of FeꢀOꢀP bonds between the sugar and the NP via a
phosphonate function. A chemical modification of the ligand
surface is then performed in an alcoholic medium to remove the
acetate protecting groups of the carbohydrates. This methodol-
ogy is general and has been applied to several sugars. It ensures
both an efficient anchoring of the sugar to the NP surface and
that the sugar points outward from the NP optimizing the cell
recognition capability.
’ AUTHOR INFORMATION
Corresponding Author
yannick.guari@univ-montp2.fr; veronique.montero@univ-montp2.fr;
claudio.sangregorio@unifi.it
’ ACKNOWLEDGMENT
We thank Dr. G. Baldi and Dr. C. Ravagli from Ce.Ri.Col—
Colorobbia Italia, C. Rebeil (PAC ICGM, University of Mon-
tpellier II, France) for assistance with the calorimetric and
magnetic measurements. We thank V. Flaud for XPS measure-
ments and spectrum analysis, the group of M. Borras
(CERETOX and Head of the Experimental Toxicology and
the ecotoxicology platform of the Barcelona Science Park) for
preliminary toxicity measurements, and the group of P. Marzola
(University of Verona) for cell-targeting experiments. We are
grateful to B. D. Howes for critical reading and improvement of
the English text. C.I., T.K., P.A., A.L., D.G., and C.S. gratefully
acknowledge the financial support of the EC through NAN-
OTHER (Grant FP7-NMP4-LA-2008-213631). J.L., L.L., Y.G,
and C.G. thank the University of Montpellier II and CNRS for
We then conducted a detailed study of the magnetic properties
of these NPs, particularly in relation to their relaxivity and
magnetic hyperthermia, in order to assess the effect of magnetic
core size on the magnetic properties.
The optimum value of SAR obtained for NPs governed by
Nꢀeel relaxation is 61 W gꢀ1 for a magnetic core size of 16 nm,
3
while it is 76 W gꢀ1 for partially blocked 35 nm NPs. The
3
magnetic field applied in these cases is Hν = 3.5 ꢁ 109 A mꢀ1 sꢀ1
,
3
3
which is within the acceptable range limitations imposed by a
potential use in the human body. In terms of relaxivity, the NPs
10470
dx.doi.org/10.1021/ja111448t |J. Am. Chem. Soc. 2011, 133, 10459–10472

Journal of the American Chemical Society
ARTICLE
financial support. L.L. thanks the UFI (Grant GF/IR/732/07,
No. 25) for financial support.
307, 538. (c) Couvreur, P.; Vauthier, C. Pharm. Res. 2006, 23, 1417. (d)
Weissleder, R. R. P. Eur. Radiol. 1993, 3, 198.
(11) Rinck, P. A., Ed. Magnetic Resonance in Medicine, 3rd ed.;
Blackwell: Oxford, U.K., 1993.
’ REFERENCES
(12) (a) Geraldes, C. F. G. C.; Laurent, S Contrast Media Mol.
Imaging 2009, 4, 1. (b) Corot, C.; Robert, Ph.; Idꢀee, J.-M.; Port, M. Adv.
Drug Delivery Rev. 2006, 58, 1471. (c) Weissleder, R.; Stardk, D. D.;
Engelastad, B. L.; Bacon, B. R.; Compton, C. C.; White, D. L.; Jacobs, P.;
Lewis, J. Am. J. Roentgenol. 1989, 152, 167. (d) Reimer, P.; Rummeny,
E. J.; Daddrup, H. E.; Balzer, T.; Tombach, B.; Berns, T.; Peters, P. E.
Radiology 1995, 195, 489. (e) Jung, C. W.; Jacobs, P. Magn. Reson.
Imaging 1995, 13, 661. (f) Josephson, L.; Tung, C.-H.; Moore, A.;
Weissleder, R. Bioconjugate Chem. 1999, 10, 186. (g) Weissleder, R.; Lee,
A. S.; Fischman, A. J.; Reimer, P.; Shen, T.; Wilkinson, R.; Callahan, R. J.;
Brady, T. J. Radiology 1991, 181, 245.
(13) (a) Jun, Y.-W.; Huh, Y.-M.; Choi, J.-s.; Lee, J.-H.; Song, H.-T.;
Kim, S.; Yoon, S.; Kim, K.-S.; Shin, J.-S.; Suh, J.-S.; Cheon, J. J. Am. Chem.
Soc. 2005, 127, 5732. (b) Huh, Y.-M.; Jun, Y.-w.; Song, H.-T.; Kim, S.;
Choi, J.-s.; Lee, J.-H.; Yoon, S.; Kim, K.-S.; Shin, J.-S.; Suh, J.-S.; Cheon, J.
J. Am. Chem. Soc. 2005, 127, 12387. (c) Jun, Y.-W.; Seo, J.-W.; Cheon, J.
Acc. Chem. Res. 2008, 41, 179. (d) Na, H.-B.; Song, I.-C.; Hyeon, T. Adv.
Mater. 2009, 21, 2133. (e) Yang, H.-M.; Lee, H.-J.; Jang, K.-S.; Park,
C. W.; Yang, H. W.; Heo, W. D.; Kim, J.-K. J. Mater Chem. 2009,
19, 4566. (f) Tromsdorf, U. I.; Bigall, N. C.; Kaul, M. G.; Bruns, O. T.;
Nikolic, M. S.; Mollwitz, B.; Sperling, R. A.; Reimer, R.; Hohenberg, H.;
Parak, W. J.; F€orster, S.; Beisigel, U.; Adam, G.; Weller, H. Nano Lett.
2007, 7, 2422. (g) Boni, A.; Marinone, M.; Innocenti, C.; Sangregorio,
C.; Corti, M.; Lascialfari, A.; Mariani, M.; Orsini, F.; Poletti, G.; Casula,
M. F. J. Phys. D: Appl. Phys. 2008, 41, 134021. (f) Lee, H.-Y.
Nanotechnology 2008, 19, 165101. Figuerola, A.; Fiore, A.; Di Corato,
R.; Falqui, A.; Giannini, C.; Micotti, E.; Lascialfari, A.; Corti, M.;
Cingolani, R.; Pellegrino, T.; Cozzoli, P. D.; Manna, L. J. Am. Chem.
Soc. 2008, 130, 1477. Corti, M.; Lascialfari, A.; Micotti, E.; Castellano,
A.; Donativi, M.; Quarta, A.; Cozzoli, P. D.; Manna, L.; Pellegrino, T.;
Sangregorio, C. J. Magn. Magn. Mater. 2008, 320, e320.
(14) (a) Rochenbertger, J.; Sher, E. C.; Alivsatos, A. P. J. Am. Chem.
Soc. 1999, 121, 11595. (b) Sun, S. H.; Zeng, H.; Robinson, D. B.; Raoux,
S.; Rice, P. M.; Wang, S. X.; Li, G. X. J. Am. Chem. Soc. 2004, 126, 273. (c)
Park, J.; An, K.; Hwang, Y.; Park, J.-G.; Hoh, H.-J.; Kim, J.-Y.; Park, J.-H.;
Hwang, N.-M.; Hyeon, T. Nat. Mater. 2004, 3, 891. (d) Jana, N. R.;
Chen, Y.; Peng, X. Chem. Mater. 2004, 16, 3931. (e) Jun, Y.; Choi, J.;
Cheon, J. Chem. Commun. (Cambridge, U. K.) 2007, 1203. (f) Cheon, J.;
Kang, N.-J.; Lee, S.-M.; Yoon, J.-H.; Oh, S. J. J. Am. Chem. Soc. 2004,
126, 1950. (g) Lu, A.-H.; Salabas, E. L.; Sch€uth, F. Angew. Chem., Int. Ed.
2007, 46, 1222. (h) Shevchenko, E. V.; Talapin, D. V.; Rogach, A. L.;
Kornowski, A.; Hasse, M.; Weller, H. J. Am. Chem. Soc. 2002, 124, 11480.
(i) Tracy, J. B.; Weiss, D. N.; Dinega, D. P.; Bawendi, M. G. Phys. Rev. B
2005, 72, 064404. (j) Sun, S. Adv. Mater. 2006, 18, 393. (k) Puntes, V. F.;
Krishnan, K. M.; Alivisatos, A. P. Science 2001, 291, 2115. (l) Song, Q.;
Zhang, Z. J. J. Am. Chem. Soc. 2004, 126, 6164. (m) Tan, Y.; Zhuang, Z.;
Peng, Q.; Li, Y. Chem. Mater. 2008, 20, 5029.
(1) (a) Jun, Y.-W.; Lee, J.-H; Cheon, J. Angew. Chem., Int. Ed. 2008,
47, 5122.(b) Gossuin, Y.; Gillis, P.; Hocq, A.; Vuong, Q. L.; Roch, A.
WIREs Nanomedecine and Nanobiotechnology; John Wiley & Sons:
New York, 2009; Vol. 1 (May/June), p 299.
(2) (a) Gil, P. R.; Parak, W. J. ACS Nano 2009, 2, 2200. (b)
Pankhurst, Q. A.; Connolly, J.; Jones, S. K.; Dobson, J. J. Phys. D: Appl.
Phys. 2003, 36, R167. (c) Cheon, J.; Lee, J.-H. Acc. Chem. Res. 2008,
41, 1630. (d) Mornet, S.; Vasseur, S.; Grasset, F.; Duguet, E. J. Mater.
Chem. 2004, 14, 2161. (e) Ito, A.; Shinkai, M.; Honda, H.; Lobayashi, T.
J. Biosci. Bioeng. 2005, 100, 1. (f) Goya, G. F.; Grazꢀu, V.; Ibarra, M. R.
Curr. Nanosci. 2008, 4, 1. (g) Fortin, J.-P.; Wilhelm, C.; Servais, J.;
Mꢀenager, C.; Bacri, J.-C.; Gazeau, F. J. Am. Chem. Soc. 2007, 129, 2632.
(3) (a) Hyeon, T. Chem. Commun. (Cambridge, U. K.) 2003, 927.(b)
Klabunde, K. J. In Nanoscale Materials in Chemistry; Klabunde, K. J., Ed.;
Wiley Intersciences: New York, 2001. (c) Diehl, M. R.; Yu, J.-Y.; Heath,
J. R.; Doyle, G. A.; Sun, S.; Murray, C. B. J. Phys. Chem. B 2001,
105, 7913. (d) Leslie-Pelescky, D. L.; Rieke, R. D. Chem. Mater. 1996,
8, 1770.(e) Schmid, G. NPs: From Theory to Applications; Wiley-VCH:
Weinheim, Germany, 2004. (f) Niederberger, M.; Garnweitner, G.; Ba,
J.; Polleux, J.; Pinna, N. Int. J. Nanotechnol. 2007, 4, 263. (g) Di Marco,
M.; Sadun, C.; Port, M.; Guilbert, I.; Couvreur, P.; Dubernet, C. Int. J.
Nanomed. 2007, 2, 609. (h) Dorman, J. L. Rev. Phys. Appl. 1981, 16, 275.
(4) Gilchrist, R. K.; Medal, R.; Shorey, W. D.; Hanselman, R. C.;
Parott, J. C.; Taylor, C. B. Ann. Surg. 1957, 146, 596.
(5) (a) Andriola Silva, A. K.; Da Silva, E. L.; Oliveira, E. E.;
Nagashima, T.; Soares, L. A. L.; Medeiros, A. C.; Araujo, J. H.; Araujo,
I. B.; Carrico, A. S.; Egito, E. S. T. Int. J. Pharm. 2007, 334, 42. (b)
ꢂ
Horꢀaka, D.; Rittichb, B.; Spanovꢀab, A. J. Magn. Magn. Mater. 2007,
331, 249.
(6) (a) Berkovsky, B. M.; Medvedev, V. F.; Krakov, M. S.; Magnetic
Fluids: Engineering Applications: Oxford University Press: New York,
1993; (b) Rosensweig, R. E.; Ferrohydrodynamics; Cambridge University
Press: New York, 1985. (c) Kang, Y. S.; Risbud, S.; Rabolt, J. F.; Stroeve,
P. Chem. Mater. 1996, 8, 2209. (d) Massart, R. IEEE Trans. Magn. 1981,
17, 1247. (e) Massart, R.; Dubois, E.; Cabuil, V.; Hasmonay, E. J. Magn.
Magn. Mater. 1995, 149, 1. (f) Shimoiizaka, J. N. K.; Fujita, T.; Kounosu,
A. IEEE Trans. Magn. 1980, 16, 368. (g) Wooding, A.; Kilner, M.;
Lambrick, D. B. J. Colloid Interface Sci. 1991, 144, 236. (h) Hyeon, T.;
Lee, S. S.; Park, J.; Chung, Y.; Na, H. B. J. Am. Chem. Soc. 2001,
123, 12798. (i) Park, J.; An, K.; Hwang, Y.; Park, J.-G.; Noh, H.-J.; Kim,
J.-Y.; Park, J.-H.; Hwang, N.-M.; Hyeon, T. Nat. Mater. 2004, 3, 891. (j)
Euliss, L. E.; Grancharov, S. G.; O’Brien, S.; Deming, T. J.; Stucky, G. D.;
Murray, C. B.; Held, G. A. Nano Lett. 2003, 3, 1489.
(7) (a) De Cuyper, M; Valtonen, S. J. Magn. Magn. Mater. 2001,
225, 89. (b) Bulte, J. W. M.; de Cuyper, M.; Despres, D.; Frank, J. A.
J. Magn. Magn. Mater. 1999, 194, 204.
(15) Yang, J.; Lee, T. I.; Lim, E.-K.; Hyung, W.; Lee, C.-H.; Song, Y. J.;
Suh, J.-S.; Yoon, H.-G.; Huh, Y.-M.; Haam, S. Chem. Mater. 2007, 19, 3870.
(16) Renshaw, P. F.; Owen, C. S.; McLaughlin, A. C.; Frey, T. G.;
Leigh, J. S. Magn. Reson. Med. 1986, 3, 217.
(17) (a) Yang, X.; Chen, Y.; Yuan, R.; Chen, G.; Blanco, E.; Gao, J.;
Shuai, X. Polymer 2008, 49, 3477. (b) Shi, X.; Wang, S. E.; Swanson,
S. D.; Ge, S.; Cao, Z.; Van Antwerp, M. E.; Landmark, K. J.; Baker, J. R.,
Jr. Adv. Mater. 2008, 20, 1671. (c) Wang, S. H.; Shi, X.; Van Antwerp, M.;
Cao, Z.; Swanson, S. D.; Bi, X.; Baker, J. R., Jr. Adv. Mater. 2007,
17, 3043.
(18) (a) Shen, T.; Weissleder, R.; Papisov, M.; Bogdanov, A.; Brady,
T. J. Magn. Reson. Med. 1993, 29, 599. (b) Reimer, P.; Jahnke, N.;
Schima, M.; Deckers, F.; Marx, F.; Holzknecht, N.; Saini, S. Radiology
2000, 217, 152. (c) Zhao, M.; Kircher, M. F.; Josephson, L.; Weissleder,
R. Bioconjugate Chem. 2002, 13, 840.
(8) Gordon, R. T.; Hines, J. R.; Gordon, D. Med. Hypotheses 1979,
5, 83.
(9) (a) Jordan, A.; Scholz, R.; Wust, P.; Fahling, H.; Krause, J.;
Wlodarczyk, W.; Sander, B.; Vogl, T.; Felix, R. Int. J. Hyperthermia 1997,
13, 587. (b) Ito, A.; Tanaka, K.; Honda, H.; Abe, S.; Yamaguchi, H.;
Kobayashi, T. J. Biosci. Bioeng. 2003, 96, 364. (c) Jordan, A.; Scholz, R.;
Maier-Hauff, K.; van Langeghem, F. K. H.; Waldoefner, N.; Teichgraeber,
U.; Pinkernelle, J.; Bruhn, H.; Neumann, F.; Thiesen, B.; Felix, R.
J. Neuro-Oncol. 2006, 78, 7. (d) Hamaguchi, S.; Tohnai, I.; Ito, A.;
Mitsudo, K.; Shigetomi, T.; Ito, M.; Honda, H.; Kobayashi, T.; Ueda, M.
Cancer Sci. 2003, 94, 834. (e) Kawai, N.; Ito, A.; Nakahara, Y.; Futakuchi,
M.; Shirai, T.; Honda, H.; Kobayashi, T.; Kohri, K. The Prostate 2005,
64, 373. (f) Le, B.; Shinkai, M.; Kitade, T.; Honda, H.; Yoshida, J.;
Wakabayashi, T.; Kobayashi, T. J. Chem. Eng. Jpn. 2001, 34, 66.
(10) (a) Bulte, J. W.; Kraitchman, D. L. NMR Biomed. 2004, 17, 484.
(b) Michalet, X.; Pinaud, F. F.; Bentolila, L. A.; Tsay, J. M.; Doose, S.; Li,
J. J.; Sundaresan, G.; Wu, A. M.; Gambhir, S. S.; Weiss, S. Science 2005,
(19) (a) Thunemann, A. F.; Schutt, D.; Kaufner, L.; Pison, U.;
Mohwald, H. Langmuir 2006, 22, 2351. (b) Jain, T. K.; Morales, M. A.;
10471
dx.doi.org/10.1021/ja111448t |J. Am. Chem. Soc. 2011, 133, 10459–10472

Journal of the American Chemical Society
ARTICLE
Sahoo, S. K.; Leslie-Pelecky, D. L.; Labhasetwar, V. Mol. Pharmaceutics
2005, 2, 194.
(48) Fabian, K.; Kirchner, A; Williams, W.; Heider, F.; Leib, T.;
Huber, A. Geophys. J. Int. 1996, 124, 89.
(20) Kim, D. K.; Mikhaylova, M.; Wang, F. H.; Kehr, J.; Bjelke,
B.; Zhang, Y.; Tsakalakos, T.; Muhammed, M. Chem. Mater. 2003,
15, 4343.
(49) Folch, J.; Larionova, J.; Guari, Y.; Molvinger, K.; Luna, C.;
Sangregorio, C.; Innocenti, C.; Caneschi, A.; Guꢀerin, Ch. Phys. Chem.
Chem. Phys. 2010, 12, 12760–12770.
(21) Huh, Y.-M.; Jun, Y.-W.; Song, H.-T.; Kim, S.; Choi, J.-S.; Lee,
J.-H.; Yoon, S.; Kim, K.-S.; Shin, J.-S.; Suh, J.-S.; Cheon, J. J. Am. Chem.
Soc. 2005, 127, 12387.
(50) Nꢀeel, L. Ann. Geophys. 1949, 5, 99.
(51) Chikazumi, S. Physics of Ferromagnetism, 2nd ed.; Clarendon:
Oxford, U.K., 1997.
(22) Yigit, M. V.; Mazumdar, D.; Lu, Y. Bioconjugate Chem. 2008,
19, 412.
(23) Xie, J.; Chen, K.; Lee, H.-Y.; Xu, C.; Hsu, A. R.; Peng, S.; Chen,
X.; Sun, S. J. Am. Chem. Soc. 2008, 130, 7542.
(52) Djurberg, C.; Svedlindh, P.; Nordblad, P.; Hansen, M. F.;
Bodker, F.; Morup, S. Phys. Rev. Lett. 1997, 79, 5154.
(53) Weast, R. C., Ed. Handbook of Chemistry and Physics, 53rd ed.;
CRC Press: Cleveland, OH, 1972; p D-128.
(24) Weissleder, R.; Moore, A.; Mahmood, U.; Bhorade, R.;
Benveniste, H.; Chiocca, E. A.; Basilion, J. P. Nat. Med. 2000, 6, 351.
(25) Lee, Y. C., Lee, R. T., Eds. Neoglycoconjugates: Preparation and
Applications; Academic Press: New York, 1994.
(26) El-Boubbou, K.; Zhu, D. C.; Vasileiou, C.; Borhan, B.; Prosperi,
D.; Li, W.; Huang, X. J. Am. Chem. Soc. 2010, 132, 4490.
(27) Barragan, V.; Menger, F. M.; Caran, K.; Vidil, C.; Morere, A.;
Montero, J.-L. Chem. Commun. (Cambridge, U. K.) 2001, 1, 85.
(28) Shimomura, M.; Ono, B.; Oshima, K.; Miyauchi, S. Polymer
2006, 47, 5785.
(54) Atkinson, W. J.; Brezovich, I. A.; Chakraborty, D. P. IEEE Trans.
Biomed. Eng. 1984, 31 (BMEE), 70.
(55) Maier-Hauff, K.; Rothe, R.; Scholz, R.; Gneveckow, U.; Wust,
P.; Thiesen, B.; Feussner, A.; von Deimling, A.; Waldoefner, N.; Felix, R.;
Jordan, A. J. Neuro-Oncol. 2007, 81, 53–60.
(56) Andr€a, W., Nowak, H., Eds. Magnetism in Medicine. A Hand-
book; Wiley-VCH Verlag: Berlin, 1998.
(57) (a) Chan, D. C. F.; Kirpotin, D. B.; Bunn, P. A. Scientific and
Clinical Applications of Magnetic Carriers; Hafeli, U., Ed.; Plenum Press:
New York, 1997; pp 607ꢀ618. (b) Kallumadil, M.; Tada, M.; Nakagawa,
T.; Abe, M.; Southern, P.; Pankhurst, Q. A. J. Magn. Mater. 2009,
321, 1509.
(29) El-Boubbou, K.; Gruden, C.; Huang, X. J. Am. Chem. Soc. 2007,
129, 13393.
(30) Astete, C. E.; Kumar, C. S. S. R.; Sabilov, C. M. Colloids Surf., A
2007, 299, 209.
(58) Rosensweig, R. E. J. Magn. Mater. 2002, 252, 370.
(59) Purushotham, S.; Ramanujan, R. V. J. Appl. Phys. 2010,
107, 114701.
(31) Horac, D.; Babic, M.; Jendelova, P.; Herynek, V.; Trchova, M.;
Pientka, Z.; Pollert, E.; Sykova, E. Bioconjugate Chem. 2007, 18, 635.
(32) Lartigue, L.; Oumzil, K.; Guari, Y.; Larionova, J.; Guꢀerin, Ch.;
Montero, J.-L.; Barragan-Montero, V.; Sangregorio, C.; Caneschi, A.;
Innocenti, C.; Kalaivani, T.; Arosio, P.; Lascialfari, A. Org. Lett. 2009,
11, 2992.
(33) (a) Barragan-Montero, V.; Winum, J.-Y.; Molꢁes, J.-P.; Juan, E.;
Clavel, C.; Montero, J.-L. Eur. J. Med. Chem. 2005, 40, 1022.(b)
Barragan, V. ; Montero, J.-L. ; Winum, J.-Y. ; Juan, E. ; Moles, J.-P.
French Patent FR 2854896, 2007.
(60) Okawa, K.; Sekine, M.; Maeda, M.; Tada, M.; Abe, M.;
Matsushita, N.; Nishio, K.; Handa, H. J. Appl. Phys. 2006, 99, 08H102.
(61) Hergt, R.; Hiergeist, R.; Zeisberger, M.; Schuler, D.; Heyen, U.;
Hilger, I.; Kaiser, W. A. J. Magn. Mater. 2005, 293, 80.
(62) Ferrante, G.; Sykora, S. Adv. Inorg. Chem. 2005, 57, 405.
(63) Gillis, P.; Moiny, F.; Brooks, R. A. Magn. Reson. Med. 2002,
47, 257. Yablonskiy, D. A.; Haacke, E. M. Magn. Reson. Med. 1994,
32, 749.
(64) (a) Bulte, J. W. M.; Vymazal, J.; Brooks, R. A.; Pierpaoli, C.;
Frank, J. A. J. Magn. Reson. Imaging 1993, 3, 641.(b) Bulte, J. W. M.;
Brooks, R. A. Scientific and Clinical Applications of Magnetic Carriers;
H€afeli, U., Ed.; Plenum Press: New York, 1997; pp 527ꢀ542.
(65) Roch, A.; Muller, R. N.; Gillis, P. J. Chem. Phys. 1999, 110, 5403.
(66) Boni, A.; Marinone, M.; Innocenti, C.; Sangregorio, C.; Corti,
M.; Lascialfari, A.; Mariani, M.; Orsini, F.; Poletti, G.; Casula, M. F.
J. Phys. D: Appl. Phys. 2008, 41, 134021.
(34) Taylor, P. R.; Gordon, S.; Martinez-Pomares, L. Trends Im-
munol. 2005, 26, 104.
(35) Hou, Y.; Xu, Z.; Sun, S. Angew. Chem., Int. Ed. 2007, 46, 6329.
(36) (a) Yu., W. W.; Falkner, J. C.; Yazuz, C. T.; Colvin, V. L. Chem.
Commun. (Cambridge, U. K.) 2004, 2306. (b) Lin, C.-R.; Chiang, R.-K.;
Wang, J.-S.; Sung, T.-W. J. Appl. Phys. 2006, 99, 08N710.
(37) Micha€elis, A.; Becker, T. Chem. Ber. 1897, 30, 1003.
(38) Montero, J.-L.; Imbach, J.-L. Eur. J. Med. Chem. 1982, 17, 97.
(39) (a) Mohapatra, S.; Pramanik, P. Colloids Surf., A 2009, 339, 35.
(b) Guerrero, G.; Mutin, P. H. A. Chem. Mater. 2001, 13, 4367.
(40) Das, M.; Mishra, D.; Dhak, P.; Gupta, S.; Maiti, T. K.; Basak, A.;
Prarmanik, P. Small 2009, 5, 2883.
(67) Casula, M. F.; Floris, P.; Innocenti, C.; Lascialfari, A.; Marinone,
M.; Corti, M.; Sperling, R. A.; Parak, W. J.; Sangregorio, C. Chem. Mater.
2010, 22, 1739.
(41) (a) Kovalenko, M. V.; Bodnarchuk, M. I.; Lechner, R. T.;
Hesser, G.; Sch€affler, F.; Heiss, W. J. Am. Chem. Soc. 2007, 129, 6352. (b)
Shavel, A.; Rodriguez-Gonzalez, B.; Spasova, M.; Farle, M.; Liz-Marzan,
L. M. Adv. Funct. Mater. 2007, 17, 3870. (c) Shavel, A.; Rodriguez-
Gonzalez, B.; Pacifico, M.; Spasova, M.; Farle, M.; Liz-Marzan, L. M
Chem. Mater. 2009, 21, 1326.
(42) Fleet, M. E. Acta Crystallogr. 1981, B37, 917.
(43) (a) Demortiꢁere, Panissod, P.; Pichon, B. P.; Pourroy, G.;
Guillon, D.; Donnio, B.; Egin- Colin, S. B. Nanoscale 2011, 3, 225. (b)
Lin, C.-R.; Chiang, R.-K.; Wang, J. S.; Sung, T. W. J. Appl. Phys. 2006,
99, 08N710.
(44) D. Parker, D.; Dupuis, V.; Ladieu, F.; Bouchaud, J.-P.; Dubois,
E.; Perzynski, R.; Vincent, E. Phys. Rev. B 2008, 77, 104428.
(45) Morrish, A. H. The Physical Principles of Magnetism; John Wiley
& Sons: New York, 1965.
(46) Hergt, R.; Dutz, S.; R€oder, M. J. Phys.: Condens. Matter 2008,
20, 385214.
(47) (a) Kneller, E. F.; Luborsky, F. E. J. Appl. Phys. 1963, 34, 656.
(b) Cullity, B. D. Introduction to Magnetic Materials; Addison-Wesley:
Reading, MA, 1972.
10472
dx.doi.org/10.1021/ja111448t |J. Am. Chem. Soc. 2011, 133, 10459–10472