Rapid microwave-assisted sol-gel preparation of Pd-substituted LnFeO 3 (Ln = Y, La): Phase formation and catalytic activity

Article abstract of DOI:10.1039/c3dt52404c

We present a rapid microwave-assisted sol-gel approach to Pd-substituted LnFeO₃ (Ln = Y, La) for applications in C-C coupling reactions. These materials could be prepared in household microwave ovens in less than 15 minutes of reaction time with the final materials displaying well-defined structure and morphology. Phase evolution was studied using time-dependent microwave heatings and then compared with the results obtained from thermogravimetric analyses. Materials were confirmed to be phase pure by laboratory and synchrotron X-ray diffraction. Substituted Pd is ionic as shown by the binding energy shift from X-ray photoelectron spectroscopy. The short heating periods required for phase purity allow these materials less time for sintering as compared to conventional solid state preparation methods, making relatively high surface areas achievable. These materials have been successfully used as catalyst precursor materials for C-C coupling reactions in which the active species is Pd⁰. Pd-substituted LnFeO₃ (Ln = Y, La) provides Pd⁰ in solution which can be complexed by the ligand SPhos, allowing for aryl chloride coupling.

Full text of DOI:10.1039/c3dt52404c

Dalton
Transactions
View Article Online
View Journal | View Issue
PAPER
Rapid microwave-assisted sol–gel preparation of
Pd-substituted LnFeO3 (Ln = Y, La): phase
formation and catalytic activity†
Cite this: Dalton Trans., 2014, 43,
2079
Lauren M. Misch, Alexander Birkel, C. Adrian Figg, Brett P. Fors, Craig J. Hawker,
Galen D. Stucky and Ram Seshadri
We present a rapid microwave-assisted sol–gel approach to Pd-substituted LnFeO3 (Ln = Y, La) for appli-
cations in C–C coupling reactions. These materials could be prepared in household microwave ovens in
less than 15 minutes of reaction time with the final materials displaying well-defined structure and mor-
phology. Phase evolution was studied using time-dependent microwave heatings and then compared
with the results obtained from thermogravimetric analyses. Materials were confirmed to be phase pure by
laboratory and synchrotron X-ray diffraction. Substituted Pd is ionic as shown by the binding energy shift
from X-ray photoelectron spectroscopy. The short heating periods required for phase purity allow these
materials less time for sintering as compared to conventional solid state preparation methods, making
relatively high surface areas achievable. These materials have been successfully used as catalyst precursor
materials for C–C coupling reactions in which the active species is Pd0. Pd-substituted LnFeO3 (Ln = Y,
La) provides Pd0 in solution which can be complexed by the ligand SPhos, allowing for aryl chloride
coupling.
Received 2nd September 2013,
Accepted 3rd November 2013
DOI: 10.1039/c3dt52404c
www.rsc.org/dalton
Perovskites have long been studied for catalytic purposes.
First reports of perovskites as automotive emissions catalysts
1. Introduction
Ionic noble metal catalysts continue to receive attention date back to the 1970s.9,10 The utility of these materials lies in
because they can provide a means to achieve enhanced reactiv- their ability to be reduced and subsequently re-oxidized mul-
ity. Substituting noble metals into oxide hosts is an attractive tiple times. Structurally, the large A site, usually occupied by
alternative to supporting noble metal nanoparticles on oxides, an alkaline earth or rare earth ion, and corner-sharing B site
with substitution allowing the use of very small amounts of octahedra make the perovskite amenable to ion substitution
expensive, precious metals while still accessing their reactivity on both cation sites. For this reason, perovskites can be
for CO oxidation, NOx reduction, and hydrocarbon conver- prepared with various substituents, notably noble metals.11
sion.1 Hegde et al. have shown that Pd-substituted binary Although primarily studied as catalysts for CO oxidation and
oxides are superior CO oxidation catalysts to the traditional NOx reduction, these materials have also been considered
metal nanoparticle catalyst on an oxide support.2,3 Recently, as catalysts for coupling reactions in organic chemistry.12
our group has reported that Pd-substituted CeO2 can be used Although LaFeO3 and YFeO3 have very similar crystal struc-
as a methane partial oxidation catalyst, and determined that tures, the significant size difference of the A site cation results
CeO2 as a host oxide is not able to maintain the stability in the much smaller Y3+ creating significantly tilted and
of ionic Pd under strongly reducing reaction conditions.4 rotated octahedra in YFeO3. This subtle difference is shown in
However, it has been previously demonstrated that Pd ions can Fig. 1. It was of interest to understand whether these differ-
be substituted into pervoskites and stabilized in complex ences in the structure would result in distinct behavior.
oxides.5–8 With this in mind, we now consider Pd-substitution
in perovskite hosts.
We have prepared Pd-substituted (La,Y)FeO3 in order to
better understand the contribution of inductive effects from
the A site cations and the role this plays in dopant ion stabiliz-
ation. Pd compounds are of particular interest as catalysts for
cross-coupling reactions (i.e. Suzuki–Miyaura,13 Negishi,14
Department of Chemistry and Biochemistry, Materials Department, and Materials
Research Laboratory University of California, Santa Barbara, CA 93106, USA
†Electronic supplementary information (ESI) available. See DOI:
10.1039/c3dt52404c
Stille,15
Sonogashira,16
Buchwald–Hartwig,17
Heck,18
Hiyama,19 etc.). More specifically, Pd-substituted perovskites
have the ability to provide Pd0 under reducing conditions, as
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 2079–2087 | 2079

View Article Online
Paper
Dalton Transactions
unambiguously through the use of the ligand 2-dicyclohexyl-
phosphine-2′,6′-dimethoxybiphenyl (SPhos) that any catalytic
activity does not stem from the perovskite itself, but from the
reduced Pd0 that is released and then bound by the ligand. We
have demonstrated that these materials are appropriate for the
coupling of aryl chlorides under more mild conditions than
has previously been reported.
2. Experimental methods
Sample preparation
All samples were prepared via a microwave-assisted reaction
pathway, employing a citric acid based sol–gel reaction.32 In a
typical synthesis, to 15 mL of MilliPore water (18 MΩ cm) and
5 mL ethanol (in a 50 mL beaker), a two-fold excess (in terms
of metal cations) of citric acid was added and the solution was
stirred until clear, which usually took about 5 minutes. Sub-
sequently, the pH was adjusted to neutral through the addition
of several drops of aqueous ammonia.
For the preparation of Ln1−xPdxFeO3−δ (with Ln = La, Y),
stoichiometric amounts of the metal nitrates Ln(NO3)3·6H2O,
Fe(III) acetylacetonate and Pd(II) acetylacetonate (all obtained
from Sigma-Aldrich) were dissolved and the solution was vigor-
ously stirred. After the addition of either La(NO3)3·6H2O or
Y(NO3)3·6H2O, a milky precipitate formed. The other reagents
were added after the dissolution of this precipitation. In the
case of Y1−xPdxFeO3−δ, a small excess (between 5 mol% and
10 mol%) of Fe(III) acetylacetonate had to be used to limit the
amount of Y2O3 impurities. The temperature of the reaction
mixture was held at 65 °C to slowly evaporate the solvent until
a gel was formed, which usually happened between 8 h and
10 h. This gel is then dried at 125 °C overnight and eventually
finely ground. A batch size of about 250 mg of dried gel is
then heated in a so-called hybrid microwave (Panasonic
NN-SN667B, 1250 W) setup, as described in previous
studies.28,29,38
Fig. 1 Crystal structures for (a) LaFeO3 and (b) YFeO3. The A site cations
are displayed as grey spheres and FeO6 octahedra in blue.
well as to regenerate with Pd cations formed in an oxidative
environment, thus providing
a
fully recyclable Pd
source.11,20,21 Previous success with Pd-perovskites as recycl-
able catalysts12 in Suzuki–Miyaura reactions has further
prompted our study of Pd-substituted (La,Y)FeO3 for these
reactions.22,23 Here we utilize the Suzuki–Miyaura reaction to
compare and understand the catalytic activity of materials with
various substitution levels of Pd (0, 5, 10 mol%), differing A
site cations and varying surface areas, which can all be easily
controlled using our rapid and facile microwave preparation
method.
One very interesting approach to rapidly prepare uniform
and phase pure materials is the use of microwave-assisted
reactions.24–27 Recently, it has been demonstrated that with
simple modification and optimization, household microwave
ovens can be employed to obtain high quality materials that
enable various types of reactions, including conventional
solid-state reactions to yield phosphors with high quantum
yields,28,29 the heating of air-sensitive materials in evacuated
fused silica tubes to access intermetallics for thermo-
electrics30,31 and the use of sol–gel based preparations.32
Along with providing stable, stoichiometric, complex com-
pounds, the microwave preparation routes are also appropriate
for preparing complex substituted oxides, in which a dopant
ion resides on a metal cation site.
In the present study we report a rapid microwave-assisted
combustion/sol–gel preparation as a method to produce noble
metal-substituted perovskites. Unsubstituted materials as well
as various synthetic pathways leading to them have been
investigated previously.6,33–37 The specific approach followed
here allows for the careful investigation of the phase formation
process due to the very short and controllable heating times.
We have applied the materials prepared using microwave-
assisted techniques to Suzuki–Miyaura coupling reactions and
determined that the substituted perovskites are actually
precursor catalysts and provide a Pd source for coupling cata-
lysis. The La containing pervoskites are more active than the
Many iterations with various heating profiles and power
levels were attempted. The reaction time and power level were
adjusted to optimize the crystallinity, phase purity, and surface
area of the investigated samples. A time-dependent study of
the phase formation was carried out at various reaction times,
using a power setting of 100%, 1250 W. The optimized reaction
time was 150 seconds at a power level setting of 70% for
La1−xPdxFeO3−δ and a multi-step heating profile (2 minutes at
a power setting of 90%, followed by several minutes at lower
power settings) for Y1−xPdxFeO3−δ
. The temperature was
measured in various intervals during and after the reaction
was finished using a portable infrared pyrometer (PalmerWahl
DHS235XEL).
Sample characterization
Laboratory powder X-ray diffraction (XRD) data were obtained
using Cu Kα radiation (Philips X’Pert) over the angular range
15° ≤ 2θ ≤ 90° with a step size of 0.016°. High resolution
synchrotron powder diffraction data were collected using
Y
containing perovskites, and it has been established
2080 | Dalton Trans., 2014, 43, 2079–2087
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
beamline 11-BM at the Advanced Photon Source (APS), with a Thermo-Nicolet Avatar-330 IR spectrometer with a
Argonne National Laboratory using an average wavelength of single-bounce attenuated total reflection (ATR) accessory with
0.413893 Å. Discrete detectors covering an angular range from a Ge crystal. Gas chromatography (GC) analyses were per-
−6° to 16° 2θ were scanned over a 34° 2θ range, with data formed on a Shimadzu GC-2014 instrument with a flame
points collected every 0.001° 2θ at a scan speed of 0.01° s−1
.
ionization detector (FID) using a Restek SHRXI-5MS column.
Rietveld fits39 of the data were obtained using the TOPAS Aca- High-resolution mass spectrometry was conducted with a
demic program suite.40 TGA was carried out using a METTLER Micromass QTOF2 Quadrupole/Time-of-Flight Tandem mass
TGA/sDTA851e ThermoGravimetric Analyzer in air in the temp- spectrometer. Flash chromatography was performed using a
erature range between 25 °C and 900 °C at a heating rate of Biotage SP4 instrument with prepacked silica cartridges.
10 °C per minute. X-ray photoelectron spectra were obtained
General procedure for a Suzuki–Miyaura reaction with
on a Kratos Axis Ultra Spectrometer with a monochromatic different (La,Y)FeO perovskites. An oven-dried test tube,
Al-Kα source (E = 1486.61 eV). Samples were mounted on a which was equipped with a magnetic stir bar and fitted with a
stainless steel holder using double-sided carbon tape. The teflon septum, was charged with SPhos (0–1 mol%), the per-
residual pressure inside the sample analysis chamber was ovskite (0–0.04 mol% Pd), phenylboronic acid (171 mg,
below 7 × 10−9 Torr. Survey spectra were collected with an ana- 1.4 mmol) and K2CO3 (193 mg, 1.4 mmol). The vessel was
lyzer pass energy of 80 eV and high-resolution Pd 3d spectra evacuated and backfilled with argon (this process was repeated
were acquired at a pass energy of 20 eV. Spectra were calibrated a total of 3 times) and then the 3-chloropyridine (95 μL,
to the C 1s peak from adventitious hydrocarbons, expected at a 1.0 mmol) and the degassed solvent mixture (1 : 1 i-PrOH–
binding energy of 285.0 eV. For peak fitting of the spin–orbit H2O, 2 mL) were added via a syringe. The solution was heated
doublets in high resolution scans, the d3/2 to d5/2 peak area at 80 °C for 20 h and then cooled to room temperature, diluted
was constrained to a ratio of 2/3 with the use of CasaXPS soft- with ethyl acetate (20 mL) and washed with water (5 mL).
ware. Field-emission scanning electron microscopy was per- Decane was added to the solution as an internal standard and
formed on a FEI XL40 Sirion FEG microscope with an Oxford the yield was determined by GC analysis.
Inca X-ray system attached for chemical analysis. SEM samples
General procedure for a Suzuki–Miyaura reaction with
were mounted on aluminum stubs using double-sided conduc- LaFe0.95Pd0.05O3−δ. An oven-dried test tube, which was
tive carbon tape and coated with a thin layer of Au/Pd in order equipped with a magnetic stir bar and fitted with a teflon
to avoid charging effects. The images were recorded with an septum, was charged with SPhos (4.1 mg,
1 mol%),
acceleration voltage of 5 kV. Energy-dispersive X-ray spectro- LaFe0.95Pd0.05O3−δ (2 mg, 0.04 mol% Pd), aryl boronic acid
scopy was performed using an acceleration voltage of 20 kV. (1.4 mmol) and K2CO3 (193 mg, 1.4 mmol). The vessel was
The samples were not sputtered with conductive coatings. evacuated and backfilled with argon (this process was repeated
Transmission electron microscopy was performed on a FEI a total of 3 times) and then the aryl halide (1.0 mmol) and the
Tecnai G2 Sphera Microscope with an acceleration voltage of degassed solvent mixture (1 : 1 i-PrOH–H2O, 2 mL) were added
200 kV. TEM samples were prepared by dispersing a small via a syringe. The solution was heated at 80 °C for 20 h and
amount of the powder in ethanol and subsequent ultrasonica- then cooled to room temperature, diluted with ethyl acetate
tion for about 15 minutes. Then, 2 to 3 drops were adminis- (20 mL), washed with water (5 mL) and concentrated in vacuo.
tered onto copper grids (lacey carbon, Type-A, Ted Pella) and The crude product was purified using a Biotage SP4 flash
the solvent was allowed to evaporate overnight. Brunauer– purification system.
Emmett–Teller (BET) surface area measurements were per-
formed on a MicroMeritics TriStar 3000 porosimeter using N2
as a probe gas.
3. Results and discussion
Catalytic activity
Evolution of perovskite products
All reactions were carried out under an argon atmosphere. The Thermogravimetric analysis including differential thermal
isopropanol was purchased from Fischer Scientific and was analysis of (La,Y)Fe0.95Pd0.05O3−δ gels was conducted after the
degassed before use (three freeze–pump–thaw cycles). Aryl drying step at 125° and before the sample was subjected to any
halides, aryl boronic acids, SPhos and potassium carbonate microwave heating (Fig. 2). This offers insight into product for-
were purchased from Sigma-Aldrich and were used as received. mation during the microwave heating steps with two signifi-
All compounds were characterized by 1H NMR, 13C NMR, IR cant heat losses being observed during the TGA analysis. The
spectroscopy, and high-resolution mass spectrometry. Nuclear first occurs between room temperature and 200 °C, when the
Magnetic Resonance (NMR) spectra were recorded on a Varian samples loses about 10% of their respective masses, due to
600 MHz instrument. 1H NMR experiments are reported in δ evaporation of re-adsorbed water and surface hydroxyl groups
units, parts per million (ppm), and were measured relative to as well as thermal decomposition of ammonium nitrate. The
the signals for residual chloroform (7.26 ppm) in the deuter- second major weight loss occurs in the temperature range
ated solvent. 13C NMR spectra are reported in δ units, ppm, between 200 °C and 400–500 °C. Here, the samples lose
relative to deuterochloroform (77.23 ppm) and all were another 70% of their mass. As described in the Experimental
obtained with 1H decoupling. FT-IR spectra were obtained section, the gels contain a two-fold excess (in terms of total
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 2079–2087 | 2081

View Article Online
Paper
Dalton Transactions
Fig. 3 Phase evolution of LaFe0.95Pd0.05O3−δ after 30 s intervals of the
reaction, using laboratory X-ray powder diffraction data. The indicated
temperatures were measured using the infrared pyrometer, aimed at the
center of the inner crucible, immediately after the microwave irradiation
has stopped. Formation begins after 650 °C and becomes more crystal-
line at higher temperatures.
Fig. 2 Thermogravimetric analysis of the decomposition of the dried
precursor gels. Panel (a) shows the decomposition for LaFe0.95Pd0.05O3
and panel (b) shows YFe0.95Pd0.05O3−δ. A 10% mass loss is observed
−δ
between room temperature and 200 °C which can be attributed to
water and surface hydroxyls while a much larger weight loss is observed
between 200 °C and 500 °C as a result of the citric acid combustion and
burning of residual organics.
pursue sol–gel preparation because we were able to achieve
metal cations) of citric acid which combusts at temperatures more phase pure samples and higher surface areas than with
exceeding 200 °C. The DTA for both samples indicates two the ceramic preparation.
sharp exothermic events at this temperature. The combustion
Here we make use of the short heating and cooling times
process is completed at about 500 °C, and only a small weight facilitated by the microwave and we focus on the evolution of
loss after 500 °C is observed, which is mostly due to the reactants in the formation of LaFe0.95Pd0.05O3−δ. Fig. 3 shows
burning of residual organics. Our findings agree well with the the laboratory X-ray diffraction data of different stages of the
results found by Qi et al.35 who prepared LaFeO3 using com- reaction, starting from the dried gel until 2 minutes, with 30
bustion synthesis. Conventional solid-state preparation of second steps. In both, the dried gel before subjecting to micro-
these materials usually starts from a mixture of oxides, gener- wave radiation as well as the sample being subjected to micro-
ally requires temperatures of 1000 °C and above, involves wave radiation for 30 s (a temperature of about 250 °C was
multiple regrinding steps in between multi-day heatings and measured after that time with fume/smoke starting to evolve
yields very low surface area products. Even most sol–gel from the specimen), weak reflections of NH4NO3 can be
based preparation methods require several hours of heating at found. These originate from the reaction of the nitrate precur-
temperatures well above 500 °C, proving again the time- and sor with the aqueous ammonia solution used to adjust the pH
energy-saving nature of this microwave-assisted reaction of the precursor solution.
pathway.
After 60 s of microwave irradiation (and a temperature of
Taking advantage of the short heating and reaction times, 475 °C), the NH4NO3 decomposed. Also, the sample’s color
the microwave-assisted sol–gel (or combustion) preparation changed from light brown to black with the evolution of
allows us to study the phase formation process of the perov- additional smoke. At this point of the reaction, the X-ray data
skites in more detail. It should be noted that although the showed only an amorphous material, consisting of the citrate
microwave heating times are very short, the sol–gel preparation complexes of the dissolved precursors. After another 30
that precedes microwave heating is completed over the course seconds (90 s total heating time), a temperature of about
of two full days. The shortened heating time allowed by the 650 °C was reached and the X-ray diffraction data showed that
microwave does significantly reduce the total time usually the phase formation is completed, as only reflections of
required for a sol–gel preparation, in which the final heating LaFe0.95Pd0.05O3−δ are visible in the trace. These reflections are
step may take up to several hours or days. The microwave- very broad and of rather low intensity, pointing to the fact that
assisted heating method can be applied to a more convention- small crystallites have been formed. Crystallite size analysis
al solid-state preparation in which precursor oxides can be using TOPAS Academic showed an average size of about
ground with a mortar and pestle in a matter of minutes fol- 30 nm. At this point, the sample body color changed from
lowed by very short microwave heating times. We chose to black to brown and the sample exhibited some fluffiness.
2082 | Dalton Trans., 2014, 43, 2079–2087
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
Fig. 4 Synchrotron X-ray powder diffraction data of (a) LaFe1−xPdxO3−δ
and (b) YFe1−xPdxO3−δ with various amounts of Pd substituted into the
host (various x values).
Further heating increases the temperature to about 800 °C,
leading to a larger crystallite size (about 75 nm) and therefore
sharper reflections. These findings correspond very well with
the thermal decomposition behavior observed from the TGA
data.
Laboratory and synchrotron X-ray powder diffraction data
were collected to confirm the phase purity of all the investi-
gated materials. Rietveld fits of high-resolution synchrotron
powder X-ray diffraction data collected at beam line 11-BM
confirm that especially the substituted YFeO3 samples contain
at least some detectable crystalline metallic palladium (Pd0),
as summarized in Fig. 4. Although the amount estimated from
a quantitative Rietveld fit does not exceed 2.5%, it is worth
Fig. 5 Scanning electron micrographs of (a) LaFe0.95Pd0.05O3−δ and (b)
YFe0.95Pd0.05O3−δ. Representative samples prepared using the optimized
heating profiles show 100 nm crystallites agglomerated together.
considering this when interpreting catalytic reactivity results. applications. High surface area materials allow access to many
In addition, LaO(OH) can be found in small quantities (less active sites at once and are more efficient in catalytic conver-
than 5 wt%) as a secondary phase in the LaFe0.9Pd0.1O3−δ sion of reactants to products than the same materials having
perovskites. We did observe some small changes in lattice lower surface area. With this in mind, the microwave heating
parameter with various levels of Pd-substitution. However, no profiles were adjusted through duration and power level to
significant trend, such as a steady increase or decrease in the ensure that the materials possessed the highest attainable
cell lengths as a function of increasing Pd content, was surface area with this method. Materials that were heated for
observed. This is not entirely unexpected, as this is an aliova- longer times and at higher power levels were subjected to sin-
lent substitution. As Pd2+ substitutes onto a 3+ site, an oxygen tering and showed fewer small particles and larger chunks.
vacancy is subsequently created for neutrality. This was This phenomenon is directly related to surface area. Energy
observed previously with Pd-substitution in YFeO3.6
dispersive X-ray spectroscopy and mapping (shown in the
Scanning electron micrographs, as presented in Fig. 5, for ESI†) confirmed increasing Pd content with Pd concentration
typical 5% Pd-substituted LaFeO3 and YFeO3 samples show during preparation and a homogeneous dispersion of Pd ions
that both substituted materials exhibit very similar in as-prepared materials. Upon closer inspection with trans-
morphologies. Near spherical particles that are 100 nm and mission electron microscopy (Fig. 6), crystallinity anticipated
smaller in size were observed and these agglomerate to form from the X-ray diffraction data was observed. Similar to obser-
larger chunks. The micrographs shown here are representative vations with the scanning electron micrographs, materials sub-
of all perovskite materials prepared using the optimized jected to more strenuous heating profiles were not comprised
heating profile, detailed in the Experimental section. Surface of small enough particles to observe uniform crystallinity in
area is
a critical parameter for materials for catalytic the TEM.
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 2079–2087 | 2083

View Article Online
Paper
Dalton Transactions
Fig. 7 X-ray photoelectron spectroscopy data of (a) LaFe0.95Pd0.05O3−δ
and (b) YFe0.95Pd0.05O3−δ. There are two contributions observed in the
case of the La perovskite and only one contribution observed for the
Y perovskite. Both materials show the Pd 3d peaks shifted to higher
binding energy than metallic Pd or PdO. Dashed reference lines are
taken from ref. 41.
scans did not show evidence for any elements aside from the
anticipated La/Y, Fe, Pd, O, and C. Quantification of the survey
scans using CasaXPS software indicated that materials
expected to contain 10% Pd were closer to 8% Pd, along with
51% La and 40% Fe. Materials expected to contain 5% Pd were
shown to contain 5% Pd with 52% La and 43% Fe. It is
expected that there is some inherent error in quantifying the
survey spectra as peak areas and background type are selected
manually within the software. High-resolution scans of the Pd
3d region shown in Fig. 7 indicate that the Pd 3d peaks are
Fig. 6 Transmission electron micrographs of (a) LaFe0.95Pd0.05O3−δ and
(b) YFe0.95Pd0.05O3−δ show uniform crystallinity.
With many iterations of various microwave heating profiles, shifted to a slightly higher binding energy than the expected
intermittent XRD analysis, and the TGA result in mind, we position for Pd2+ in PdO, 336.8 eV for the 3d5/2 peak, and Pd0
were able to achieve phase-pure, crystalline materials, with in which the 3d5/2 appears at 335.4 eV.41 This suggests that Pd
surface areas near 25 m2 g−1 for LaFeO3 materials and near is in a very ionic environment, more so than if PdO domains
5 m2 g−1 for YFeO3 materials. The sol–gel microwave-assisted are forming along with the perovskite host.
preparation method for substituted oxide catalysts has been
It is clear that Pd substituted into LaFeO3 is different from
proven to produce superior materials in an extremely time- Pd substituted into YFeO3. In the case of La as the A-site
efficient manner compared to traditional solid-state methods. cation, we see two distinct Pd species. This is in contrast to the
Due to different heating profiles required for La and Y contain- single Pd species observed when Y is on the A-site. This
ing oxides, their surface areas differ slightly. The lower surface anomaly may be a result of differing inductive contributions
area that is achieved for the Y analogue is attributed to the from La and Y or Pd occupying more than one site in the case
longer reaction time required for achieving phase purity. of La. The difference in catalytic behavior, discussed below,
Though these surface areas are less than those of high surface may be a result of the difference in Pd occupation in the two
area microporous catalysts, such as zeolitic materials, a 20-fold perovskites. We hypothesize that the La perovskite may make
increase in surface area (in the La perovskite) over convention- the Pd more accessible for catalysis and cycling to metallic Pd,
ally prepared solid-state materials is appreciable and suggests while the Pd in the Y perovskite may be less accessible. This
that this microwave-assisted method may be amenable to may explain the difference in catalytic reactivity for the two
further modification for surface area enhancement.
perovskites, as the difference in surface area is not significant
X-ray photoelectron survey spectra and high-resolution enough to cause a noticeable effect. One future application of
spectra for the Pd 3d region were collected over the range of this work is to probe ionic Pd catalysis, which requires the
1000 eV to 0 eV and 350 eV to 330 eV, respectively. Survey preparation of materials that stabilize ionic Pd under various
2084 | Dalton Trans., 2014, 43, 2079–2087
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
reaction conditions. It was shown in our previous work4 that a
binary oxide like CeO2 is not a sufficient ion stabilizing host to
maintain Pd ions under reducing reaction conditions. Here we
show that Pd ions substituted onto the B-site of a perovskite
host may experience increased ion stabilization as a result of
A-site cation induction. The fact that dopant ionicity is
affected by differing A-site cations further advances this stabi-
lizing host oxide concept.
Table 1 Impact of the perovskite surface area, Pd-substitution levels
and A-site cation identity in Suzuki–Miyaura cross-coupling reactions
Entry
Perovskite
Surface area (m2 g−1
)
GC yield (%)
a
1
2
3
4
5
6
7
LaFeO3
≈15
≈25
<5
≈15
≈10
<5
0
93
50
23
0
8
0
4. Catalytic activity
LaFe0.95Pd0.05O3−δ
LaFe0.95Pd0.05O3−δ
LaFe0.90Pd0.10O3−δ
The Suzuki–Miyaura coupling was chosen as a model reaction
to demonstrate the potential of Pd-substituted (La,Y)FeO3 perov-
skites as catalyst precursors in cross-coupling reactions. Of
equal importance, this reaction was selected to understand the
effects of Pd-substitution levels, the contribution of the A-site
cation and the particle surface area of our microwave syn-
thesized perovskites on the catalytic activity of these systems.
Recently, Martin and coworkers demonstrated that Pd-con-
taining perovskites are suitable catalysts for Suzuki–Miyaura
reactions.22,23 They elegantly demonstrated that under their
reaction conditions, Pd species desorb from the perovskite to
a
YFeO3
YFe0.95Pd0.05O3−δ
YFe0.90Pd0.10O3−δ
<5
Reaction conditions: perovskite (0.04 mol% Pd), SPhos (0 to 1 mol%),
3-chloropyridine (1.0 mmol), phenylboronic acid (1.4 mmol), K2CO3
(1.4 mmol), 1 : 1 i-PrOH–H2O (2 mL), 80 °C for 20 h. There was no
yield for reactions in which the ligand (SPhos) was omitted. a For
perovskites with 0% Pd substitution the same mass was used in the
reaction as the materials with 5% substitution.
give a soluble active catalyst. However, aryl and heteroaryl catalyst. Further, increasing the Pd-substitution from 5% to
chlorides were poor substrates using this method, requiring 10% in these materials substantially decreased the yield to
harsh reaction conditions (i.e. 135 °C) and resulting in modest 23% (Table 1, entry 4). This trend is similar to what was
yields. Currently, the most general catalyst systems for coup- observed for Pd-substituted in CeO2, where 5% Pd-substitution
ling these difficult substrates are based on discrete phosphine is more active than 2.5% Pd and 7.5% Pd substitution.4 More-
ligated Pd0 complexes.13 In order to demonstrate that perov- over, using a material with no Pd-substitution, LaFeO3, pro-
skites can be general catalysts for cross-coupling reactions, we vided no desired product (Table 1, entry 1), confirming that Pd
set out to develop a method that would efficiently couple aryl is responsible for catalysis. Lastly, substituting La for Y in
and heteroaryl chlorides under mild conditions.
these materials caused nearly complete loss in activity
We commence with the reaction of a heteroaryl chloride (Table 1, entries 6–8). We postulate that the differing A-site
with an aryl boronic acid using LaFe0.95Pd0.05O3−δ as the cata- cation will contribute a varying inductive effect in the perovs-
lyst. We hypothesized that by adding a biarylphosphine ligand kite, influencing the stability of the Pd ions in the material, as
developed by Buchwald for Suzuki–Miyaura couplings, well as the rate of desorption under reducing conditions.
SPhos,42 to the reaction mixture, the Pd0 species would be
The scope of this process was further examined utilizing
ligated as it is desorbed from the perovskite, leading to the for- the high surface area LaFe0.95Pd0.05O3−δ as the Pd source and
mation of a highly active catalyst. In support of this, using SPhos as
supporting ligand (Fig. 8). Employing only
a
LaFe0.95Pd0.05O3−δ (0.04 mol% Pd), SPhos (1 mol%), and 0.04 mol% Pd, heteroaryl chlorides and bromides were
K2CO3 in i-PrOH–H2O at 80 °C, 3-chloropyridine was reacted coupled with aryl boronic acids in good to excellent yields
with phenylboronic acid to give the desired biaryl product in (Fig. 8a–d). Moreover, heteroaryl boronic acids proved to be
93% yield (Table 1, entry 2). Interestingly, removing SPhos proficient coupling partners under these conditions. For
from this same reaction gave no product, demonstrating example, 2-thienylboronic acid was reacted with 2-bromopyri-
that the ligand is playing a critical role in the reaction dine to give the desired product in 78% isolated yield (Fig. 8).
and suggesting that the Pd0 is indeed being ligated as it is Aryl chlorides that did not contain any heteroatoms proved to
desorbed from the LaFe0.95Pd0.05O3−δ
.
be acceptable substrates under these conditions (Fig. 8e–g);
We next investigated the impact of surface area, Pd-substi- however, they resulted in moderately decreased yields com-
tution levels, as well as the contribution of the A-site cation in pared to heteroaryl chlorides. The above results clearly demon-
the perovskite, on the catalyst activity of this system. Reducing strate that Pd-containing LaFeO3 perovskites can be effective
the surface area of the particles from 28.5 m2 g−1 to 3.4 m2 g−1 Pd sources in challenging cross-coupling reactions and can
led to a decrease in yield from 93% to 50% (Table 1, entry 3); provide general and efficient catalytic activity when used in
this result suggests that surface area has some influence on combination with supporting ligands. We envisage that this
catalysis for this system. We hypothesize that the increased strategy will be applicable to a wide range of processes and has
surface area enables more efficient desorption of the Pd from the potential to afford highly active, long-lived and recyclable
the perovskite leading to higher concentrations of active catalysts.
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 2079–2087 | 2085

View Article Online
Paper
Dalton Transactions
of the NSF-funded Materials Research Facilities Network
(http://www.mrfn.org). The use of the Advanced Photon Source
at Argonne National Laboratory was supported by the U.S.
Department of Energy, Office of Science, Office of Basic Energy
Sciences, under contract no. DE-AC02-06CH11357. We thank
Dr Christina Birkel for help with the TGA analysis.
References
1 M. S. Hegde, G. Madras and K. C. Patil, Acc. Chem. Res.,
2009, 42(6), 704–712.
2 A. Gupta, A. Kumar, U. V. Waghmare and M. S. Hegde,
Chem. Mater., 2009, 21(20), 4880–4891.
3 T. Baidya, A. Gupta, P. A. Deshpandey, G. Madras and
M. S. Hegde, J. Phys. Chem. C, 2009, 113(10), 4059–4068.
4 L. M. Misch, J. A. Kurzman, A. R. Derk, Y.-I. Kim,
R. Seshadri, H. Metiu, E. W. McFarland and G. D. Stucky,
Chem. Mater., 2011, 23(24), 5432–5439.
5 U. Singh, J. Li, J. Bennett, A. Rappe, R. Seshadri and
S. Scott, J. Catal., 2007, 249(2), 349–358.
Fig. 8 Substrate scope for Suzuki–Miyaura cross-coupling reactions
using LaFe0.95Pd0.05O3 as the Pd source (isolated yields).
6 J. Li, U. G. Singh, T. D. Schladt, J. K. Stalick, S. L. Scott and
R. Seshadri, Chem. Mater., 2008, 20(3), 6567–6576.
7 J. A. Kurzman, X. Ouyang, W. B. Im, J. Li, J. Hu, S. L. Scott
and R. Seshadri, Inorg. Chem., 2010, 49(10), 4670–4680.
8 J. A. Kurzman, J. Li, T. D. Schladt, C. R. Parra, X. Ouyang,
R. Davis, J. T. Miller, S. L. Scott and R. Seshadri, Inorg.
Chem., 2011, 50(17), 8073–8084.
9 W. F. Libby, S. Url and S. Levine, Science, 1971, 171(3970),
499–500.
10 D. B. Meadowcroft, Nature, 1970, 226, 847–848.
11 Y. Nishihata, J. Mizuki, T. Akao, H. Tanaka, M. Uenishi,
M. Kimura, T. Okamoto and N. Hamada, Nature, 2002, 418
(6894), 164–167.
12 S. P. Andrews, A. F. Stepan, H. Tanaka, S. V. Ley and
M. D. Smith, Adv. Synth. Catal., 2005, 347(5), 647–654.
13 R. Martin and S. L. Buchwald, Acc. Chem. Res., 2008, 41(11),
1461–1473.
14 E. Negishi, Q. Hu, Z. Huang, M. Qian and G. Wang, Aldri-
chimica Acta, 2005, 38, 1–48.
15 K. Fugami and M. Kosugi, Top. Curr. Chem., 2002, 219,
87–130.
16 R. Chinchilla and C. Nájera, Chem. Soc. Rev., 2011, 40(10),
5084.
5. Conclusions
A rapid microwave-assisted sol–gel preparation method yields
Pd-substituted perovskite materials Ln1−xPdxFeO3−δ (with Ln =
La, Y). The very short heating times in microwave reactions
help us elucidate the phase formation process and have shown
that the reaction for forming LaFeO3 is completed after
90 s. In contrast to furnace-based sol–gel or combustion
methods, the microwave heating method has allowed the easy
and reproducible tuning of material parameters such as crys-
tallite size and surface area. Using a combination of character-
ization methods, we have shown that materials prepared with
this method are phase pure and contain well-dispersed ionic
Pd. We have also shown that changing the A site cation affects
the structure and properties of the perovskite hosts. The appli-
cability of the prepared perovskites as Pd0 catalyst sources in
Suzuki cross-coupling reactions has been demonstrated for
aryl chlorides. This method is amenable to further tuning of
the reaction parameters and may allow for the preparation of
materials with even higher surface areas, thereby impacting
their properties.
17 D. S. Surry and S. L. Buchwald, Chem. Sci., 2010, 2(1), 27.
18 I. P. Beletskaya and A. V. Cheprakov, Chem. Rev., 2000, 100,
3009–3066.
19 T. Hiyama and E. Shirakawa, Top. Curr. Chem., 2002, 219,
61–85.
Acknowledgements
This work is supported by the Department of Energy through
DE-FG02-10ER16081. L.M.M. acknowledges support from the 20 I. Hamada, A. Uozumi, Y. Morikawa, A. Yanase and
ConvEne IGERT Program (NSF-DGE 0801627). B.P.F. thanks
the California NanoSystems Institute for the Elings Prize
H. Katayama-Yoshida, J. Am. Chem. Soc., 2011, 133(46),
18506–18509.
Fellowship in Experimental Science. The research carried out 21 S. Yanagisawa, A. Uozumi, I. Hamada and Y. Morikawa,
here made extensive use of shared experimental facilities J. Phys. Chem. C, 2013, 117(3), 1278–1286.
of the Materials Research Laboratory supported by the 22 M. D. Smith, A. F. Stepan, C. Ramarao, P. E. Brennan and
NSF-MRSEC Program (DMR 1121053). The MRL is a member
S. V. Ley, Chem. Commun., 2003, 2652.
2086 | Dalton Trans., 2014, 43, 2079–2087
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
23 S. P. Andrews, A. F. Stepan, H. Tanaka, S. V. Ley 32 A. Birkel, N. A. DeCino, N. C. George, K. A. Hazelton,
and M. D. Smith, Adv. Synth. Catal., 2005, 347(5),
647–654.
B.-C. Hong and R. Seshadri, Solid State Sci., 2013, 19,
51–57.
24 R. Roy, S. Komarneni and L. Yang, J. Am. Ceram. Soc., 1985, 33 Z. Zhong, K. Chen, Y. Ji and Q. Yan, Appl. Catal., A, 1997,
68(7), 392–395.
156, 29–41.
25 J. D. Katz, Annu. Rev. Mater. Sci., 1992, 22, 153–170.
26 D. Baghurst, A. Chippindale and D. M. P. Mingos, Nature,
1988, 332, 311.
27 K. Biswas, S. Muir and M. Subramanian, Mater. Res. Bull.,
2011, 46(12), 2288–2290.
34 P. Ciambelli, S. Cimino, L. Lisi, M. Faticanti, G. Minelli,
I. Pettiti and P. Porta, Appl. Catal., B, 2001, 33(3), 193–203.
35 X. Qi, J. Zhou, Z. Yue, Z. Gui and L. Li, Mater. Chem. Phys.,
2002, 78, 25–29.
36 X. Zhang, Y. Li, H. Li and W. Shen, J. Nat. Gas Chem., 2012,
21(2), 113–118.
28 A. Birkel, L. E. Darago, A. Morrison, L. Lory, N. C. George,
A. A. Mikhailovsky, C. S. Birkel and R. Seshadri, Solid State 37 J. Zhang and H. Li, Perovskite: Crystallography, Chemistry
Sci., 2012, 14(6), 739–745.
and Catalytic Performance, Nova Science Pub Incorporated,
29 A. Birkel, K. A. Denault, N. C. George, C. E. Doll, B. He,
2013.
A. A. Mikhailovsky, C. S. Birkel, B.-C. Hong and 38 P. D. Ramesh, D. Brandon and L. Schächter, Mater. Sci.
R. Seshadri, Chem. Mater., 2012, 24, 1198–1204. Eng., A, 1999, 266(1–2), 211–220.
30 C. S. Birkel, W. G. Zeier, J. E. Douglas, B. R. Lettiere, 39 H. M. Rietveld, J. Appl. Crystallogr., 1969, 2(2), 65–71.
C. E. Mills, G. Seward, A. Birkel, M. L. Snedaker, Y. Zhang, 40 A. A. Coelho, Topas Academic V4.1 Software, Coelho Soft-
G. J. Snyder, T. M. Pollock, R. Seshadri and G. D. Stucky,
Chem. Mater., 2012, 24(13), 2558–2565.
31 C. S. Birkel, J. E. Douglas, B. R. Lettiere, G. Seward,
ware, Brisbane, Australia, 2007.
41 M. Brun, A. Berthet and J. Bertolini, J. Electron Spectrosc.
Relat. Phenom., 1999, 104(1–3), 55–60.
N. Verma, Y. Zhang, T. M. Pollock, R. Seshadri and 42 T. E. Barder, S. D. Walker, J. R. Martinelli and
G. D. Stucky, Phys. Chem. Chem. Phys., 2013, 15(18), 6990–
S. L. Buchwald, J. Am. Chem. Soc., 2005, 127(13), 4685–
6997.
4696.
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 2079–2087 | 2087