Monitoring the formation of carbide crystal phases during the thermal decomposition of 3d transition metal dicarboxylate complexes

Article abstract of DOI:10.1039/c4dt01207k

Single molecule precursors can help to simplify the synthesis of complex alloys by minimizing the amount of necessary starting reagents. However, single molecule precursors are time consuming to prepare with very few being commercially available. In this study, a simple precipitation method is used to prepare Fe, Co, and Ni fumarate and succinate complexes. These complexes were then thermally decomposed in an inert atmosphere to test their efficiency as single molecule precursors for the formation of metal carbide phases. Elevated temperature X-ray diffraction was used to identify the crystal phases produced upon decomposition of the metal dicarboxylate complexes. Thermogravimetric analysis coupled with an infrared detector was used to identify the developed gaseous decomposition products. All complexes tested showed a reduction from the starting M²⁺ oxidation state to the M⁰ oxidation state, upon decomposition. Also, each complex tested showed CO₂ and H ₂O as gaseous decomposition products. Nickel succinate, iron succinate, and iron fumarate complexes were found to form carbide phases upon decomposition. This proves that transition metal dicarboxylate salts can be employed as efficient single molecule precursors for the formation of metal carbide crystal phases.

Full text of DOI:10.1039/c4dt01207k

Dalton
Transactions
View Article Online
View Journal | View Issue
PAPER
Monitoring the formation of carbide crystal phases
during the thermal decomposition of 3d transition
metal dicarboxylate complexes†
Cite this: Dalton Trans., 2014, 43,
12236
Zachary J. Huba and Everett E. Carpenter*
Single molecule precursors can help to simplify the synthesis of complex alloys by minimizing the
amount of necessary starting reagents. However, single molecule precursors are time consuming to
prepare with very few being commercially available. In this study, a simple precipitation method is used
to prepare Fe, Co, and Ni fumarate and succinate complexes. These complexes were then thermally
decomposed in an inert atmosphere to test their efficiency as single molecule precursors for the
formation of metal carbide phases. Elevated temperature X-ray diffraction was used to identify the crystal
phases produced upon decomposition of the metal dicarboxylate complexes. Thermogravimetric analysis
coupled with an infrared detector was used to identify the developed gaseous decomposition products.
All complexes tested showed a reduction from the starting M2+ oxidation state to the M0 oxidation state,
upon decomposition. Also, each complex tested showed CO2 and H2O as gaseous decomposition
products. Nickel succinate, iron succinate, and iron fumarate complexes were found to form carbide
phases upon decomposition. This proves that transition metal dicarboxylate salts can be employed as
efficient single molecule precursors for the formation of metal carbide crystal phases.
Received 24th April 2014,
Accepted 22nd May 2014
DOI: 10.1039/c4dt01207k
www.rsc.org/dalton
available, and can be time consuming and expensive to
synthesize.2
Introduction
The metal precursors most commonly employed for the syn-
For the synthesis of metal carbides, carbonyl containing
thesis of metal nanoparticles are low oxidation state metal complexes are a commonly used single molecule precursor.6–8
salts with chloride, nitrate, sulfate, and acetate groups as Recently, it has been demonstrated that a cobalt fumarate
ligands.1 These precursors are inexpensive and offer high solu- complex can also be used as an efficient precursor for the syn-
bility in a wide variety of polar solvents. To increase solubility thesis of cobalt carbide nanoparticles.9,10 In the current
in nonpolar solvents, complexes containing long chained report, we prepare Fe, Co, and Ni fumarate and succinate com-
carboxylate ligands (e.g. metal oleate complexes) can be syn- plexes using a simple precipitation technique. The thermal
thesized and used as a metal precursor.1 By increasing the decomposition of Fe, Co, and Ni fumarate and succinate com-
chain length of the carboxylate ligand, monodispersity and plexes was then investigated in order to develop a better under-
decreased particle sizes can be achieved. Along with long standing of their decomposition process and to assess their
chained carboxylates, organic molecules containing P, S, and ability to form metal carbide structures.
Se can be used to complex metal atoms and be used as a pre-
cursor to control size and shape.2 Ligands containing C, N, P,
S, and Se can also serve to promote the formation of metal
Experimental
carbides, nitrides, phosphides, sulfides, and selenides from a
single molecular precursor.3–5 One drawback to the use of
Iron(II) Fumarate 98%, Co(NO3)2·6H2O 97.7%, Ni(NO3)2·6H2O
98%, and Na2Succinate·4H2O 99% were purchased from Alfa
Aesar, and used without further purification. FeCl2·4H2O 99%
and Na2Fumarate 99% were purchased from Sigma Aldrich Co.
and used without further purification.
single molecular precursors is that many are not commercially
Department of Chemistry, Virginia Commonwealth University, 1001 W. Main Street,
Richmond, VA 23284, USA. E-mail: ecarpenter2@vcu.edu
†Electronic supplementary information (ESI) available: FT-IR spectra and DSC
Synthesis of dicarboxylate complexes
thermograms for the dicarboxylate complexes. TGA-IR plots of Na fumarate and
Na succinate starting materials. Also, the crystallographic parameters for the
Ni phases identified in Fig. 6. See DOI: 10.1039/c4dt01207k
For the precipitation of the Fe, Co, and Ni fumarate and succinate
complexes, FeCl2·4H2O, Co(NO3)2·6H2O, and Ni(NO3)2·6H2O were
12236 | Dalton Trans., 2014, 43, 12236–12242
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
Fig. 1 XRD scans of synthesized fumarate (left) and succinate (right) precursor complexes.
used as the metal precursor. To start, 0.800 g (5.0 mmol) of temperature to 600 °C, at a ramp rate of 5 °C min−1. The TGA
Na2Fumarate or 1.350 g (5.0 mmol) of Na2Succinate·4H2O was sample chamber was purged with N2 gas throughout the ana-
added to 5 mL of DI H2O. The solution was then magnetically lysis at a flow rate of 100 mL min−1. Detection of exhaust gas
stirred and heated to 50 °C. In a separate flask, 5.0 mmol of was conducted with a Thermo Scientific Nicolet 6700 model
the desired transition metal precursor (Fe, Co, or Ni) was FT-IR. Exhaust gas was transferred to the FT-IR detector
added to 5 mL of DI H2O, and vortexed until all solids were through a stainless steel transfer tube that was heated to
dissolved. Once the sodium dicarboxylate solution was at 200 °C. The FT-IR detector employed a TGA-FTIR interface
50 °C and all solids dissolved, the transition metal solution stage equipped with a gas flow cell, also heated to 200 °C.
was added dropwise to the sodium dicarboxylate solution. The FT-IR scans from 500 cm−1 to 4000 cm−1 were collected every
mixed solution was then allowed to react for 15 minutes. In 180 seconds during the TGA analysis. This equates to a scan
this 15 minute time frame precipitation of the transition metal being collected at 15 °C intervals.
dicarboxylate complex could be observed. The solution was
then allowed to naturally cool to room temperature with con-
Complimentary analysis techniques
Analysis of ETXRD data was performed using XPert Highscore
Plus data analysis software. Crystal grain sizes were calculated
using Scherrer Analysis with instrument broadening deter-
mined using a Si wafer. Differential Scanning Calorimetry
(DSC) was conducted on a Thermal Analysis DSC Q200 (ESI†).
Scans were conducted under N2 atmosphere, from 25 °C to
600 °C at a ramp rate of 5 °C per minute. Raman spectra were
collected on a Thermo Scientific DXR Smart Raman using an
excitation wavelength of 532 nm.
tinued stirring. Once at room temperature the solution was
centrifuged at 5000 RPM for 5 minutes. After the first centrifu-
gation step, the collected metal dicarboxylate complex was
washed with a mixture of 50 : 50 ethanol and DI H2O, and cen-
trifuged. This cycle was repeated three times. After washing,
the product was allowed to dry in a vacuum oven at room
temperature for 48 hours. X-Ray diffraction patterns for the
metal dicarboxylate complexes are shown below in Fig. 1;
Fourier Transform-Infrared (FT-IR) spectra for the complexes
can be found in the ESI (S1†).
Results
Elevated temperature X-ray diffraction analysis (ETXRD)
ETXRD analysis was performed using a Panalytical X’Pert Pro
MPD X-ray diffractometer (Cu Kα, λ = 1.54 Å). Coupled with the
diffractometer was an Anton Paar HTK-1200N heating stage
with an Anton Paar TCU1000 heating control unit. All metal
dicarboxylate complexes were analyzed employing an alumina
(Al2O3) sample holder, under flowing N2 atmosphere. The
samples were analyzed at temperatures ranging from room
temperature to 900 °C. At each temperature, the sample height
was adjusted to account for thermal expansion and volume
loss due to sample decomposition.
The use of metal complexes bearing monobasic carboxylate
ligands and their decomposition behaviors has been widely
studied.11,12 However dibasic carboxylate complexes, such as
fumarate and succinate complexes, have been studied to a
lesser extent. The thermal decomposition of lanthanide,
manganese, cobalt, nickel, copper, and zinc fumarates has
been previously reported.13–17 However, most reports were con-
ducted under an air atmosphere, causing the decomposition
product formed to be oxide phases. Analysis of the gaseous
decomposition products showed CO2 and H2O to be produced
upon decomposition of the metal organic complexes.17 The
thermal decomposition of manganese fumarate and succinate
in a CO2 atmosphere resulted in the formation of manganese
Thermogravimatric analysis coupled with FT-IR detection
(TGA-IR)
TGA-IR analysis was conducted under identical conditions for oxide and carbon as the major products.16 In the present
each metal dicarboxylate complex using a Thermal Analysis study, the decomposition of Fe, Co, and Ni fumarate and
Q5000 model TGA. Thermograms were collected from room succinate complexes under an N2 atmosphere resulted in a
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 12236–12242 | 12237

View Article Online
Paper
Dalton Transactions
reduction of the M+2 to M0 particles. Also, for three of the com- sizes increased, along with the disappearance of diffraction
plexes, metal carbide phases were observed. intensities corresponding to HCP-Co. Co metal undergoes a
Cobalt fumarate and cobalt succinate showed very similar martensitic transition from HCP to FCC at temperatures
decomposition pathways. The first decomposition event was between 400 °C and 500 °C, and helps explain the loss of the
the dehydration of the tetrahydrate complexes. Cobalt succi- hexagonal close packed Co phase at high temperatures.18
nate lost 30.0% weight with an onset temperature of 82 °C.
Nickel fumarate decomposed in a similar manner to the
Cobalt fumarate lost 29.3% weight starting at 95 °C. This cobalt dicarboxylate complexes. It first showed dehydration at
dehydration was confirmed by the detection of H2O bending 107 °C, resulting in a 29.2% weight loss. Decomposition of the
vibrations in collected IR spectra of the gaseous decompo- nickel fumarate complex occurred at 324 °C and produced CO2
sition products (Fig. 2). Also, the loss of water resulted in a and H2O as the gaseous decomposition products. ETXRD
loss of crystallization in collected ETXRD scans at 200 °C. At showed FCC-Ni to be formed at 350 °C. The crystal grain size
300 °C, cobalt fumarate crystallized into an anhydrous crystal for FCC-Ni at 350 °C was calculated to be 3.5 nm. Increasing
structure, but cobalt succinate showed no significant recrystal- the temperature to 600 °C resulted in only slight growth of
lization. Increasing temperatures further, caused cobalt fuma- the FCC-Ni crystals to 7.7 nm. At temperatures above 600 °C,
rate and cobalt succinate complexes to decompose at 378 °C crystal growth occurred more rapidly, resulting in crystal grain
and 349 °C, respectively. The gaseous decomposition products sizes of 14.6 nm, 34.8 nm, and 47.2 nm at 700 °C, 800 °C, and
at these temperatures were CO2 and H2O. At 400 °C, ETXRD 900 °C, respectively (Fig. 3).
scans showed diffraction intensities consistent with metallic
The thermal decomposition of nickel succinate resulted in
Co phases (face centered cubic (FCC) and hexagonal close dramatic differences in decomposition products, when com-
packed (HCP)) for cobalt fumarate. Cobalt succinate showed pared to nickel fumarate and cobalt dicarboxylate complexes.
metallic Co phases at 350 °C. Comparing the crystalline pro- Nickel succinate underwent dehydration at 96.6 °C. At 200 °C
ducts upon decomposition of each cobalt complex, cobalt suc- and 300 °C, no significant diffraction peaks were observed in
cinate showed higher intensity and narrower peaks for the Co collected ETXRD scans. The onset of decomposition for the
phases when compared to cobalt fumarate. The average calcu- nickel succinate complex occurred at 336 °C, and resulted in 3
lated crystal grain sizes for the metallic Co phases at 400 °C crystal phases forming; FCC-Ni, HCP-Ni, and Ni3C phases were
were 3.7 nm and 11.6 nm for cobalt fumarate and cobalt succi- all observed at 350 °C. The formation of Ni3C does show that
nate, respectively. At temperatures above 400 °C crystal grain dicarboxylate ligands are capable as a carbon source. HCP-Ni
Fig. 2 ETXRD scans and TGA-IR contour plots for the decomposition of cobalt fumarate (top) and cobalt succinate (bottom).
12238 | Dalton Trans., 2014, 43, 12236–12242
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
Fig. 3 ETXRD scans and TGA-IR contour plots for the decomposition of nickel fumarate (top) and nickel succinate (bottom).
is not a thermodynamically stable crystal structure of Ni, and
Fe Succinate showed a dehydration process beginning at
is only observed under conditions with high levels of stress or 77.5 °C resulting in a 28.04% weight loss, similar to all the
strain, like those during ion bombardment.19 Hence, the for- other tetrahydrate complexes tested. ETXRD scans showed that
mation of HCP-Ni could be a strained transition state between after the dehydration process, no crystalline components were
the more stable FCC-Ni and Ni3C phases. At 400 °C only present until the onset of the decomposition process at
HCP-Ni and FCC-Ni phases were identified, implying the 277 °C. The decomposition process led to the liberation of CO2
decomposition of the Ni3C phase. This corresponds well to lit- and H2O, and the formation of FeO. FeO was the major crystal-
erature values of Ni3C decomposition occurring around line component present in ETXRD scans collected from 300 °C
400 °C.20,21 Above 400 °C, an increase in FCC-Ni composition to 450 °C. At 450 °C, peaks consistent with Fe3C could be
corresponding to a decrease in HCP-Ni phase was observed.
observed resulting in a phase composition of 24.1% Fe3C and
Iron fumarate was the only dicarboxylate complex investi- 75.9% FeO. Increasing to 500 °C caused the reduction of the
gated that started with an anhydrous hydration state. The lack FeO phase, with 30.6% BCC-Fe and 69.4% Fe3C being present.
of coordinated water molecules resulted in no observable Increasing temperatures further produced ETXRD results very
events in collected ETXRD and TGA-IR spectra from 25 °C to similar to those observed with Fe fumarate; Fe3C phase %
350 °C. With an onset temperature of 391 °C, decomposition slowly decreased at the expense of BCC-Fe, with BCC-Fe
of the Fe fumarate complex took place, resulting in a weight showing increased crystal grain sizes with increasing tempera-
loss of 48.5% and liberation of CO2. ETXRD collected at 400 °C ture. At 750 °C, a mixture of Fe3C, BCC-Fe, and FCC-Fe was
showed only weak intensities consistent with the Fe fumarate identified, while at 900 °C only FCC-Fe and Fe3C were
precursor complex. However, increasing to 450 °C resulted in observed.
the observance of 3 phases: FeO, BCC-Fe, and Fe3C. At 500 °C
In all complexes analyzed, the observed weight percentages
the FeO was completely reduced, resulting in a crystal phase at 600 °C were higher than would be predicted for the
composition of 42.3% BCC-Fe and 57.3% Fe3C. The Fe3C com- decomposition of the metal dicarboxylate complexes into pure
position further increased to 72.3% at 550 °C but dropped to metal particles. For example, the decomposition of Co fuma-
29.8% at 600 °C, with the balance phase being BCC-Fe. The rate would be predicted to result in 24% pure Co0. However,
apparent decomposition of Fe3C to BCC-Fe agrees well with a the remaining weight percent observed at 600 °C was 31.92%.
weight loss of 10.73%, beginning at 550 °C. At temperatures Raman Spectroscopy was used to investigate the possible pres-
above 600 °C, BCC-Fe transitioned to FCC-Fe (Fig. 4).
ence of surface carbon residue or amorphous metal oxides.
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 12236–12242 | 12239

View Article Online
Paper
Dalton Transactions
Fig. 4 ETXRD scans and TGA-IR contour plots for the decomposition of iron fumarate (top) and iron succinate (bottom).
Raman analysis of samples from the ETXRD study resulted in
the observance of D and G bands of carbon in collected
Raman spectra at 1355 cm−1 and 1575 cm−1, respectively. Also
a peak at 2330 cm−1 was detected in each spectrum which
corresponds to the stretching mode of N2. Most complexes
showed no presence of contaminant metal oxides; metal
oxides commonly show Raman active modes between
500 cm−1 to 800 cm−1 22 The presence of the D and G bands of
.
carbon shows a significant amount of carbon residue to exist,
helping to explain the higher than expected weights resulting
from decomposition of the dicarboxylate complexes. D and G
band ratios were around 1 for each product tested and implies
the resulting carbon structure to be poorly ordered.23 However,
the presence of carbon does substantiate that decomposition
of the fumarate and succinate ligands to produce free carbon
atoms (Fig. 5).
Fig. 5 Raman spectra for the metal dicarboxylate complexes heated to
900 °C.
C–C, or CvC stretch was observed for any complex tested,
implying that the central ethane and ethylene moieties of the
succinate and fumarate ligands are oxidized to CO2, and serve
as a possible reduction source for the metals. Also, ethylene
and ethane are commonly used as gaseous carbon sources for
Discussion
From the ETXRD and Raman data, it is shown that the metal catalyzed production of ordered carbon structures.24
thermal decomposition of all of metal dicarboxylate complexes Hence, the decomposition of the fumarate and succinate
tested cause a reduction of the M2+ atoms to M0 and led to the ligands supplies a reduction and carbon source for the for-
formation of amorphous carbon. During the decomposition/ mation of metal carbide structures. However, only three of the
reduction process, CO2 was the major gaseous product complexes investigated formed metal carbide crystal phases
observed in TGA-IR studies. No significant signal for C–H, (Fe fumarate, Fe succinate, and Ni succinate).
12240 | Dalton Trans., 2014, 43, 12236–12242
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
Fig. 6 (a) XRD scan taken during the decomposition of Ni succinate at
350 °C. Fits and phase percentages are shown for FCC-Ni, HCP-Ni, and
Ni3C. Lattice parameters, space groups, and Wyckoff positions for each
phase can be found in the ESI† (b) Average crystal grain sizes as a func-
tion of temperature for phases produced from the decomposition of Ni
fumarate and Ni succinate.
Fig. 7 (a) XRD scan of Fe3C samples as a result of decomposition of Fe
dicarboxylate complexes. (b) Magnetization vs. applied field curve for
phase pure Fe3C.
Both Fe complexes formed carbide phases upon decompo-
sition. However, with increasing temperature the carbide
phases decomposed to BCC-Fe. During the heating process, no
pure phase Fe3C was observed. Now to be able to label Fe
fumarate or succinate as a true single molecule precursor for
FexC phases, a pure phase carbide should be attained. At
temperatures above 600 °C, the BCC-Fe phase transitioned to
FCC-Fe. FCC-Fe possesses carbon solubility values that are sig-
nificantly higher than BCC-Fe (on the order of 10 times
higher).27 This difference in carbon solubility of the two iron
crystal phases was exploited to achieve the pure phase Fe3C. By
heating Fe fumarate quickly to 900 °C to form FCC-Fe and
then cooling at a rate of 100 °C min−1, phase pure Fe3C can be
formed (Fig. 7). The phase pure Fe3C was ferromagnetic in
nature at room temperature, possessing a magnetic saturation
value of 124 emu g−1. This magnetization is lower than the
The decomposition of the cobalt complexes showed no crys-
talline carbide phases. The lack of carbide formation for the
tested cobalt complexes can be explained by looking at the
decomposition temperatures for the stable Co carbide phases:
Co3C and Co2C. Co3C decomposes at 315 °C and Co2C decom-
poses at 300 °C.25 Both cobalt complexes decomposed well
above these temperatures, and were energetically favored to
form metallic cobalt crystal structures.
While none of the cobalt dicarboxylate complexes decom-
posed to a carbide phase, Ni succinate formed 15.8% Ni3C at
350 °C. Ni3C has a higher decomposition temperature than
the CoxC phases, falling between 400 °C and 425 °C.20,21 Ni
succinate and Ni fumarate decomposed at 336 °C and 324 °C,
respectively. However only Ni succinate formed Ni3C and Ni
fumarate did not.
Analysis of ETXRD scans for the decomposition of both Ni
complexes showed dramatic differences in average crystal
grain sizes for the crystalline decomposition products of each
Ni complex. The crystals phases formed as a result of the
decomposition of Ni Fumarate showed significantly smaller
crystal grain sizes than Ni Succinate (Fig. 6(a)). Carbon solubi-
lity in transition metal crystal structures is highly dependent
on the crystal grain size; decreasing grain sizes typically show a
decrease in carbon solubility.26 Hence, lower carbon solubility
theoretical maximum magnetization of 143 emu g−1 28,29
This
.
deviation in magnetization can be attributed to the excess
carbon as observed in collected Raman spectra. The successful
synthesis of phase pure Fe3C shows that late row 3d transition
metal dicarboxylates can act as a single molecule precursor for
the synthesis of pure phase metal carbide crystal phases under
appropriate synthesis conditions.
Conclusion
due to small grain sizes for the Ni crystal phase formed upon The purpose of this investigation was to examine the ability of
decomposition of Ni fumarate is one possible explanation for
its inability to form the Ni3C phase.
Fe, Co, and Ni fumarate and succinate complexes to act as
single molecule precursors for the synthesis of endothermic
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 12236–12242 | 12241

View Article Online
Paper
Dalton Transactions
carbide phases. In all cases, the decomposition of the metal
dicarboxylate complex resulted in a reduction of the M2+ pre-
cursor to an M0 state. In three of the complexes, metal carbide
phases were observed. Nickel succinate decomposed into a
8 C. Yang, H. B. Zhao, Y. L. Hou and D. Ma, J. Am. Chem.
Soc., 2012, 134, 15814–15821.
9 A. A. El-Gendy, T. Almugaiteeb and E. E. Carpenter,
J. Magn. Magn. Mater., 2013, 348, 136–139.
mixture of FCC-Ni, HCP-Ni, and Ni3C at 350 °C. Fe succinate 10 A. A. El-Gendy, M. Qian, Z. J. Huba, S. N. Khanna and
and Fe fumarate formed Fe3C at 500 °C and 550 °C, but E. E. Carpenter, Appl. Phys. Lett., 2014, 104, 023111.
decomposed to BCC-Fe at higher temperatures. At 550 °C, 11 J. Mu and D. D. Perlmutter, Thermochim, Acta, 1981, 49,
both iron complexes formed between 60–70% Fe3C, with Fe3C 207–218.
grain sizes of 107 nm for Fe fumarate and 40 nm for Fe succi- 12 M. R. Buck, A. J. Biacchi and R. E. Schaak, Chem. Mater.,
nate. For all complexes, the gaseous decomposition products 2014, 26, 1492–1499.
were CO2 and H2O. This report shows that dicarboxylate 13 B. S. Randhawa and M. Kaur, J. Therm. Anal. Calorim.,
ligands can serve as carbon sources for the synthesis of
endothermic 3d metal carbide crystal phases.
2007, 89, 251–255.
14 P. S. Bassi, B. S. Randhawa, C. M. Khajuria and S. Kaur,
J. Therm. Anal., 1987, 32, 569–577.
15 Y. Suzuki, Thermochim. Acta, 1995, 255, 155–170.
16 Y. Suzuki, K. Muraishi and H. Ito, Thermochim. Acta, 1995,
258, 231–241.
Acknowledgements
Financial support from the ARPA-e REACT program 1574-1674 17 E. Y. Ionashiro, F. J. Caires, A. B. Siqueira, L. S. Lima and
is greatly acknowledged. The help of the staff of the VCU Nano-
C. T. Carvalho, J. Therm. Anal. Calorim., 2012, 108, 1183–
materials Core Characterization Facility is also greatly appreci-
1188.
ated. This work was also partially funded through a graduate 18 F. Frey, W. Prandl, J. Schneider, C. Zeyen and K. Ziebeck,
research fellowship through Altria Group Inc.
J. Phys. F-Metal Phys., 1979, 9, 603–616.
19 E. Johnson, T. Wohlenberg and W. A. Grant, Phase
Transitions, 1979, 1, 23–33.
20 Y. Leng, L. Xie, F. Liao, J. Zheng and X. Li, Thermochim.
Acta, 2008, 473, 14–18.
Notes and references
1 M. A. Willard, L. K. Kurihara, E. E. Carpenter, S. Calvin and 21 Z. L. Schaefer, K. M. Weeber, R. Misra, P. Schiffer and
V. G. Harris, Int. Mater. Rev., 2004, 49, 125–170. R. E. Schaak, Chem. Mater., 2011, 23, 2475–2480.
2 C. Amiens, B. Chaudret, D. Ciuculescu, V. Colliere, 22 C. W. Tang, C. B. Wang and S. H. Chien, Thermochim. Acta,
K. Fajerwerg, P. Fau, M. Kahn, A. Maisonnat, K. Soulantica
and K. Philippot, New J. Chem., 2013, 37, 3374–3401.
3 J. Park, B. Koo, K. Y. Yoon, Y. Hwang, M. Kang, J. G. Park
and T. Hyeon, J. Am. Chem. Soc., 2005, 127, 8433–8440.
2008, 473, 68–73.
23 A. C. Ferrari and J. Robertson, Phys. Rev. B: Condens.
Matter, 2000, 61, 14095–14107.
24 C. Park and R. T. K. Baker, J. Catal., 1998, 179, 361–374.
4 W. Maneeprakorn, M. A. Malik and P. O’Brien, J. Mater. 25 K. J. Carroll, Z. J. Huba, S. R. Spurgeon, M. Qian,
Chem., 2010, 20, 2329–2335.
S. N. Khanna, D. M. Hudgins, M. L. Taheri and
E. E. Carpenter, Appl. Phys. Lett., 2012, 101, 012409.
26 P. A. Chernavskii, V. I. Zaikovskii and G. V. Pankina,
Russ. J. Phys. Chem. A, 2012, 86, 1274–1280.
5 W. Maneeprakorn, C. Q. Nguyen, M. A. Malik, P. O’Brien
and J. Raftery, Dalton Trans., 2009, 2103–2108.
6 A. Meffre, B. Mehdaoui, V. Kelsen, P. F. Fazzini, J. Carrey,
S. Lachaize, M. Respaud and B. Chaudret, Nano Lett., 2012, 27 D. E. Jiang and E. A. Carter, Phys. Rev. B: Condens. Matter,
12, 4722–4728. 2003, 67.
7 M. D. Shultz, S. Calvin, F. Gonzalez-Jimenez, V. Mujica, 28 J. M. D. Coey, Magnetism and Magnetic Materials, Cam-
B. C. Alleluia and E. E. Carpenter, Chem. Mater., 2009, 21,
5594–5600.
bridge University Press, 2010.
29 J. M. D. Coey, IEEE Trans. Magn., 2011, 47, 4671–4681.
12242 | Dalton Trans., 2014, 43, 12236–12242
This journal is © The Royal Society of Chemistry 2014