Structurally characterized erbium alkoxides for use as an amphoteric dopant in PErZT ceramic thin film and nanoparticles

Article abstract of DOI:10.1002/ejic.200700314

A family of erbium alkoxides [Er(OR)₃] was prepared from the reaction of Er[N(SiMe₃)₂]₃ with a series of alcohols (HOR) in selected solvents and crystallographically characterized as: [Er(μ-Onep)₂(Onep)]₄, (1), [Er₄(μ ₄-O)(μ₃-Onep)(μ-Onep)₅-(Onep) ₄(py)₃] (3), Er₃(μ₃-OtBu) ₂(μ-OtBu)₃(OtBu)₄(HOtBu)₂ (4), Er₃(μ₃-O)(μ-OtBu)₄(OtBu) ₃(py)₃ (7), Er(dmp)₃(solv)₃ [solv = thf (9), py (10)], [Er(η-dip)(dip)₂]₂ (11), Er(dip)₃(thf)₂ (12), [Er(μ-OH)(dbp)₂(thf)] ₂ (15), Er(dbp)₃(py)₂ (16), [Er(μ-tps)-(tps)₂]₂ (17), Er(tps)₃(solv) ₃ [solv = thf (18) and (py) (19)] where Onep = neopentoxide (OCH ₂CMe₃), OtBu = tert-butoxide (OCMe₃), dmp = 2,6-dimethylphenoxide [OC₆H₃-(Me)₂-2,6], dip = 2,6-diisopropylphenoxide [OC₆H₃-(CHMe₂) ₂-2,6], dbp = 2,6-di-tert-butylphenoxide [OC₆H ₃-(CMe₃)₂-2,6], tps = triphenylsiloxide [OSi(C₆H₅)₃], tol = toluene, thf = tetrahydrofuran, and py = pyridine. The structures observed and data collected for the Er family of compounds are, in general, consistent with those reported previously for the Dy congeners over the range of solvated mono-, di-, tri-, and tetranuclear species. Representative members of the Er(OR)₃ precursors were used for the production of a PErZT precursor solution, which was subsequently used to generate thin films and nanoparticles. Further, these select Er precursors were used to generate Er₂O₃ nanoparticles. The full characterization of this family of compounds is reported along with the results of the materials investigations. Wiley-VCH Verlag GmbH & Co. KGaA, 2007.

Full text of DOI:10.1002/ejic.200700314

FULL PAPER
DOI: 10.1002/ejic.200700314
Structurally Characterized Erbium Alkoxides for Use as an Amphoteric
Dopant in PErZT Ceramic Thin Film and Nanoparticles
Timothy J. Boyle,*[a] Leigh Anna M. Ottley,[a] Luke N. Brewer,[b] Jennifer Sigman,[b]
Paul G. Clem,[b] and Jacob J. Richardson[c]
Keywords: Ceramics / Perovskites / Nanomaterials / Lanthanide alkoxides / PZT
A family of erbium alkoxides [Er(OR)3] was prepared from
the reaction of Er[N(SiMe3)2]3 with a series of alcohols (HOR)
in selected solvents and crystallographically characterized
as: [Er(µ-Onep)2(Onep)]4, (1), [Er4(µ4-O)(µ3-Onep)(µ-Onep)5-
(Onep)4(py)3] (3), Er3(µ3-OtBu)2(µ-OtBu)3(OtBu)4(HOtBu)2
(4), Er3(µ3-O)(µ-OtBu)4(OtBu)3(py)3 (7), Er(dmp)3(solv)3 [solv
= thf (9), py (10)], [Er(η-dip)(dip)2]2 (11), Er(dip)3(thf)2 (12),
[Er(µ-OH)(dbp)2(thf)]2 (15), Er(dbp)3(py)2 (16), [Er(µ-tps)-
(tps)2]2 (17), Er(tps)3(solv)3 [solv = thf (18) and (py) (19)]
where Onep = neopentoxide (OCH2CMe3), OtBu = tert-
butoxide (OCMe3), dmp = 2,6-dimethylphenoxide [OC6H3-
ene, thf = tetrahydrofuran, and py = pyridine. The structures
observed and data collected for the Er family of compounds
are, in general, consistent with those reported previously for
the Dy congeners over the range of solvated mono-, di-, tri-,
and tetranuclear species. Representative members of the
Er(OR)3 precursors were used for the production of a PErZT
precursor solution, which was subsequently used to generate
thin films and nanoparticles. Further, these select Er precur-
sors were used to generate Er2O3 nanoparticles. The full
characterization of this family of compounds is reported
along with the results of the materials investigations.
(Me)2-2,6], dip
=
2,6-diisopropylphenoxide [OC6H3-
(CHMe2)2-2,6], dbp = 2,6-di-tert-butylphenoxide [OC6H3-
(CMe3)2-2,6], tps = triphenylsiloxide [OSi(C6H5)3], tol = tolu-
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2007)
compounds) to determine the optimal Dy-dopant precur-
sor.[8] From this novel family of compounds, the 2,6-diiso-
propylphenoxide (dip) derivative yielded the highest quality
PDyZT films with impressive fatigue resistance. In an
effort to further explore this phenomenon we were inter-
ested in determining where the edges of the ACE stopped.
Fortunately, the Ln series possesses a systematic change in
the +3 Ln cations, which allows for a controlled probing of
the ACE phenomenon in PZT ceramic thin films. However,
for the smaller lanthanides, there are surprisingly few struc-
turally characterized alkoxides available in the litera-
ture.[9–11]
Therefore, to further develop this neglected area and
identify new sol-gel precursors, we undertook the synthesis
and characterization of a systematically varied family of
Er(OR)3 in a variety of solvents. These compounds were
isolated following Equation (1), where the alcohol (H-OR)
ranged from H-Onep = neopentyl alcohol (H-OCH2CMe3),
H-OtBu = tert-butyl alcohol (H-OCMe3), H-dmp = 2,6-
dimethylphenol [H-OC6H3(Me)2-2,6], H-dip = 2,6-diiso-
propylphenol [H-OC6H3(CHMe2)2-2,6], H-dbp = 2,6-di-
Introduction
The perovskite phase of lead zirconium titanate (PZT)
has found utility in a wide range of applications including
ferroelectric nonvolatile random access memories, sports
equipment, window tinting, sensors, smart cards, and nu-
merous other electronic applications.[1–6] However, PZT ce-
ramic materials demonstrate rapid fatigue when cycled on
standard platinized silica (Pt/Si) supports. Recently, we re-
ported on the “aliovalent cation effect” (ACE) where the
PZT thin films fatigue properties were improved through
selective doping with lanthanide (Ln) cations that could oc-
cupy either the A or B site.[7,8] In particular, the Dy-doped
PZT films (PDyZT) were found to possess high remanent
polarization (Pr), high dielectric constant (εЈ), and reduced
fatigue when cycled on Pt/Si wafers.[7,8] Due to the void of
acceptable dysprosium alkoxide [Dy(OR)3] precursors (see
Table 1), it was necessary to synthesize and characterize a
series of systematically varied Dy(OR)3 compounds (21 new
[a] Sandia National Laboratories, Advanced Materials Laboratory,
1001 University Boulevard SE, Albuquerque, New Mexico
87105, USA
tert-butylphenol [H-OC6H3(CMe3)2-2,6], H-tps
= tri-
Fax: +1-505-272-7336
E-mail: tjboyle@Sandia.gov
phenylsilanol [H-OSi(C6H5)3] using a variety of solvents: tol
= toluene, thf = tetrahydrofuran, and py = pyridine. The
general matrix used in this study led to the synthesis of
compounds 1–19 and is shown in Table 1 along with a com-
parison to the Dy(OR)3 compounds that were previously
[b] Sandia National Laboratories, New Mexico,
P. O. Box 5800, Albuquerque, NM 87185-1411, USA
[c] University of California – Santa Barbara, Materials Depart-
ment,
Santa Barbara, CA 93106-5050, USA
Eur. J. Inorg. Chem. 2007, 3805–3815
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3805

T. J. Boyle et al.
FULL PAPER
Table 1. List of structure types isolated for the matrix of ligands characterized including Er5(µ5-O)(µ3-OiPr)4(µ-OiPr)4-
[10,11]
and Er(dip)3(thf)2.[9] The remainder of alkoxy-
and solvents used to explore Er(OR)3 and Dy(OR)3 structures. The
(OiPr)5
nuclearity is listed and in parentheses the type of interaction used
by the ligands to form the molecule and/or the number of solvent
molecules bound.
like species possess other metals,[17–21] halides,[22] or poly-
functional ligands[23] which prevents use for generation of
high quality ceramic oxide materials due to the potential
cross cation, anion, and/or residual ligand contamination.
Due to the lack of available Er(OR)3 compounds, it was
necessary to synthesize a series of precursors to determine
the optimal species. A study of a matrix of sterically varied
precursors in different solvents was undertaken (see
Table 1). For this report, only those compounds that were
crystallographically characterized will be discussed.
Er
tol
thf
py
HOR
[a]
[a]
[a]
Onep 1 tetra (µ, 0)
3 tri (µ and µ3; 3)
[a]
OtBu
dmp
7 tri (µ and µ3; 3) 4 tri (µ and µ3; 2)
[a]
[a]
9 mono (3)
12 mono (2)
15 di (2)
10 mono (3)
[a]
[a]
[a]
dip
dbp
tps
11 di (π; 0)
[a]
16 mono (2)
19 mono (3)
17 di (µ, 0)
18 mono (3)
Dy
tol
thf
py
NH3
Onep tetra (µ, 0)
OtBu tri (µ and µ3) tri (µ and µ3; 2) tri (µ and µ3; 2)
tri (µ and µ3; 2) tri (µ and µ3; 2)
hex
[a]
Synthesis
dmp di (π; 0)
dip di (π; 0)
dbp mono (0)
tps di (µ)
mono (3)
mono (2)
mono (1)
mono (3)
mono (3)
mono (3)
mono (2)
mono (3)
linear tri (2)
mono (2)
On the basis of our previous experience with the prob-
lems associated with the ammoniacal route, we elected to
use the amide/alcohol metathesis route that proved benefi-
mono (1)
[a]
[a] No acceptable crystal solutions reported.
cial for the Dy system.[8] The Er(NR2)3, where NR2
=
N(SiMe3)2, was synthesized according to established routes
involving the metathesis reaction between ErCl3 and
3 equiv. of KNR2 in thf.[8,24–29] The resultant powder iso-
lated from the metathesis route was sublimed and crys-
tallized to ensure pure starting material. It is of note that
purification of the Er(NR2)3 is critical to the successful iso-
lation of crystalline material. Once isolated, this precursor
was treated with a variety of alcohols as shown in Equation
(1) (Table 1).
The majority of samples remained clear after stirring for
12 h; however, those that formed a precipitate, typically the
more sterically demanding ligands, were warmed slightly to
redissolve the powder. Crystals were either isolated by slow
evaporation or rotary evaporation of the volatile portion of
the reaction mixture followed by cooling the reaction mix-
ture to –35 °C, if necessary.
isolated.[8] From this matrix, the following Er(OR)3 com-
pounds were crystallographically characterized as: [Er(µ-
Onep)2(Onep)]4 (1), [Er4(µ4-O)(µ3-Onep)(µ-Onep)5(Onep)4-
(py)3] (3), Er3(µ3-OtBu)2(µ-OtBu)3(OtBu)4(HOtBu)2 (4),
Er3(µ3-O)(µ-OtBu)4(OtBu)3(py)3 (7), Er(dmp)3(solv)3 [solv
= thf (9), py (10)], [Er(η-dip)(dip)2]2 (11), Er(dip)3(thf)2
(12), [Er(µ-OH)(dbp)2(thf)]2 (15), Er(dbp)3(py)2 (16), [Er(µ-
tps)(tps)2]2 (17), Er(tps)3(thf)3 (18), Er(tps)3(py)3 (19). The
synthesis and full characterization of these compounds are
discussed. Suitable crystals were not isolated for com-
pounds 2 (Onep/thf), 5 (OtBu/tol), 6 (OtBu/thf), 8 (dmp/
tol), 13 (dip/py), 14 (dbp/tol) and additional discussion of
these compounds will not be pursued since their structures
were not unequivocally established.
Er[N(SiMe3)2]3 + 3 H-ORЈǞEr(ORЈ)3 + 3 H-N(SiMe3)2
(1)
FTIR data of the resultant powders revealed no amide
stretches and, for the majority of samples no –OH stretches,
which implies complete substitution had occurred between
the amide and the alcohol. Two exceptions to this were the
spectra noted for compound 4 and 15, which showed broad
IR stretches around 3000 cm–1 associated with the bound
H-OtBu and the -OH, respectively. In general, elemental
analyses were found to be consistent with the single-crystal
X-ray structures isolated (vide infra) for these compounds.
Those that were off slightly from expected values were sol-
vated species and the high volatility of the bound solvent
often causes significant problems in obtaining accurate
analyses.[8,30] Due to the paramagnetic nature of the Er
metal center, we were unable to get meaningful NMR spec-
troscopic data either in solution or solid-state. Therefore,
single-crystal X-ray diffraction studies were undertaken to
assist in elucidating the identity of the final products iso-
lated.
Representative species from this novel family of com-
pounds were used to explore their utility for the production
of ceramic materials. The ACE for PErZT thin films was
probed using 11 as the dopant precursor due to the similar-
ity with the Dy counterpart (Table 1).[7,8] Further, since new
properties were expected on the nanoscale, attempts were
made to generate Er2O3 and PErZT nanoceramics. With
this wide variety of structurally varied species, it was pos-
sible to further explore the “precursor structure argument”
(PSA)[12–16] for nanoparticle morphology predictions. A
solution route to nanoparticles of Er2O3 was discovered
using 1 and 11. Full details of the synthesis, characteriza-
tion, and subsequent properties of these precursors and ma-
terials are presented.
Results and Discussion
A systematically varied series of Er(OR)3 was required in
order to initiate the material evaluation of Er ceramics with
controlled properties. A search of the literature reveals only
Crystal Structures
As mentioned only two structures of Er(OR)3 have pre-
two Er(OR)3 had been previously crystallographically viously been reported.[9–11] Obviously, no general trends can
3806
www.eurjic.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2007, 3805–3815

Erbium Alkoxides as Dopants in PErZT Ceramic Thin Film and Nanoparticles
FULL PAPER
be discerned from this limited set of precursors but com-
parisons to literature compounds will be made when appro-
priate. Figures 1, 2, 3, and 4 are the structure plots of repre-
sentative crystal structures for this family of compounds.
The various compounds are discussed below in order of the
ligand increasing steric bulk of the ligand.
Figure 2. Structure plot of 4. Thermal ellipsoids of heavy atoms
are drawn at the 30% level and carbon atoms are shown as stick
and ball for clarity. Selected bond lengths [Å]: Er(1)–O(1) 2.305(3);
Er(1)–O(2) 2.371(3); Er(1)–O(3) 2.269(3); Er(1)–O(4) 2.525(4);
Er(1)–O(5) 2.067(3); Er(1)–O(6) 2.250(3). Selected bond angles [°]:
O(5)–Er(1)–O(6) 106.17(13); O(5)–Er(1)–O(3) 104.93(13); O(6)–
Er(1)–O(3) 143.78(12); O(5)–Er(1)–O(1) 109.45(13); O(6)–Er(1)–
O(1) 76.35(12); O(3)–Er(1)–O(1) 76.16(12); O(5)–Er(1)–O(2)
177.71(13); O(6)–Er(1)–O(2) 74.47(12); O(3)–Er(1)–O(2) 73.78(12);
O(1)–Er(1)–O(2) 68.48(11); O(5)–Er(1)–O(4) 74.73(14); O(6)–
Er(1)–O(4) 101.31(13); O(3)–Er(1)–O(4) 104.40(12); O(1)–Er(1)–
O(4) 175.59(12); O(2)–Er(1)–O(4) 107.37(12).
Figure 1. Structure plot of 1. Thermal ellipsoids of heavy atoms
are drawn at the 30% level and carbon atoms are shown as stick
and ball for clarity. Selected bond lengths [Å]: Er(1)–O(1) 2.033(5);
Er(1)–O(2) 2.241(6); Er(1)–O(3) 2.244(6). Selected bond angles [°]:
O(1)–Er(1)–O(3) 122.8(3); O(1)–Er(1)–O(2) 112.9(3); O(3)–Er(1)–
O(2) 72.7(2); O(1)–Er(1)–O(3) 118.3(3); O(2)–Er(1)–O(3) 89.5(3);
O(1)–Er(1)–O(2) 105.8(3); O(3)–Er(1)–O(2) 86.9(2); O(3)–Er(1)–
O(2) 71.7(2).
Figure 3. Structure plot of 11. Thermal ellipsoids of heavy atoms
are drawn at the 30% level and carbon atoms are shown as stick
and ball for clarity. Selected bond lengths [Å]: Er(1)–O(1)
2.096(18); Er(1)–O(3) 2.069(17); Er(1)–O(2) 2.095(18); Er(1)-ring
2.81 Å. Selected bond angles [°]: O(2)–Er(1)–O(1) 109.3(7); O(2)–
Er(1)–O(3) 102.5(7); O(3)–Er(1)–O(1) 106.3(7); O(1)–Er(1)–ring
108.1(9); O(2)–Er(1)–ring 83.9(9); O(3)–Er(1)–ring 140.6(8).
Using the more sterically demanding Onep ligand re-
duced the nuclearity and prevented the formation of an oxo
ligand as noted for the smaller OiPr derivative Er5(µ5-
O)(µ3-OiPr)4(µ-OiPr)4(OiPr)5.[10,11] The resultant species,
identified as [Er(µ-Onep)2(Onep)]4 (1), is consistent with the
other Ln- and Group 3 neopentoxide species (Figure 1)[31]
wherein, each of the four trigonal bipyramidal (TBP) Er Lewis basic solvents (i.e., thf, py, or HOtBu) however, the
metal centers possess two µ-Onep and one Onep ligand. Er corollary could not be isolated without bound solvents
Attempts to isolate this compound using the Lewis basic as noted for the pyridine adduct, Er3(µ3-O)(µ-OtBu)4-
solvents thf (2) did not lead to X-ray quality crystals; how- (OtBu)3(py)3 (7). Again, the py adduct appears to produce
ever, from py the tetramer observed in 1 was reduced to a the oxo which argues for water present in the py; however,
trinuclear species surprisingly forming an oxo species, numerous structures using the same solvent did not yield
[Er(µ3-O)(µ-Onep)6(Onep)4(py)3] (3). The source of the oxo the oxide, so the origin of the oxo is still unproven. This
species is not known as of yet and typically not observed complex is similar to 4 and other trinuclear species where
for the Onep derivatives, unless esterification mechanisms the terminal ligands are now replaced by these stronger.
from side reactions with carboxylic acids occur but may be
Aryl oxides are often used to control nuclearity because
the steric bulk around the metal center can be easily al-
due to adventitious water in the solvent.[30–32]
Increasing the steric bulk to the OtBu groups, led to tered.[8,12,13,30,33–36] We routinely use the most sterically de-
crystallization of the trinuclear 4 which had additional manding di-ortho-substituted phenoxide ligands to induce
HOtBu bound to it, as shown in Figure 2. Previously, we as large an effect as possible. The resulting products are
isolated a Ce/OtBu tetramer[30] that did not possess any again discussed in increasing steric demand (dmp, dip,
Eur. J. Inorg. Chem. 2007, 3805–3815
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
3807

T. J. Boyle et al.
FULL PAPER
and 19, for thf and py. The bulky tps ligands allow for tetra-
hedral (Td) arrangements for the metal centers of 17 and
octahedral (Oh) geometries for 18 and 19 through the bind-
ing of three solvent molecules.
In general, the structures noted for the Er system were
consistent with the Dy(OR)3 precursors previously re-
ported;[8] however, 15 does prove that unaccounted varia-
tions are possible, and probable. Surprisingly, a number of
compounds could not be successfully isolated for the Er
system that were easily isolated for the Dy system.[8] Subtle
effects that dictate crystal growth are at play and require
further investigations.
Figure 4. Structure plot of 16. Thermal ellipsoids of heavy atoms
are drawn at the 30% level and carbon atoms are shown as stick
and ball for clarity. Selected bond lengths [Å]: Er(1)–O(2) 2.095(4);
Er(1)–O(1) 2.161(3); Er(1)–O(1A) 2.161(3); Er(1)–N(1) 2.469(4);
Er(1)–N(1A) 2.469(4). Selected bond angles [°]: O(2)–Er(1)–O(1)
112.54(8); O(2)–Er(1)–O(1A) 112.54(8); O(1)–Er(1)–O(1A)
134.92(16); O(2)–Er(1)–N(1) 101.96(9); O(1)–Er(1)–N(1) 81.96(12);
O(1)–Er(1)–N(1A) 88.91(13); O(2)–Er(1)–N(1A) 101.96(9); O(1A)–
Materials
With these compounds successfully characterized, a suf-
ficient number and variety of species were available for sys-
Er(1)–N(1A) 88.91(13); O(1A)–Er(1)–N(1) 88.96(12); N(1)–Er(1)– tematic materials structure–property investigations. Two
N(1A) 156.08(18).
systems were investigated: Er-doped PZT (or PErZT) thin
films for electronic properties and nanomaterials of Er2O3
and PErZT.
dbp). For the dmp derivative, large clear crystals were easily
isolated from toluene but the diffraction was poor and an
Film Characteristics
acceptable structure solution could not be garnered for 8.
Therefore, we employed the Lewis basic solvents to assist [PErZT(4/30/70)]
The precursor solutions of Pb0.98(Er)0.04(Zr0.3Ti0.7 0.98
)
O3
O3
and Pb0.98(Er)0.04(Zr0.52Ti0.48 0.98
)
in growing acceptable crystals. For both thf and py, mono- [PErZT(4/52/48)] were prepared as described in the experi-
nuclear species with three solvents to form an octahedral mental section. Selection of the Er dopant on the B site of
geometry were isolated as Er(dmp)3(solv)3 [solv = thf (9) the perovskite is based on previous studies and a dopant
and py (10)].
concentration of 4% Er was selected to maximize the pos-
Increasing the steric bulk of the ortho substituent to iso- sible effects of the dopant without creating unwanted sec-
propyl groups led to the isolation of 11 (Figure 3). The η- ondary phases. Dektak profilometer film thickness mea-
interaction noted for 11 is the same as previously noted for surements were 150–180 nm after the deposition of 4 layers.
the La,[37] Nd,[9] and Dy[8] systems. Again, the η-interaction X-ray diffraction patterns and hysteresis loops are shown in
was easily disrupted using Lewis basic solvents forming the Figure 5, a–d. Both films were found to be polycrystalline,
trigonal bipyramidal (TBP) monomeric species 12 for the single-phase perovskite films (Figure 5, b and d) with an
thf adduct.
enhanced (111) orientation, attributed to the templating ef-
Using the most sterically demanding dbp ligand, crystals fects of the Pt(111)/Ti/SiO2/Si substrates. The PErZT(4/30/
of the unsolvated species 14 could not be isolated in high 70) films also showed a more pronounced (100) peak than
quality; however, in the presence of a Lewis base, two very PErZT(4/52/48) films. This is consistent with the room tem-
different species were characterized. Out of thf, a dinuclear perature tetragonal structure of a PZT(30/70) film com-
thf adduct was isolated as 15. Surprisingly, the dinuclear pared to the room temperature morphotropic structure of a
species possesses a µ-OH group which was not identified for PZT(52/48) film. The previously reported PZT(30/70) films
the Dy series[8] but have been reported for other lanthanides isolated under identical conditions used for the PErZT
species including, [(µ-OH)Yb(OAr)2(thf)]2 (OAr = dbp,[38] films, had narrow and square loops with reasonable proper-
dbp-But-4,[38,39]) and [(µ-OH)Sm(dbp-Me-4)2(OPPh3)]2.[40] ties: spontaneous polarization (Pr) = 24 µC/cm2, coercive
Increasing the Lewis basicity of the solvent to py led to the field (Ec) = 93 kV/cm2, a dielectric constant (εЈ) = 396, and
TBP monomer 16 (Figure 4). It is most likely not due to tan δ = 0.007 (10 kHz, 100 mV). For the PDyZT(4/30/70)
adventitious water in the solvent or it would have been ob- films, comparable values were obtained: Pr = 26 µC/cm2, Ec
served in the Dy system.[8]
Siloxide ligands offer the potential for the production of 100 mV).[8]
= 126 kV/cm, εЈ = 343, and tan δ = 0.009 (10 kHz,
Si-containing materials and slightly different ligand proper-
After top-electrodes were sputter-deposited, the ferro-
ties both sterically and electronically. Therefore, we added electric properties were discerned for PErZT films. Hystere-
the tps ligand into our structural investigation. The tps de- sis loops for the PErZT(4/30/70) and PErZT(4/52/48) films
rivatives isolated for the Er system are consistent with the are shown in Figure 5a and Figure 5c, respectively. In agree-
Dy(tps)3 structure[8] wherein a dinuclear complex was iso- ment with the Dy-doped films, the PErZT(4/30/70) films
lated from tol (17) with µ-tps and terminal tps in standard displayed a Pr = 30 µC/cm2, Ec = 171 kV, εЈ = 366 and
edge-shared tetrahedral-bound metal centers. Subsequent tan δ = 0.03. For the PErZT(4/52/48) films slightly lower
monomers were isolated from Lewis basic solvents as 18 ferroelectric property values were obtained, with a Pr =
3808
www.eurjic.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2007, 3805–3815

Erbium Alkoxides as Dopants in PErZT Ceramic Thin Film and Nanoparticles
FULL PAPER
cessing temperatures than the Ln-acac route.[43] The results
for solution precipitation of 1 and 11 are presented, fol-
lowed by the results of the solvothermal routes.
Solution Precipitation
The reaction pathway for both of the alkoxides appeared
to follow identical pathways, wherein the pale pink solution
of the precursor immediately formed a white precipitate
upon injection into the refluxing solution. The subsequent
isolation also proceeded in an identical manner. Therefore,
all of the observed differences in the generated manner can
only be attributed to the variations in pendant hydrocarbon
chain of the alkoxide ligand.
For 1, the resulting powder was found to be made of 7–
8 nm highly crystalline Er2O3 (see Figure 6) particles, as
found by TEM imaging, the observed SAED pattern, and
only Er and O observed in the EDS spectrum (Cu is back-
ground from the grid). The PXRD spectrum was fairly
broad but consistent with Er2O3 nanoparticles of 1.7 nm,
based on the Scherrer equation calculations. In comparison,
using 11 (Figure 7) 3–4 nm nanoparticles of Er2O3 were
formed as characterized by TEM and EDS analyses with a
noted decrease in crystallinity by SAED. For select grains,
HRTEM images clearly show the lattice fringes of Er2O3.
The PXRD was broad peaks consistent with Er2O3 wherein
calculations indicated nanoparticles of 1.9 nm were formed.
Figure 5. Characterization of PErZT(4/30/70): (a) hysteresis loop,
(b) XRD; and characterization of PErZT (4/52/48): (c) hysteresis
loop, (d) XRD.
20 µC/cm2 and an Ec = 137 kV; however, the dielectric
properties improved: εЈ = 564 and tan δ = 0.01. The higher
Pr observed in the PErZT(4/30/70) is attributed to the room
temperature tetragonal composition of this film while the
higher εЈ observed in the PErZT(4/52/48) is attributed to
the filmЈs room temperature morphotropic boundary com-
position. Unfortunately, the fatigue testing for the Er-doped
films did not show the significant improvement noted for
the Dy doping.[7,8] However, this sets a lower limit for the
ACE phenomenon and the slightly larger Y and Ho cations
are now under investigation.
Nanoparticles
With this family of well-characterized compounds in
hand, it was decided to investigate their utility as precursors
for nanomaterials. Nanoparticles are of interest due to the
novel properties expected for materials on this size scale. In
the literature, 5 to 30 nm sized erbium(III) oxide nanocrys-
tals dispersed in titania have been previously isolated by
employing Er(NO3)3 in an inverse microemulsion process
followed by a heat treatment at 600 °C.[41,42] The resulting
materials were identified as Er2O3 by powder XRD and
photoluminescent measurements.[41,42] Recently, a solution
route using decomposition of lanthanide acetylacetonate in
oleic acid, oleylamine, and octadecene at 310 °C yielded
Er2O3 nanoparticles.[43] We are interested in controlling the
properties of nano-Er2O3 for electrooptic applications.
Therefore, we undertook two different approaches to gener-
ate nanoparticles that were more amenable to large-scale
syntheses: (a) solution precipitation and (b) solvothermal
using our novel representative Er(OR)3 precursors 1 and
Figure 6. TEM images of Er2O3 generated from 1 using the solu-
tion precipitation route: (a) TEM image (50 nm scale bar), (b)
HRTEM (5 nm scale bar), and (c) PXRD.
The difference in the observed particle sizes between the
11. Both of the processing routes are greatly simplified in TEM and XRD measurements may be due to several fac-
comparison to the microemulsion route,[41,42] are amenable tors in the X-ray diffraction measurement. Particle size af-
to large scale synthesis, and require significantly lower pro- fects the PXRD pattern by broadening the peak. Several
Eur. J. Inorg. Chem. 2007, 3805–3815
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
3809

T. J. Boyle et al.
FULL PAPER
Figure 7. TEM images of Er2O3 generated from 11 using the solu-
tion precipitation route: (a) TEM image (40 nm scale bar), (b)
HRTEM (5 nm scale bar), and (c) PXRD.
Figure 8. Nanoparticles from Parr bomb using 1: (a) TEM image
(50 nm scale bar), (b) HRTEM(10 nm scale bar), and (c) PXRD.
factors contribute to this peak broadening, including in-
strument line broadening, reduction in particle size, varia-
tion in particle composition, and microstrain. Of these pos-
sible effects, variation in particle composition and some in-
strument line broadening may account for the systematic
underestimation of the particle size from the PXRD mea-
surement. Despite this difference, the PXRD measurement
of particle size is useful for efficient measurement of many
samples and for particle size measurement on materials
with strong agglomeration.
In general, from this study the formation of crystalline
Er2O3 was more easily achieved with 1 than 11. Upon intro-
duction of a Lewis base, both compounds alter their origi-
nal structure forming a trinuclear solvated species for 1 (as
evidenced by 4, and 7) and a solvated monomer for 11 (9,
10, 12, 16, 18, and 19). Using the PSA, the size of the result-
ant nanoparticles could be reasoned; however, the morpo-
hologies did not follow any trends. The faster and more
uniform the nanoparticles form, the longer they will have
to crystallize under the conditions we used. The particle
precursor for 1 is more likely to have a larger nucleation
seed in comparison to that of monomeric 11. Thus, 1 yields
higher quality nanoparticles as determined by the nature of
the solvated structure.
Solvothermal Synthesis
Under solvothermal conditions, a Parr digestion bomb
of the same solution was investigated to determine pro-
cessing variations for the same compounds. For 1, spheroi-
Figure 9. Nanoparticles from Parr bomb using 11: (a) TEM image
dal nanoparticles of Er2O3 were isolated ranging from 10– (50 nm scale bar), (b) HRTEM (5 nm scale bar), and (c) PXRD.
3810
www.eurjic.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2007, 3805–3815

Erbium Alkoxides as Dopants in PErZT Ceramic Thin Film and Nanoparticles
FULL PAPER
15 nm in size verified by EDS and PXRD (Figure 8). Sur- ever, the PXRD pattern shows only ErTiO3 is present as a
prisingly, the PXRD pattern was much sharper with calcu- crystalline material. The absent Pb and Zr cations are prob-
lated sizes of 7.0 nm. For 11, more spherical shaped nano- ably phase separated as amorphous PbZrO3, which requires
particles were isolated in the same approximate size range significantly higher temperatures to crystallize. Alterna-
of 10–15 nm (Figure 9). The particles were more uniform in tively, the solvothermal route led to the isolation of par-
morphology and size in comparison to 1 but the particles ticles on the size of 5–10 nm (Figure 11) with each of the
appear less crystalline by PXRD pattern; however, HRTEM cations present in the EDS. Interestingly, highly crystalline
reveals the crystalline lattice planes. The broad peaks of particles (HRTEM with d spacing ca. 3.0 Å) were observed
PXRD pattern were calculated to generate particles of in the PbTiO3 structure that is consistent with PZT (3.1 Å).
3.7 nm. The TEM images and those calculated from the Further work to improve the uniformity and consistency of
XRD pattern have a significant difference again the issues the multication nanomaterial in order to explore the prop-
discussed previous (vide infra) may account for this vari- erties of these electroceramic materials is underway.
ance.
In general, the structures of the precursors do not appear
to play as large a role in determining the morphology of
the final Ln-based nanomaterials isolated as noted for tran-
sition metals.[12–16] However, there are significant differ-
ences based on the structural considerations of 1 vs. 11. The
crystallinity of the final material appears to be higher with
the bomb preparative routes vs. the solution precipitation
methods.
PErZT Nanoparticles
Due to the success with the simple Er system, it was of
interest to determine if more complex nanoparticles could
be generated from the same solution under similar condi-
tions. Attempts to generate PErZT nanomaterials were un-
dertaken using the same precursor thin film solution with
1 as the dopant and the process for preparing nano-Er2O3
as discussed above. The solution precipitation route yielded
large nanoparticles on the order of 30 nm in size (Fig-
ure 10). EDS indicates each of the cations is present; how-
Figure 11. PErZT nanoparticles bomb prep: calcined at 650 °C: (a)
TEM image (50 nm scale bar), (b) HRTEM (5 nm scale bar), and
(c) PXRD.
Conclusions
We have synthesized and fully characterized a family of
Er(OR)3 from the reaction of Er[N(SiMe3)2]3 and a variety
of commercially available alcohols that systematically var-
ied in steric bulk. The isolated compounds are structurally
similar to the Dy(OR)3 species[8] isolated previously with
nuclearity ranging from mono- (9, 12, 16, 18, 19), to bis-
(11, 15, 17), to tri- (6) to tetranuclear (1) species. From this
set of Er(OR)3 compounds 11 was used to generate PErZT.
The ferroelectric properties were consistent with the Dy
doped PZT; however, the fatigue properties were not im-
Figure 10. PErZT nanoparticles MeIM/H2O prep: calcined at
650 °C: (a) TEM image (50 nm scale bar), (b) HRTEM (5 nm scale
proved as noted for the Dy species.[7,8] Additional work to
bar), and (c) PXRD.
elucidate the ACE demarcation line is in progress with
Eur. J. Inorg. Chem. 2007, 3805–3815
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
3811

T. J. Boyle et al.
FULL PAPER
2345 (w), 1593 (s), 1469 (s), 1441 (s), 1428 (S). 1374 (m,sh), 1290
(s), 1269 (m,sh), 1226 (m), 1153 (m), 1092 (s), 1068 (m), 1039 (m),
1007 (m), 978 (m), 917 (m), 871 (w,sh), 849 (s), 757 (s), 748 (s), 623
(m), 569 (w), 532 (s), 489 (w), 462 (w) cm–1. C36H51ErO6 (747.06):
calcd. C 57.9, H 6.90; found C 58.31, H 6.92.
other large Ln cations. Nanoparticles of Er2O3 were suc-
cessfully synthesized and isolated; however, the more com-
plex ternary PErZT yielded mixed size particles of various
compositions.
Er(dmp)3(py)4 (10): Used KNR2 (0.500 g, 0.771 mmol), H-dmp
(0.282 g, 0.231 mmol) and ca. 10 mL of py. Yield 0.450 g (68.9%).
Experimental Section
FT-IR (KBr): ν = 2960 (m,sh), 2918 (s), 2850 (s), 2363 (w), 2345
˜
All compounds described below were handled with rigorous ex-
clusion of air and water using standard Schlenk line and glove box
techniques. All solvents were stored under argon and used as re-
ceived (Aldrich) in sure seal bottles, including hexanes (hex), tol,
thf, and py. The following chemicals were used as received (Al-
drich): ErCl3, KN(SiMe3)2 (KNR2), H-Onep, H-OtBu, H-dmp, H-
dip, H-dbp, H-tps, [Zr(OiPr)4(HOiPr)]2. Ti(OiPr)4 was distilled im-
mediately prior to use and stored under and inert atmosphere.
Er(NR2)3 was synthesized from the reaction of ErCl3 and three
equivalents of LiNR2 in thf and purified by sublimation and subse-
quent crystallization from hexanes.[8] FT-IR data were obtained on
a Bruker Vector 22 Instrument using KBr pellets under an atmo-
sphere of flowing nitrogen. Elemental analysis was performed on a
Perkin–Elmer 2400 CHN-S/O Elemental Analyzer.
(w), 1592 (m), 1466 (s), 1443 (w,sh), 1428 (w,sh), 1378 (w), 12889
(w,sh), 1264 (s), 1220 (w), 1092 (s), 1070 (w,sh), 1038 (w), 918
(w,sh), 874 (m), 847 (m), 801 (s), 756 (m), 720 (w), 704 (s), 623 (w),
534 (m) cm–1. C44H47ErN4O3 (847.14): calcd. C 2.4, H 5.59, N 6.61;
found C 62.1, H 5.67, N 6.68.
[Er(η-dip)(dip)2]2 (11): Used KNR2 (0.500 g, 0.771 mmol), H-dip
(0.412 g, 0.231 mmol) and ca. 10 mL of tol. Yield 0.450 g (84.7%).
1.0618 g FT-IR (KBr): ν = 3063 (w,sh), 3020 (w,sh), 2961 (s), 2869
˜
(m), 2362 (m), 2344 (w), 1601 (m,sh), 1591 (m), 1460 (w,sh), 1433
(s), 1382 (m), 1361 (m), 1333 (s), 1267 (s), 1202 (m), 1155 (w), 1069
(m), 1068 (w), 1059 (w), 1042 (m), 1007 (m), 978 (w), 924 (s), 889
(m), 864 (m,sh), 843 (*s), 797 (m), 692 (w,sh). 669 (w), 650 (w), 627
(w), 567 (m), 489 (w), 470 (w) cm–1. C71H93Er2O6 (1377.03): calcd.
C 61.9, H 6.82; found C 62.4, H 6.89.
General Synthesis: The appropriate HOR was added by a pipette
to a stirring mixture of Er(NR2)3 that was dissolved in desired sol-
vent. After 12 h, if a precipitate formed, the solution was warmed
until the precipitate redissolved. After cooling to room tempera-
ture, the volatile material was allowed to slowly evaporated or the
volume of the reaction was drastically reduced via rotary evapora-
tion. The reaction was then set aside or cooled to –35 °C, if neces-
sary until crystals formed.
Er(dip)3(thf)2 (12):[9] Used KNR2 (0.500 g, 0.771 mmol), H-dip
(0.412 g, 0.231 mmol) and ca. 10 mL of thf. Yield 0.320 g (49.2%).
FT IR (KBr): ν = 3057 (w), 2960 (s), 2890 (w,sh), 2362 (w), 2345
˜
(w), 1588 (s), 1450 (w,sh), 1432 (s), 1381 (m), 1357 (m,sh), 1333 (s),
1274 (s), 1210 (s), 1174 (w), 1109 (m), 1097 (m), 1042 (s), 1016 (m),
953 (w), 934 (w,sh), 888 (m,sh), 864 (s), 804 (m), 754 (s), 691 (s),
617 (w), 566 (s) cm–1. C44H67ErO5 (843.27): calcd. C 62.7, H 8.03;
found C 61.9, H 7.98.
[Er(µ-Onep)2(Onep)]4 (1): See ref.[31]
.
[Er(µ-OH)(dbp)2(thf)]2 (15): Used KNR2 (0.500 g,0.771 mmol), H-
Er(µ-O)(µ3-Onep)(µ-Onep)5(Onep)4(py)3 (3): Used KNR2 (0.500 g,
dbp (0.477 g, 0.231 mmol) and ca. 10 mL of thf. Yield 0.380 g
0.771 mmol), H-Onep (0.204 g, 0.231 mmol) and ca. 10 mL of py.
(73.9%). FT-IR (KBr): ν = 2957 (s), 2917 (w,sh), 2872 (w,sh), 2362
˜
Yield 0.215 g (62.1%). FT-IR (KBr): ν = 2957 (s), 2906 (w), 2902
˜
(s), 2344 (s), 1476 (w), 1459 (m), 1407 (s), 1385 (m,sh), 1316 (w),
1244 (s), 1200 (m), 1122 (m), 1104 (s), 1045 (w), 1007 (m), 922 (w),
865 (s), 821 (s), 796 (w), 748 (s), 708 (m), 656 (s), 527 (m), 429 (m)
cm–1. C64H100Er2O8 (1332.01): calcd. C 57.7, H 7.56; found C 57.1,
H 7.37.
(m), 2809 (m), 2747 (w), 2681 (m), 2368 (m), 2348 (m), 1481 (s),
1462 (m,sh), 1394 (s), 1364 (s), 1334 (w), 1266 (m), 1216 (w), 1059
(s), 1061 (s), 1017 (s), 974 (w), 935 (m), 896 (m), 826 (m), 701 (m),
700 (m), 670 (w), 606 (s), 517 (m), 481 (s), 418 (s) cm–1.
C65H126Er4N3O11 (1794.76): calcd. C 43.5, H 7.08, N 2.34, for
C75H136Er4N5O11 (3 + 2py): calcd. C 46.1, H 7.02, N 3.59; found
C 46.5, H 7.20, N 3.05.
Er(dbp)3(py)2 (16): Used KNR2 (0.500 g, 0.771 mmol), H-dbp
(0.477 g, 0.231 mmol) and ca. 10 mL of py. Yield 0.390 g (53.7%).
FT-IR (KBr): ν = 2959 (s), 2366 (w), 2345 (s), 1442 (m,sh), 1408
˜
Er3(µ3-OtBu)2(µ-OtBu)3(OtBu)4(HOtBu)2 (4): Used KNR2
(0.500 g, 0.771 mmol), H-OtBu (0.171 g, 0.231 mmol) and ca.
(s), 1384 (m,sh), 1360 (w), 1251 (s), 1219 (m,sh), 1200 (m), 1154
(w), 1104 (m), 1068 (w), 1042 (m), 866 (s), 822 (m), 749 (s), 703 (s),
655 (s), 629 (m), 545 (w), 431 (s) cm–1. C52H73ErN2O3 (941.42):
calcd. C 66.3, H 7.82, N 2.98; found C 66.8, H 8.10, N 2.42.
10 mL of tol. Yield 0.220 g (65.5%). FT-IR (KBr): ν = 2968 (s),
˜
2362 (s), 2344 (s,sh), 1375 (m), 1356 (w,sh), 1203 (s), 1153 (w), 1142
(w), 1071 (w), 1039 (w,sh), 1004 (s), 938 (m), 920 (m,sh), 845 (m),
766 (w), 753 (w), 700 (m), 681 (m), 636 (w), 527 (m), 479 (m) cm–1.
C44H99Er3O11 (1306.04): calcd. C 40.47, H 7.64, for C40H89Er3O11
(4–1 H-OtBu): calcd. C 38.49, H 7.18; found C 38.6, H 6.93.
[Er(µ-tps)(tps)2]2 (17): Used KNR2 (0.500 g, 0.771 mmol), H-tps
(0.639 g, 0.231 mmol) and ca. 10 mL of tol. Yield 0.480 g (59.9%).
FT-IR (KBr): ν = 3069 (s), 3043 (m,sh), 3021 (w), 3008 (w), 2998
˜
(m), 2958 (s), 2366 (m), 2342 (m), 1975 (w,sh), 1958 (m), 1902
(w,sh), 1886 (m), 1818 (s), 1773 (m), 1652 (m), 1588 (s), 1560 (w),
1482 (s), 1426 (s), 1329 (w), 1301 (w,sh), 1259 (s), 1251 (w,sh), 1115
(s), 1093 (s,sh), 1027 (m,sh), 995 (w,sh), 857 (w,sh), 842 (s), 746 (s),
710 (s,sh), 698 (s), 621 (w), 571 (m), 486 (s) cm–1. C115H98Er2O6Si6
(2079.07): calcd. C 66.4, H 4.75; found C 66.6, H 5.26.
Er3(µ3-O)(µ-OtBu)4(OtBu)3(py)3 (7): Used KNR2 (0.500 g,
0.771 mmol), H-OtBu (0.171 g, 0.231 mmol) and ca. 10 mL of py.
Yield 0.25 g (76.5%). FT-IR (KBr): ν = 2961 (s), 2362 (s), 2344
˜
(m), 1618 (s), 1561 (m), 1509 (m), 1378 (w), 1242 (m), 1206 (m).
1108 (s), 1061 (s), 1016 (m), 933 (w,sh), 829 (s), 748 (m), 669 (s),
594 (w), 474 (w) cm–1. C43H78Er3N3O8: cald. C 40.76, N 3.316, H
6.20. C28H63Er3O8 (7–3 py): calcd. C 32.70, H 6.17; found C 32.4,
H 6.81.
Er(tps)3(thf)3 (18): Used KNR2 (0.500 g,0.771 mmol), H-tps
(0.639 g, 0.231 mmol) and ca. 10 mL of thf. Yield 0.720 g (72.9%).
FT-IR (KBr): ν = 3063 (m), 3045 (m), 2994 (w), 2882 (w), 1459
˜
Er(dmp)3(thf)3 (9): Used KNR2 (0.500 g, 0.771 mmol), H-dmp (m), 1450 (m), 1428 (s), 1342 (w), 1331 (w), 1296 (w), 1257 (m),
(0.282 g, 0.231 mmol) and ca. 10 mL of thf. Yield 0.350 g (60.8%). 1185 (s), 1111 (s), 1068 (m), 1047 (w,sh), 1027 (s), 1000 (w,sh), 984
FT-IR (KBr): ν = 3621 (s), 3066 (w,sh), 3039 (w,sh), 3010 (w,sh). (s), 916 (w), 875 (s), 742 (s), 704 (s), 672 (w), 519 (s), 508 (w,sh),
˜
2957 (s), 2920 (w,sh), 2853 (w,sh), 2722 (w), 2585 (w), 2368 (w), 452 (m), 432 (m), 413 (m) cm–1. C66H69ErO6Si3·C4H8O: calcd. C
3812
www.eurjic.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2007, 3805–3815

Erbium Alkoxides as Dopants in PErZT Ceramic Thin Film and Nanoparticles
FULL PAPER
65.6, H 6.05; for C66H69ErO6Si3 (18–1 thf): calcd. C 65.5, H 5.76;
found C 65.1, H 5.96.
sor mixture. The appropriate Er(OR)3 precursor was dissolved in
py and then added to the PZT mixture to form the ca. 0.4  PErZT
precursor solution.
Er(tps)3(py)3 (19): Used KNR2 (0.500 g, 0.771 mmol), tps(0.639 g,
0.231 mmol) and ca. 10 mL of py. Yield 0.680 g (71.7%). FT-IR
Multilayered films of the desired composition were spin-coat de-
posited, in air, onto Pt(1700 Å)/Ti(300 Å) coated, thermally oxid-
ized SiO2/Si substrates using a photoresist spinner (3000 rpm for
30 s). After each deposition, the films were heated on a hot plate
(300 °C for ca. 5 min) and cooled to room temperature before the
deposition of the next layer. After the final layer received the
300 °C treatment (for this study 2–4 layers were used), the film was
crystallized in a tube furnace at 650 °C for 30 min in air. X-ray
diffraction (XRD) was used to confirm the phase purity and orien-
tation of the final films.
(KBr): ν = 3063 (m), 3044 (m,sh), 3006 (w), 2363 (m), 2345 (m),
˜
1599 (s), 1579 (w, sh), 1524 (w), 1483 (s), 1442 (s), 1427 (s), 1273
(w), 1257 (w), 1231 (w), 1217 (m), 1185 (m), 1154 (w), 1109 (s),
1067 (s), 1038 (s), 1001 (m,sh), 985 (s), 745 (s), 701 (s), 622 (s),
602 (w, sh), 520 (s), 507 (w,sh), 453 (m), 433 (m), 411 (m) cm–1.
C69H60ErN3O3Si3 (1230.77): calcd. 67.3C, 4.91H, 3.42N.
C64H55ErN2O3Si3 (19–1 py): C 66.7, H 4.81, N 2.43; found C 67.0,
H 4.91, N 2.80.
[44]
General X-ray Crystal Structure Information:
Crystals were
mounted onto a glass fiber from a pool of Fluorolube and imme-
diately placed in a cold N2 vapor stream, with a Bruker AXS dif-
fractometer equipped with a SMART 1000 CCD detector using
graphite-monochromatized Mo-Kα radiation (λ = 0.7107 Å). Lat-
tice determination and data collection were carried out using
SMART Version 5.054 software. Data reduction was performed
using SAINTPLUS Version 6.01 software and corrected for ab-
sorption using the SADABS program within the SAINT software
package.
Ferroelectric Testing: To measure the electrical properties of the
films, gold top electrodes (ca. 250ϫ250 µm) were sputter-deposited
onto the films using a shadow mask to create a parallel-plate capac-
itor geometry. Capacitor top electrode areas were individually mea-
sured with a calibrated optical microscope and film thicknesses
were measured using a Dektak profilometer. The low-field dielec-
tric properties were measured at 10 kHz (100 mV p–p) using an
HP 4284A impedance analyzer (Palo Alto, CA). The ferroelectric
properties of the PLnZT films were measured using an RT66A
ferroelectric tester from Radiant Technologies (Albuquerque, NM).
Accelerated ferroelectric fatigue tests were performed using a 1-
MHz bipolar square wave pulse, with the applied electric field nor-
malized to ca. 700 kV/cm.
Structures were solved by direct methods that yielded the heavy
atoms, along with a number of the lighter atoms or by using the
PATTERSON method, which yielded the heavy atoms. Subsequent
Fourier syntheses yielded the remaining light-atom positions. The
hydrogen atoms were fixed in positions of ideal geometry and re-
fined using SHELXS software. The final refinement of each com-
pound included anisotropic thermal parameters for all non-hydro-
gen atoms. It is of note that crystal structures of M(OR)x often
contain disorder within the atoms of the ligand chain causing
higher than normal final correlations.[24–29]
Nanoparticle Synthesis: Nanoparticles of Er2O3 were formed fol-
lowing one of two methods using 1 or 11 which were chosen as
representative members of the alkyl- and aryloxide derivatives,
respectively. The BRP method was also used to generate the PErZT
precursor solution for the multication nanoparticles.
Two general routes were explored to synthesize nanoparticles: (a)
solution precipitation and (b) solvothermal routes. (a) Solution Pre-
cipitation. The first route used a mixture of 1-methyl imidazole and
water (MeIm/H2O) having a 14.5/0.5 mL ratio. This solution was
placed in a two-neck round bottom equipped with a reflux con-
denser and a septum. Under flowing argon, the MeIm/H2O solu-
tion was brought to reflux conditions (ca. 200 °C) and the desired
precursor dissolved in tol was injected via syringe. A white precipi-
tate formed immediately and the mixture was stirred for 30 min at
197 °C then cooled to room temperature. The as prepared particles
were collected by centrifugation and rinsed with ethanol and then
calcined in a tube furnace under flowing argon at 600 °C for 1 h.
(b) Solvothermal. The second route used was a 45 mL Parr Acid
Digestion Bomb. In the glove box, the desired precursors were dis-
solved in ca. 15 mL of benzyl alcohol (BzOH) and the clear solu-
tions were transferred to the TeflonTM liners and placed in the
bomb. The bomb was taken out of the glove box and the reaction
was heated to 200 °C for 48 h. The white precipitate was collected
by centrifugation and rinsed with ethanol.
Data collection parameters for 1–19 are given in Table 2. Specific
issues associated with individual structures are discussed below.
Fully refined structures of 3 and 7 could not be obtained due to
significant disorder in the Onep and OtBu ligands, respectively;
however, the connectivity of both compounds were unequivocally
established and are in agreement with the Dy(OR)3 structures re-
ported. Structure 11 had significant disorder in the isopropoxide
ligands that resulted in oblong thermal ellipsoids and a slightly
higher R1 value; however, the overall structure observed is in agree-
ment with [Dy(η-dip)(dip)2]2 in both unit cell and connectivity. For
17, it was necessary to squeeze disorder electron density of 317.1 A3
which is in agreement with ca. 1 tol molecule. Noncentrosymmetric
settings were chosen for 1, 7, 12, and 18 even though a mirror plane
appears to exist for these molecules. By further investigation of
these species, disruption in the symmetry due to nonrigid ligands,
such as Onep (1 and 7) or disorder in coordinating solvent thf
(12 and 18) ultimately requires refinement in noncentrosymmetric
settings.
CCDC-641110 to -641120 contain the supplementary crystallo-
graphic data for this paper. These data can be obtained free of
charge from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif.
The resulting nanopowders were slurried in tol followed by placing
a drop of the solution onto a holey-carbon copper grid and al-
lowing it to air dry. Upon complete drying for the samples were
then examined with a transmission electron microscope (TEM)
Philips CM30 with acceleration voltage of 300 KeV to obtain
images, as well as selected area electron diffraction (SAED) data
when necessary. Elemental composition information was obtained
Film Production: Standard spin-coat deposition routes[45] were used
to generate the PErZT thin films with a brief description of the
preparation and processing described below. The PZT (30/70 or 48/
52) films were produced following the established “basic route to with a Noran energy dispersive X-ray detector (EDX) using system
PZT” (BRP) route,[45,46] in which Pb(OAc)2·(HOAc) (Aldrich,
95%) dissolved in pyridine (py) was added to a mixture of Ti(OiPr)
4 and [Zr(OiPr)4(HOiPr)]2 dissolved in tol to form the PZT precur-
6 software. In addition, the powder was characterized using a PAN-
alytical powder diffractometer using Cu-Kα radiation with step size
0.0167 degree, with 53.340 seconds dwelling time.
Eur. J. Inorg. Chem. 2007, 3805–3815
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
3813

T. J. Boyle et al.
FULL PAPER
Table 2. Crystallographic data collection parameters for 1–19.
Compound
1
3
4
7
9
Empirical formula
Formula weight
Temp. [K]
C15H33ErO3
428.67
168(2)
C65H126Er4O11N3
1794.74
203(2)
monoclinic
P21/c
C44H99Er3O11
1306.01
168(2)
monoclinic
P21/c
C43H78Er3N3O8
1266.86
203(2)
hexagonal
P63/mc
C36H51ErO6
747.03
203(2)
monoclinic
P21/n
Space group
tetragonal
¯
P421/c
a [Å]
b [Å]
c [Å]
20.1985(14)
20.1985(14)
12.0226(16)
16.801(3)
13.371(2)
35.136(6)
90.479(3)
7893(2)
19.2913(13)
10.7831(8)
27.2513(19)
99.2430(10)
5595.2(7)
4
1.550
4.506
4.41(6.33)
8.45(9.25)
17.239(4)
17.239(4)
11.269(6)
9.8147(7)
27.6997(19)
13.1323(9)
93.4720(10)
3563.6(4)
4
1.392
2.395
3.03(4.34)
7.18(7.69)
β [°]
V [Å3]
4905.0(8)
8
1.161
3.424
4.10(4.77)
2900.2(17)
2
1.421
4.341
7.772(9.999)
19.99(22.21)
Z
4
Dcalcd. [Mg/m3]
4.530
4.259
12.05(12.76)
26.57(26.86)
µ (Mo-Kα) [mm–1]
[a]
R1 (%) (all data)
[b]
wR2 (%) (all data) 14.60(15.26)
Compound
10
11
12
15
16
Empirical formula
Formula weight
Temp. [K]
C44H47ErN4O3
847.12
203(2)
C71H93Er2O6
1376.97
203(2)
monoclinic
P21/c
C44H67ErO5
843.24
203(2)
monoclinic
P21
C64H100Er2O8
1331.96
203(2)
orthorhombic
Pbca
C52H73Er N2O3
941.38
203(2)
orthorhombic
Pbcn
Space group
triclinic
¯
P1
a [Å]
b [Å]
c [Å]
α [°]
10.8524(11)
14.0287(15)
27.954(3)
81.924(2)
87.581(2)
74.401(2)
4058.4(7)
4
1.386
2.110
4.65 (7.24)
8.87 (9.63)
17.399(9)
21.796(12)
9.796(5)
9.6910(6)
19.3993(12)
12.2166(8)
19.419(4)
15.633(3)
20.710(4)
20.394(3)
13.736(2)
17.450(3)
β [°]
105.827(7)
109.6430(10)
γ [°]
V [Å3]
3574(3)
2
1.280
2.377
14.33 (24.99)
28.59 (36.04)
2149(3)
2
1.295
1.980
2.29 (2.34)
6.17 (6.20)
6287(2)
4
1.407
2.701
2.32 (2.94)
5.77 (6.22)
4888.4(14)
4
1.279
1.758
4.67 (9.25)
8.21 (9.50)
Z
Dcalcd. [Mg/m3]
µ (Mo-Kα) [mm–1]
R1 (%) (all data)
wR2 (%) (all data)
Compound
17
18
19
Empirical formula
Form. weight
Temp. [K]
C115H98Er2O6Si6
2079.16
168(2)
C66H69ErO6Si3·C4H8O
1281.85
203(2)
monoclinic
P21
C69H60ErN3O3Si3
1230.73
203(2)
rhombohedral
R3c
Space group
triclinic
¯
P1
a [Å]
b [Å]
c [Å]
α [°]
13.6370(15)
14.0918(15)
14.4596(15)
75.603(2)
70.595(2)
71.970(2)
2459.0(5)
1
1.342
1.820
3.49 (4.11)
7.66 (7.87)
14.5349(15)
16.4587(17)
14.9029(15)
20.8803(5)
20.8803(5)
20.8803(5)
66.54
66.54
66.54
7343.8(3)
4
1.113
1.232
7.08 (9.50)
23.48 (26.26)
β [°]
115.687(2)
γ [°]
V [Å3]
3212.8(6)
2
1.251
1.409
6.69 (8.77)
17.15 (18.62)
Z
Dcalcd. [Mg/m3]
µ (Mo-Kα) [mm–1]
R1 (%) (all data)
wR2 (%) (all data)
2
2 2
[a] R1 = Σ | |Fo| – |Fc| |/Σ |Fo|ϫ100. [b] wR2 = [Σ w (Fo – Fc ) /Σ (w |Fo|2)2]1/2 ϫ100.
dia Corporation, a Lockheed Martin Company, for the United
States Department of Energy under contract DE-AC04–
94AL85000.
Acknowledgments
This work was partially funded by the Office of Basic Energy Sci-
ences and the National Institutes of Health through the NIH Road-
map for Medical Research, Grant #1 R21 EB005365–01. Infor-
mation on this RFA (Innovation in Molecular Imaging Probes) can
be found at http://grants.nih.gov/grants/guide/rfa-files/RFA-RM-
04–021.html. Sandia is a multiprogram laboratory operated by San-
[1] P. Muralt, J. Micromech. Microeng. 2000, 10, 136.
[2] S. Trolier-McKinstry, P. Muralt, J. Electroceram. 2004, 12, 7.
[3] P. Muralt, IEEE Transactions on Ultrasonics, Ferroelectrics and
Frequency Control 2000, 47, 903.
3814
www.eurjic.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2007, 3805–3815

Erbium Alkoxides as Dopants in PErZT Ceramic Thin Film and Nanoparticles
FULL PAPER
[4]
[26] N. Y. Turova, E. P. Turevskaya, V. G. Kessler, M. I. Yanov-
skaya, The Chemistry of Metal Alkoxide, Kluwer Academic
Publishers, Boston, 2002.
T. J. Zhu, L. Lu, M. O. Lai, Appl. Phys. A: Mater. Sci. Process.
2005, 701.
Y. Arimoto, H. Ishiwara, MRS Bull. 2004, 29, 823.
C. J. Taylor, G. N. Washington, Smart Structures and Materials
2002, Smart Structures and Integrated Systems, San Diego,
CA, 2002.
T. J. Boyle, P. G. Clem, B. A. Tuttle, G. L. Brennecka, J. T.
Dawley, M. A. Rodriguez, T. D. Dunbar, W. F. Hammetter, J.
Mater. Res. 2002, 17, 871.
T. J. Boyle, S. D. Bunge, P. G. Clem, J. Richardson, J. T. Daw-
ley, L. A. M. Ottley, M. A. Rodriguez, B. A. Tuttle, G. Avilu-
cea, R. G. Tissot, Inorg. Chem. 2005, 44, 1588.
D. M. Barnhart, D. L. Clark, J. C. Gordon, J. C. Huffman, In-
org. Chem. 1994, 33, 3487.
M. Kritikos, M. Moustiakimov, M. Wijk, G. Westin, J. Chem.
Soc. Dalton Trans. 2001, 1931.
G. Westin, M. Kritikos, M. Wijk, J. Solid State Chem. 1998,
141, 168.
T. J. Boyle, S. D. Bunge, T. M. Alam, G. P. Holland, T. J. Head-
ley, G. Avilucea, Inorg. Chem. 2005, 44, 1309.
T. J. Boyle, S. D. Bunge, N. L. Andrews, L. E. Matzen, K. Sieg,
M. A. Rodriguez, T. J. Headley, Chem. Mater. 2004, 16, 3279.
H. Gerung, T. J. Boyle, L. J. Tribby, S. D. Bunge, C. J. Brinker,
S. M. Han, J. Am. Chem. Soc. 2006, 128, 5244.
[5]
[6]
[27] K. G. Caulton, L. G. Hubert-Pfalzgraf, Chem. Rev. 1990, 90,
969.
[28] C. D. Chandler, C. Roger, M. J. Hampden-Smith, Chem. Rev.
1993, 93, 1205.
[29] L. G. Hubert-Pfalzgraf, New J. Chem. 1987, 11, 663.
[30] T. J. Boyle, L. J. Tribby, S. D. Bunge, Eur. J. Inorg. Chem. 2006,
4553.
[31] T. J. Boyle, L. A. M. Ottley, S. D. Daniel-Taylor, L. J. Tribby,
S. D. Bunge, A. L. Costello, T. M. Alam, J. C. Gordon, M.
McCleskey Inorg. Chem. 2007, 46, 3705.
[32] T. J. Boyle, R. P. Tyner, T. M. Alam, B. L. Scott, J. W. Ziller,
B. G. Potter, J. Am. Chem. Soc. 1999, 121, 12104.
[33] T. J. Boyle, M. A. Rodriguez, D. Ingersoll, T. J. Headley, S. D.
Bunge, D. M. Pedrotty, S. M. DeЈAngeli, S. C. Vick, H. Fan,
Chem. Mater. 2003, 15, 3903.
[34] T. J. Boyle, N. L. Andrews, M. A. Rodriguez, C. Campana, T.
Yiu, Inorg. Chem. 2003, 42, 5357.
[35] T. J. Boyle, D. M. Pedrotty, T. M. Alam, S. C. Vick, M. A. Rod-
riguez, Inorg. Chem. 2000, 39, 5133.
[36] S. D. Bunge, T. J. Boyle, H. D. Pratt, T. M. Alam, M. A. Rodri-
guez, Inorg. Chem. 2004, 43, 6035.
[37] R. J. Butcher, D. L. Clark, J. C. Gordon, J. G. Watkin, J. Or-
ganomet. Chem. 1999, 577, 228.
[38] G. B. Deacon, G. Meyer, D. Stellfeldt, G. Zelesny, B. W. Skel-
ton, A. H. White, Z. Anorg. Allg. Chem. 2001, 627, 1652.
[39] G. B. Deacon, T. Feng, S. Nickel, M. I. Ogden, A. H. White,
Aust. J. Chem. 1992, 45, 1992.
[40] Y.-M. Yao, K.-D. Zhang, L.-B. Yu, Q. Shen, Y.-Y. Lei, J. Sun,
Jiegou Huaxue 2003, 2, 143.
[41] W. Que, Y. Zhou, Y. L. Lam, K. Ppita, Y. C. Chan, C. H. Kam,
Appl. Phys. A; Mater. Sci. Process. 2001, 73, 209.
[42] W. Que, Y. Zhou, Y. L. Lam, J. Zhou, Y. C. Chan, C. H. Kam,
Appl. Phys. A; Mater. Sci. Process. 2001, 89, 3058.
[43] R. Si, Y.-W. Zhang, H.-P. Zhou, L.-D. Sun, C.-H. Yan, Chem.
Mater. 2007, 19, 18.
[44] The listed versions of SAINT, SMART, XSHELL, XPOW in
SHELXTL, and SADABS Software from Bruker Analytical X-
ray Systems Inc., 6300 Enterprise Lane, Madison, WI 53719
were used in analysis.
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15] H. Gerung, S. D. Bunge, T. J. Boyle, C. J. Brinker, S. M. Han,
Chem. Commun. 2005, 1914.
[16] T. J. Boyle, B. A. Hernandez-Sanchez, C. M. Baros, L. N.
Brewer, M. A. Rodriguez, Chem. Mater. 2007, 1, 2016.
[17] D. L. Clark, J. C. Gordon, J. C. Huffman, R. L. Vincent-Hollis,
J. G. Watkin, B. D. Zwick, Inorg. Chem. 1994, 33, 5903.
[18] W. J. Evans, M. A. Johnston, M. A. Greci, M. A. Ansari, J. C.
Brady, J. W. Ziller, Inorg. Chem. 2000, 39, 2125.
[19] M. Kritikos, M. Wijk, G. Westin, Acta Crystallogr. Sect. C
1998, 54, 576.
[20] G. Westin, R. Norrestam, M. Nygren, M. Wijk, J. Solid State
Chem. 1998, 135, 149.
[21] M. Wijk, R. Norrestam, M. Nygren, G. Westin, Inorg. Chem.
1996, 35, 1077.
[22] Y.-M. Yao, Q. Shen, Y. Shang, M.-Q. Xue, Polyhedron 2001,
20, 3201.
[23] M. R. Burgstein, P. W. Roesky, Angew. Chem. Int. Ed. 2000, 39,
549.
[24] D. C. Bradley, R. C. Mehrotra, D. P. Gaur, Metal Alkoxides,
Academic Press, New York, 1978.
[25] D. C. Bradley, R. C. Mehrotra, I. P. Rothwell, A. Singh, Alkoxo
and Aryloxo Derivatives of Metals, Academic Press, New York,
2001.
[45] T. J. Boyle, H. N. Al-Shareef, J. Mater. Sci. 1997, 32, 2263.
[46] Conquest Version 1.8, Cambridge Crystallographic Data Cen-
tre: support@ccdc.cam.ac.uk or www.ccdc.cam.ac.uk (2004,
AUG 2006 update).
Received: March 21, 2007
Published Online: July 3, 2007
Eur. J. Inorg. Chem. 2007, 3805–3815
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
3815