Fast synthesis of nanocrystalline Mg2Si by microwave heating: A new route to nano-structured thermoelectric materials

Article abstract of DOI:10.1039/c0dt00519c

The ultra fast synthesis of nanocrystalline Mg₂Si was carried out using microwave radiation. The elemental precursors were first milled together under dry conditions to get fine particles. The resulting mixture of powders of Mg and Si was cold pressed before being heated by microwave irradiation. Precursors and products were analyzed by X-ray diffraction and scanning electron microscopy. The high energy ball milling parameters utilized to prepare the reactive powders have quite an influence on the behavior of the mixture under irradiation. Moreover, SEM imaging demonstrates that the power and time of irradiation are crucial for the grain growth of the Mg₂Si and must be adequately controlled in order to avoid the decomposition of the phase. Our results show that we successfully managed to easily and quickly synthesize homogeneous nanocrystalline Mg₂Si with particle size smaller than 100 nm using a microwave power of only 175 W for two minutes on powders ball milled for two hours. The Royal Society of Chemistry 2010.

Full text of DOI:10.1039/c0dt00519c

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/dalton | Dalton Transactions
Fast synthesis of nanocrystalline Mg2Si by microwave heating: a new route to
nano-structured thermoelectric materials
Etienne Savary, Franck Gascoin* and Sylvain Marinel
Received 21st May 2010, Accepted 7th October 2010
DOI: 10.1039/c0dt00519c
The ultra fast synthesis of nanocrystalline Mg2Si was carried out using microwave radiation. The
elemental precursors were first milled together under dry conditions to get fine particles. The resulting
mixture of powders of Mg and Si was cold pressed before being heated by microwave irradiation.
Precursors and products were analyzed by X-ray diffraction and scanning electron microscopy. The
high energy ball milling parameters utilized to prepare the reactive powders have quite an influence on
the behavior of the mixture under irradiation. Moreover, SEM imaging demonstrates that the power
and time of irradiation are crucial for the grain growth of the Mg2Si and must be adequately controlled
in order to avoid the decomposition of the phase. Our results show that we successfully managed to
easily and quickly synthesize homogeneous nanocrystalline Mg2Si with particle size smaller than
100 nm using a microwave power of only 175 W for two minutes on powders ball milled for two hours.
merit higher than unity whereas many efforts have been devoted
Introduction
to increasing the performance of p-type Mg2Si.3,4
Thermal to electric energy conversion is nowadays clearly iden-
Mg2Si can be synthesized by mechanical alloying, by conven-
tified as a source of alternative “clean” energy. The reduction
tional high temperature reaction or by directly processing powders
of the elements by spark plasma sintering.3,4 However, these
techniques never provide nanocrystalline materials because they
use metallic precursors whose sizes are micrometric or because
they use long periods of time of heating that are favorable to grain
growth. In contrast, microwave heating can be extremely fast and if
applied on small enough particles, this method can limit the grain
growth to a minimum. To the best of our knowledge, microwave
synthesis has never been employed to produce magnesium silicide
whereas it has been shown in many reports that it is indeed
possible to heat metals using microwave irradiation whether for
synthetic purposes or sintering applications.5,6 Therefore, using a
microwave synthetic method for the fabrication of thermoelectric
material, not only can reduce the price of the synthesis by reducing
its time but also could lead to nanostructured materials with
potentially reduced thermal conductivity and therefore improved
thermoelectric performance. Here we report on the fast synthesis
of nanocrystalline Mg2Si by fast microwave heating of high energy
ball milled powders of the elements and we discuss the possibility
offered by such a combination of techniques on material science.
of fuel consumption is a way to reduce CO2 emissions in the
automotive industry by scavenging the heat loss on exhaust
systems for instance. Evidently, the relatively high cost combined
with the limited efficiency of these thermoelectric generators is
a real drawback to their large scale development. The efficiency
of such a converter is directly proportional to the thermoelectric
figure of merit of materials constituting the device. In turn, this
figure of merit, labeled ZT, defines the thermoelectric quality of a
2
given material. This dimensionless value is equal to a T/rk, where
T represents the absolute temperature, a the seebeck coefficient,
r the electrical resistivity and k the total thermal conductivity.1
Usually, the quantities a, r, and k for conventional 3D crystalline
materials are interrelated in such a way that increasing ZT by
manipulating these variables independently is unfeasible. Whereas
a and r vary inversely with the carrier concentration, r and k
are related by the Weidemann Franz law kel = LT /r, where L
is the Lorenz constant and kel = k - klat with klat representing
the heat transported by the phonon. However, nanostructuring
of bulk materials has shown that major reduction in thermal
conductivities can be achieved by the multiplication of grain
boundaries and therefore by the creation of extensive interfacing
between the particles enhancing the phonon scattering at the grain
boundaries, evidently decreasing klat.2
Magnesium silicide is now accepted to be an extremely promis-
ing material for thermoelectric application.3 Indeed, it is made of
widely abundant, cheap, light and moreover non-toxic elements, all
critical criteria that must be addressed in order to envisage a large
scale production of thermoelectric generators for heat scavenging
applications. Furthermore, the n-type doped version of Mg2Si is
already a state of the art material,3 with a thermoelectric figure of
Experimental
Powders processing
The powders containing elemental magnesium and silicon in a 2 : 1
ratio are prepared by the high energy ball milling of magnesium
turnings (Alfa Aesar Puratronic 99.98%) and silicon lumps (Alfa
Aesar 99.9999%) both used as received. Due to the reactivity
of magnesium towards air and moisture, all manipulations were
performed under nitrogen in a glove box or sealed tubes. Six grams
of a stoichiometric amount of the pure elements were placed in
a 20 ml tungsten carbide bowl containing six 10 mm diameter
tungsten carbide balls. The grinding bowls were equipped with a
fitting lid and a silicone o-ring that insured that the assembly was
Laboratoire CRISMAT ENSICAEN UMR CNRS 6508, 6 boulevard du
Mare´chal Juin, 14050, Caen Cedex 04. E-mail: franck.gascoin@ensicaen.fr;
Fax: +33 (0) 231 951 600; Tel: +33 (0) 231 452 605
11074 | Dalton Trans., 2010, 39, 11074–11080
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
airproof so that the milling occured under a nitrogen atmosphere
(always done in order to avoid the oxidation of the reactants).
These assemblies were then placed in a high energy ball mill
(Fritsch pulverisette 7 Premium line). The powders were processed
for 2, 4, or 6 times 30 min at a speed of 450 rounds per minute
with a pause of 1 min in between periods of 30 min and this led to
batches B1, B2, and B3, respectively.
same density as elemental silicon so that the pucks resulting from
the mixture Mg/SiO2 and Mg/Si were very similar in terms of
mass and volume. Furthermore, since the hardness of Si and fused
SiO2 is very comparable, it seems reasonable to assume that the
size of the magnesium particles is about the same in both types of
puck.
After completion of the milling program, the grinding bowls
were opened in a glove box and the resulting powder was cold
pressed under 3 tons in order to obtain 500 mg “pucks”. Each puck
was subsequently inserted into a 12 mm outside diameter fused
silica tube equipped with a valve so that the microwave heating
of the puck could be conducted under nitrogen. Evidently, the
cold pressing was performed in order to facilitate the solid-state
reaction between the magnesium and the silicon particles.
Differential thermal analysis and thermogravimetry measurements
(DTA/TG)
A Netzsch thermal analysis system STA 449F3 cell equipped
with a TASC414/4 controller was used to evaluate the thermal
properties of Mg2Si powder under different atmospheres (air,
regular nitrogen, and purified argon) from room temperature to
up to 1250 K. Typically, between 12 and 20 mg of polycrystalline
sample were placed in an alumina crucible and heated under
flowing gas at a rate of 20 K min-1. After a baseline was
established, data were acquired using the software provided with
the instrument.
Microwave heating apparatus
The microwave device consisted of a microwave generator working
at 2.45 GHz (SAIREM, 20 KSM) that delivers microwave
radiation with a tunable power of up to 2 kW. The radiation was
guided along a commercial TE10 waveguide equipped with two
circulators that absorb the reflected power and therefore protect
the generator. Immediately placed after the circulators, a three-
stub tuner was used to adapt the impedance of the wave guide
with respect to the cavity and to concentrate the radiation into the
applicator. Finally, the line was ended by a TE10m rectangular
single mode microwave cavity.7
Two modes were excited, namely TE102 and TE103, the first
enabled the study of the effect of the magnetic field on the sample’s
absorption while the latter was used to probe the effect of the
electric field on the sample’s absorption (Fig. 1). In order to
assess the microwave absorption of pure silicon, a lump and ball
milled powders were tested. For elemental magnesium, due to
its malleable nature, it is impossible to ball mill it without the
addition of another element. Accordingly, pucks of a mixture of
magnesium and fused silica (SiO2) were prepared exactly like the
pucks made of elemental silicon and magnesium. Fused silica was
chosen because it is microwave transparent and it has about the
Characterization
The crystalline phases were identified by X-ray diffraction (XRD)
using Cu Ka radiation (Philips X’Pert diffractometer). In order
to evaluate the size of the particles in the precursor ball milled
mixture, scanning electron microscopy (Zeiss Supra 55 operated at
5 kV) for the cold pressed pucks and of the raw powders extracted
from the ball mill was performed. Unfortunately, the malleable
nature of the Mg makes both type of sample look alike as the
Si grains seems to be embedded in a magnesium “mush” and no
grain size analysis can be conducted on such pucks. However,
the magnesium could be dissolved in diluted nitric acid solution
and the remaining Si particles rinsed and filtered for analysis
by SEM. The microwaved powders were observed by SEM and
energy dispersive spectroscopy (EDAX-EDS) was used to check
the chemical composition.
Results and discussion
Microwave heating mode
As stated above, in our microwave apparatus, two modes were
excited, namely TE102 and TE103. Whether in the TE102 or
TE103 mode, no sign of heating could be observed on the
puck prepared with elemental magnesium and fused silica (see
Experimental) using a power of 200 W. In contrast, an elemental
silicon lump or ball milled powders could be heated to a red
hot color by using low power in the TE102 mode while no
heating could be observed using the TE103 mode. This shows
that silicon highly absorbs microwaves in an H field whereas it
is rather transparent in E mode. This behavior can be related
to the semiconducting property of silicon which favors the H
coupling.7,8 Furthermore, the fact that the puck prepared with
elemental magnesium and fused silica does not significantly absorb
microwaves (in both modes) leads us to conclude that silicon will
act as a microwave coupler in the pucks of mixed Mg and Si
powders. Consequently, in the following, it was chosen to use only
the TE102 mode for the processing of the pucks of mixed Mg and
Si powders.
Fig. 1 Electric and magnetic field amplitude profiles in the cavity
depending on the excited mode (TE102 or TE103). Arrows indicate the
ideal positioning of the sample for each mode.
This journal is
The Royal Society of Chemistry 2010
Dalton Trans., 2010, 39, 11074–11080 | 11075
©

View Article Online
Powder preparation
traces of Mg2Si. More importantly no sign of oxidation could
be observed, although an amorphous oxide layer could be present
and not detectable by way of X-ray diffraction. After removing the
magnesium from the ball milled mixture (see Experimental), the
silicon particles could be inspected by SEM. As already reported,
the Si particles are readily fragmented to smaller particle upon
increasing the milling time.4 Due to the high reactivity of fine
powders of magnesium, no other analysis could be conducted.
The cold pressed pucks of B1, B2 and B3 were then heated using
microwave irradiation in the TE102 mode. The three batches could
indeed be heated, however several differences could be observed
depending on the batch.
The use of a high energy ball mill to grind a mixture of Mg
and Si (in a ratio of 2 : 1) can be really problematic. In fact,
the same method is usually utilized to synthesize Mg2Si directly
from the milling of the corresponding elements, but because of the
difference of brittleness between the hard and easy to crush silicon
and the soft and very malleable magnesium, this synthetic method
is usually performed under wet conditions (addition of hexane in
the bowls) to prevent agglomeration of magnesium on the walls
of the bowls and on the balls and therefore to assure the good
completion of the reaction. However, a major drawback of using
wet conditions is the removal of the liquid medium in order to
recover the powdery products. Furthermore, the “wet” synthesis
complicates the preparation method, hence an increased cost and
time of production. Thus our primary target was to be able to
produce fine powders of a mixture of magnesium and silicon in
dry conditions without having the problem of agglomeration of
the powder. We find that depending on the classical parameters
associated with ball milling, namely, milling time and speed, ball
to powder ratio and diameter of the balls, different products can
be obtained. However, we repeatedly obtained fine powders and
upon opening of the grinding bowls, no sign of agglomeration
could be found using the cycling program described above (see
Experimental). The three batches B1, B2 and B3 were analyzed
by X-ray diffraction and only pure magnesium and pure silicon
could be detected for B1 (Fig. 2) and B2 while B3 also contained
Influence of the grinding time on the necessary power to initiate the
reaction
The minimum power necessary to successfully heat up B1 is 125
W, while it is 150 W and 300 W for B2 and B3, respectively.
Furthermore, the heating of B3 is clearly inhomogeneous as it
starts on the periphery of the sample and slowly propagates to
the heart of the sample. As stated above, B3 is the only mixture
that contains not only pure Mg and Si but also the alloy Mg2Si
so that the phenomena associated with the microwave heating
of such a mixture could be very complex as the behaviors of
Mg, Si and Mg2Si towards microwave radiation are certainly
different. Therefore, only the different results and observations
between B1 and B2 will be discussed hereafter. At the beginning
Fig. 2 Powder diffraction patterns of unreacted ground powders obtained by high energy ball milling (top) and of microwaved B1 pucks at t0 for powers
of 125 W, 150 W, 175 W and 200 W, respectively (from the second from the top to the bottom). Note that the diffractograms all show the characteristic
peaks of Mg2Si and traces of elemental silicon.
11076 | Dalton Trans., 2010, 39, 11074–11080
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
of the process, the compact is a mixture of silicon and magnesium
grains that are randomly distributed. At such low power and
according to our experiment (see above) only silicon seems to
absorb the microwaves, it is thus assumed that eddy currents are
individually distributed within the silicon grains. This occurs at
the very beginning of the heating process (Fig. 3) and as the
heating progresses, the absorbing Mg2Si phase is synthesized and
percolation paths are created between grains. The consequence
is that eddy currents can flow all around the cylindrical puck
(Fig. 3). Evidently, it is the outer surface of the sample that will
first be inductively heated due to the screening effect of the eddy
currents.8 The heat is then transported from grain to grain towards
the center of the sample. However, the velocity of the reaction
is such that it is merely impossible to visualize any temperature
difference between the periphery and the center of the pucks.
Since the only distinction between B1 and B2 is the grinding time,
the major difference observed upon heating should be reasonably
associated with the difference in size between the powders of the
two batches. Fig. 4 shows micrographs of silicon particles extracted
from the pucks of B1 (top) and B2 (bottom). They demonstrate
the difference in size of the Si particles present in the pucks B1
and B2. As in smaller particles less heat can be carried or stored, it
seems reasonable that for B2 (containing the smaller Si particles)
a higher microwave power is needed than for B1 to initiate the
reaction between elemental silicon and magnesium.
Fig. 4 Representative electron microprobe images of elemental Si parti-
cles extracted from ball milled B1 (top) and B2 (bottom) pucks.
that the reaction is nearly instantaneous. Moreover, according
to the X-ray diffraction patterns there is no noticeable difference
between the resulting Mg2Si powders (Fig. 2) as they all show the
antifluorite Mg2Si and traces of elemental silicon. These traces can
be attributed either to the loss of magnesium by evaporation during
the heat treatment before the reaction synthesis takes place, or to
the beginning of decomposition of Mg2Si into Si and Mg followed
by the evaporation of the magnesium. This second hypothesis is
supported by the appearance of a morphological change as the
grains appear to have coalesced together under the action of the
elevated temperature as indicated by SEM imaging (Fig. 5) of the
surface of the B1 sample obtained after a longer heat treatment
(t0 + 10 min) using a power of 200 W. The same micrograph also
shows islands of silicon-rich regions detected by EDS. This result is
also directly correlated to the thermogravimetric analysis (Fig. 6)
performed under dry inert gas that does not show any sign of
decomposition up to 850 ◦C, the temperature at which a slight
loss of mass can be observed that corresponds to the evaporation
of the magnesium. This phenomenon is more visible on the
corresponding DSC curve and evidences the decomposition of
Mg2Si around 850 ◦C under an inert atmosphere. For comparison,
the same experiment was conducted under air and a mass increase
characteristic of an oxidation starts at a temperature of only
400 ◦C.
Fig. 3 Scheme depicting the assumed eddy currents distribution within
the puck at the initial stage of the heating (t0) and once the Mg2Si phase
has been synthesized (t0 + e).
Influence of the power on the reaction
As state above, a power of 125 W is necessary to initiate the
reaction for B1. In fact it has been observed that the reaction
starts after about three minutes of irradiation, including the time
to correctly adjust the resonance. When increasing the power, this
initiation time t0 (defined as the necessary time to visualize the
incandescence of the sample) is reduced by a factor of two for
every increase of 25 W in power, thus for a power of 200 W (the
highest we used for B1) t0 is about 30 s. The total reaction time
can therefore be reduced to less than a minute. It is noteworthy
that the time necessary for the reaction between elemental silicon
and magnesium to form Mg2Si is extremely small, not to say
This journal is
The Royal Society of Chemistry 2010
Dalton Trans., 2010, 39, 11074–11080 | 11077
©

View Article Online
Fig. 5 Representative electron microprobe image of B1 powder irradiated
with a power of 200 W for 10 min. Note the silicon islands formed on the
surface of the grains and the “fused” aspect of the Mg2Si matrix.
Fig. 6 DSC signal (mV/mg) and TG (mass%) traces as a function of
temperature for Mg2Si powder under air and dry inert gas.
Influence of the time of synthesis on the grain size
As the use of a fast synthetic technique like the microwave
heating can provide a facile route to micro and/or nanostructured
materials, the time of synthesis has a real influence on the grain
size. Fig. 7 shows the evolution of the grain size of a B1 puck
irradiated with a power of 150 W at t0 (top), t0 + 5 min (center) and
t0+10 min (bottom). All photographs show that the samples are
relatively homogeneous and composed of rather spherical grains
that build up aggregates. In all cases, EDS analysis confirms the
stoichiometry of the phase (Mg2Si) and no traces of elemental
silicon or magnesium can be detected neither by X-ray diffraction
analysis nor by SEM/EDS mapping analysis. For a t0 treatment,
the grains are about 200 nm in diameter while after only five
minutes of heating they have grown to a size of between 1 and 2
microns. After 10 min under 150 W, they are even larger as their
size reaches 4 microns. Interestingly, even after 10 min of exposure
at 150 W there is no evidence of decomposition of Mg2Si in Mg
Fig. 7 Micrographs showing the evolution of the grain size of a B1 puck
irradiated with a power of 150 W at t0 (top), t0 + 5 min (center) and t0 +
10 min (bottom).
and Si, as indicated by the absence of morphological change, Si-
rich regions and coalesced grains, as is the case for the t0 + 10 min
at a power of 200 W as shown above (Fig. 5).
Towards nanostructured Mg2Si
The ultimate aim of using a fast synthesis technique is to
easily produce nanosized materials. This is especially crucial for
11078 | Dalton Trans., 2010, 39, 11074–11080
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
Fig. 8 B2 sample after t0 + 2 min of microwave irradiation at a power of 175 W.
the possible development and improvement of thermoelectric
properties through the lowering of the thermal conductivity due
to the increased number of grain boundaries in nanostructured
materials. Fig. 8 shows the nano character of a B2 sample at t0 +
2 min irradiated with a power of 175 W. As evidenced by the left
micrograph, the sample’s composition is homogeneous and it is
composed of cauliflower-like aggregates of smaller particles. The
size of these particles is smaller than 100 nm as shown on the right
picture and it seems that the grain size is very homogeneous. It
is absolutely outstanding that without optimization of neither the
grinding parameters of the elements nor the microwave irradiation
parameters we are able to obtain nanosize grains of Mg2Si in a
very short time (4 periods of 30 min of grinding and merely 2 min
of microwave heating). Forthcoming work will focus on using the
same synthetic technique on optimized composition in order to
assess the effect of the particle size on the transport properties.
We successfully synthesized aggregates of Mg2Si nanoparticles
of less than 100 nanometres. Evidently, further optimization of
the synthesis parameters is underway and might lead to even
smaller particles of Mg2Si. More important is the validation of
this synthetic process that couples high energy ball milling with
microwave irradiation for producing nanostructured materials.
This new process could be very valuable for the improvement of
the thermoelectric materials by nanostructuration, a very “hot”
topic in the thermoelectric community,9 especially when coupled
with fast densification by spark plasma sintering. Perhaps more
pragmatically, the short time and the little power needed for the
microwave preparation of Mg2Si reduces the costs of production
of a very promising and environmentally benign thermoelectric
material.
Acknowledgements
The authors would like to thank Ronan Macanjo for technical
support. The ANR SONDE (grant no. ANR-06-BLAN-0331) is
cordially acknowledged for supporting in part the cost of the ball
milling equipment. Etienne Savary thanks the French Ministry of
Research for the financial support. F.G. wishes to thank Walter
De Olivera (Fritsch) for helpful discussions concerning the ball
milling.
Conclusion
For the first time, Mg2Si has been synthesized through the
microwave irradiation of mixtures of elemental powders of Mg
and Si prepared by high energy ball milling under dry conditions.
Elemental silicon easily absorbs the microwave radiation and a
fairly low power is enough to warm it up while it is not possible
to heat magnesium powder pressed in puck at such low power,
thus, the reaction between magnesium and silicon can be initiated
by a microwave irradiation through the heating of the silicon
particles. However, the high energy ball milling parameters utilized
to prepare the reactive powders have quite an influence on the
behavior of the mixture under irradiation. With increasing the
ball milling time, the size of the silicon particles is decreased
and influences the microwave heating parameters. In fact, for
smaller particles, a higher power is necessary to initiate the reaction
between silicon and magnesium. Nonetheless, the power must be
finely controlled in order to avoid the sublimation of magnesium
References
1 H. J. Goldsmid, Thermoelectric Refrigeration, Plenum Press, New York,
1964.
2 M. S. Dresselhaus, G. Chen, M. Y. Tang, R. Yang, H. Lee, D. Wang, Z.
Ren, J.-P. Fleurial and P. Gogna, Adv. Mater., 2007, 19, 1043–1053.
3 K. Mars, H. Ihou-Mouko, G. Pont, J. Tobola and H. Scherrer, J. Electron.
Mater., 2009, 38, 1360–1364; R. B. Song, Y. Z. Liu and T. Aizawa, Phys.
Status Solidi RRL, 2007, 1, 226–228; V. K. Zaitsev, M. I. Fedorov,
E. A. Gurieva, I. S. Eremin, P. P. Konstantinov, A. Yu. Samunin
and M. V. Vedernikov, Phys. Rev. B: Condens. Matter Mater. Phys.,
2006, 74, 045207; G. S. Nolas, D. Wang and M. Beekman, Phys. Rev.
B: Condens. Matter Mater. Phys., 2007, 76, 235204; J. Tani and H.
Kido, Intermetallics, 2007, 15, 1202–1207; M. I. Fedorov, V. K. Zaitsev,
I. S. Eremin, E. A. Gurieva, A. T. Burkov, P. P. Konstantinov, M. V.
Vedernikov, A. Yu. Samunin, G. N. Isachenko and A. A. Shabaldin,
Phys. Solid State, 2006, 48, 1486–1490; R. Song, Y. Liu and T. Aizawa,
Mater. Sci. Technol., 2005, 21, 618; L. M. Zhang, Y. G. Leng, H. Y.
◦
that triggers the decomposition of Mg2Si at around 850 C. The
microwave synthesis is indeed afast technique and the grain growth
can actually also be very fast, the time of irradiation is therefore
another crucial parameter that needs to be correctly adjusted.
This journal is
The Royal Society of Chemistry 2010
Dalton Trans., 2010, 39, 11074–11080 | 11079
©

View Article Online
Jiang, L. D. Chen and T. Hirai, Mater. Sci. Eng., B, 2001, 86, 195–199;
J. M. Munoz Palos, M. D. Cristina and P. Adeva, Mater. Trans., 1996,
37, 1602.
4 R. B. Song, T. Aizawa and J. Q. Sun, Mater. Sci. Eng., B, 2007, 136,
111–117; T. Aizawa and R. Song, Intermetallics, 2006, 14, 382–391.
5 Reviews: M. P. Mingos, Adv. Mater., 2004, 5, 857; C. Gabriel, S. Gabriel,
E. H. Grant, B. S. J. Halstead and D. P. Mingos, Chem. Soc. Rev.,
1998, 27, 213; K. J. Rao, B. Valdhyanathan, M. Ganguli and P. A.
Ramakrishnan, Chem. Mater., 1999, 11, 882.
6 C. Mastrovito, J. W. Lekse and J. A. Aitken, J. Solid State Chem., 2007,
180, 3262–3270; J. W. Lekse, T. J. Stagger and J. A. Aitken, Chem. Mater.,
2007, 19, 3601–3603; J. W. Lekse, A. M. Pischera and J. A. Aitken, Mater.
Res. Bull., 2007, 42, 395; G.-T. Zhou, V. G. Pol, O. Palchik, R. Kerner,
E. Sominski, Y. Koltypin and A. Gedanken, J. Solid State Chem., 2004,
177, 361–365; O. Palchik, A. Gedanken, V. Palchik, M. A. Slifkin and
A. M. Weiss, J. Solid State Chem., 2002, 165, 125–130; Q. Xia, J. Yi, Y.
Peng, S. Luo and L. Li, Mater. Lett., 2008, 62, 4006–4008.
7 In a TE10m single mode microwave cavity, it is possible to preferentially
select the H field or the E field to irradiate the sample, depending
especially on the material absorption mechanisms. Otherwise, in such
a cavity, the required amount of power to heat the sample is very low
owing to the amplification of the electromagnetic fields provided by the
resonance. In a multimode cavity, the electromagnetic fields distribution
is averaged in the space and it is difficult to distinguish the contribution
on the heating of the H or the E fields. See for example: R. Peelamedu, R.
Roy, L. Hurtt, D. Agrawal, A. W. Fliflet, D. Lewis III and R. W. Bruce,
Mater. Chem. Phys., 2004, 88, 119–129.
8 S. Marinel and G. Desgardin, Adv. Mater., 1998, 10, 1448–
1452.
9 M. G. Kanadzidis, Chem. Mater., 2010, 2, 648–659.
11080 | Dalton Trans., 2010, 39, 11074–11080
This journal is
The Royal Society of Chemistry 2010
©