Multidomain iron nanoparticles for the preparation of polyacrylamide

ferrogels

Ajay Shankar a*, Alexander P. Safronova,b, Ekaterina A. Mikhnevicha, Igor V. Beketova,b

a
Ural Federal University, 19 Mira Str., 620002, Yekaterinburg, Russia.
b
Institute of Electrophysics UB RAS, 106 Amundsen Str., 620016, Yekaterinburg, Russia.

Abstract: Ferrogels (FG) based on poly(acrylamide) (PAAm) with embedded multidomain iron magnetic
nanoparticles (MNPs) were synthesized by radical polymerization in water. Iron MNPs prepared by the electrical
explosion of wire were spherical in shape and have an average diameter around 100 nm. MNPs were modified by a
surfactant oleic acid to improve their dispersion in water. DLVO theoretical consideration was done to understand
the stability of dispersions. By microcalorimetry it was shown that the oleic layer on the surface of MNPs prevents
their interaction with PAAm network of FG. Mechanical testing of the compression modulus and the deformation of
FGs in magnetic field show up their prospectiveness as a material for magnetically sensitive MEMS and actuators.
Keywords: Iron nanoparticles, extended DLVO theory, poly(acrylamide), ferrogels, magnetodeformation, interfacial
interaction.
*
Corresponding author: ashankar@urfu.ru
Corresponding address: Chair of macromolecular compounds, Institute of natural sciences, Ural federal university,
Yekaterinburg, Russia.
Contact no.: +7 343 261 6046
1. Introduction. Ferrogels (FGs) are a soft type of
highly elastic polymer composites which have
magnetic particles as a constituent. The particles can
be multidomain, single domain or superparamagnetic
in nature. Their significant response through
controllable deformation in magnetic field has
attracted considerable interest for their application as
smart materials[1]. Their biological applications,
namely, viz. targeted delivery, labelling, radio
frequency methods and contrast enhancements in
imaging are few examples where size of filler particles
are needed to be ~ 1 m or less [2]. This dimension
usually corresponds to multidomain regime of
applicable materials like Fe, Ni, Fe3O4 etc. To use
them in preparation of FGs, a stable concentrated
dispersion is required. Making stable aqueous
dispersion of multidomain iron magnetic nanoparticles
(MNPs) is a challenging task [3]. Moreover,
functionalization procedure for these MNPs, post
synthesis, further complicates this problem due to high
particle aggregation. In present study a room
temperature grinding method was employed to modify
the surface of Fe MNPs to achieve their stable
dispersion. The dispersion was later used in
preparation of FGs with poly(acrylamide) (PAAm)
network
2. Experimental section.
Materials The iron MNPs were synthesized by the
electric explosion of wire technique [4]. Surface
functionalization of MNPs was carried out under
ambient conditions by grinding air-dry MNPs with
surfactant, oleic acid. In a typical procedure, 0.5 g of
MNPs and 1.5 g of KOH were ground by mixing in
agate mortar for 5 min. After homogenization, 4.5 g of
surfactant was added followed by grinding for ~80 min
till highly viscous films are formed. The mixture was
washed till neutral pH and then dispersed in distilled
water for desired concentration of MNPs. The visual
assessment of sedimentation stability of dispersion
was ~15-20 min.
FGs were synthesized at room temperature by free
radical polymerization of acrylamide (AAm,
AppliChem, Darmstadt, Germany). Aqueous solutions
of a cross-linker N,N methylene diacrylamide
(MDAA, Merck, Schtuchardt, Germany) (48 mM),
initiator ammonium persulphate (APS) (50 mM) and
aqueous dispersion of MNPs (6.3% w/w) were
prepared prior to the reaction. Reactants were used as
received. In a typical synthesis procedure, 1 mL of
MDAA solution was added to the 0.341 g of AAm and
homogenized. 0.3 mL APS solution was added
followed by addition of 3 mL of water or dispersion of
MNPs or both in specified amount for the preparation
of gels with varying MNPs content. Then, 20 L of
catalyst - N,N,N,N-Tetramethylethane-1,2-diamine
(Merck) was added to the reaction mixture with
vigorous stirring for few seconds. Polymerization took
place in poly(ethylene) cylindrical tubes for 4 hours
[5]. The gels were taken out of mold and washed for
two weeks with distilled water renewal at a two-day
interval. The equilibrium swelling degree () of gels
was calculated by = (ms - md)/ md, where ms is mass
of swollen gel and md is mass of dried gel. This value
was further used in calculation of volume fraction of
oleate coated MNPs () in swollen FGs.
Methods. XRD diffractorgrams were obtained using a
diffractometer Bruker D8 Discover in copper radiation
(Cu K1,2 = 1,542) with a graphite monochromator
in the diffracted beam. The Rietveld refinement of
XRD pattern was performed using Topas-3 software.
Transmission electron microscopy (TEM) was
performed were performed using a JEOL JEM2100
microscope operating at 200 kV. Dynamic light
scattering (DLS) and zeta-potential measurements
were performed using Brookhaven instruments BI-90
Plus Particle Size Analyzer. Bright field optical
microscopy was done using BX51 Olympus
microscope. Specific surface area of MNPs was
measured using Micromeritics TriStar3000 sorption
analyzer. Microcalorimetry studies of the enthalpy of
interfacial adhesion between PAAm and MNPs were
performed using Calvet micro calorimeter of a
laboratory construction [6]. Youngs modulus was
measured using a laboratory setup provided
compressive loading of gel samples and optical
registration of their deformation. Deformation of
ferrogels in magnetic field was monitored by an
optical system in a laboratory setup. A magnetic
system of NdFeB permanent magnets provided a
uniform field 420 mT in the central zone 10 X 10 X 10
mm3[7].
3. Results and discussion.
 Fig. 1. M-H loop of bare MNPs at 298 K. Inset 1a shows
magnified view of magnetization curve. Inset 1b shows the
TEM micrograph of bare FNPs. Inset 1c shows size
distribution of MNPs from several TEM images. The red
curve is lognormal fit on distribution. Inset 1d shows the
rietveld refinement of XRD pattern of oleate coated MNPs.
The structural and magnetic characterization of bare
and oleate modified MNPs is shown in fig.1.
= 282.46 g mol-1, vaccum permeability, o= 1.25 X 10-6 H m-1 and
remanent magnetization, Mr= 157.7 kAm-1 (from fig. 1). Arrows
indicate respective ordinates of curves.
The van der Waals interactions were calculated in a
classical manner excluding retardation effects. This
attractive interaction energy term (VvdW) can be
expressed as

The room temperature magnetization loop of MNPs

shows typical ferromagnetic behavior as visible from A 2a 2 2a 2 (4a s) (1)
VvdW ln s 2
high coercivity and retentivity (fig. 1 and inset a). The 6 s 4a s (2a s) 2
(2a s)
(TEM) images of bare MNPs shows that particles have
spherical shape and are polydisperse (fig.1 inset b). The electrostatic repulsion between coated MNPs due
The statistical size distribution analysis was carried to the charged outer layer of fatty acid was evaluated
out using lognormal distribution function by counting as:
over 1000 particles (inset c). The mean average
particle diameter was 65.8 nm. The X-ray diffraction Ve 2 a r o o2 ln 1 e s (2)
patterns (XRD) of oleate modified MNPs show Bragg
peaks at 44.66o, 65.03o and 82.35o which are is the inverse Debye length and can written as
characteristic of -Fe phase, as shown in inset d. The (2 ( 2 )/ )1/2 . The magnitude of zi,
specic surface area of MNPs was found to be 7 m2/g
which corresponds to average diameter of 108.1 nm, ci, and p for inverse Debye length calculation were
which correlates well with TEM results. taken in the range of similar studied systems [11,12]

The colloidal stability of MNPs dispersion employed The short range steric interactions among MNPs.
in preparation of FGs was qualitatively modeled by imparted by surfactants including osmotic effect and
extended DLVO approach [3]. Interaction energy elastic repulsions were taken as [13]
curves simulated for current aqueous dispersion are 4 ak B T 2 1 s 1 s
p 2 ln
shown in fig. 2. v 2 2 4

2 ak B T s ln s 3 ( s / ) 6 ln 3 ( s / ) 3 1 s
2 2

for s <

p
Vstr Mw 2 2

4 ak B T 2 1
2
s
v p for < s < 2
2 2
0 for s > 2

Fig. 2. Normalized interaction energy curves for two identical
MNPs as a function of surface-to-surface separation (s). The
parameters used for the calculation of energy curves are: the radius
of particles, a = 54.5 nm, Hamaker constant, A = 2.94 X 10-19 J
[3,8,9], relative permittivity of carrier liquid, r = 80.4, vaccum
permittivity, o = 8.85 X 10-12 Fm-1, o is the surface potential which
is approximated here with measured zeta potential, = -35 mV, T =
298 K, kB = 1.38 X 10-23 JK-1, elementary charge, q = 1, NA = 6.023
X 1023 mol-1, valency of ions (approximated here with ionized
constituents of free surfactants i.e. K+ and oleate), zi= 1, ionic
concentration, ci = 0.001 M, volume of a water molecule, = 0.03
nm3, volume fraction of coated surfactant, p= 0.05, Flory-Huggins
parameter, = 0.42, adsorbed layer thickness, = 3.0 nm [10],
density of surfactant, = 0.895 g cm-3, molar mass of surfactant, Mw
(3)
Although no magnetic field was applied to a
suspension the multidomain Fe MNPs could still get
magnetized due to their numerous collisions in the
preparation. The maximum magnetic attraction energy
(VM) among multidomain MNPs in the absence of
field was estimated as
8o M r2 a3
(4)
VM 3
s
9 2
a
The overall particle-particle interaction energy (VT)
can be written as VT = VvdW + Ves + Vst + VM. An small
energy barrier of ~ 7 kBT imparted by non-magnetic
forces can be seen. In addition, an order of magnitude
difference between magnetic and non-magnetic forces
can be noticed (fig. 2). The magnetic forces are greater
than the sum of van der Waals, electrostatic and steric
forces. Thus, the aggregation of MNPs and subsequent
sedimentation process cannot be avoided completely
in such dispersions
Stability of dispersion was monitored using DLS
measurements. Figure 3 shows multimodal
distribution of hydrodynamic sizes of MNPs in a
freshly prepared dispersion. Here, bimodal
distribution indicates the presence of two types of
species in dispersion. The first mode at 200 nm may
correspond to surfacted MNPs while the second mode
at 800 nm obviously represent aggregates. The relative
volume fractions of individual MNPs and their
aggregates show that the sufficient stability of the
dispersion for the preparation of ferrogels can be
maintained at least at short time scale.
Fig.3 Volume weighted distribution of hydrodynamic
diameters of aqueous dispersion. Stock dispersion was
diluted to 4.2 mg/L for DLS measurements.
To determine the extent of aggregation of MNPs in
FGs, the optical images were taken from thin slices.
(fig. 4). A homogeneous distribution of aggregated
MNPs was observed rather than individual particles.
Fig. 4. A typical bright field optical microscopic image of
ferrogel. Bar scale corresponds to 200 m. Inset shows
histogram of diameters of MNP aggregates embedded in
polymer network obtained using ImageJ by counting over
600 particles. The white curve shows two parameter Weibull
fit on size distribution of aggregates.
It was clear from the optical images that formation of
aggregates could not be avoided during gelation. The
mean size of aggregates with = 0.71 X 10-4 and 1.56
X 10-4 were found to be 2.6 m and 3.1 m
respectively. These dimensions are larger than
hydrodynamic size of aggregates as observed from
DLS. Larger aggregates can be related to reaction
conditions as it might have also perturbed the stability
of MNPs by unravelling of outer surfactant layer.
The interaction between MNPs and polymer network
in FGs can be understood on the basis of the enthalpy
of mixing of PAAm-MNP composites. As the
enthalpy of mixing of a polymer with MNPs cannot be
measured directly, an indirect route was followed. [6]
To make a comparison, separate composites of both
bare and oleate modified MNPs were prepared with
linear PAAm (MW. = 56.9 kDal; determined by
viscometry). The filler compositions were varied in
the entire range. The enthalpy of mixing of polymer
composites (Hm) can be written as Hmix= fill
Hwet+ pol Hdis pol - Hdis comp, where fill and pol are
weight fractions of filler (i.e., bare or coated MNPs)
and linear PAAm respectively. Hwet, Hdis pol and
Hdis comp are heat of dissolution of filler, linear PAAm
and composites in solvent, respectively. They were
measured directly and used for the calculation of the
enthalpy of mixing. The concentration dependence of
Hmix for the composites is shown in fig. 5.
Fig. 5. Calculated enthalpies of mixing of PAAm composites
with bare and coated MNPs. Dotted lines are a guide for an
eye only.
Hmix values for composites with bare MNPs are
negative over the entire concentration range. It means
strong interfacial adhesion of PAAm chains to the
surface of bare MNPs. Meanwhile, in case of modified
MNPs Hmix values are close to zero within the
experimental error of calorimetry measurements. It
means that the shell of the oleate on the surface of iron
MNPs blocks the interaction with PAAm chains likely
due to poor compatibility between polar groups of
PAAm monomeric units and outer layer of adsorbed
surfactant. The zero values of Hmix indicate that
oleate coated MNPs are not bonded to the polymeric
matrix of PAAm FG. Meanwhile they can not leave
the network as they are entrapped in the FG due to the
special limitations. It was shown [14] that PAAm gels
with similar composition had a mesh size of
approximately 3-5 nm which is much lower than the
size of MNPs and their aggregates.
Poor interfacial adhesion in FGs explains the trend in
their Youngs modulus determined in a compression
mode. It was found that modulus gradually decreased
with the increase in MNPs content in FG. As
embedded MNPs are not bonded to the polymeric
network of FG they act as weak points of the matrix
and diminish its mechanical strength.
In the magnetodeformation studies cylindrical FGs
with aspect ratio ~1 were kept in aqueous medium and
placed in uniform magnetic field 420 mT. The
direction of field was parallel to the main axis of
cylinder. The sample was monitored for changes in its
relative dimensions using programmable camera by
capturing images at different time intervals. The
kinetics of magnetic field induced deformation of a FG
with = 1.56 X 10-4 is shown in fig. 6. The dynamic
increase of volume and enhancement in relative
dimensions was found to be nearly linear at first 60
min of observation. After 6 hrs, a saturation in the
swelling behavior was reached. A total of more than
8% increase in FG volume was observed during this
time period.
Fig. 6. Magnetodeformation of ferrogel with = 1.56 X 10-
4. Green symbols represent the data observed at t = 0 in
absence of magnetic field. Inset shows a typical FG sample
as visualized from camera during measurement.
4. Conclusions. The possibility of the synthesis
of aqueous poly(acrylamide) (PAAm) ferrogels (FGs)
with embedded multidomain iron MNPs was
demonstrated. To provide sufficient colloidal stability
of iron MNPs in water dispersion they were modified
by a surfactant (oleic acid). Although DLVO
theoretical consideration did not show the long-term
stability of the dispersion of modified iron MNPs in
water, a short-term stability was observed. It was
sufficient for the polymerization of FG. Due to the
general instability of the dispersion, iron MNPs are
aggregated both in the dispersion and in FG, with
micron sized aggregates uniformly dispersed.
Microcalorimetry measurements of the enthalpy of
interfacial adhesion showed that the modification of
MNPs by oleic acid blocks the interaction of PAAm
chains with particles. It results in the diminishing trend
of the compression modulus of FGs upon the content
of MNPs. PAAm FGs with embedded multidomain
iron MNPs show up the increase of their total volume
if placed in the uniform magnetic field. It makes them
a prospective material for the further testing in
magnetically responsive sensors and actuators. The
grinding methodology described in this work can be
extended for the preparation of dispersions of other
metal nanoparticles, e.g. Ni, Cu, Al, etc.
5. Acknowledgements. The work was
supported by the Russian Science Foundation project
no 14-19-00989. Authors thank Dr. Andrey Galyas for
bright field optical measurements.
References
[1] M. Zrnyi, L. Barsi, D. Szabo, H.-G. Kilian, Direct
observation of abrupt shape transition in ferrogels
induced by nonuniform magnetic field, J. Chem.
Phys. 106 (1997) 56855692.
[2] G. Filipcsei, I. Csetneki, A. Szilgyi, M. Zrnyi,
Magnetic Field-Responsive Smart Polymer
Composites, in: Oligomers Polym. Compos. Mol.
Imprinting, Springer Berlin Heidelberg, Berlin,
Heidelberg, 2007: pp. 137189.
[3] M. Sanchez-Dominguez, C. Rodriguez-Abreu,
Nanocolloids: A meeting point for scientists and
technologists, Elsevier, Netherlands, 2016.
[4] I.V. Beketov, A.P. Safronov, A.V. Bagazeev, A.
Larraaga, G.V. Kurlyandskaya, A.I. Medvedev, In
situ modification of Fe and Ni magnetic nanopowders
produced by the electrical explosion of wire, J. Alloys
Compd. 586 (2014) S483S488.
doi:10.1016/j.jallcom.2013.01.152.
[5] A.P. Safronov, T.V. Terziyan, Formation of chemical
networks of acrylamide and acrylic acid hydrogels
initiated by ammonium persulfate, Polym. Sci. Ser.
B. 57 (2015) 481487.
doi:10.1134/S1560090415050140.
[6] A.P. Safronov, A.S. Istomina, T.V. Terziyan, Y.I.
Polyakova, I.V. Beketov, Influence of interfacial
adhesion and the nonequilibrium glassy structure of a
polymer on the enthalpy of mixing of polystyrene-
based filled composites, Polym. Sci. Ser. A. 54
(2012) 214223. doi:10.1134/S0965545X12030108.
[7] A.P. Safronov, T.V. Terziyan, A.S. Istomina, I.V.
Beketov, Swelling and contraction of ferrogels based
 on polyacrylamide in a magnetic field, Polym. Sci.
Ser. A. 54 (2012) 2633.
doi:10.1134/S0965545X12010063.
[8] D. Blundell, S. Wai, Measurement of van der Waals
Forces Between Iron/Iron Surfaces in Water by
Atomic Force Microscopy, Microsc. Microanal. 11
(2005). doi:10.1017/S1431927605504331.
[9] Theoretical values of Hamaker constants of water (Aw =
3.675 X 10-20 J) and iron (AI = 5.39 X 10-19 J) across vacuum
were used in calculation. The Hamaker constant of iron
across water as a medium (A) can be approximated by A
2
(AI -Aw ) .
[10] L. Rodrguez-Arco, M.T. Lpez-Lpez, J.D.G.
Durn, A. Zubarev, D. Chirikov, Stability and
magnetorheological behaviour of magnetic fluids
based on ionic liquids, J. Phys. Condens. Matter. 23
(2011) 455101. doi:10.1088/0953-
8984/23/45/455101.
[11] J. de Vicente, A.V. Delgado, R.C. Plaza, J.D.G.
Durn, F. Gonzlez-Caballero, Stability of Cobalt
Ferrite Colloidal Particles. Effect of pH and Applied
Magnetic Fields, Langmuir. 16 (2000) 79547961.
doi:10.1021/la0003490.
[12] J.L. Viota, J. de Vicente, J.D.G. Durn, A.V.
Delgado, Stabilization of magnetorheological
suspensions by polyacrylic acid polymers, J. Colloid
Interface Sci. 284 (2005) 527541.
doi:10.1016/j.jcis.2004.10.024.
[13] P. Kralchevsky, R. Miller, F. Ravera, Colloid and
interface chemistry for nanotechnology, CRC Press,
2013.
[14] G.V. Kurlyandskaya, E. Fernndez, A.P. Safronov,
A.V. Svalov, I. Beketov, A.B. Beitia, A. Garca-
Arribas, F.A. Blyakhman, Giant magnetoimpedance
biosensor for ferrogel detection: Model system to
evaluate properties of natural tissue, Appl. Phys. Lett.
106 (2015) 193702. doi:10.1063/1.4921224.