One-step synthesis of core-shell (Ce0.7Zr0.3O 2)x(Al2O3)1-x [(Ce 0.7Zr0.3O2)@Al2O3] nanopowders via liquid-feed flame spray

Article abstract of DOI:10.1021/ja9017545

We report here the synthesis of Ce_xZr_1-xO₂ and (Ce_0.7Zr_0.3O₂)_x(Al ₂O₃)_1-x core-shell nanopowders in a single step by liquid-feed flame spray pyro

Full text of DOI:10.1021/ja9017545

Published on Web 06/15/2009
One-Step Synthesis of Core-Shell (Ce0.7Zr0.3O2)x(Al2O3)1-x
[(Ce0.7Zr0.3O2)@Al2O3] Nanopowders via Liquid-Feed Flame
Spray Pyrolysis (LF-FSP)
Min Kim and Richard M. Laine*
Department of Materials Science and Engineering, UniVersity of Michigan, Ann Arbor,
Michigan 48109-2136
Received March 7, 2009; E-mail: talsdad@umich.edu
Abstract: We report here the synthesis of CexZr1-xO2 and (Ce0.7Zr0.3O2)x(Al2O3)1-x core-shell nanopowders
in a single step by liquid-feed flame spray pyrolysis (LF-FSP) of the metalloorganic precursors,
Ce(O2CCH2CH3)3(OH), alumatrane [N(CH2CH2O)3Al], and Zr(O2CCH2CH3)2(OH)2. Solutions of all three
precursors in ethanol with ceramic yields of 2.5 wt% were aerosolized with O2, combusted at temperatures
above 1500 °C, and rapidly quenched at ∼1000 °C/ms to form CexZr1-xO2 and (Ce0.7Zr0.3O2)x(Al2O3)1-x
nanopowders of selected compositions, at rates of 50-100 g/h. The resulting, as-processed, materials are
unaggregated nanopowders with average particle sizes (APSs) < 20 nm and corresponding specific surface
areas of 30-50 m2/g. The as-processed powders were characterized in terms of phase, particle size, specific
surface area, compositions, and morphology by XRD, BET, DLS, SEM, TEM, XPS, TGA-DTA, and FT-IR.
LF-FSP provides access to binary CexZr1-xO2 nanopowders and ternary (Ce0.7Zr0.3O2)x(Al2O3)1-x nanopowders
in one step. The obtained Ce0.7Zr0.3O2 powders are solid solutions with
a cubic phase. In
contrast, LF-FSP of mixtures of the three precursors at specific compositions [x ) 0.5, 0.7 for
(Ce0.7Zr0.3O2)x@(Al2O3)1-x] provide core-shell nanopowders in a single step. The most reasonable explanation
is that there are differences in the rates of condensation, nucleation and miscibility between the gas phase ions
that form the CexZr1-xO2 solid solutions and those that condense to δ-Al2O3 during processing. These as-
produced materials are without microporosity at surface areas of g30 m2/g. Evidence is presented suggesting
the formation of (Ce/Zr)3+ species in the as-processed (Ce0.7Zr0.3O2)x(Al2O3)1-x core-shell materials. An
accompanying paper indicates that these materials offer significant and novel catalytic activities for hydrocarbon
oxidation and deNOx processes without using platinum as a co-catalyst.
Introduction
the late 50s and early 60s by researchers exploring methods of
making single quantum magnetic particles and related materials
The past two decades have witnessed an explosion in efforts
directed toward the synthesis, characterization and application
of nanoparticles in very diverse fields. Indeed, there are now
fields within fields in the study of nanoparticles with major areas
of study concerned with metal, metal oxide, metal chalcogen,
and metal pnictogen particles among others, and of course
studies at the interface between these materials centered for
example around core-shell systems.1-7
for recording media, catalysts and pigments as partially reviewed
by Luborsky.1,3,4
If we limit discussion to the synthesis of nanoparticles rather
than nanorods, whiskers, or nanotubes, there are still numerous
preparative approaches described in the literature.1-10 Although a
majority of this research focuses on solution syntheses via direct
reactions in high-boiling solvents5-8 or sol-gel approaches;12-14
commercial scale synthesis of ultrafine and nanopowders is mostly
via gas phase reaction chemistry.15,16 Thus, many thousands of
tons of titania, silica, alumina, and zirconia powders are produced
annually primarily by reaction of the volatile metal chlorides in a
hydrogen/oxygen flame, reaction 1, in a process called flame spray
pyrolysis, FSP. Depending on the conditions the resulting products
are either ultrafine <500 nm average particle sizes (APSs) or
nanosized <100 nm.
Although the synthesis and characterization of metal and
metal oxide nanopowders have received tremendous recent
attention, early work on nanoparticles was first conducted in
(1) Reviews: (a) Luborsky, F. E. J. Appl. Phys. 1961, 32, S171. (b)
Cushing, B. L.; Kolesnichenko, V. L.; O’Connor, C. J. Chem. ReV.
2004, 104, 3893. (c) Swihart, M. T. Curr. Opin. Colloid Interface
Sci. 2003, 8, 127–133. (d) Kruis, F. E.; Fissan, H.; Peled, A. 2004,
29, 511. (e) Strobel, R; Pratsinis, S. E. J. Mater. Chem. 2007, 17,
4743.
MClx + H2 + O2 1200°C8 MOy + HCl + Cl2 + H2O, etc.
(2) Klabunde, K. J. Free Atoms, Clusters, and Nanoscale Particles;
Academic Press: New York, 1994.
(3) (a) Fabian, R. W.; Sienczyk, J. L.; U.S. Patent 2,884,319. April 1959.
(b) Thomas, R.; U.S. Patent 3,281,344. Oct. 1966. (c) Smith, T. W.;
U.S. Patent 4,252,671. Feb. 1981.
(4) (a) Dan, N.; Zubris, M.; Tannenbaum, R. Macromolecules 2005, 38,
9243–9250. (b) Tannenbaum, R.; Zubris, M.; Goldberg, E. P.; Reich,
S.; Dan, N. Macromolecules 2005, 38, 4254.
(1)
Unfortunately, the number of potential ultrafine and nano-
powder oxides that can be synthesized by FSP is limited by the
number of volatile metal chlorides. Another problem noted in
9
9220 J. AM. CHEM. SOC. 2009, 131, 9220–9229
10.1021/ja9017545 CCC: $40.75  2009 American Chemical Society

One-Step Synthesis of Core-Shell Nanopowders via LF-FSP
A R T I C L E S
the literature17,18 has been the great difficulty in using FSP to
synthesize mixed-metal oxide powders. This problem arises
because of the disparate rates of hydrolysis/oxidation of the
individual chlorides even when the resulting mixed-metal oxides
are expected to be completely miscible, e.g., SiO2 and Al2O3
readily form aluminosilicates.
It is important to note at the outset that there are multiple
solution chemistry routes to core-shell metal oxide nano-
powders;8,10,11,27,28 however, these routes often lead to amor-
phous materials that must be heated to crystallize one or both
components. Furthermore, this heating typically causes particle
aggregation limiting their utility for further processing. In
contrast, LF-FSP provides access to fully crystalline core-shell
and easily processed nanomaterials. It also offers access to
unusual kinetic phases and in some instances by repassing the
nanopowders through the LF-FSP system, it provides access to
nanopowders of thermodynamically stable phases that are
difficult to obtain by traditional processing techniques.15,16,24-26
In this paper, we describe efforts to develop routes to binary
CexZr1-xO2 solid solution and ternary (Ce0.7Zr0.3O2)x@(Al2O3)1-x
core-shell nanostructured nanoparticles targeting the develop-
ment of novel catalyst systems for emission control of hydro-
carbons and NOx. In an accompanying paper, we describe high
throughput, combinatorial testing of sets of these nanopowders
to assess their catalytic activity for promoting oxidation of
propane and coincident reduction of NOx. In the accompanying
paper, we find that the LF-FSP produced CexZr1-xO2 and
(Ce0.7Zr0.3O2)x(Al2O3)1-x core-shell nanopowders offer catalytic
activities for deNOx reactions approaching those of Pt containing
catalysts but without the need for Pt. Here we present evidence
In the past decade, we and others have resolved this problem
through the development of liquid-feed flame spray pyrolysis
(LF-FSP) wherein alcohol solutions of organometallic and
preferably metalloorganic precursors (e.g., carboxylates, ꢀ-dike-
tonates, and alkoxides) are aerosolized with oxygen and ignited.
The combustion process generates flames of 1500-2000 °C and
if the combustion derived metal oxide ions are quenched rapidly
enough, ∼1000 °C/ms, it is possible to produce a wide variety
of unaggregated (therefore easily dispersed) nanopowders whose
compositions are determined almost completely by the composi-
tions of the precursors in solution.17-26 Furthermore, it is
possible to produce 100 g/h quantities of mixed-metal oxide
nanopowders in the laboratory with APSs <100 nm and
frequently <20-30 nm, which equates to specific surface areas
(SSAs) of up to 100 m2/g without internal porosity. Alternately,
it is possible to produce up to five different samples/day at
20-30 g quantities.23-26
(5) (a) Figlarz, M.; Fie´vet, F.; Lagier, J. P. U.S. Patent 4,539,041, Sept.
31985. (b) Fie´vet, F.; Lagier, J. P.; Blin, B.; Beaudoin, B.; Figlarz,
M. Solid State Ionics, 1989, 32/33, 198. (c) Kurihara, L. K.; Chow,
G. M.; Schoen, P. E. Nanostruct. Mater. 1995, 6, 607. (e) Laine, R. M.;
Sellinger, A.; U.S. Patent 6,551,960. April 22, 2003.
(6) (a) Suh, W. H.; Suslick, K. S. J. Am. Chem. Soc. 2005, 127, 12007.
(b) Skrabalak, S. E.; Suslick, K. S J. Am. Chem. Soc. 2005, 127, 9990.
(c) Suh, W. H. J. Am. Chem. Soc. 2005, 127, 12007.
for the possible formation of zirconia suboxide species, ZrO2-x
,
which may be responsible for this catalytic behavior.
Experimental Section
Materials. Cerium carbonate [Ce2(CO3)3 ·x(H20), 99%] and
zirconium carbonate [2ZrO2(CO2)·x(H2O), 99%] were purchased
from PIDC Inc. (Ann Arbor, MI). EtOH (99%) was purchased
from standard sources and used as received. Alumatrane
[N(CH2CH2O)3Al] was prepared as described elsewhere.25a Pro-
pionic acid (C2H5CO2H, 99%) was purchased from Aldrich and
used as received.
Precursor Preparation. Precursors with ceramic compositions
[(CeO2)0.7(ZrO2)0.3]x(Al2O3)1-x (x ) 0.9, 0.7, 0.5, 0.3, and 0.1) were
prepared from mixtures of the following metalloorganics.
Al2O3 Precursor. In all cases, alumatrane [N(CH2CH2O)3Al] was
used as the Al2O3 source. An alumatrane/EtOH solution was
prepared containing 10.8 wt % Al2O3 per TGA.
(7) (a) Puntes, V. F.; Zanchet, D.; Erdonmez, C. K.; Alivisatos, A. P.
J. Am. Chem. Soc. 2002, 124, 12874. (b) Liu, H.; Guo, J.-H.; Yin, Y.;
Augustsson, A.; Dong, C.; Nordgren, J.; Chang, C.; Alivisatos, P.;
Thornton, G.; Ogletree, D. F.; Requejo, F. G.; de Groot, F.; Salmeron,
M. Nano Lett. 2007, 7, 1919.
(8) (a) Tsay, J. M.; Pflughoefft, M.; Bentolila, L. A.; Weiss, S. J. Am.
Chem. Soc. 2004, 126, 1926. (b) Kim, S.; Fisher, B.; Eisler, H.-J.;
Bawendi, M. J. Am. Chem. Soc. 2003, 125, 11466.
(9) (a) Grass, R. N.; Athanassiou, E. K.; Stark, W. J. Angew. Chem. 2007,
46, 4909. (b) Athanassiou, E. K.; Grass, R. N.; Osterwalder, N.; Stark,
W. J. Chem. Mater. 2007, 19, 4847. (c) Ponce, A. A.; Klabunde, K. J.
J. Molec. Catal. A 2005, 225, 1.
(10) Nyutu, E. K.; Conner, W. C.; Auerbach, S. M.; Chen, C.-H.; Suib,
S. L. J. Phys. Chem. C. 2008, 112, 1407.
Cerium Propionate Precursor: Ce(O2CCH2CH3)3(OH). Ce-
rium carbonate [Ce2(CO3)3 ·x(H2O), 99%, 70 g, 0.15 mol] was
reacted with excess propionic acid (400 mL, 5.44 mol) in a 1 L
flask equipped with a still head and an addition funnel. N2 was
(11) Chen, T-Y.; Somasundaran, P. J. Am. Ceram. Soc. 1998, 81, 140.
(12) (a) Chen, Q.; Boothroyd, C.; Tan, G. H.; Sutanto, N.; Soutar, A. M.;
Zeng, Z. T. Langmuir 2008, 24, 650. (b) Yeshchenko, O. A.; Dmitruk,
I. M.; Alexeenko, A. A.; Dmytruk, A. A. J. Phys. Chem. Sol. 2008,
69, 1615.
(13) Masala, O.; Seshadri, R. J. Am. Chem. Soc. 2005, 127, 9354.
(14) Aslan, K.; Wu, M.; Lakowicz, J. R.; Geddes, C. D. J. Am. Chem.
Soc. 2007, 129, 1524–1525.
(25) (a) Hinklin, T.; Toury, B.; Gervais, C.; Babonneau, F.; Gislason, J. J.;
Morton, R. W.; Laine, R. M. Chem. Mater. 2004, 16, 21. (b) Kim,
M.; Hinklin, T. R.; Laine, R. M. Chem. Mater. 2008, 20, 5154. (c)
Marchal, J.; Johns, T.; Baranwal, R.; Hinklin, T.; Laine, R. M. Chem.
Mater. 2004, 16, 822. (d) Kim, S.; Gislason, J. J.; Morton, R. W.;
Pan, X. Q.; Sun, H. P.; Laine, R. M. Chem. Mater. 2004, 16, 2336.
(e) Azurdia, J. A.; Marchal, J.; Laine, R. M. J. Am. Ceram. Soc. 2006,
89, 2749. (f) Azurdia, J. A.; Marchal, J.; Shea, P.; Sun, H.; Pan, X. Q.;
Laine, R. M. Chem. Mater. 2006, 18, 731.
(15) Kodas, T. T.; Hampden-Smith, M. J. Aerosol Processing of Materials;
Wiley-VCH: New York, 1999.
(16) Winterer, M. Nanocrystalline Ceramics, Synthesis and Structure;
Springer: New York, 2002.
(17) Ulrich, G. D. Chem. Eng. News 1984, 62, 22.
(18) Gurav, A.; Kodas, T.; Pluym, T.; Xiong, Y. Aerosol Sci. Technol.
1993, 19, 411.
(19) Burgos-Montes, O.; Moreno, R.; Colomer, M. T.; Farin˜as, J. C. J. Am.
Ceram. Soc. 2006, 89, 484.
(26) (a) Laine, R. M.; Marchal, J. C.; Sun, H. P.; Pan, X. Q. Nat. Mater.
2006, 5, 710. (b) Laine, R. M.; Marchal, J.; Sun, H. J.; Pan, X. Q.
AdV. Mater. 2005, 17, 830. (c) Hinklin, T. R.; Rand, S. C.; Laine,
R. M. AdV. Mater. 2008, 20, 1270.
(20) McCormick, J. R.; Zhao, B.; Rykov, S. A.; Wang, H.; Chen, J. G. J.
Phys. Chem. B. 2004, 108, 17398.
(21) (a) Strobel, R.; Metz, H. J.; Pratsinis, S. E. Chem. Mater. 2008, 20,
6346. (b) Teoh, W. Y.; Setiawan, R.; Ma¨dler, L.; Grunwaldt, J.-D.;
Amal, R.; Pratsinis, S. E. Chem. Mater. 2008, 20, 4069.
(22) Li, D.; Teoh, W. Y.; Selomulya, C.; Woodward, R. C.; Amal, R.;
Rosche, B. Chem. Mater. 2006, 18, 6403.
(27) (a) Sakai, H.; Kanda, T.; Shibata, H.; Ohkubo, T.; Abe, M. J. Am.
Chem. Soc. 2006, 128, 4944. (b) Xu, Z.; Hou, Y.; Sun, S. J. Am. Chem.
Soc. 2007, 129, 8698. (c) Aslan, K.; Wu, M.; Lakowicz, J. R.; Geddes,
C. D. J. Am. Chem. Soc. 2007, 129, 1524. (d) Xu, Z.; Hou, Y.; Sun,
S. J. Am. Chem. Soc. 2007, 129, 8698.
(23) Laine, R. M.; Waldner, K.; Bickmore, C.; Treadwell, D. R. U.S. Patent
5,958,361, September 28, 1999.
(28) (a) Teng, X.; Yang, H. J. Am. Chem. Soc. 2003, 125, 14559. (b) Lai,
J.; Shafi, K. V. P. M.; Ulman, A.; Loos, K.; Popovitz-Biro, R.; Lee,
Y.; Vogt, T.; Estournes, C. J. Am. Chem. Soc. 2005, 127, 5730. (c)
Deng, Y.; Qi, D.; Deng, C.; Zhang, X.; Zhao, D. J. Am. Chem. Soc.
2008, 130, 28.
(24) (a) Kim, M.; Laine, R. M. J. Cer. Proc. Res. 2007, 8, 129. (b) Kim,
M.; Hinklin, T. R.; Laine, R. M. Chem. Mater. 2008, 20, 5154. (c)
Kim, M., Laine, R. M. submitted to J. Am. Ceram. Soc.
9
J. AM. CHEM. SOC. VOL. 131, NO. 26, 2009 9221

A R T I C L E S
Kim and Laine
sparged directly through the solution (2 psi pressure) as the solution
was heated at 120 °C/2 h to distill off ∼150 mL of liquid (water
and propionic acid). The ceramic loading of the resulting precursor
was 9 wt % as determined by TGA. We describe the isolation and
characterization of this precursor elsewhere.24b
Zirconium Propionate Precursor: Zr(O2CCH2CH3)2(OH)2.
Zirconium carbonate [ZrO2(CO2)·x(H2O), 99%, 150 g, 0.34 mol]
was reacted with excess propionic acid (500 mL, 6.80 mol) in a 1
L flask equipped with a still head and an addition funnel. N2 was
sparged directly through the solution (2 psi pressure) as the solution
was heated at 120 °C/2 h with magnetic stirring to distill off ∼150
mL of liquid (water and propionic acid). The ceramic loading of
the resulting precursor was 11 wt % as determined by TGA. We
describe the isolation and characterization of this precursor else-
where.22
Analytical Studies. Information concerning the experimental
procedures used for the following analytical tools are provided in
the experimental section in the Supporting Information including
X-ray powder diffraction (XRD), dynamic laser light scattering
(DLS), thermal gravimetric and differential thermal analysis (TGA/
DTA), specific surface area (SSA, BET) analyses, scanning electron
microscopy (SEM), transmission electron microscopy (SEM), FTIR
spectral analyses, and X-ray photoelectron spectroscopy (XPS).
Liquid Feed-Flame Spray Pyrolysis (LF-FSP). The apparatus
used for LF-FSP is comprised of an aerosol generator, a combustion
chamber, and an electrostatic powder collection system described
elsewhere.24-26 The precursor solution was pumped through the
aerosol generator at a rate controlled by the ceramic yield of the
solution. Typically, more concentrated (5 wt % ceramic yield)
solutions were pumped at 100 mL/min to avoid producing large
particles (200-1000 nm).24-26 Solutions with lower ceramic yields
were pumped at 400 mL/min. Here we used precursor solutions
with 2.5 wt % ceramic yield for each batch at the pumping rate of
300 mL/min. The solution was atomized with oxygen to form an
aerosol and ignited by two methane/oxygen pilot torches, while
the pressure was kept at 20 psi. Combustion produces temperatures
>1500 °C24-26 and nanosized oxide powders are collected in
electrostatic precipitators (ESP).
Initial studies on LF-FSP benchmarked the process against
FSP for production of nano-Al2O3.25a,b Both methods produce
similar high surface area, micropore free δ- rather than R-Al2O3
nanopowders despite flame temperatures exceeding the normal
phase transformation temperature near 1200 °C.25 The possible
reasons for this are addressed elsewhere.26a These same studies
demonstrated that any impurities in the starting precursor are
incorporated into the resulting δ-Al2O3 nanopowders.23-26,29
The potential to purposely introduce impurities (dopants)
combined with the fact that the density of δ-Al2O3 is 3.5 g/cm3
(vs 3.99 g/cm3 for R-Al2O3) indicating a more open crystal lattice
suggested efforts to incorporate large dopant ions, especially
rare earths (REs). The motivation for this work was two-fold.
One was to use LF-FSP to replace Al3+ with RE3+ ions to
develop novel phosphors. The second was to inhibit the
transformation temperature of δ- to R-Al2O3 to stabilize it as a
potential catalyst support given that REs are known to inhibit
both phase transformations and grain growth in alumina.30,31
An important aspect of this work was the fact that most dopant
studies coat alumina particles with RE dopants whereas LF-
FSP offers the opportunity to place the dopant directly in the
lattice perhaps greatly affecting (slowing) transformation rates.
Efforts to incorporate REs into the δ-Al2O3 lattice were quite
successful as LF-FSP provides nanopowders that retain the
δ-Al2O3 crystal structure yet contain up to 5 wt % Ce3+ ions
before generating Ce-magnetoplumbite (CeAl11O18).24b Thus,
LF-FSP provides access to a δ-Al2O3 wherein a significant
fraction of M3+ lattice sites normally occupied by Al3+ are
instead occupied by Ce3+. These results were extended to a wide
variety of RE3+ ions.32 In contrast, the solubility of RE3+ in
R-Al2O3 is limited to ∼10 ppm.31
Surprisingly, not only did these new nanopowders act as
phosphors,32 they also offered unique properties acting as “laser
paints” and exhibiting incoherent lasing by trapping light within
virtual cavities.33 In particular, cathodoluminscence experiments
using Ce3+ doped δ-Al2O3 nanopowders demonstrate continuous
wave UV lasing.32b These same Ce3+ doped δ-Al2O3 nanopo-
wders also offer considerable resistance to sintering coincident
with elevated phase transformation temperatures (to R-Al2O3).33
As noted above, because LF-FSP permits facile processing
of multiple samples, efforts were made to extend these studies
to their logical conclusion by producing samples along the entire
tie-line both for completeness but also to explore opportunities
to make novel kinetic materials leading to the discovery of
core-shell or nanostructured nanoparticles, CeOx@δ-Al2O3.24b
Results and Discussion
The LF-FSP process, described in detail elsewhere,24-26
aerosolizes solutions of single and mixed-metal alkoxides and/
or carboxylates dissolved in alcohol (typically EtOH) using
oxygen. The aerosol is ignited in a quartz combustion chamber
generating flame temperatures of 1500-2000 °C depending on
the processing conditions, the precursor, and the alcohol used.
The temperature drops rapidly to less than 400 °C in ∼1.5 m
leading to very rapid quenching of the gas entrained ceramic
“soot.” This soot is then collected in electrostatic precipitators
downstream from the combustion chamber and consists of
unaggregated nanopowders of the exact composition found in
the original solution. Typical production rates are 50-100 g/h
depending on the precursor. The resulting APSs are 10-150
nm depending on processing conditions giving SSAs of 20-100
m2/g. They are most frequently single crystal particles that
disperse readily in common solvents.
We have used LF-FSP to produce a wide variety of single
and binary-metal oxide nanopowders24-26 along the
(MOz)x(Al2O3)1-x tielines were M ) Ni, Ti, Co, Mg, Y, Cu, Zr,
and Ce. The as-produced binary nanopowders are also unag-
gregated with SSAs of 30-60 m2/g and APSs of 15-30 nm.
However, we recently discovered that under some conditions,
it is possible to make crystalline core-shell nanoparticles with
novel properties the subject of this and recent papers.24 To frame
the current work properly, it is pertinent to provide a brief
background.
(29) Sutorik, A. C.; Neo, S. S.; Hinklin, T.; Baranwal, R.; Treadwell, D. R.;
Narayanan, R.; Laine, R. M. J. Am. Ceram. Soc. 1998, 81, 1477–
1486.
(30) Ragan, D. D.; Mates, T.; Clarke, D. R. J. Am. Ceram. Soc. 2003, 86,
541–545.
(31) Thompson, A. M.; Soni, K. K.; Chan, H. M.; Harmer, M. P.; Williams,
D. B.; Chabala, J. M.; Levi-Setti, R. J. Am. Ceram. Soc. 1997, 80,
373–376.
(32) (a) Williams, G. R; Bayram, S. B; Rand, S. C; Hinklin, T; Laine,
R. M. Phys. ReV. A. 2001, 65, 013807. (b) Ultraviolet nanophosphors,
B.; Li, T; Hinklin, R; Laine, S J. Lumin. 2007, 122-123, 345-347.
(c) Li, C. B; Williams, G; Rand, S. C; Hinklin, T; Laine, R. M. Opt.
Lett. 2002, 27, 394–396. (d) Laine, R. M; Hinklin, T; Williams, G;
Rand, S. C. Mater. Sci. Forum 2000, 343-346, 500-10. (e) Li, D.
B; Oliveira, S. L.; Rand, S. C. Azurdia; J, A; Hinklin, T. R; Marchal,
J. C.; Laine, R. M. J. Appl. Phys. 2007, 101, 053534.
(33) Redmond, S. M; Armstrong, G. L; Chan, H.-Y; Mattson, E; Mock,
A; Li, B; Potts, J. R; Cui, M; Rand, S. C; Oliveira, S. L; Marchal, J.;
Hinklin, T.; Laine, R. M J. Opt. Soc. Am. B. 2004, 21, 214–222.
(34) Hinklin, T; Ph, D. Dissertation, UniVersity of Michigan 2006, 6.
9
9222 J. AM. CHEM. SOC. VOL. 131, NO. 26, 2009

One-Step Synthesis of Core-Shell Nanopowders via LF-FSP
A R T I C L E S
The discovery of core-shell nanoparticles prompted efforts to
also examine zirconia analogues but for completely different
reasons.
ZrOx is added to form a solid solution that is generally
considered to prevent CeOx sintering during operation allowing
the TWC to maintain its catalytic activity over extended periods.
The entire process of preparing the washcoat system including
the alumina consists of some 10 steps.39 We realized that it
might be possible to use LF-FSP processing to produce the same
combination of components in a single step and even coat the
honeycomb monolith by placing it directly in the flame. If
possible, we would save multiple steps, reducing costs in terms
of equipment needed and waste byproducts generated.
This then motivated the current efforts to use LF-FSP to
produce high SSA, nonporous Ce1-xZrxO2/Al2O3 materials in
one step as a prelude to exploring their utility for emission
treatment catalysts as described in an accompanying paper.24,48
Despite the extensiVe literature on Ce0.7Zr0.3O2 materials,39-47
our findings below and in the accompanying paper point to Very
noVel properties, not seen preViously.
Of particular importance is the fact that the quality of the
nanopowders produced depends strongly on the type of precur-
sor used. In our studies on the LF-FSP synthesis of δ-Al2O3,
we determined that metal nitrates although relatively inexpensive
are actually very poor precursors for LF-FSP processing because
they tend to form large (200-2000 nm) hollow particles
(by a spray pyrolysis mechanism), whereas alumatrane
N(CH2CH2O)3Al provides access to powders with SSAs ≈
60-100 m2/g.25 Furthermore, combustion flame temperatures
typically have no influence on the quality of the nanopowders
produced or the phase.24-26 It appears that the quenching rate
is the primary variable that controls what phases form, the degree
of particle aggregation, and to some extent particle sizes.24-26
Sutorik and Baliat briefly reported using LF-FSP to make
Ce1-xZrxO2 solid solutions.49 More detailed recent studies made
by Stark, Jossen et al.50 also discuss the oxygen storage capacity
of Ce0.5Zr0.5O2 nanopowders made using laminar flow LF-FSP.50
The current work goes one step further than these previous
studies in that it explores the coincident introduction of an
immiscible alumina component in an attempt, as noted above,
to develop a single-step “washcoat” process. Thus, we illustrate
an approach to the one step production of CexZr1-xO2 solid
solutions coated with imperfect δ-Al2O3 shells with essentially
any composition desired. The imperfect shells are necessary to
obtain good catalytic properties.
Tetragonal, zirconia (t-ZrO2) toughened R-Al2O3 (ZTA) is
one of the more important new, hard and “tough” ceramic
materials receiving a great deal of attention especially for
prosthetic implants, cutting tools, kitchen knives and even
photonic microdevices.35-38 The normally very high R-Al2O3
sintering temperatures required to produce fully dense, optimal
ZTA materials can require hot isotatic pressing at temperatures
exceeding 1400 °C and pressures of up to 100 MPa to achieve
>99.8% densities.
To stabilize the tetragonal phase, which provides the toughen-
ing properties, both yttria and ceria dopants are added in up to
3 wt %. We now find it possible to escape the high temperatures,
pressures and dopants required to densify ZTA by using LF-
FSP to produce a number of t-ZrOx@δ-Al2O3 nanopowders.24a
These powders allow pressureless sintering to completely dense
monoliths at temperatures just above 1100 °C. Basically the
δ-Al2O3 shell sinters at much lower temperatures even than
nano-R-Al2O3.26a The resulting dense materials have final grain
sizes <200 nm, the δ-Al2O3 shell transforms to R-Al2O3 while
the encapsulated t-ZrO2 particle remains without the need to
add a stabilizing agent and the resulting monoliths offer
transformation toughening.26c Thus, the baseline studies that
allow us to understand the work reported here were originally
directed toward other objectives than those reported here.
Thus the current study, while building on these previous
studies, has completely different objectives. Catalytic control
of exhaust emissions from both internal combustion engines and
more recently diesel engines has been the subject of an
enormous number of studies with multiple, highly cited reviews
already written.39-47 Intense efforts to optimize three-way auto
exhaust catalysts, TWCs have occurred over the past 30 years
for auto exhaust emission control only. Current TWCs used in
more than 10 million new automobiles annually in the U.S.
alone, are catalyst systems largely based on high surface area
ceria-zirconia solid solutions applied to monolithic honeycomb
supports (typically cordierite, 2MgO ·2Al2O3 ·5SiO2) through the
use of a “washcoating” method that employs sol-gel derived
alumina to affix the Ce1-xZrxO2 powders and a precious metal,
usually Pd, to the support. The most commonly used TWC
Ce1-xZrxO2 solid solution is Ce0.7Zr0.3O2. The CeOx component
stores and then delivers oxygen needed to oxidize residual
hydrocarbons and CO found in the exhaust whereas the Pd
component promotes reduction of NOx species in the exhaust.
Both are thought to also promote the water-gas shift reaction.
For this study, 70:30 Ce/Zr ratios of their propionate
precursors were used with varying amounts of an alumatrane
precursor (see Experimental Section for details). The total
ceramic loadings in EtOH solutions were kept to 2.5 wt % to
minimize viscosity problems. These solutions were aeroso-
lized with O2 and combusted to produce Ce1-xZrxO2 and
[(CeO2)0.7(ZrO2)0.3]x@(Al2O3)1-x (x ) 0.1-0.9) composition
nanopowders at rates of 50-100 g/h. The resulting materials
are unaggregated nanopowders with SSAs of 45 ( 5 m2/g
without microporosity.
(35) Piconi, C.; Maccauro, G. Biomaterials 1999, 20, 1–25.
(36) Ighodaro, O. L.; Okoli, O. I. Int. J. Appl. Ceram. Technol 2008, 5,
313–323.
(37) Li, X. S.; Low, I. M. J. Mater. Sci. Lett. 1993, 12, 1916–1919.
(38) Chen, W.; Kirihara, S.; Miyamoto, Y. Int. J. Appl. Ceram. Technol.
2008, 5, 353–359.
In the following sections we begin by discussing the Al2O3,
CeO2, and ZrO2 LF-FSP precursors. A discussion of LF-FSP
processing follows and thereafter we describe the detailed
(39) Flytzani-Stephanopoulos, M. Mater. Res. Soc. Bull. 2001, 885–9.
(40) Jen, H. W.; Graham, G. W.; Chun, W.; McCabe, R. W.; Cuif, J. P.;
Deutsch, S.; Touret, O. Cat. Today 1999, 50, 309.
(41) Ge´lin, P.; Primet, M. Appl. Catal. B 2002, 39, 1–37.
(42) Kaspar, J.; Fornasiero, P. J. Solid State Chem. 2003, 171, 19–29.
(43) Ozawa, M. J. Alloys Cmpds. 1998, 275-277, 886-890.
(44) Sugiura, M. Catal. SurV. Asia 2003, 7, 77–87.
(45) Shelef, M.; McCabe, R. W. Catal. Today 2000, 62, 35.
(46) Ghandi, H. S.; Graham, G. W.; McCabe, R. W. J. Catal. 2003, 216,
433.
(48) Weidenhof, B.; Reiser, M.; Sto¨we, K.; Maier, W. F.; Kim, M.; Azurdia,
J.; Gulari, E.; Sekers, E.; Barks, A.; Laine, R. M. J. Am. Chem. Soc.
2009, 131, http://dx.doi.org/ja8091345.
(49) Sutorik, A. C.; Baliat, M. S. Mat. Sci. For. 2002, 386, 371.
(50) (a) Stark, W. J.; Maciejewski, M.; Madler, L.; Pratsinis, S. E.; Baiker,
A. J. Catal. 2003, 220, 35. (b) Jossen, R.; Heine, M. C.; Pratsinis,
S. E.; Akhtar, M. K. Chem. Vapor Dep. 2006, 12, 614–619. (c) Jossen,
R.; Heine, M. C.; Pratsinis, S. E; Akhtar, M. K; Jossen, b.; Pratsinis,
R.; Stark, S. E.; Madler, W. J L. J. Am. Ceram. Soc. 2005, 88, 1388.
(47) (a) Marecot, P.; Pirault, L.; Mabilon, G.; Prigent, M.; Barbier, J. Appl.
Catal. B 1994, 5, 57. (b) Juan, R. G.; Miguel, A. G.; Jean-Louis, M.;
Pilar, G.; Gilbert, B. Appl. Catal. B 2000, 25, 19.
9
J. AM. CHEM. SOC. VOL. 131, NO. 26, 2009 9223

A R T I C L E S
Kim and Laine
(111) peak position, using Vegard’s law, as 28.92° 2θ. This
matches the 28.91° 2θ peak position found for our (CeO2)0.7-
(ZrO2)0.3 composition within the limits of XRD resolution
((0.02° 2θ).
For the samples Ce/Zr ) 1:1 and 3:7, we observe phase
separation of the two possible phases (cubic ceria, t-zirconia)
from the data analysis (see Experimental Section). This contrasts
with the work of Sutorik and Baliat49 and with Stark, Jossen et
al.50 where at 1:1 and 3:7 only the solid solutions were observed.
The SSAs of Sutorik and Baliat were ∼15 m2/g, suggesting
some influence of particle size and/or processing conditions on
the product phase. In contrast, the SSAs of Stark et al.50 were
higher than those reported here but were made under laminar
rather than turbulent flow conditions and with production rates
at 3-4 vs 50-100 g/h here. Thus, this does not seem to be the
explanation. Another possible explanation is presented below
on the basis of the XPS studies.
The Ce/Zr ) 7:3 composition is that often used in commercial
three-way auto exhaust catalysts or TWCs39-47 and is the reason
we chose Ce0.7Zr0.3O2 for the work reported here. Thus, only
the alumina content was varied in these studies with four
separate samples made at 10, 30, 50, or 70 mol % Al2O3.
Therefore the powder sample XRDs shown in Figure S2 were
produced using a propionate precursor system formulated to a
Ce0.7Zr0.3O2 composition and then mixed with [Al(OCH2CH2)3N]
to obtain ternary compositions [Ce0.7Zr0.3O2]x(Al2O3)1-x (x )
0.1-0.9).
By comparison with the XRD of Ce0.7Zr0.3O2 in Figure S1,
we reproducibly generate the same cubic (CeO2)0.7(ZrO2)0.3 solid
solution with alumina. The (111) peak position of the four
different [Ce0.7Zr0.3O2]x(Al2O3)1-x samples in Figure S2
(28.93° 2θ) is close to the calculated (111) peak position of
Ce0.7Zr0.3O2 (28.92° 2θ) within the limits of the XRD resolution
((0.02° 2θ).
Figure 1. TGA of Zr(O2CCH2CH3)2(OH)2 ramped at 10 °C/min in synthetic
air.24a
characterization of the Ce1-xZrxO2 and [(CeO2)0.7-
(ZrO2)0.3]x@(Al2O3)1-x (x ) 0.1-0.9) materials using various
analytical tools including XRD, BET, DLS, FTIR, TGA-DTA,
SEM, TEM, and XPS.
Precursors and Precursor Formation. We previously reported
the characterization of alumatrane [N(CH2CH2O)3Al], cerium
propionate Ce(O2CCH2CH3)3(OH), zirconium propionate
Zr(O2CCH2CH3)2(OH)2, and their use as precursors in LF-FSP for
the synthesis of δ-alumina, (CeOx)1-x(Al2O3)x and (ZrO2)1-x(Al2O3)x
nanopowders.24,26 These precursors have thermal decomposition
patterns similar to other metal carboxylate precursors studied
previously.24-26
Samples with higher Ce0.7Zr0.3O2 loadings make it difficult
to observe the δ-Al2O3 component in XRD patterns due to the
relative peak intensity differences from the higher “z” Ce/Zr
materials.
For example, Figure 1 shows a TGA trace for Zr(O2CCH2-
CH3)2(OH)2.24 Initial mass losses (8%) are due to propionic acid
of recrystallization. Thereafter, mass loss events are attributed
to the decomposition of the propionate ligands as suggested in
reactions 1-3.24-26
Average Particle Sizes (APSs). The APSs for these materials
were estimated from Debye-Scherer line broadening, their
SSAs (Table 1) and via DLS for the (CeOx)0.5@(Al2O3)0.5 and
(Ce0.7Zr0.3O2)0.5@(Al2O3)0.5 nanopowders. The first two
methods give very similar results. APS values for all of the
powders generated are 10-15 nm. The average SSAs for the
Ce/Zr samples are ∼33 ( 2 m2/g, and those for the Ce/Zr/Al
samples are 45 ( 2 m2/g. Figure S3 DLS data show APSs of
40-50 and 60-80 nm for (CeOx)0.5@(Al2O3)0.5 and
(Ce0.7Zr0.3O2)0.5@(Al2O3)0.5, respectively.
Note that the higher z value of the Ce/Zr component might
be expected to make it difficult to determine particle sizes
from X-ray line broadening. Fortunately, the BET derived
values are similar. Furthermore, for the higher alumina loaded
samples, the alumina might be expected to dominate the
observed surface areas and therefore the particle sizes. The
fact that the data are uniform for both methods over all types
of powders suggests that both measurement methods are
valid. However, DLS, SEM, and TEM provide further
substantiation.
Zr(O2CCH2CH3)2(OH)2 f Zr(O2CCH2CH3)(OH)3 +
CH3CHdCdO
Calcd (Found) Mass Losss ) 19.01% (19%)
Zr(O2CCH2CH3)(OH)3 f Zr(OH)4+ CH3CHdCdO
Calcd (Found) Mass Losss ) 19.01% (19%)
(2)
Zr(OH)4 f ZrO2+ 2H2O
(3)
Calcd (Found) Mass Losss ) 12.22% (12%)
Final ceramic yields (42% for ZrO2) are within experimental
error of the calculated value (41.75%) from the decomposition
of the precursor [Zr(CH3CH2COO)2(OH)2] to ZrO2 and are as
expected on the basis of previous studies.24-26
XRD Studies. XRD was used to characterize the phase
compositions of as-processed LF-FSP nanopowders. The XRD
patterns (Figure S1, Supporting Information) show that the (111)
peak of cubic ceria shifts from its standard position 28.55° 2θ
to 28.91° 2θ in the Ce0.7Zr0.3O2 composition. We attribute this
(111) peak shift to the formation of a solid solution. Because
the standard peak position for the ceria (111) peak (PDF file:
43-1002) is 28.55° 2θ, and that for the zirconia (111) peak (PDF
file: 42-1164) is 29.80° 2θ, we can calculate the Ce0.7Zr0.3O2
Although the DLS data (Figure S3) offer particle sizes
roughly double those determined by BET and line broadening,
the DLS data rely on standards (see Supporting Information,
experimental section) to estimate particle sizes none of which
are smaller than 300 nm. As such, the precision of the DLS
9
9224 J. AM. CHEM. SOC. VOL. 131, NO. 26, 2009

One-Step Synthesis of Core-Shell Nanopowders via LF-FSP
A R T I C L E S
Figure 2. SEM images of (a) (CeO2)0.7(ZrO2)0.3, (b) (CeO2)0.3(ZrO2)0.7, (c) [(CeO2)0.7(ZrO2)0.3]x (Al2O3)1-x for x ) 0.7, and (d) [(CeO2)0.7(ZrO2)0.3]x (Al2O3)1-x
for x ) 0.5.
3+
Table 1. APSs and SSAs of as-produced LF-FSP Samples
Table 2. Possible Maximum Residual (Ce0.7Zr0.3
)
Species in
δ-Alumina
Debye-Scherer BET-derived
particle size (nm) particle size (nm)
sample
SSA (m2/g)
3+
mass gain as O2
content (wt %)
possible maximum (Ce0.7Zr0.3)
species (mol (0.1%)a
sample
[(CeO2)0.7(ZrO2)0.3]x (Al2O3)1-x
for x ) 0.9
[(CeO2)0.7(ZrO2)0.3]x (Al2O3)1-x
for x ) 0.7
[(CeO2)0.7(ZrO2)0.3]x (Al2O3)1-x
for x ) 0.5
[(CeO2)0.7(ZrO2)0.3]x (Al2O3)1-x
for x ) 0.3
(CeO2)0.1(ZrO2)0.9
(CeO2)0.3(ZrO2)0.7
(CeO2)0.5(ZrO2)0.5
16
13
13
11
20
22
23
24
42
45
46
47
CZ30A70
CZ50A50
CZ70A30
0.16
0.18
0.20
0.3
0.3
0.4
a Assumes all reduced species are (Ce0.7Zr0.3
) .
3+
to demonstrate that the particle populations produced by LF-
FSP do not include any obvious micrometer-size particles;
although there are some weak agglomerates. Bell and Rodriguez
demonstrated that LF-FSP δ-Al2O3 nanopowders disperse
perfectly without any evidence of aggregates.51
TEM Images. Discussions of actual size/size distributions are
not appropriate if based solely on TEM micrographs, unless
combined with the XRD, BET, and/or DLS results. To begin
with, Figure S4 shows a TEM image of as-shot Ce0.7Zr0.3O2.
Faceted crystals are observed, similar to those seen by Stark,
Jossen, et al.50
Figure 3 TEMs of (Ce0.7Zr0.3O2)0.5@(Al2O3)0.5 and (Ce0.7Zr0.3-
O2)0.3@(Al2O3)0.7 particles reveal well-defined lattice fringes
indicating formation of single crystal cores. The (111) plane
d-spacings for both are the same, 3.07 Å, confirming the
formation of a cubic Ce0.7Zr0.3O2 solid solution per Figure S2.
Moreover, the combined XRDs match those of the individual
component XRDs of Ce0.7Zr0.3O2 solid solutions and δ-Al2O3.24a
16
16
16
16
16
28
28
28
26
22
32
32
32
33
36
(CeO2)0.7(ZrO2)0.3
(CeO2)0.9(ZrO2)0.1
data cannot be confirmed. One might argue that the DLS data
suggest the formation of agglomerates or even aggregates;
however, then the interpretation would be that these are two
or three particle agglomerates or aggregates. Given the small
size distribution, this argument seems inappropriate. How-
ever, it is important to note that the DLS do not show a
bimodal distribution of particle sizes. In addition, the TEM
data shown below suggest that the DLS particle size data
may be high.
SEM. SEM images (Figures 2) are presented to demonstrate
the uniformity of both powder systems. SEM resolution is
insufficient to carefully characterize individual particles, but it
does provide a view of the general population. The goal here is
(51) Bell, N. S.; Rodriguez, M. A. J. Nanosci. Nanotechnol. 2004, 4, 283.
9
J. AM. CHEM. SOC. VOL. 131, NO. 26, 2009 9225

A R T I C L E S
Kim and Laine
Figure 3. (a) TEM images of (Ce0.7Zr0.3O2)0.5(Al2O3)0.5 and (b) (Ce0.7Zr0.3O2)0.3(Al2O3)0.7
.
Figure 4. TGA of as-processed (Ce0.7Zr0.3O2)x(Al2O3)1-x ramped at 10 °C/
min/air.
Figure 5. DTA of as-processed (Ce0.7Zr0.3O2)x(Al2O3)1-x nanopowders.
FTIR Studies. Once particle morphologies were characterized,
a detailed picture of particle surface chemistries and thermal
behavior was developed through FTIR examination per Figure
S5.
All of the materials exhibit weak νO-H absorptions in the
3700-2500 cm-1 region, attributable to surface hydroxyls
arising from both physi- and chemisorbed water.54 For the two
Al2O3-rich samples (CZ5050, CZ30A70), νC-H bands were
observed between 3000 and 2700 cm-1, indicating traces of
hydrocarbons on the surfaces.54
It is important to note that because the Tm’s of both oxides
are well above LF-FSP flame temperatures, particle growth must
occur by gas phase nucleation followed by gas to solid addition
of ions with preferential growth on selected crystal planes.
Because Al2O3 has a lower vaporization temperature (3000 °C)
than CeO2 (3906 °C) or ZrO2 (5155 °C),52,53 ZrO2 and CeO2
should co-condense and nucleate first followed by Al2O3 in LF-
FSP processing. Thus, we initially assumed that Ce0.7Zr0.3O2
solid-solution single crystal nanoparticles form first during
quenching, then δ-alumina wets the Ce0.7Zr0.3O2 nanoparticles
forming core-shell nanopowders relatively uniformly.
However, the LF-FSP derived [Ce0.7Zr0.3O2]x@(Al2O3)1-x
nanopowders shown in Figure 3, in contrast to pure CeO2,
Ce0.7Zr0.3O2, or Ce0.5Zr0.5O2,50 are spherical with APSs typically
<30 nm with the vast majority <20 nm. Furthermore, the single
crystal cores are also spherical. One conclusion discussed in
more detail below is that the particle growth mechanisms for
the two sets of materials appears to be quite different, greatly
influenced by the presence of AlOx ions in the gas phase which
may also influence their catalytic behavior.
From 1800 to 1400 cm-1, peaks attributable to traces of
surface confined CO2 and carbonates are observed again in
accord with those seen for pure δ-Al2O3.25a Peak intensities for
the CO2 and carbonates are proportional to the amount of Al2O3.
Absorption bands in the 1000-400 cm-1 region correspond to
two νAl-O bands at 810 (stretching vibrations of tetrahedrally
coordinated Al-O) and 610 cm-1 (octahedral coordination).54,55
56
The νCe-O band (470 cm-1
the limits of the instrument.
)
is not seen clearly because of
TGA-DTA Studies. Figure 4 records the mass loss events
for the (Ce0.7Zr0.3O2)x@(Al2O3)1-x nanopowders. All as-
(52) CRC Handbook of Chemistry and Physics, 80th ed.; CRC Press: Boca
Raton, FL, 1999.
(53) Peri, J. B. J. Phys. Chem. 1965, 69, 211.
(54) Tarte, P. Spectrochim. Acta 1967, 23A, 2127.
(55) Saniger, J. M. Mater. Lett. 1995, 22, 109.
(56) Harrison, P. G.; Daniell, W. Chem. Mater. 2001, 13, 1708.
9
9226 J. AM. CHEM. SOC. VOL. 131, NO. 26, 2009

One-Step Synthesis of Core-Shell Nanopowders via LF-FSP
A R T I C L E S
indicate the presence of significant amounts of Ce3+ ions.57-60
For comparative purposes the (CeO2)x(Al2O3)1-x data are
provided in Figure 7. The XPS patterns for Ce3+/Ce4+ in
(Ce0.7Zr0.3O2)x@(Al2O3)1-x nanopowders are very similar to our
previous (CeO2)x(Al2O3)1-x studies. The Ce3+/4+ ratios obtained
by quantification of the XPS data (CasaXPS program) are given
in Table 3.
Because significant amounts of Ce3+ were observed in the
(CeOx)x(Al2O3)1-x study with XPS patterns similar to the
(Ce0.7Zr0.3O2)x@(Al2O3)1-x nanopowders, we assume that
similar concentrations of Ce3+ are present in our (Ce0.7Zr0.3-
O2)x@(Al2O3)1-x nanopowders. These results are unexpected
given that the LF-FSP process occurs in a highly oxidizing
environment and may relate to core-shell effects wherein the
shell prevents oxidation of the Ce3+ to Ce4+ in the highly
oxidizing environment. Alternately, it suggests that some
significant fraction of Ce3+ ions may be present in the δ-Al2O3
shell, which in retrospect is not surprising given our baseline
studies.24
Consequently, we also sought to determine if there were
significant alterations to the Zr3+ contents versus those expected
based on the above calculations. Unfortunately, because of the
relatively higher amounts and peak intensities of ceria in XPS,
the intensities of the Zr ion peaks were negligible in the
(Ce0.7Zr0.3O2)x@(Al2O3)1-x XPS data.60-62
As an alternative, we re-examined samples from our previ-
ously published work on (ZrO2)1-x(Al2O3)x.24a The XPS studies
for these materials are shown in Figure 8 along with NIST
standards.63
The Figure 7a,b data coupled with the NIST XPS database
and related studies on Zr ions51-54 support the existence of
mixtures of Zr2+/3+/4+ ions in our (ZrOx)x(Al2O3)1-x materials.
According to the NIST database and previous studies,63-66 the
XPS peak centered at 181.2 eV indicates the presence of
zirconium suboxide containing significant amounts of Zr2+/3+
ions. In the 13 and 50 mol % ZrOx data in Figure 7, we observe
mixtures of typical Zr4+ peaks (182 and 179.8 eV) and Zr2+/3+
peaks (181.2 eV) as broad shapes.
The coexistence of Ce3+ and Zr3+ ions provides one possible
explanation for the differences in the XRD data we report here
and those of Sutorik and Baliat, or Stark, Madler, et al.45,50 It
may be that (Ce3+/Zr3+)2O3 solid solutions form in LF-FSP
processing that are more stable than previous expectations for
these materials in oxidizing environments. A further point to
be made is that the presence of Zr2+/3+ species may be the reason
these materials offer good catalytic activity for a variety of
oxidizing reactions and the water gas shift reaction normally
requiring Pt or Pd as a co-catalyst as found in the accompanying
paper.48 The development of Pt- or Pd-free catalysts for such
Figure 6. XPS of Ce3+/4+ in CZ50A50 sample.
Table 3. Quantification of (CeOx)x(Al2O3)1-x XPS Data24
sample
Ce3+/Ce4+ ratio
6 mol % CeOx in Al2O3
75 mol % CeOx in Al2O3
0.17
0.19
processed powders exhibit 1-1.5 wt % mass loss up to ∼300
°C, attributed to evolution of both physi- and chemisorbed water.
The mass loss differences between air and nitrogen in the
TGA (Figure S6a-c) and peak intensities around 300 °C in
the DTA (Figure 5) are proportional to the amount of alumina
in three different samples. From the FTIR data (Figure S5), we
see hydrocarbon species from νC-H bands in alumina rich
samples (CZ30A70, CZ50A50).
Thus, we presume that the mass differences between air and
nitrogen in the TGA and peak intensities around 300 °C in the
DTA are due to oxidation of organics trapped on the surfaces
of the alumina rich samples.
At higher temperatures, we observe mass gains (TGA at
g1200 °C) with a corresponding exotherm in the DTA (Figure
5). We believe this is attributable to the oxidation of (Ce/Zr)3+
species substituting for Al3+ species in the δ-Al2O3 lattice during
LF-FSP processing, as observed in previous studies.24,25a It
seems reasonable to suggest that the (Ce/Zr)3+ species in the
δ-Al2O3 lattice segregate out during the δ to R-Al2O3 transfor-
mation and oxidize to (Ce/Zr)O2 given the low solubility of
these ions in R-Al2O3.31
In Table 2, we use these TGA mass gains to calculate
approximate amounts of residual (Ce/Zr)3+ species that oxidize
during the δ- to R-Al2O3 transformation. Similar calculations
were done for Zr3+ ions present in the δ-Al2O3 lattice in related
work on (ZrO2)x(Al2O3)1-x core-shell nanopowders during
transformation to R-Al2O3.24 In these studies it was determined
that the amounts of Zr3+ present are much higher than
anticipated on the basis of the literature where its solubility in
alumina was reported to be negligible.24 These initial observa-
tions are now supported by the following XPS studies.
XPS Studies. XPS (see Experimental Section) was used to
further quantify the amounts of Ce3+/Zr3+ ions in our nanopo-
wders. The XPS data for the CZ50A50 sample (Figure 6)
(57) Zhu, H. Y.; Hirata, T. J. Mater. Sci. Lett. 1993, 12, 749.
(58) Gigola, C. E.; Moreno, M. S.; Costilla, I. Appl. Surf. Sci. 2007, 254,
325.
(59) Brenier, R.; Mugnier, J.; Mirica, E. Appl. Surf. Sci. 1999, 143, 85.
(60) Galtayries, A.; Sporken, R.; Riga, J.; Blanchard, G.; Caudano, R. J.
Electron. Spectrosc. Relat. Phenom. 1998, 88-91, 951.
(61) Wang, X.; Lu, G.; Guo, Y.; Xue, Y.; Jiang, L.; Guo, Y.; Zhang, Z.
Cat. Today 2007, 126, 412.
(62) Nelson, A. E.; Schulz, K. H. Appl. Surf. Sci. 2003, 210, 206.
(63) NIST X-Ray Photoelectron Spectroscopy Database. (NIST Standard
Reference Database 20, Version 3.5).
(64) Kumar, L.; Sarma, D. D.; Krummacher, S. Appl. Surf. Sci. 1988, 32,
309–319.
(65) Bastiannon, A.; Braicovich, L.; Michelis, B. D. Surf. Sci. 1992, 264,
423–428.
(66) De Gonzalez, C. O.; Garcia, E. A. Surf. Sci. 1988, 193, 305.
9
J. AM. CHEM. SOC. VOL. 131, NO. 26, 2009 9227

A R T I C L E S
Kim and Laine
Figure 7. XPS of (a) 6 mol % CeOx in Al2O3 (b) 75 mol % CeOx in Al2O3.24
Figure 8. XPS analysis of (ZrO2)1-x(Al2O3)x (x ) 0-0.5) nanopowders where (a) x ) 0.5, (b) x ) 0.77, and (c) x ) 0.96.24 (d) Equivalent to the NIST
standard for ZrO2. (e) NIST standard for ZrOx.63-66
reactions would be of tremendous commercial importance and
deserves further study.
Like many of our previous studies, LF-FSP as-produced
nanopowders generated by rapid quenching provide access to
novel kinetic products not expected from traditional processing
methods that typically drive formation through thermodynamic
control. Thus, these materials may offer unique opportunities
for varieties of applications.
As mentioned above, the single-crystal Ce0.7Zr0.3O2 cores in
these materials are spherical rather than faceted, as seen in
Figure S4. This difference most likely arises because of the
presence of Al3+ species during particle growth in the gas phase.
Thus, a further argument for the distinct differences in catalytic
behavior between our studies,48 and the current literature on
the behavior of these systems might result from the incorporation
of Al3+ ions into the Ce0.7Zr0.3O2 cores again changing their
chemical reactivity.
Conclusions
The above results suggest that LF-FSP processing can provide
low-cost, efficient routes to well-known catalyst materials. The
9
9228 J. AM. CHEM. SOC. VOL. 131, NO. 26, 2009

One-Step Synthesis of Core-Shell Nanopowders via LF-FSP
A R T I C L E S
accompanying paper48 suggests that these materials offer novel
properties not expected based on the performance of traditional
catalysts.
other conventional processing method. These nanopowders may
offer novel potential for catalytic applications.
Acknowledgment. We would like to thank the Air Force Office
of Scientific Research, through Contracts No. F49620-03-1-0389
and subcontract from UES Inc. F074-009-0041/AF07-T009, Maxit
Inc., and Diamond Innovations Inc. for support of this work. We
also thank Dr. George W. Graham (refs 40 and 46) for extensive
discussions.
LF-FSP provides access to (CeOx)x(ZrO2)1-x and
(Ce0.7Zr0.3O2)0.x(Al2O3)1-x mixed-metal oxide nanopowders with
exceptional control of stoichiometry and phase purity. We were
able to produce nanopowders of any composition in the
Ce-Zr-O and Ce-Zr-Al-O systems with specific surface
area of g30 m2/g at rates of 50-100 g/h. We were able to
produce nano Ce0.7Zr0.3O2 solid solution single crystals in the
(CeOx)x(ZrO2)1-x and (Ce0.7Zr0.3O2)0.x(Al2O3)1-x systems.
Finally, we succeeded in producing core-shell nanoparticles
in the (Ce0.7Zr0.3O2)0.x(Al2O3)1-x system in a single step with
the correct choice of metalloorganic precursors. Because LF-
FSP offers rapid quenching of the combustion species, it
provides access to new, kinetic materials not accessible by any
Supporting Information Available: Experimental analytical
methods, including BET, XRD, SEM, TEM, DLS, FTIR, TGA-
DTA, and XPS and selected corresponding data. This material
is available free of charge via the Internet at http://pubs.acs.org.
JA9017545
9
J. AM. CHEM. SOC. VOL. 131, NO. 26, 2009 9229